Miljø- og Fødevareudvalget 2015-16
L 68
Offentligt
1601764_0001.png
JRC Scientific and Technical Reports
Water Framework Directive intercalibration
technical report
Part 3: Coastal and Transitional waters
Edited by Alessandro Carletti and Anna-Stiina Heiskanen
EUR 23838 EN/3 - 2009
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0003.png
JRC Scientific and Technical Reports
Water Framework Directive intercalibration
technical report
Part 3: Coastal and Transitional waters
Edited by Alessandro Carletti and Anna-Stiina Heiskanen
EUR 23838 EN/3 - 2009
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0004.png
The mission of the Institute for Environment and Sustainability is to provide scientific-
technical support to the European Union’s Policies for the protection and sustainable
development of the European and global environment.
European Commission
Joint Research Centre
Institute for Environment and Sustainability
Contact information
Address: Via Enrico Fermi 2749
I-21020 Ispra (VA), Italy
E-mail: [email protected]
Tel.: +39 0332 789955
Fax: +39 032 789352
http://ies.jrc.ec.europa.eu/
http://www.jrc.ec.europa.eu/
Legal Notice
Neither the European Commission nor any person acting on behalf of the Commission is
responsible for the use which might be made of this publication.
Europe Direct is a service to help you find answers
to your questions about the European Union
Freephone number (*):
(*)
Certain mobile telephone operators do not allow access to 00 800 numbers or these calls may be billed.
A great deal of additional information on the European Union is available on the Internet.
It can be accessed through the Europa server http://europa.eu/
EUR 23838 EN/3
ISBN 978-92-79-12568-3
ISSN 1018-5593
DOI 10.2788/19561
Luxembourg: Office for Official Publications of the European Communities
© European Communities, 2009
Reproduction is authorised provided the source is acknowledged
Printed in Italy
00 800 6 7 8 9 10 11
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
Contents
Section 1 - Introduction
......................................................................................................
1 Introduction
...........................................................................................................................
2 Background
............................................................................................................................
3 Geographical Intercalibration Groups (GIGs) and common intercalibration types
......
3.1 Geographical intercalibration groups ...............................................................................
3.2 Common intercalibration types ........................................................................................
4 Methodology and Results
......................................................................................................
5 Discussion
...............................................................................................................................
5.1 Comparability between quality elements .........................................................................
5.2 Open issues and need for further work.............................................................................
6 Summary and Conclusions
...................................................................................................
7
7
7
11
11
11
15
17
17
18
18
19
19
19
19
19
21
24
26
34
34
35
35
35
35
35
42
43
43
43
44
44
75
3
Section 2 – Benthic Invertebrates
..................................................................................
1 Introduction
...........................................................................................................................
2 Methodology and results
.......................................................................................................
2.1 Baltic GIG ........................................................................................................................
2.1.1 Intercalibration approach
.....................................................................................
2.1.2 National methods that were intercalibrated..........................................................
2.1.3 Reference criteria and class boundary setting
......................................................
2.1.4 Results of the comparison
.....................................................................................
2.1.5 Results of the harmonization – Boundary EQR values
.........................................
2.1.6 Open issues and need for further work
.................................................................
2.2 Black Sea GIG..................................................................................................................
2.2.1 Intercalibration approach
.....................................................................................
2.2.2 National methods that were intercalibrated..........................................................
2.2.3 Reference conditions and class boundary setting
.................................................
2.2.4 Results of the comparison
.....................................................................................
2.2.5 Results of the harmonisation – Boundary EQR values
.........................................
2.2.6 Open issues and need for further work
.................................................................
2.3 Mediterranean GIG ..........................................................................................................
2.3.1 Intercalibration approach
.....................................................................................
2.3.2 National methods that were intercalibrated..........................................................
2.3.3 Reference conditions and class boundary setting
.................................................
2.3.4 Results of the comparison
.....................................................................................
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
2.3.5 Results of the harmonisation – Boundary EQR values
.........................................
2.3.6 Open issues and need for further work
.................................................................
2.4 NE Atlantic GIG ...............................................................................................................
2.4.1
2.4.2
2.4.3
2.4.4
2.4.5
2.4.6
75
75
76
Intercalibration approach
..................................................................................... 76
National methods intercalibrated
......................................................................... 77
Reference conditions and class boundary setting
................................................. 77
Results of the comparison
..................................................................................... 91
Results of the harmonisation – Boundary EQR values
......................................... 102
Open issues and need for further work
................................................................. 103
3 Discussion
............................................................................................................................... 104
3.1 Comparability between GIGs ........................................................................................... 104
3.2 Open issues and need for further work............................................................................. 105
4 References
............................................................................................................................... 105
5 Annex
...................................................................................................................................... 108
Section 3 – Phytoplankton
................................................................................................. 109
1 Introduction
........................................................................................................................... 109
2 Methodology and results
....................................................................................................... 109
2.1 Baltic GIG ........................................................................................................................ 109
2.1.3
2.1.4
2.1.5
2.1.6
2.1.7
2.1.8
2.2.1
2.2.2
2.2.3
2.2.4
2.3.1
2.3.2
2.3.3
2.3.4
2.3.5
4
Intercalibration approach
.....................................................................................
National methods that were intercalibrated..........................................................
Reference conditions and boundary setting
..........................................................
Results of the comparison and harmonisation
......................................................
Results of the harmonization – Boundary EQR values
.........................................
Open issue and need for further work
...................................................................
Intercalibration approach
.....................................................................................
Reference conditions and class boundary setting
.................................................
Results of the harmonization – Boundary EQR values
.........................................
Open issues and need for further work
.................................................................
Intercalibration approach
.....................................................................................
National methods that were intercalibrated..........................................................
Reference conditions and class boundary setting
.................................................
Results of the comparison and harmonization
......................................................
Open issues and need for further work
.................................................................
109
111
111
117
121
122
123
123
132
132
132
135
136
137
138
2.2 Black Sea GIG.................................................................................................................. 123
2.3 Mediterranean GIG .......................................................................................................... 132
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
2.4 NE Atlantic GIG ...............................................................................................................
2.4.1 Intercalibration approach
.....................................................................................
2.4.2 National methods that were intercalibrated..........................................................
2.4.3 Reference conditions and class boundary setting
.................................................
2.4.4 Results of the comparison
.....................................................................................
2.4.5 Results of the harmonization – Boundary thresholds and EQR values
................
2.4.6 Open issues and need for further work
.................................................................
138
138
140
147
161
161
169
3 Discussion
............................................................................................................................... 170
3.1 Comparability between GIGs ........................................................................................... 170
3.2 Open issues and need for further work............................................................................. 171
4 References
............................................................................................................................... 171
5 Annex
...................................................................................................................................... 174
Section 4 – Macroalgae
...................................................................................................... 175
1 Introduction
........................................................................................................................... 175
2 Methodology and results
....................................................................................................... 175
2.1 Mediterranean GIG ..........................................................................................................
2.1.1 Intercalibration Approach
.....................................................................................
2.1.2 National methods that were intercalibrated..........................................................
2.1.3 Reference conditions and class boundary setting
.................................................
2.1.4 Intercalibration
.....................................................................................................
2.1.5 Results of the comparison
.....................................................................................
2.1.6 Results of the harmonisation – Boundary EQR values
.........................................
2.1.7 Open issues and need for further work
.................................................................
2.2 NE Atlantic GIG ...............................................................................................................
2.2.1 Intercalibration approach
.....................................................................................
2.2.2 National methods that were intercalibrated..........................................................
2.2.3 Reference conditions and class boundary setting
.................................................
2.2.4 Results of the comparison
.....................................................................................
2.2.5 Results of the harmonisation – Boundary thresholds and EQR values
................
2.2.6 Open issues and need for further work
.................................................................
175
175
176
178
186
188
189
190
191
191
192
194
205
208
213
3 Discussion
............................................................................................................................... 213
3.1 Comparability between GIGs ........................................................................................... 213
3.2 Open issues and need for further work............................................................................. 213
4 References
............................................................................................................................... 214
Section 5 – Angiosperms
................................................................................................... 219
1 Introduction
........................................................................................................................... 219
2 Methodology and Results
...................................................................................................... 219
5
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
2.1 Baltic GIG ........................................................................................................................
2.1.1 Intercalibration approach
.....................................................................................
2.1.2 National methods that were intercalibrated..........................................................
2.1.3 Reference conditions and class boundary setting
.................................................
2.1.4 Results of the comparison
.....................................................................................
2.1.5 Results of the harmonisation – Boundary EQR values
.........................................
2.1.6 Open issues and need for further work .................................................................
2.2 NE Atlantic GIG ...............................................................................................................
2.2.1 Intercalibration approach
.....................................................................................
2.2.2 National methods that were intercalibrated..........................................................
2.2.3 Reference conditions and class boundary setting
.................................................
2.2.4 Results of the comparison
.....................................................................................
2.2.5 Results of the harmonisation – Boundary thresholds and EQR values
................
2.2.6 Open issues and need for further work
.................................................................
219
219
221
221
225
226
226
227
227
228
229
237
237
239
3 References
............................................................................................................................... 240
6
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0009.png
Section 1 – Introduction
1 Introduction
This Technical Report gives an overview of the technical and scientific work that has been carried
out in the Coastal and Transitional Waters intercalibration of ecological classification systems across
the European Union as required by the Water Framework Directive (WFD).
The results of this exercise were published in the Official Journal of the European Union as
“Commission Decision 2008/915/EC of 30 October 2008
1
The Coastal and Transitional Waters intercalibration exercise was carried out within 4 Geographical
Intercalibration Groups (GIGs) – Baltic, Black Sea, Mediterranean and North East Atlantic.
Common intercalibration types shared by Member States within each GIG were defined for the
intercalibration exercise. The results of the first intercalibration exercise are the status boundaries
for the benthic invertebrate fauna quality element (all GIGs), metrics and boundaries representing
the phytoplankton quality element (all GIGs), metrics representing the macroalgae and angiosperms
quality elements (Baltic, Mediterranean and NE Atlantic GIGs) and provisional boundaries for the
fish quality element (NE Atlantic GIG only). These boundaries are based on definitions of reference
criteria and the application of the Boundary Setting Protocol (BSP) to set the high-good and good-
moderate boundaries in line with the normative definitions for status class boundaries for each
quality specified in the WFD.
This report includes descriptions common and national coastal and transitional water types, national
methods, common and national boundary setting protocols, the results of harmonisation of these
boundaries between Member States as well as discussion of problems and way forward.
This report is available electronically at the following internet address:
http://circa.europa.eu/Public/irc/jrc/jrc_eewai/library?l=/intercalibration_2&vm=detailed&sb=Title
Annexes are not included in the printed version, but can be downloaded from the above address.
2 Background
The
Water Framework Directive
(WFD) establishes a framework for the protection of all
waters (including inland surface waters, transitional waters, coastal waters and groundwater). The
environmental objectives of the WFD set out that good ecological status
2
of natural water bodies and
good ecological potential
3
of heavily modified and artificial water bodies should be reached by 2015.
One of the key actions identified by the WFD is to carry out a European benchmarking or
intercalibration (IC) exercise to ensure that good ecological status represents the same level of
1
2
http://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri=OJ:L:2008:332:0020:0044:EN:PDF
‘Ecological status’ is an expression of the quality of the structure and functioning of aquatic ecosystems associated
with surface waters, classified in accordance with Annex V WFD; ‘Good ecological status’ is the status of a body of sur-
face water so classified in accordance with Annex V.
3
‘Good ecological potential’ is the status of a heavily modified or artificial body of water, so classified in accordance
with the relevant provision of Annex V.
7
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
ecological quality everywhere in Europe (Annex V WFD). It is designed to ensure that the values
assigned by each Member State (MS) to the good ecological class boundaries are consistent with
the Directive’s generic description of these boundaries and comparable to the boundaries proposed
by other MS. The intercalibration of surface water ecological quality status assessment systems is a
legal obligation.
Intercalibration is carried out under the umbrella of Common Implementation Strategy (CIS)
Working Group A - Ecological Status (ECOSTAT), which is responsible for evaluating the results
of the IC exercise and making recommendations to the Strategic Co-ordination Group or WFD
Committee. The IC exercise aims at consistency and comparability in the classification results of
the monitoring systems operated by each MS for biological quality elements (CIS WFD Guidance
Document No. 14; EC, 2005). In order to achieve this, each MS is required to establish Ecological
Quality Ratios (EQRs) for the boundaries between high (H) and good (G) status and for the
boundary between good (G) and moderate (M) status, which are consistent with the WFD normative
definitions of those class boundaries given in Annex V of the WFD.
All 27 MS of the European Union are involved in this process, along with Norway, who has joined
the process on a voluntary basis. Expert groups have been established for lakes, rivers and coastal/
transitional waters, subdivided into 14 Geographical Intercalibration Groups (GIGs -groups of MSs
that share the same water body types in different sub-regions or ecoregions).
The IC exercise aims to ensure that the H/G and the G/M boundaries in all MS’s assessment
methods for biological quality elements correspond to comparable levels of ecosystem alteration
(EC, 2005). Intercalibration guidance produced by CIS (WFD Guidance Document No. 14) warns
that the process will only work if common EQR boundary values are agreed for very similar
assessment methods or where the results for different assessment methods are normalised using
appropriate transformation factors (EC, 2005). Different assessment methods (e.g. using different
parameters indicative of a biological element) may show different response curves to pressures and
therefore produce different EQRs when measuring the same degree of impact (EC, 2005).
In each GIG, the IC exercise will be completed for those MS that already have data and (WFD
compliant) assessment methods to set boundary EQR values for some of the biological quality
elements. Countries that do not have data or assessment methods already available, or do not
actively participate in the current IC exercise, need to agree with the outcome of the IC exercise and
harmonise their assessment methods, taking into account the results of the current exercise, when
their data/methods becomes available.
The WFD refers to an ‘intercalibration network’, comprising sites selected from a range of surface
water body types present within each ecoregion, as the basis for intercalibration (Annex V; 1.4.1).
For each surface water body type selected, the WFD specifies that at least two sites corresponding
to the boundary between high and good status, and between good and moderate status should be
submitted by each Member State for intercalibration. However, as the IC exercise evolved, this
network has become redundant, as these datasets were too small to permit robust intercalibration.
This Technical Report provides a detailed description of the work that was carried out in the
framework of the EU Water Framework Directive intercalibration exercise. harmonising the
classification scales of national methods for ecological classification scales for rivers across the
European Union. The technical work was carried from 2004 to 2007 by groups of experts from all
EU Member States, within the framework of the Common Implementation Strategy working group
(2)A on Ecological Status, facilitated by a steering group lead by the European Commission Joint
Research Centre (JRC) (Figure 1.1).
8
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0011.png
WG 2A
Intercalibration Steering Group
JRC
Lake Expert Group representative
River Expert Group representative
Coast Expert Group representative
Lake experts/GIGs
C
N
AL
M
AT
River experts/GIGs
N
AL
C
EC
M
Coast experts/GIGs
BA
NEA
BS
M
EC
Figure 2.1:
Overview of the organisational structure of the intercalibration process (from EC 2005).
Before the start of the intercalibration exercise a guidance document (EC 2005) was agreed
describing the key principles and process options for the intercalibration exercise. The key principles
of the intercalibration process as described in the guidance document are reproduced below.
Key principles of the intercalibration process (from Guidance on the Intercalibration
Process, EC 2005)
1. The intercalibration process is aimed at consistency and comparability of the classification results of the monitoring systems
4
operated by each Member State for the biological quality elements
5
. The intercalibration exercise must establish values for
the boundary between the classes of high and good status, and for the boundary between good and moderate status, which are
consistent with the normative definitions of those class boundaries given in Annex V of the WFD
6
.
2. The essence of intercalibration is to ensure that the high-good and the good-moderate boundaries in all Member State’s
assessment methods for biological quality elements correspond to comparable levels of ecosystem alteration. Intercalibration
is not necessarily about agreeing common ecological quality ratio (EQR) values for the good status class boundaries as
measured by different assessment methods. Common EQR values only make sense, and are only possible, where very similar
assessment methods are being used or where the results for different assessment methods are normalised using appropriate
transformation factors. This is because different assessment methods (e.g. using different parameters indicative of a biological
element) may show different response curves to pressures and therefore produce different EQRs when measuring the same
degree of impact.
3. The first phase of the process is the establishment of an intercalibration network for a limited number of water body types
consisting of sites representing boundaries between the quality classes High-Good and Good-Moderate, based on the WFD
normative definitions. The WFD requires that selection of these sites is carried out “using expert judgement based on joint
inspections and all available information
7
”.
4
The term ‘monitoring system’ in the way it is commonly used includes the whole process from sampling, measure-
ment and assessment including all quality elements (biological and other). In the context of WFD Annex V, 1.4.1, the term
‘monitoring system’ only refers to a biological assessment method, applied as a classification tool, the results of which
can be expressed as ecological quality ratios. This guidance uses the term ‘WFD assessment method’ in place of the term
‘monitoring system’ that may be misleading in this context.
5
The WFD intercalibration as described in Annex V, 1.4.1 does not concern the monitoring systems themselves, nor the
biological methods, but the classification results
6
WFD Annex V, 1.4.1 (ii), (iii), (iv), (vi)
7
WFD Annex V, 1.4.1 (v)
9
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0012.png
4. The Intercalibration Guidance states that “some artificial or heavily modified water bodies could be considered to be
included in the intercalibration network, if they fit in one of the natural water body types selected for the intercalibration
network. Artificial and heavily modified water bodies that are not comparable with any natural water bodies should only
be included in the intercalibration network, if they are dominant within a water category in one or more Member States; in
that case they should be treated as one or several separate water body types”. An artificial or heavily modified water body is
considered to fit in a natural water type if the maximum ecological potential of the artificial or heavily modified water body
is comparable to the reference conditions of the natural type for those quality elements considered in the intercalibration
exercise
8
.
5 In the second phase of the process, each Member State’s assessment method must be applied to those sites on the register
that are both in the ecoregion (or, as pointed out in section 2.8, in the Geographical Intercalibration Group (GIG)) and of a
surface water body type to which the system will be applied. The results of the second phase must be used to set the EQR
values for the relevant class boundaries for each Member States’ biological assessment system. The results of the exercise
will be published by the Commission by 22 December 2006 at the latest.
6. Intercalibration sites are selected by the Member States, and represent their interpretation of the WFD normative definitions
of high, good and moderate status. There is no guarantee that different Member States will have the same views on how the
normative definitions should be interpreted. Differences in interpretation are reflected in the intercalibration network
9
. A
common interpretation of the normative definitions should be the main outcome of the intercalibration exercise. At the end
of the intercalibration exercise the intercalibration network may need to be revised according to this common interpretation.
7. The Intercalibration Exercise is focused on specific type/biological quality element/pressure combinations
10
. The selection
of these combinations is based on the availability of adequate data within the time constraints of the exercise. This means
that the exercise will not identify good status boundary EQR values for all the type/biological quality element/pressure
combinations relevant for the implementation of the WFD. However, the Intercalibration Exercise will identify, and test the
use of, a procedure and criteria for setting boundaries in relation to any such combinations
11
.
8. The intercalibration process described in this guidance is aimed at identifying and resolving:
(a) Any major/significant inconsistencies between the values for the good ecological status class boundaries established
by Member States and the values for those boundaries indicated by the normative definitions set out in Section 1.2
of Annex V of the WFD; and,
(b) Any major/significant incomparability between the values established for the good status class boundaries by
different Member States.
9. The process will identify appropriate values for the boundaries of the good ecological status class applicable to the
ecological quality ratio EQR scales produced by the Member States’ assessment methods.
10. The Intercalibration Exercise will be undertaken within GIGs rather than the ecoregions defined in Annex XI of the WFD.
This is to enable intercalibration between a maximum number of Member States.
11. The Intercalibration Exercise assumes that all Member States will have developed their national WFD assessment methods
to a sufficient extent to enable the consistency with the normative definitions, and the comparability between Member
States, of the good status boundary EQR values for those methods to be assessed during 2005. It was recognized however
that this assumption might be problematic. An inventory on the state-of-the-art in the developments of WFD compliant
methods is carried out during the process of finalisation of the intercalibration network
12
.
This is not the case for those quality elements that are significantly impacted by the hydromorphological alteration that
has led to the water body to be designated as heavily modified.
9
Intercalibration Guidance, section 3.5
10
as described in the document’ Overview of common Intercalibration types’ (available at the intercalibration site sub-
mission web pages, http://wfd-reporting.jrc.cec.eu.int/Docs/typesmanual)
11
If the results of the method are significantly affected by biogeographical or other ecological differences within the
intercalibration type, different boundary EQR values may be appropriate for different parts of the type
12
The metadata questionnaire is available at the intercalibration site submission web pages, http://wfd-reporting.jrc.
cec.eu.int/Docs/ metadata
8
10
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0013.png
3 Geographical Intercalibration Groups (GIGs)
and common intercalibration types
3.1 Geographical intercalibration groups
For coastal and transitional waters, four geographical intercalibration groups were agreed upon:
Baltic, (see chapter 1.1.3), including Denmark, Germany, Finland, Sweden, Latvia, Lithuania,
Estonia and Poland.
Black Sea (see chapter 1.1.4) includes Bulgaria and Romania.
Mediterranean (see chapter 1.1.5), including Spain, France, Italy, Slovenia, Greece, Cyprus, Malta.
Croatia as a potential accession country also participates.
North East Atlantic (see chapter 1.1.6), including Portugal, Spain, France, Ireland, UK, Belgium,
Netherlands, Germany, Denmark, Sweden and Norway.
3.2 Common intercalibration types
The common coastal intercalibration types are characterised broadly by the descriptors of the WFD
System B typology: geographical position (for latitude and longitude), tidal range and salinity as
obligatory factors plus optional factors such as, exposure, depth, mixing characteristics, substratum
composition, current velocity, residence time and ice cover where applicable.
Baltic
The Baltic GIG includes the whole or parts of the coastline of the following countries: DE= Germany,
DK=Denmark, EE=Estonia, FI= Finland, LT=Latvia, LI=Lithuania, PL=Poland and SE=Sweden.
The common intercalibration types were agreed using the basic factors of salinity and exposure with
further delineation based on depth and number of ice cover days plus the identification of one lagoon
type. The countries which have each type within their coastal waters are shown in the table below.
All countries agreed to intercalibrate quality elements that respond to eutrophication pressures. The
chosen elements in each type are also shown in the table below.
Table 2.2.1:
Common intercalibration types.
Type
CW B0
Salinity 0.5-3, sheltered,
shallow, >150 ice days
CW B2
Salinity 3-6, sheltered, shallow,
90-150 ice days
CW B3
Salinity 3-6, sheltered, shallow,
90 ice days
Pressure
Eutrophication
Eutrophication
Eutrophication
Quality element
Phytoplankton:
Chlorophyll a
Benthic Fauna:
National indices
Phytoplankton:
Chlorophyll a
Benthic Fauna:
National indices
Phytoplankton:
Chlorophyll a
Benthic Fauna:
National indices
Countries involved
(number of sites)
SE
FI , SE
FI , SE
FI, SE
FI
FI, SE
Phytoplankton:
Chlorophyll a
CW B12
Eutrophication
Angiosperms:
Eelgrass depth limit (DK + DE) DE, DK, EE, SE
Salinity 6-22, sheltered, shallow
Benthic Fauna:
National indices
EE, LV, PL
CW B13
Phytoplankton:
Chlorophyll a
Eutrophication
Salinity 6-22, exposed, shallow
Benthic Fauna:
National indices
DK, LT, LT, PL
CW B14
PL
Salinity 6-22, sheltered, shallow Eutrophication
Phytoplankton:
Chlorophyll a
lagoons
DK, PL
11
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0014.png
Black Sea
Bulgaria and Romania are the only two countries participating in the Black Sea GIG and identified
one common intercalibration type, BS1: based on salinity (mesohaline), depth (shallow waters)
and substratum (mixed). In this phase of the intercalibration only the benthic invertebrates and
phytoplankton quality elements have been considered, as these are the most developed indicators in
the two countries.
Mediterranean
The Mediterranean GIG includes seven Member States, Spain, France, Italy, Slovenia, Greece,
Cyprus and Malta plus Croatia as an accession country. The Mediterranean GIG has at this
stage confined its work to coastal waters. Transitional waters were not included in this phase of
the intercalibration exercise due to lack of sites and data in participating countries. Preliminary
discussions on transitional water have taken place with a view to undertaking further assessments in
phase two.
Mediterranean Coastal IC types were defined primarily on the substratum composition and the
depth profile. Salinity was not seen as a primary discriminating factor as it is very similar across the
whole Mediterranean basin. The Mediterranean GIG agreed four basic coastal water types as shown
below:
Table 2.2.2:
Mediterranean Coastal Waters Types.
Type
CW - M1
CW - M2
CW - M3
CW - M4
Name of Type
Rocky shallow coast
Rocky deep coast
Sedimentary shallow coast
Sedimentary deep coast
Substratum (1)
rocky
rocky
sedimentary
sedimentary
Depth (2)
shallow
deep
shallow
deep
(1) In many cases different seabed substrata will occur within one waterbody type. The dominant
substratum should be selected.
(2) Depth division is based on 40 m depth at 1 mile distance from the coastline.
The following quality elements with respective countries’ participation have been included in
this phase of the intercalibration exercise:
Phytoplankton:
CY, ES (Balearic Islands, Catalonia,
Valencia), F, GR, IT, SL and Croatia participating,
Benthic macroinvertebrates:
CY, ES (Balearic Islands, Catalonia, F, GR, IT, SL participating,
Macroalgae:
CY, ES (Catalonia, Valencia), F, GR and I, participating
Angiosperms
(P.oceanica): ES (Catalonia, Valencia), F, GR, IT, Malta participating.
Typology
Not all types are included for each quality element: in some cases the type distinctions were not
relevant for the IC exercise (e.g. Angiosperm) and new types have been defined for the specific
analysis of phytoplankton, as shown below:
12
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0015.png
Table 2.2.3:
details about typology.
All types included?
Phytoplankton
Revision of Typologies
Benthic invertebrate fauna
Only M2 & M3
Macroalgae
M1, M2 & M3
Yes
Yes
Yes
All pressures included?
Yes
Angiosperms:
P.oceanica
No types distinction
Phytoplankton Typologies
Details on the process that was followed for this types’ revision are found in the specific part Section
3. Phytoplankton.
For general information:
Three different water types, in an ecological perspective, have been described as follows:
Type 1
coastal sites highly influenced by freshwater inputs
Type 2
coastal sites not directly affected by freshwater inputs
Type 3
coastal sites not affected by freshwater inputs
Further discrimination within types based on geographical/ecological differences (e.g. eastern,
western Mediterranean basins) was used, in some cases in order that the exercise produced
biologically meaningful results.
North East Atlantic
The North East Atlantic GIG involves eleven countries, Portugal, Spain, France, Ireland, UK,
Belgium, Netherlands, Germany, Denmark, Sweden and Norway. Common intercalibration types
were agreed based on the obligatory factors salinity and tidal range, plus optional factors, depth,
current velocity, exposure, mixing and residence time. After consideration of the relevance of the
original types within the NE Atlantic complex, based solely on the above factors, it was decided that
in some cases there was no biological differences between types in relation to the chosen quality
element or metric(s) being intercalibrated and that some could be merged together. This resulted in
the adoption of the following grouped types in this intercalibration exercise:
13
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0016.png
Table 2.2.4:
NEA Coastal Waters Types.
New Type ID
CW –NEA1/26a
CW –NEA1/26b
CW –NEA1/26c
CW –NEA1/26d
CW –NEA1/26e
CW – NEA3/4
CW – NEA7
Name
Open oceanic, exposed
or sheltered, euhaline,
shallow
Enclosed seas, exposed
or sheltered, euhaline,
shallow
Enclosed seas, exposed
or sheltered, partly
stratified
Scandinavian coast,
exposed or sheltered,
shallow
Areas of upwelling,
exposed or sheltered,
euhaline, shallow
Polyhaline, exposed or
moderately exposed
(Wadden Sea type)
Deep, low current,
sheltered
Polyhaline, microtidal,
sheltered, shallow
(Skagerrak inner arc
type)
Fjord with a shallow
sill at the mouth with
a very deep maximum
depth in the central
basin with poor
deepwater exchange.
Polyhaline, microtidal
exposed, deep
(Skaggerak outer arc
type)
Transitional waters
Salinity
(PSU)
Fully saline
(> 30)
Fully saline
(> 30)
Fully saline
(> 30)
Fully saline
(> 30)
Fully saline
(> 30)
Polyhaline
(18 - 30)
Fully saline
(> 30)
Polyhaline
(18 - 30)
Tidal range
(m)
Mesotidal
(1 - 5)
Mesotidal
(1 - 5)
Microtiadl/
Mesotidal
(1 - 5)
Microtidal
(<1)
Mesotidal
(1 - 5)
Mesotidal
(1 - 5)
Mesotidal
(1 - 5)
Microtidal
(< 1)
Depth
(m)
Shallow
(< 30)
Shallow
(< 30)
Shallow
(< 30)
Shallow
(< 30)
Shallow
(< 30)
Shallow
(< 30)
Deep
(> 30)
Shallow
(< 30)
Current
velocity
Medium
(1 - 3
knots)
Medium
(1 - 3
knots)
Medium
(1 - 3
knots)
Low
(<1 knot)
Medium
(1 - 3
knots)
Medium
(1 - 3
knots)
low
(< 1
knot)
low
(< 1
knot)
low
(< 1
knot)
low
(< 1
knot)
Medium
Exposure
Exposed or
sheltered
Exposed or
sheltered
Exposed or
sheltered
Exposed or
moderately
exposed
Exposed or
sheltered
Exposed or
moderately
exposed
Sheltered
Mixing
Fully mixed
Fully mixed
Partly
stratified
Partly
stratified
Fully mixed
Fully mixed
Fully mixed
Partially
Stratified
Residence
time
Days
Days
Days to
weeks
Days to
weeks
Days
Days
Days
Days-
Weeks
CW – NEA8
Sheltered
CW – NEA9
Polyhaline
(18 - 30)
Microtidal
(< 1)
Deep
(> 30)
Sheltered
Permanently
Stratified
Weeks
CW – NEA10
Polyhaline
(18 - 30)
Oligo-
Euhaline
(0 - 35)
Microtidal
(< 1)
Mesotidal
(1 – 5 )
Deep
(> 30)
Shallow
(< 30)
Exposed
Sheltered
or
moderately
Exposed
Permanently
Stratified
Partially- or
Permanently
Stratified
Days
TW – NEA11
Days-
Weeks
The above types occur in Member State’s waters as detailed below:
Table. 2.2.5:
Occurrence of coastal water types in NEA MSs.
Type
CW – NEA1/26a
CW – NEA1/26b
CW – NEA1/26c
CW – NEA1/26d
CW – NEA1/26e
CW – NEA3/4
CW – NEA7
CW – NEA8
CW – NEA9
CW – NEA10
TW – NEA11
x
X
X
X
X
x
x
x
x
x
x
x
X
X
X
x
x
X
X
X
X
X
X
x
BE
DK
FR
X
X
x
DE
IE
x
x
NL
NO
x
PT
ES
X
SE
UK
x
x
14
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0017.png
In each type relevant biological quality elements have been included in this stage of the
intercalibration process, linked to specified pressures. These are detailed in the table below:
Table 2.2.6:
Quality elements and pressures included per types in the intercalibration.
Type
CW – NEA1/26, NEA3/4, NEA7,
NEA8, NEA9, NEA10
Biological Quality Elements
Benthic Invertebrates
Phytoplankton
Macroalgae/Angiosperms
Benthic Invertebrates
Macroalgae/Angiosperms
Fish
Pressures
Organic enrichment
Hazardous substances
Nutrient enrichment
Morphological alteration
Organic enrichment
Hazardous substances
Morphological alteration
Organic enrichment (DO)
TW – NEA11
This does not mean that intercalibration for all quality elements (or metrics) has been completed in
each type at this stage of the process.
4 Methodology and Results
Following the CIS guidance on intercalibration options the coastal GIGs selected the most
appropriate for each quality element and metric selected. The following describes the options chosen
within each GIG:
Baltic
The following intercalibration options were chosen for the relevant biological quality elements and
metrics.
Phytoplankton:
a) Hybrid between option 2 and 3 (Denmark, Estonia, Finland, Germany, Latvia and Sweden)
A common metric – summer mean of chlorophyll a from May/June to September - was
agreed. In some parts of the Baltic Sea the summer period can be shorter. Intercalibration was
performed by comparison of the results from the national assessment tools. Within each type
member state essentially agreed on common reference and classifications values for the whole
type or for sub areas in the type. (However, Type B3 in Sweden and Finland differ a lot from
each other by their natural conditions, which complicates intercalibration).
b) Option 3
A combined data set from most member states participating in the Baltic GIG has been
compiled and an overall relation between chl. a and TN established. Further and future work on
establishing type or site specific chl. a-TN relationships and ranges of nutrient concentrations
related to the different ecological classes may provide a useful tool for setting boundaries.
Benthic fauna:
Option 3 (Estonia, Denmark, Finland, Germany, Latvia and Sweden)
Intercalibration is performed in three steps, comparison of species sensitivity classifications,
comparison of indices and comparison of classification of water bodies based on national methods
for assessment. Work is still ongoing.
15
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
Angiosperms:
Hybrid between option 2 and 3. Bilaterally between Denmark and Germany
Reference levels are based on historical data, expert judgement and modeling. Two approaches have
been used for classification: 1) percent deviation (3 scenarios) and 2) modeling.
Denmark used the maximum depth of 5 % eelgrass cover to define the depth limit. Germany used
historical records of depth limit to define reference and light modeling to define depth limits.
Black Sea
Taking into account that both countries are aiming to use at national level the same methods
for phytoplankton and macrozoobenthos, Option 1 was initially chosen however the final
intercalibration used option 3 in the first instance as only common metrics for each quality element
could be agreed on rather than common whole classification systems covering the whole element.
Mediterranean
For each biological quality element considered, different options were used, mainly hybrids.
Phytoplankton
A hybrid between options 1 and 2 is used. No national methods were intercalibrated. Only common
statistical analysis on chl-α, nutrients and physico-chemical data and some multivariate techniques
have been performed as a starting point (exploratory data analysis) for the intercalibration process.
Benthic invertebrates
A hybrid between options 2 and 3 is used.
Macroalgae
Option 3 has been used; i.e. Member States have set national boundaries using their own nationally
developed assessment methods. These systems are then compared between Member States by
application of these methods on each country’s data.
Angiosperms (P.oceanica)
Option 3 has been used; i.e. Member States have set national boundaries using their own nationally
developed assessment methods. These systems are then compared between Member States by
application of these methods on each country’s data.
North East Atlantic
The following intercalibration options have been used for the biological quality elements.
Benthic Invertebrates
Option 3
has been used in all types i.e. Member States have set national boundaries using their
own nationally developed assessment methods. These systems are then compared between Member
States by application of these methods on each country’s data. Initial appraisal is made of the
16
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
level of agreement on status assessments between countries methods when data is exchanged, and
national boundaries adjusted post this comparison until a satisfactory degree of agreement is reached
(e.g. > 80 % comparability).
Phytoplankton
Option 2
has been used in all types. Metrics have been selected that form part of the overall quality
element assessment that are used in all Member States own classification systems where a common
approach can be adopted to setting boundaries for these metrics at the GIG level. These boundaries
may differ according to the type selected and in some cases region specific boundaries within types
have been adopted. At this stage the selected metrics are, chlorophyll-a (90
th
percentile over a six
year period), frequency of
Phaeocystis
cell counts above 10
6
cells/l over a six year period and
frequency of microphytoplanton cell counts above 10
5
cells/l.
Macroalgae and Angiosperms
A hybrid approach between option 2 and 3 has been used for the macroalgae and angiosperms
metrics. Metrics have been selected that form part of the overall quality element assessment that
are used in all Member States own classification systems where a common approach can be adopted
to setting boundaries for these metrics at the GIG level. These boundaries may differ according to
the type selected and in some cases region specific boundaries within types have been adopted. At
this stage the selected metrics are, opportunistic macroalgae (areal extent and/or biomass), perennial
intertidal and subtidal algae (extent and species composition), intertidal and subtidal seagrass (areal
extent, density and species composition) and saltmarsh (areal extent and species composition). For
seagrass a combination of options 2 and 3 was used because UK and NL use broadly the same
metrics but derived boundary conditions separately which were then compared and harmonized.
Fish
Option 3
has been used i.e. Member States have set national boundaries using their own nationally
developed assessment methods. These systems are then compared between Member States
by application of these methods on each country’s data. Initial appraisal is made of the level of
agreement on status assessments between countries methods when data is exchanged, and national
boundaries adjusted post this comparison until a satisfactory degree of agreement is reached (e.g.
> 80 % comparability). The original intention was to pursue an
option 1
solution. This is still the
preferred option and may be possible in Phase II.
5 Discussion
5.1 Comparability between quality elements
There has not been an exercise in phase one of the intercalibration process to directly compare the
results for common quality elements between the GIGs. A range of intercalibration options have
been used for each quality element in the different GIGs. Therefore it is not possible at this time
to demonstrate that where similar metrics or classification systems have been used that they show
the same level of comparability between the GIGs as within individual GIGs. However in many
17
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
cases the biology of the geographical regions is very different e.g. it is very difficult to compare the
Baltic with the Mediterranean. However further evaluation is needed where the same basic metric
is being used e.g. depth limitation, to investigate if there is more scope for further harmonization of
approaches.
5.2 Open issues and need for further work
The main area where further work is needed in the second phase of the process is in transitional
waters. Almost none of the results presented in phase one of the process are applicable in transitional
waters, except for a few metrics developed for coastal waters that can be used in the higher salinity
areas. One key quality element in transitional waters is fish. No agreed results were produced in phase
one, making this a key priority in phase two. However a lot of work has been done on fish in the North
East Atlantic GIG which may be applicable in other GIGs. It is anticipated that results for fish can
be produced quite early in the next phase. Another issue is whether the basic typology needs to be
redefined in some GIGs. There have been some difficulties in phase one with the broad nature of the
typology and further subdivisions have been necessary in order to produce results. Transitional waters
will definitely need a new typology for all quality elements. Further work is also needed to cover the
key pressures that have not been assessed in phase one. Some methods and metrics for which results
have been agreed so far respond to specific pressures. It is also important to look in future at how
biological indicators respond to different combinations of pressures. Another issue is how assessments
of the status of whole waterbodies are done using the agreed methods and metrics. Further work is
needed to ensure a common approach is adopted to classification of waterbodies across all GIGs. This
work should be linked to the development of classification rules, data aggregation and assessment and
common approaches to sampling and analysis strategies.
6 Summary and Conclusions
In summary results have been produced in all GIGs for at least two quality elements. In many
cases these results for which intercalibrated results have been agreed are only part of the overall
national classification systems. Therefore a main conclusion is that while common values have been
agreed principally at the metric level much more work is required in the next phase to compare
Member States’ whole quality element methods and apply these to the status assessment of whole
waterbodies. Only after this has been done can it be truly demonstrated that there is an equal level
of ambition across GIGs and within GIGs.
18
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0021.png
Section 2 – Benthic Invertebrates
1 Introduction
Three of the four coastal water GIGs have been able to produce results for the benthic invertebrate
quality element. Each GIG has several typologies that are found in these waters. Not all countries
within each GIG have all the types within its borders. Information about the types and countries
with each type in each GIG is described in the sections below.
All the GIGs have chosen Option 3 in this phase of the intercalibration process. Therefore Member
States have developed their own classification schemes, these have been assessed against each other
through the exchange and evaluation of data and the boundaries in each scheme harmonized to give
an acceptable level of agreement.
It is important to note that the methods intercalibrated in this phase are habitat specific. All methods
are for use in soft sediment habitats. Methods for other habitats are largely still under development
and cannot be assessed at this stage.
2 Methodology and results
2.1 Baltic GIG
2.1.1 Intercalibration approach
The Baltic Sea Geographical Intercalibration Group (GIG) carried out intercalibration of the
macroinvertebrate assessment systems for coastal waters developed in four Member States:
Denmark, Finland, Germany and Sweden. The results from the intercalibration between Sweden
and Finland are presented in this technical report. Comparability between the Danish, German,
and Swedish macroinvertebrate methods could not be demonstrated and thus those results are not
included in the technical report. In addition, macroinvertebrate assessment methods have been
developed in Estonia, but the intercalibration exercise could not be completed as the other countries
sharing the same coastal types with Estonia had not yet their national methods ready
13
.
The Option 3, as described in the Guidance of the Intercalibration Process
14
, was used for the
Benthic Invertebrate fauna quality element. The Member States have developed their own
classification schemes, these have been assessed against each other through the exchange and
evaluation of data and the boundaries in each scheme harmonized to give an acceptable level of
agreement. The setting of the reference conditions and high-good and good-moderate boundaries
for each Member States’ assessment systems was first carried out separately and then the outcomes
of these systems were compared on water body (or sample) level against each others by applying
13
Report of the Estonian assessment method and the preliminary evaluation of the comparability with methods from
other Baltic Sea countries is presented in the Baltic Sea Milestone Report 6. Quality element: Benthic Fauna. Version 16
June 2006. Rev. 3, 30 March 2007. Available at: http://circa.europa.eu/Public/irc/jrc/jrc_eewai/library, in folder: GIG Mile-
stone Reports
14
Guidance on Intercalibration Process, No 14. Available at: http://circa.europa.eu/Public/irc/env/wfd/library
19
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0022.png
different national assessment methods to local data sets from countries sharing the same common
types. At this stage the assessment systems of four countries, Denmark, Finland, Germany and
Sweden could be compared bilaterally against each others using data from the coastal types that
these countries share. Finally, only the results between Sweden and Finland were concluded to be
sufficiently comparable and could be included in the final decision.
Intercalibration was performed in three steps, 1) comparison of species sensitivity classifications,
2) comparison of indices, and 3) comparison of water body classifications using different national
assessment methods applied to local datasets.
Baltic Sea Common intercalibration types
The Baltic Sea Geographical Intercalibration Group (GIG) includes the whole or parts of the
coastline of the following countries: Germany, Denmark, Estonia, Finland, Latvia, Lithuania, Poland
and Sweden (Table 2.1.1).
The common coastal water types are characterised by the descriptors of the System B typology.
The typology factors are based on the common typology framework presented in the guidance
on the typology for the coastal and transitional waters
15
. In the Baltic Sea GIG, the common
intercalibration types were characterized using basic salinity and exposure with further delineation
based on depth and number of ice cover days (Table 2.1.1). One transitional water type (TW B 13)
was identified. All countries agreed to focus the intercalibration on the quality elements that are
sensitive to eutrophication pressures.
Table 2.1.1:
Description of Baltic Sea Common intercalibration types that have included in the intercalibration exercise.
Type
CW B0
CW B2
CW B3 a
CW B3 b
CW
B12 b
Western
Baltic Sea
CW B13
Salinity
psu
0.5- 3
3-6
3-6
3-6
8 - 22
Exposure
Sheltered
Sheltered
Sheltered
Exposed
Sheltered
Depth
Shallow
Shallow
Shallow
Shallow
Shallow
Ice days
> 150
90-150
∼90
∼90
-
Other Characteristics
Sites in Botnian Bay (Northern Quark)
Sites in Bothnian Sea
Sites in the area extending from the southern
Bothinian Sea to the Archipelago Sea and
the western Gulf of Finland
Sites at the Southern Swedish coast and the
South western Baltic Sea open coast along
Denmark and Germany
Sites along the coast of the Estonia, Latvia
and Lithuania, the Polish coast and the
Danish island “Bornholm”
Lagoons
Transitional water. Sites along the coast of
Lithuania and Poland
6-22
Exposed
Shallow
-
CW B 14
TW B 13
6-22
6-22
Sheltered
Exposed
Shallow
Shallow
-
15
Guidance document No. 5
‘Transitional and Coastal Waters - Typology, Reference conditions, and Classification sys-
tems’.
Common Implementation Strategy of the Water Framework Directive, Available at: http://forum.europa.eu.int/Pub-
lic/irc/env/wfd/library
20
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0023.png
The common intercalibration types were characterised by the following descriptors:
– Salinity (using practical salinity scale): low (0,5-3) and high (3-6) oligohaline, mesohaline (6-22)
– Depth: all shallow (<30 m)
– Exposure (using agreed Pan-European scale
16
):
exposed, sheltered and very sheltered
– Duration of ice cover: >150 days/ year, 90-150 days/ year, no or very short ice cover
At this stage four countries, Denmark, Finland, Germany and Sweden had developed benthic
macroinvertebrate assessment systems that could be intercalibrated for five Baltic Sea common
types:
Types CWB0, CWB2, CWB3a, CWB3b:
Finland, Sweden.
Type CWB12b:
Germany, Denmark, Sweden.
At this stage, intercalibration was completed for the four types shared between Finland and Sweden
in the Bothnian Sea and Bothnian Bay. The list of the two national methods intercalibrated is
presented below (Table 2.1.2.).
Table 21.2:
The national assessment methods for the classification of ecological quality based on benthic
macroinvertebrates that were compared during the intercalibration exercise.
Country
Finland
Sweden
Assessment Method
BBI- Finnish Brackish Water Benthic Index
BQI–Swedish multimetric biological quality index (soft sediment infauna)
2.1.2 National methods that were intercalibrated
Macroinvertebrates have for decades been an integrated part of monitoring programmes setup to
detect marine pollution or eutrophication. Soft bottom macrofauna is a well-suited parameter to use
since it is stationary, relatively long-lived, easy to collect quantitatively, restricted to very limited
vertical distribution in or just below the sediment surface, and they can be identified to species also
after unlimited preservation. The impact of pollution in the water column is not necessarily the same
as in the bottom and at the sediment-water interface, in which the long-term effects of discharged
pollutants, may be better monitored by using sessile or sedentary organisms as indicators.
Changes in abundance, biomass and species composition of the benthic communities are signs of
eutrophication. Some of these species will respond to changes in food supply and/or sedimentation
rates and/or lowered oxygen concentrations (Diaz & Rosenberg 1995; Gray et al. 2002). The
complex benthic communities respond to anthropogenic loading and stress by establishing a new
community structure more tolerant to the increasingly unfavorable physio-chemical conditions
(Leppäkoski 1975). Karlson et al. (2002) gives a good review on eutrophication and oxygen
deficiency and their effects on the benthic community in Baltic coastal waters. Many indices and
approaches have been developed for the assessment of the status of zoobenthos communities in
marine waters. However, many of them are not applicable in the brackish Baltic Sea ecoregion due
to the low biodiversity. The gradually decreasing salinity and diminishing species richness towards
16
According to the definitions of the common European exposure categories; Guidance document No. 5
21
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0024.png
north- and east in the Baltic Sea make only region-specific comparisons of data and assessment
systems meaningful.
Finland:
The Finnish Brackish Water Benthic Index, BBI, follows the theory that biodiversity increases with
increasing distance from a pollution source along a gradient of disturbance (Pearson & Rosenberg
1978). The index is very similar to other multimetric indices developed for marine conditions
(Danish DKI-index and British BMI-index) (Borja et al. 2007), but it has additional features which
make it apt to brackish water conditions, typical for the Baltic Sea (Perus et al. 2007). The index is
a construction of the Baltic Sea-adapted BQI-index (Rosenberg et al. 2004, Blomqvist et al. 2007)
with additional biodiversity and abundance factors incorporated into the matrix. Therefore the
multimetric BBI index meets the specification of the Annex V in the WFD. The BBI Index values
are continuous and therefore useful for further analysis and interpretations in classification systems
(Perus et al. 2007).
The index compare observed BQI- and Shannon-Weaver (H´) values against highest recorded values
within individual types (serving as reference values) and further deducts the value for stations
showing low biodiversity or abundance. A value of zero (0) indicates conditions without any benthic
macroinvertebrates (azoic) and value one (1) indicate unpolluted bottom conditions.
The multimetric BBI-index includes relative abundance ( %) of sensitive or tolerant species. The
list of the species sensitivity or tolerance is same as used for the Swedish BQI-index (Rosenberg et
al. 2004; Blomqvist et al. 2007). The evaluation of sensitivity and tolerance of each species is based
on literature information (Anger 1975, 1977; Borja et al. 2000; Helawell 1986; Järvekülg 1970;
Landner et al. 1977; Leppäkoski 1975; Mandaville 2002; Wiederholm 1973) and expert judgment.
The levels of sensitivity/tolerance are:
1 – Very tolerant to pollution
5 – Tolerant
10 – Pollution sensitive
15 – Very pollution sensitive
The BBI is calculated as following:
where,
BQI
is the Swedish Biological Quality Index (Rosenberg et al. 2004, Blomqvist et al.
2007),
is the (log2-base) Shannon-Weaver diversity,
AB
is species abundance, and
S
the species
richness. A detailed presentation of the BBI index and its application for the national Finnish
classification and boundary setting is described in Perus et al. (2007).
Sweden:
The Benthic Quality Index (BQI) is based on the distribution of sensitive and tolerant species, the
number of species and the number of individuals. The complete classification method is described
by Blomqvist et al. (2007), and the original background paper presenting the characteristics and
composition of the BQI is published by Rosenberg et al. (2004).
22
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0025.png
The distribution between sensitive and tolerant species comprises the base of the BQI index. This
value generally varies between 1 and 15. Low values indicate high proportion of tolerant species
and high values indicate high proportion of sensitive species. The factor enumerating the number
of species will increase the index value when there are more than 9 species per sample and will
decrease the index value when there are less than 9 species per sample. The factor enumerating
the number of individuals has been derived in order to handle situations when there are only a few
individuals per sample. In a few instances, these individuals may represent sensitive species, and
thus produce an unjustifiably high index value. When fewer than about 20 individuals are present in
a sample (0.1 m
2
), the index is significantly depressed by this factor.
where,
S
is the number of taxa,
S
classified
is the number of taxa having a sensitivity value,
N
i
is the
number of individuals of taxon I,
N
totalclassified
is the total number of individuals of taxa having a
sensitivity value,
N
total
is the total number of individuals per 0.1 m
2
and the
Sensitivityvalue
i
is the
sensitivity value for taxa i.
The Baltic Sea coastal fauna is already under physiological stress due to low salinity. The fauna was
classified based on national expert knowledge and literature information. Each macroinvertebrate
species was given a value of 1, 5, 10 or 15 depending on its’ sensitivity for anthropogenic
disturbance. A high value indicates high sensitivity for disturbance and a low value high tolerance
against disturbance. The major pressure in the Swedish coastal waters is excess nutrient loading
leading to eutrophication and resulting in increased organic load to the seafloor and decreased
oxygen content in the sediments and deep water. This classification of different species into
sensitive and non-sensitive species is shared between Finland and Sweden. The Sensitivity values
for different taxa are presented in an Excel-file at:
http://www.naturvardsverket.se/sv/Arbete-med-naturvard/Vattenforvaltning/Handbok-20074/
Since benthic invertebrates display large natural spatial variation, we chose to base the estimate for
determining status using the entire water body, instead of basing it on individual sample results. For
this estimate, we chose to follow the precautionary principle and used the 20th percentile instead of
the median of the BQI values from a water body, when comparing with class boundaries per water
body type. The 20th-percentile was calculated using a special method based on 9,999 randomly
selected mean values from the existing index values in a water body. Naturally, the status estimate
becomes more robust the greater the number of sampling stations in a water body. As a rule of
thumb we recommend five or more sampling stations per water body.
A complete classification requires the following calculations in order to derive a value for a water
body in order to compare with the class boundaries for the type of that water body:
1. Calculate the BQI based on species and abundance information from each individual sample
2. Calculate the mean BQI for each station
3. Calculate the 20th-percentile using randomization based on the mean BQI values from
all stations
4. Compare the value for the 20th percentile with the class boundaries
23
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
2.1.3 Reference criteria and class boundary setting
Finland:
Reference conditions and boundary setting
Areas not influenced by human activities are regarded not to be present in the Baltic Sea today.
Historic reference data is almost completely lacking from Finnish coastal areas and use of old data
is therefore not an option in determining reference conditions. Number of stations for the Baltic
common intercalibration types B0, B2, B3a, and B3b in Finland are presented in Table 2.1.3.
Bulk of the benthic macroinvertebrate data in the national database are sampled between 1990
and present. In order to determine reference conditions, a method that is used in lakes and running
waters (Vuori et al. 2006) was modified to be applicable for the coastal waters.
Bäck et al. (2006) concludes on the Finnish coastal zoobenthos monitoring that “Long-term
zoobenthos monitoring has been carried out only in one area with two depth zones since 1964, with
few observations from the 1920s (Kangas et al. 2001). The zoobenthos monitoring is a part of the
HELCOM monitoring for the assessment of the state of the Baltic Sea. In addition to the national
monitoring, there is some regional zoobenthos monitoring carried out along the Finnish coast.
However, all available benthic invertebrate (or zoobenthos) data for setting the WFD reference
conditions and good-moderate boundary is gathered from the pollution control monitoring system,
which covers all the essential Finnish coastal areas impacted by nutrient loading with several
hundreds sampling stations near the sites that are recipients for the loading.”
Baltic Sea common intercalibration types B0, B2, B3a, and B3b are spatially larger than the
Finnish national coastal type areas. The environmental conditions (e.g. hydromorphological and
topographic) characteristics of these large intercalibration types vary considerably within respective
types and in order to produce more reliable analysis types were subdivided into smaller subtypes. In
the macroinvertebrate analyses a depth separation of the water column (0 to 10 m and deeper than
10 m) was used in relating the ecological meaningfulness on macroinvertebrate communities to the
typology. This was done to improve the interpretation of environmental status in individual types
(Perus et al. 2004).
The Finnish Brackish water Benthic Index (BBI) incorporates all parameters required in Annex V of
the WFD. The index takes into account relative abundance of sensitive/tolerant species, abundance
and biodiversity (Shannon-Wiener and Species Richness) and can thus identify data of different
ecological status.
The determination of the reference conditions for the BBI was carried out as following:
1) The median of the 10 % highest BBI-values were chosen to represent reference conditions for
each type and depth interval.
2) The EQR was calculated by dividing Observed values with the Reference values
3) The value for the high-good ecological status boundary is set at the 10 %-percentile of
reference EQR-values. Values below this boundary are divided into 5 equal classes Good 2/5,
Moderate 1/5, Poor 1/5 and Bad 1/5) to represent classes below high ecological status.
Boundary values were further validated by checking species richness, abundance, diversity values
and community composition of tolerant/sensitive species. The species tolerance or sensitivity was
determined according to species classification list made for the Swedish BQI-index in the Baltic
Sea, which was also used for the Finnish BBI (Perus et al. 2007). Definition criteria for high, good,
and moderate status in the coastal waters were following description in Annex V of the Directive.
24
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0027.png
Table 2.1.3:
The classification boundary values for the Finnish Brackish Water Benthic Index (BBI). Lower boundary
values for High (H), Good (G), Moderate (M) and Poor (P) status as BBI- and EQR values. Number of stations (Stations)
and reference stations (Ref.) visited. These are the final boundary values adjusted after intercalibration with Sweden.
#
Type
B3
Middle
B3
Outer
B2
Inner
B2
Outer
B0
Outer
Depth
0-10
10+
0-10
10+
0-10
10+
0-10
10+
0-10
10+
Stations
52
290
45
210
280
96
55
58
41
46
#
Ref.
5
29
5
21
28
10
6
6
4
5
BBI-EQR
H
0,93
0,89
0,92
0,90
0,94
0,95
0,88
0,92
0,94
0,98
G
0,56
0,53
0,55
0,54
0,56
0,57
0,53
0,55
0,56
0,59
M
0,37
0,36
0,37
0,36
0,38
0,38
0,35
0,37
0,38
0,39
P
0,19
0,18
0,18
0,18
0,19
0,19
0,18
0,18
0,19
0,20
B
<0,19
<0,18
<0,18
<0,18
<0,19
<0,19
<0,18
<0,18
<0,19
<0,20
BBI-values
H
>,70
>,53
>,74
>,62
>,52
>,71
>,67
>,60
>,71
>,64
G
0,42
0,32
0,44
0,37
0,31
0,42
0,40
0,36
0,43
0,38
M
0,28
0,21
0,29
0,25
0,21
0,28
0,27
0,24
0,28
0,25
P
0,14
0,11
0,15
0,12
0,10
0,14
0,13
0,12
0,14
0,13
B
<0,14
<0,11
<0,15
<0,12
<0,10
<0,14
<0,13
<0,12
<0,14
<0,13
For each individual national type comprehensive species lists have been compiled and reference criteria
can be set as a percentage value of species (alternatively number of sensitive species) that need to be
present for reference criteria to be met. At present such a percentage value have not been set.
Boundary values are further validated by checking species richness, abundance, diversity values and
community composition of tolerant/sensitive species. The same species list and species classification
is used in both Sweden and Finland (Blomqvist et al. 2007; Perus et al., in 2007).
Regional expertise has been contacted to determine whether or not the selected station visits are
appropriate (e.g. located in area where periodic hypoxia/anoxia may have occurred in periods when
no sampling has taken place.)
The example in Figure 2.1.1, illustrates how the boundary values of the BBI are linked with
sensitivity of the macroinvertebrate species groups for one common intercalibration type B0. The
species sensitivity groups 15 and 10 should be dominant or sub-dominant above the Good-Moderate
Figure 2.1.1:
Distribution of the relative abundance
of the four macroinvertebrate sensitivity groups as
a function of the BBI values for the common the
intercalibration type B0 (Data from the corresponding
national Finnish type). The distributions are illustrated
with the least square smoothing lines (DWLS,
stiffness=0.5). Group 15 is the most sensitive and group 1
the most tolerant group of species.
25
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0028.png
boundary (BBI value 0.43 from Table 3). The species sensitivity groups 5 and 1 should be dominant
or sub-dominant below the Good-Moderate boundary (Fig. 1). Additionally, diversity was higher
above Good-Moderate boundary than below.
Sweden:
Class boundary setting procedure and reference criteria
It was not possible to describe reference conditions using spatial approach or historical data,
since there are no unaffected areas in the Baltic nowadays. Also historical data was not collected
using comparative methodology as today. Instead the best available data from areas without local
discharges were used as a proxy for reference conditions in the boundary setting. Data from these
regions was assumed represent at least good ecological status.
In deriving the type-specific class boundaries, the greatest emphasis was placed on the good–
moderate boundary. This boundary was primarily determined using data from high–good areas.
Comparative data was chosen for each national type from regions lacking local discharges; in
practice from areas with the highest mean BQI values existing for that type. Sequential tests
identifying the level of BQI (20
th
percentile) where a water body significantly differs from the
comparison material were made to assist in the setting of good-moderate boundary. In those types
where the amount of data was insufficient, expert assessments were conducted based on existing
data and data from nearby types with similar properties.
Once the good–moderate boundary was determined, it was deemed acceptable to consider the area
from the good–moderate boundary up to the highest observed index value in the existing type as
mainly constituting a status of good. Two-thirds (2/3) of the span exceeding the good–moderate
boundary was assigned a status of good, while the upper third was reserved for the status of high
(Fig. 2.1.2). The area below the good–moderate boundary was divided into three equal intervals for
the remaining boundaries.
For calculation of EQR we have used the 20
th
percentile of BQI in a water body divided by the
highest observed BQI-value in a type.
The low salinity in the Baltic put a big natural stress on the fauna resulting in low number of taxa.
For this reason one can not expect a perfect fit to the normative definition.
More details on this approach are available in Blomqvist et al. (2007).
2.1.4 Results of the comparison
Three steps were carried out in the intercalibration exercise: 1) comparison of species sensitivity
classifications between the methods, 2) comparison of index values when applied for different
coastal types, and 3) comparison of ecological status classifications on water body level.
Comparisons between Swedish and Finnish methods
Due to differences in monitoring and assessment methods
17
, two separate comparisons were made.
Firstly, the comparison of the Finnish BBI and Swedish BQI macroinvertebrate classification
systems was made by applying both assessment tools using Swedish monitoring data. Secondly,
both methods were applied to Finnish monitoring data.
In Sweden macroinvertebrate monitoring is carried out using 0.1 m
2
van Veen grab and 1 mm sieve, while in Finland
0.025 m
2
Ekman grab and 0.5 mm sieve are used. The Swedish BQI index values are calculated separately for each grab
while the Finnish BBI values are calculated for five (5) pooled grab samples.
17
26
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0029.png
National type 8
100
Sensitivity class 1
Sensitivity class 5
Sensitivity class 10
80
Sensitivity class 15
60
% abundance
40
20
0
0
2
4
6
BQI
8
10
12
14
Figure 2.1.2:
Distribution of the relative abundance of the four macroinvertebrate sensitivity
groups as a function of the BQI values. Group 15 is the most sensitive and group 1 the most
tolerant group of species. Good-moderate boundary is indicated with a green and the High-good
boundary with a blue vertical line.
Comparison based on Swedish data
The first comparison of the two national methods was carried out by Sweden using Swedish national
and regional monitoring data from the Gulf of Bothnia. The data included in this exercise comes
from the national types No. 16 (Coastal waters of South Bothnian Sea, inner parts), 18 (Coastal
waters of North Bothnian Sea, Höga kusten, inner parts), 20 (Coastal waters of the Quark, inner
parts) and 22 (Coastal waters of North Bothnian Bay, inner parts). Only data deeper than 5 m were
used from these types. Due to methodological differences between Sweden and Finland, it was not
possible to apply the Finnish national boundaries on Swedish data. The intercalibration is based on
applying the Finnish method for setting of boundaries on Swedish data and comparing this with the
Swedish boundaries derived on the same Swedish data.
The intercalibration exercise was carried out in two steps: 1) Comparison of index values in
different environments, and 2) Comparison of classification status on water body level using both
BBI and BQI methods.
1. Comparison of index values
In the first step the Swedish BQI and The Finnish BBI indices were calculated for each grab sample
(~0.1 m
2
) separately and the results where plotted against each other (Fig. 2.1.3). The Swedish BQI
and the Finnish BBI indices were relatively well correlated, but the scatter increased in the two
northern types (types 20 and 22).
27
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0030.png
0,8
0,7
0,6
0,5
BBI
Typology 16
0,8
0,7
0,6
0,5
BBI
Typology 18
0,4
0,3
0,2
0,1
0,0
0
2
4
6
8
10
12
14
0,4
0,3
0,2
0,1
0,0
0
2
4
6
BQI
0,8
0,7
0,6
0,5
Typology 22
BQI
8
10
12
14
0,8
0,7
0,6
0,5
BBI
Typology 20
0,4
0,3
0,2
0,1
0,0
0
2
4
6
8
10
12
14
BBI
0,4
0,3
0,2
0,1
0,0
0
2
4
6
BQI
8
10
12
14
BQI
Figure 2.1.3:
Correlation between the BQI and BBI using Swedish national and regional monitoring data from the
types 16, 18, 20 and 22 (inner coastal waters, south to north, corresponding to common intercalibration types B3, B2
and B0, respectively).
2. Comparison of classification status on water body level
The classification status of water bodies was calculated applying both BBI and BQI on Swedish
monitoring data. The boundary setting was done following both Finnish and the Swedish
approaches for identifying reference status and the boundary setting principles for each station and
year separately.
The comparison was made using time series data from seven water bodies from the Swedish
national types 16, 18, 20 and 22. The Swedish method states that at least five stations per year
should be present in a water body for an assessment to be made (Blomqvist et al 2007). This rule has
been applied to both assessment methods. Both the Swedish and Finnish methods (Jens Perus, pers.
comm. March 2007) use the 20
th
percentile for water body assessment. The results of comparison of
the two methods are shown in (Fig. 2.1.4).
28
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0031.png
Swedish National type 16
16
Swedish National type
14,0
12,0
10,0
BQI
Edeboviken
Edeboviken
0,6
0,5
0,4
BBI
0,3
0,2
0,1
8,0
6,0
4,0
2,0
0,0
1981
1986
1991
1996
2001
2006
0,0
1981
1986
1991
1996
2001
2006
Swedish National type 18
Swedish National type 18
Svartviksfjärden
12,0
10,0
8,0
BQI
6,0
4,0
2,0
0,0
1984
1989
1994
Gaviksfjärden
12,0
10,0
8,0
BQI
6,0
4,0
2,0
0,0
1993
1998
Husumbukten
12,0
10,0
8,0
BQI
6,0
4,0
2,0
0,0
1983
1985
1987
1989
1991
BBI
0,6
0,5
0,4
0,3
0,2
0,1
0,0
1993
1998
2003
2003
BBI
0,6
0,5
0,4
0,3
0,2
0,1
0,0
1993
1998
Husumbukten
2003
1999
BBI
0,6
0,5
0,4
0,3
0,2
0,1
0,0
1984
1989
1994
Gaviksfjärden
1999
Svartviksfjärden
29
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0032.png
Swedish National type 20
Swedish National type 20
Örefjärden
12,0
10,0
8,0
BQI
6,0
4,0
2,0
0,0
1982
1987
1992
1997
2002
BBI
0,6
0,5
0,4
0,3
0,2
0,1
0,0
1982
1987
1992
1997
2002
Örefjärden
Swedish National type 22
22
Swedish National type
Kinnbäcksfjärden
12,0
10,0
8,0
BQI
6,0
4,0
2,0
0,0
1994
1999
Bergöfjärden
12,0
10,0
8,0
BQI
6,0
4,0
2,0
0,0
1994
1999
2004
BBI
0,7
0,6
0,5
0,4
0,3
0,2
0,1
2004
BBI
0,7
0,6
0,5
0,4
0,3
0,2
0,1
0,0
1994
1999
Bergöfjärden
2004
Kinnbäcksfjärden
0,0
1994
1999
2004
Figure 2.1.4:
Comparisons of the water body assessments using Swedish (BQI; left column) and Finnish (BBI; right
column) macroinvertebrate methods applied to Swedish monitoring data from the national Swedish types No. 16
(Coastal waters of South Bothnian Sea, inner parts), 18 (Coastal waters of North Bothnian Sea, Höga kusten, inner
parts), 20 (Coastal waters of the Quark, inner parts) and 22 (Coastal waters of North Bothnian Bay, inner parts). Error
bars denotes 20
th
and 80
th
percentile. Note that the 20
th
percentile (lower error bar) is used to determine the calssification
status. Boundaries hose are from March 2007 (Jens Perus, pers. comm. and Blomqvist et al. 2007).
In most cases the two methods show the same general pattern for the water body assessments. In
some years there are differences in the two assessment methods, e.g. the collapse of the dominating
Monoporeia affinis
in Kinnbäcksfjärden year 2004 and Örefjärden year 2000 is more clearly
detected by the Swedish assessment method.
30
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0033.png
Comparison based on Finnish data
The other comparison of the Finnish BBI and Swedish BQI macrozoobenthos assessment methods
was done by using Finnish national and regional monitoring data from the Baltic Sea. Data used in
the comparison was representative for the four national Finnish water body types (Table 2.1.4). Only
data from below 10 m depth was included in the exercise. The national Finnish typology separates
coastal types into two depth zones: above 10 m and below 10 m depth intervals. The national
boundaries are set separately for different depth intervals within the same type.
Table 2.1.4:
Correspondence of the common intercalibration types for the Northern Baltic Sea with the Finnish
national types (all deeper than 10 m; Perus et al. 2004) and Swedish national types (NFS 2006:11) included in the
intercalibration exercise. All the common types were characterized to be deeper than 10 m.
Intercalibration type
B0
B2
B3a
B3b
Finnish type
Mu
SeS
Lv
Lu
Swedish type
20
16
12
14
The Finnish BBI and Swedish BQI assessment systems were compared by setting the boundaries
for the Finnish water bodies belonging to the intercalibration types B0, B2 and B3 and using the
Swedish assessment method. The Swedish calculation method cannot be applied directly on Finnish
data due to methodological differences
18
. However, this problem was overcome using the boundary
setting principle of the Swedish method (pers. comm. Mats Blomqvist, Hafok AB). According to
this approach 80 % (20
th
percentile) of the natural variation within a data set, originating from an
area without local discharges, can be regarded as high or good status i.e. thus being above the good-
moderate boundary. This good-moderate boundary value was set for each of the four intercalibration
types (Fig. 2.1.5).
A common species sensitivity classification list is used for both indices and no comparison of
differences in species sensitivity classification is thus needed.
After comparison of the Finnish-Swedish classification results, a change in the Finnish boundaries
was proposed. The original Finnish classification boundaries for the classes Good-Moderate-Poor-
Bad were set as four equally sized classes beneath the High-Good boundary. The proposal is that
the boundary for High-Good remains the same, but that the remaining BBI-index spectra should be
divided in to five equally sized classes out of which the Good status class should occupy two fifths
(2/5) and classes Moderate, Poor and Bad one fifth each (1/5). A new specification on the Finnish
assessment of status is that the use of the 20
th
percentile value as assessment value would be taken
into practice (identical to Swedish practice).
18
In Finland 5 pooled (mean of abundances (ind/m
-2
) samples taken by Ekman-Birge grab samples (0.5mm mesh)
are required and total species richness needs to be determined while in Sweden individual (0,1 m
2
) Van-Veen grab sam-
ples (1mm mesh) are required to be taken. Also the Swedish method requires at least five stations per year for assess-
ment
31
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0034.png
Rönnskär (B0 outer 10+m)
1
0,8
BBI
0,6
0,4
0,2
0
1994
1995
1996
1997
1999
2000 All data
BQI(2006)
Rönnskär (B0 outer 10+m)
15
12
9
6
3
0
1994
1995
1996
1997
1999
2000 All data
Replotfjärden (B0 outer 10+m)
1
0,8
0,6
BBI
0,4
0,2
0
1994
1995
1996
1997
All data
BQI(2006)
15
12
9
6
3
0
1994
Replotfjärden (B0 outer 10+m)
1995
1996
1997
All data
B2 inner (10+m)
B2 inner (10+m)
1
0,8
BBI
0,6
0,4
0,2
BQI(2006)
1990
1994 1995
1996
1997
1998 All data
6
5
4
3
2
1
0
1990
1994 1995
1996
1997
1998 All data
0
1
0,8
BBI
0,6
0,4
0,2
0
B3 middle 10+ (Da10+m)
16
14
12
10
8
6
4
2
0
B3 middle 10+ (Da10+m)
BQI(2006)
1991 1994 1995 1996 1997 1998 1999 All data
Kasnas East (B3 middle 10+)
1
1991 1994 1995 1996 1997 1998 1999 All data
Kasnas East (B3 middle 10+)
16
14
12
10
8
6
4
2
0
1991 1994 1995 1996 1997 1998 1999 2000 All data
0,8
BBI
0,6
0,4
0,2
0
1991 1994 1995 1996 1997 1998 1999 2000 All data
BQI(2006)
32
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0035.png
Kustavi straits (B3 middle 10+)
1
0,8
0,6
BBI
0,4
0,2
0
1995
1998
All data
BQI(2006)
16
14
12
10
8
6
4
2
0
Kustavi straits (B3 middle 10+m)
1995
1998
All data
B3 (outer, 10+m)
1
0,8
0,6
BBI
0,4
0,2
0
1991
1993
1995
1997
All data
BQI(2006)
16
14
12
10
8
6
4
2
0
1991
B3 (outer, 10+m)
1993
1995
1997
All data
Figure 2.1.5:
Comparisons of the water body assessments using Swedish (BQI; left column) and Finnish (BBI; right
column) macroinvertebrate methods applied to national Finnish monitoring data from water bodies belonging to common
intercalibration types B0, B2, B3a (B3 middle, Kasnäs East, Kustavi Straits) and B3b (B3 outer, WB Jussarö at the
entrance of the Gulf of Finland (GOF), WB Rosala-Hitis). Error bars denotes 20
th
and 80
th
percentile. Note that the 20
th
percentile (lower error bar) is used to determine the classification status. Detailed results are presented in Annex 2.1.
Table 2.1.5:
The intercalibrated Finnish and Swedish EQR-values for the high-good (HG) and good-moderate (GM)
boundaries calculated for the national coastal types representative for the common intercalibration types B0, B2 B3a,
and 3b.
Type
B0
B2
B3a
B3b
Country
Finland
Sweden
Finland
Sweden
Finland
Sweden
Finland
Sweden
National type
Mu 10+m
20
Ses 10+m
16
Lv 10+m
12
Lu 10+m
14
EQR HG
0,99
0.77
0,95
0.76
0,89
0.76
0,90
0.76
EQR GM
0,59
0.31
0,57
0.29
0,53
0.29
0,54
0.29
33
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0036.png
2.1.5 Results of the harmonization – Boundary EQR values
Biological Quality Element
Benthic invertebrate fauna
Results: Ecological quality ratios of the national classification systems intercalibrated
Type and country
National classification
system intercalibrated
Ecological Quality Ratios
High-Good boundary
CW B0
Finland
Sweden
CW B2
Finland
Sweden
CW B3 a
Finland
Sweden
CW B3 b
Finland
Sweden
BBI- Finnish Brackish
water Benthic Index
BQI–Swedish multimetric
biological quality index
(soft sediment infauna)
0.90
0.76
0.54
0.29
BBI- Finnish Brackish
water Benthic Index
BQI–Swedish multimetric
biological quality index
(soft sediment infauna)
0.89
0.76
0.53
0.29
BBI- Finnish Brackish
water Benthic Index
BQI–Swedish multimetric
biological quality index
(soft sediment infauna)
0.95
0.76
0.57
0.29
BBI- Finnish Brackish
water Benthic Index
BQI–Swedish multimetric
biological quality index
(soft sediment infauna)
0.99
0.77
0.59
0.31
Good-Moderate boundary
2.1.6 Open issues and need for further work
Intercalibration need to be extended to cover all Baltic Sea countries and all common types.
Currently results are completed at quality element level only for two countries (FI, SE) covering
three (3) common types in the Northern Baltic Sea (Bothnian Sea and Bothnian Bay).
Denmark, Sweden, and Germany carried out intercalibration of their national methods for the
common type B12. However, at this stage comparability was not sufficiently demonstrated, and the
results are not included in the final intercalibration decision. New comparisons have to be carried
out when more monitoring data will be available for common intercalibration types in the Southern
Baltic Sea.
34
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
Other countries still need to develop and intercalibrate WFD compliant assessment methods for
benthic invertebrates for the Baltic coastal waters.
Currently most national methods have been developed for the soft bottom sediment habitats only,
with the exception the German method that is a multi-habitat method. Assessment methods for other
coastal habitats (sandy and rocky bottoms. littoral zone) have to be developed. methods for the
transitional water types in the Baltic have to be developed.
2.2 Black Sea GIG
2.2.1 Intercalibration approach
Typology
The countries participating at this intercalibration exercise into the Black Sea region are Bulgaria
and Romania. One common water body type CW-BL 1 was defined, as is shown below:
Salinity: Mesohaline
Tidal Range: Microtidal
Depth: Shallow
Subsratum: Mixed
Wave Exposure: Moderately Exposed
2.2.2 National methods that were intercalibrated
Boundaries were agreed in this intercalibration exercise for the Shannon diversity index H’, AMBI,
M-AMBI, applicable to both Bulgaria and Romania.
2.2.3 Reference conditions and class boundary setting
As Option 3 has been selected, both countries have used the same methods for sampling, laboratorial
analyses, reference conditions and EQRs. Those were derived using literature, historical data and
expert judgment.
The criteria for establishment of biological reference conditions and the boundary setting procedure
are described below separately for each country.
2.2.4 Results of the comparison
The method of sampling and analytical methods are as provided by the Manual “Quantitative
sampling and sample treatment of marine soft bottom – macrozoobentos” (Tsenka Konsulova,
Valentina Todorova – 2005), agreed by the Black Sea Commission. The species names used in the
process of AMBI calculation should be in accordance with the valid names from the same manual.
The macroinvertebrates experts, from both countries, agreed to use three indices to assess the
ecological status of the selected common type: Shannon Diversity Index H’, AMBI and M-AMBI.
Both countries agreed to use “one out all out principle” to establish the ecological status.
35
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0038.png
BULGARIA
Reference conditions and EQR values were developed for Benthic invertebrate fauna in Bulgarian
waters by the following scientists from the Marine Biology and Ecology Department; INSTITUTE
OF OCEANOLOGY, BAS, Varna:
Assoc. Researcher Antoaneta Trayanova;
Dr. Assoc. Researcher Valentina Todorova;
Dr. Assoc. Professor Tsenka Konsulova;
For the Bulgarian Black Sea coastal waters, no reliable data of the “pristine” period (before the
1970s) collected with the same methodology as we use today, exist. Use of historical data is
therefore not an option in determining reference conditions. Another approach is screening for un-
impacted areas, i.e. communities not affected by human activities. Unaffected areas are regarded
as absent in the Bulgarian Black Sea coastal waters nowadays. The remaining option to establish
condition comparable to a reference is to derive it from areas as unaffected by human activities
as possible using the data of slightly disturbed benthic communities (having at least good status).
An expert judgment and knowledge of the conditions under evaluation were applied in obtaining of
“virtual” reference conditions.
The study area covers the one nautical mile zone along the entire Bulgarian Black Sea coast.
Nineteen sites have been sampled for macrozoobenthos in the period 2002-2006 (Figure 2.2.1).
Figure 2.2.1:
Monitoring sampling
network of the Bulgarian Black
Sea coastal waters (used for
intercalibration exercise)
36
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0039.png
Table 2.2.1:
Number of samples per station by sampling years, months and totally.
Station (1nm) / Sampling
year and month
10 (100) Krapetz
20 Shabla
30 Tjulenovo
2002
8
1
11
1
3
2003
6
1
6
1
2004
9
1
3
1
6
1
2005
9
1
11
1
2006
7
1
Number
of
samples
per
station
10
1
1
1
1
40 (111) Rusalka Resort
200 Kaliakra
(in front of the cape)
50 (211) Kaliakra
(under the cape)
60 Albena Resort
1
1
1
1
1
1
1
1
1
1
1
7
4
1
1
1
1
1
1
1
1
1
1
1
1
1
1
7
1
13
1
1
5
70 (B3) Varna Bay
90 Kamchia River
411 Irakli
80 (301) cape Galata
100 Dvoinitsa River
110 (500) Slanchev Briag
120 Burgas Bay
140 - Sozopol
160 - Varvara
130 (501) Burgas
150 (600) Achtopol
170 (611) Veleka River
2
2
1
1
1
1
1
1
1
1
1
1
3
1
1
1
1
1
1
1
2
1
1
2
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
15
11
1
8
1
3
1
4
Total number of samples
9
7
4
7
7
7
8
10
10
9
17
95
a. Shannon diversity index H’
Although the identification of reference and bad status values rely on expert assessment the limited
number of samples and/or replicates results in high degree of uncertainty.
When deriving the type-specific class boundaries, the greatest emphasis was placed on the good–
moderate boundary. This boundary was primarily determined with the aid of data from high–good
areas.
The approach in identifying reference conditions is based on the hypothesis that the high ecological
status represents certain percentage of the good status.
The reference value for water bodies with sandy and mixed sediments is derived with the
assumption that the average community diversity index of stations reaching good ecological status
constitutes 75 % from the high status. The reference value for water bodies with muddy sediments
have been determined from the average community diversity index of stations reaching good status
and having more than 25 species per sample, than the reference value is extrapolated from current
status deemed as 70 % from reference. The percentage is lower due to the specificity of the muddy
sediments characterized by higher sensitivity of the developing communities to environmental
disturbances.
37
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0040.png
The boundaries between the ecological classes are set as percentage from average reference values.
The interval of 0.2 between the ecological status boundaries is equally scaled (80 % for good,
60 % for moderate, 40 % for poor and 20 % for bad status). For determination of EQRs each
class boundary value has been divided by the average reference value – 3.6 for water bodies with
muddy sediments and 4.5 for water bodies with sandy and mixed sediments (Table 2.2.2). The data
available do not allow developing of separate classification schemes for water bodies with sandy
and for water bodies with mixed sediments. It is highly recommended the boundary values to be
readjusted after new data are available as well classification scale for water bodies with mixed
sediments to be developed.
Table 2.2.2.
Classification scheme for Shannon community diversity index (H’).
Ecological status
High
H’ ≥ 3.3
≥ 0.92
High
H’ ≥ 4
4.5
3.6
Water bodies with muddy sediments
Good
2.9
Moderate
2.2
Poor
1.8 > H’ ≥ 1.1
0.31
1.5
Bad
H’ < 1.1
< 0.31
Bad
H’ < 1.3
< 0.29
0.9
0.7
Range
EQR
H’ average
3.3 > H’ ≥ 2.5
0.69
2.5 > H’ ≥ 1.8
0.50
Range
EQR
H’ average
Ecological status
Water bodies with sandy and mixed sediments
Good
3.6
Moderate
2.7
Poor
1.8
≥ 0.89
4 > H’ ≥ 3.1
0.69
3.1 > H’ ≥ 2.2
0.49
2.2 > H’ ≥ 1.3
0.29
b. A Marine Biotic Index (AMBI)
The method is based on the assignment of species to five ecological groups, according to the
sensitivity to an increasing stress gradient (Gray and Glemarec (1997):
• Group I – species very sensitive to organic enrichment and present under unpolluted conditions
(initial state). They include the specialist carnivores and some deposit feeding tubicolous
polychaetes;
• Group II – species indifferent to enrichment, always present in low densities with non-
significant variations with time. These includes suspension feeders, less sensitive carnivores
and scavengers;
• Groups III – species tolerant to excess organic matter enrichment. These species may occur
under normal conditions, but their populations are stimulated by organic enrichment. They are
surface deposit – feeding species, as tubiculous spionids;
• Group IV – Second order opportunistic species. Mainly small sized polychaetes; subsurface
deposit feeders, such as cirratulids;
• Group V – First order opportunistic species. These are deposit feeders, which proliferate in
reduced sediments;
The reference value (AMBI ≤ 1.2) is derived as dominance of sensitive and indifferent taxa in
the abundance. The bad status value (AMBI > 5.5) is achieved when second order and first order
opportunists dominate. The boundaries between the ecological classes are those identified by Borja
et al.,
(2000, 2003) and Muxica
et al.,
(2005). EQR is determined by subtraction of the boundary
value divided by the maximal value 7 from 1 (Table 2.2.3). AMBI is not sediment dependant which
allows to be applied for all WBT regardless of their typology
38
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0041.png
Table 2.2.3:
Classification scheme for A Marine Biotic Index (AMBI).
Ecological status
EQR
Range
0.2 < AMBI ≤ 1.2 1.2 < AMBI ≤ 3.3 3.3 < AMBI ≤ 4.3 4.3 < AMBI ≤ 5.5 5.5 < AMBI ≤ 7.0
≥ 0.83
0.53
0.39
0.21
< 0.21
High
Good
Moderate
Poor
Bad
c. Multivariate AMBI (M-AMBI)
The default EQR boundaries of M-AMBI are set through intercalibration exercise for North Atlantic
ecoregion (Borja
et al.,
2006). Because no proper intercalibration is carried out for the Black Sea
ecoregion we adopted the default boundaries between the ecological classes. Further changes of
these values for each WBT are advisable after intercalibration procedure and harmonisation of these
boundaries between Member States took place. The default EQR boundaries of high and bad status
have been accepted for all WBT
Table 2.2.4.
Classification scheme for Multivariate AMBI (M-AMBI).
Ecological status
Range
EQR
High
M-AMBI ≥ 0.85
≥ 0.85
0.85 > M-AMBI
≥ 0.55
0.55
Good
0.55 > M-AMBI
≥ 0.39
0.39
Moderate
0.39 > M-AMBI
≥ 0.20
0.20
Poor
Bad
0.20 > M-AMBI
< 0.20
The hierarchical clustering of stations based on the Bray-Curtis similarities of log (x + 1)
transformed abundance was employed to reveal the similarity pattern of macrozoobenthic
community. Five main groups have been differentiated on the dendrogram corresponding to the type
of the sediment (Figure 2.2.2). In some cases the WB typology differentiates from the sediment type
defined by the clustering technique, which is indication of misclassification of sediments resulting
in inaccurate water bodies’ typology.
Because the sediment type determine the development of specific macrozoobenthic communities,
it is reasonable the assignment of high and bad status values for the selected metrics to be done
according to the cluster groupings.
Figure 2.2.2:
Dendrogram for
hierarchical clustering (group-
average linking) of WBT and
stations.
39
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0042.png
ROMANIA
The study area covers the southern part of the Romanian Black Sea coast, from Singol Cape to
Vama Veche (Fig.1). To assess the ecological status it was use the data from four sampling stations,
during the period 2002-2006 (Table 2.2.4).
Table 2.2.4
Sampling stations / Number of samples per station.
Station/
Sampling year
and month
Costinesti
Vama Veche
Mangalia
Eforie
2002
May
1
1
2
-
-
Aug
1
1
2
-
-
2003
May
1
1
2
1
-
-
May
1
2004
Aug
1
1
4
1
1
May
1
1
1
4
1
2005
Oct
1
-
2006
Apr
1
1
1
3
-
Number of
samples /
station
5
3
6
Total number
of samples
1
3
1
7
21
Figure.2.2.3:
Monitoring sampling
network of the Romanian Black
Sea coastal waters (used for
intercalibration exercise)
40
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0043.png
The sampling and analytical methods used were those provided in the manual Quantitative sampling
and sample treatment of marine soft bottom – macrozoobentos” (Tsenka Konsulova, Valentina
Todorova – 2005), approved by the Black Sea Commission.
a. A Marine Biotic Index (AMBI)
The AMBI was calculated using the data from the period 2002 – 2006. Results are shown in
Table 2.2.5.
Table 2.2.5:
AMBI values for the Romanian sampling network.
Sites
2002
2003
2004
2005
2006
Average
Eforie South
-
-
3.04
3.12
3.31
3.2
Costinesti
4.12
3.85
4.17
3.13
2.86
3.6
3
2.96
3.01
Mangalia
Vama Veche
3.06
3.25
2.89
2.14
1.58
2.6
It was estimated that the general status was moderate (3.3<AMBI≤4.3).
The boundaries between ecological classes are those identified by Borja et al. (Table 2.2.6).
b. Multivariate AMBI (M-AMBI)
The M-AMBI was calculated using the data exported by the AMBI software for each of the stations.
Based on the results and on expert judgment it was decided that the boundaries identified by Borja
will be considered in assessment of the ecological status.
The boundaries values were considered as were identified by Borja et al. (Table 2.2.7).
c. Shannon diversity index H’
For the Romanian coastal waters, no reference sites were available for macro invertebrates.
Although, the experts were used data from scientific literature and based on expert judgment, they
set the values for reference conditions.
Using the AMBI software, the Shannon diversity index was calculated and based on acknowledge
of the experts regarding the actual ecological status, the boundaries were considered as shown in
Table 2.2.8.
Table 2.2.6:
H/G and G/M boundaries for AMBI
Ecological
status
Value
EQR
H/G
1.2
0.83
G/M
3.3
0.53
Table 2.2.7:
H/G and G/M boundaries for M-AMBI
Ecological
status
Value
EQR
H/G
0.85
0.85
G/M
0.55
0.55
41
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0044.png
Table 2.2.8:
H/G and G/M boundaries for Shannon diversity index H.
Ecological status
Value
EQR
H/G
0.8
4
G/M
0.6
3
The evaluation was made processing the data from the period 2002-2006.
2.2.5 Results of the harmonisation – Boundary EQR values
AMBI
In was established that the AMBI should be calculated according to the “Guidelines for the use of
AMBI (AZTI’s Marine Biotic Index) in the assessment of the benthic ecological quality”.
The species names used with AMBI should be according to the valid names from the manual
“Quantitative sampling and sample treatment of marine soft bottom – macrozoobentos” (Tsenka
Konsulova, Valentina Todorova – 2005).
In the assessment of the ecological status, both countries decided to use the boundaries defined by
Borja (Table 2.2.9).
Table 2.2.9:
AMBI boundary values.
Ecological status
Range
EQR
AMBI <1.2
> 0.83
High
1.2<AMBI<3.3
>0.53
Good
H/G
0.83
1.2
G/M
0.53
3.3
Shannon H’
Both countries agreed that the Shannon diversity index should be calculated using log
2,
as Bulgarian
experts did in their assessment:
Table 2.2.10:
Shannon Diversity boundary values
Ecological status
Range
EQR
H’ > 4
≥ 0.89
High
4> H’ > 3.1
≥ 0.69
Good
H/G
4
0.89
G/M
3.1
0.69
M-AMBI
The process of evaluation of M-AMBI was made taking into consideration some values for high
status and bad status for richness and diversity (required by the program), established by common
agreement as being characteristic to the Black Sea common type.
Table 2.2.11:
M-AMBI boundary values.
Range
EQR
Ecological status
High
M-AMBI>0.85
Good
0.85>M-AMBI>0.55
H/G
0.85
0.85
G/M
0.55
0.55
> 0.85
> 0.55
42
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0045.png
In the process of calculation, the reference and bad status values for Species richness (S), diversity
(H’) and AMBI were:
Table 2.2.12:
Reference and bad status values for species richness, Shannon diversity, and AMBI.
high
Richness
Shannon
AMBI
50
4
1.2
bad
15
1.3
5.5
For macroinvertebrates, both countries agreed to use the “one out, all out” principle to establish the
ecological status.
2.2.6 Open issues and need for further work
Refining of the typology of water bodies is needed after the sediment types are accurately classified.
The determination of the reference values for the community diversity index (H’) was derived
from actual data for good ecological status and based on expert judgement and knowledge. The
limited number of stations and/or replicates in one nautical mile zone allowed partial application
of statistical approaches and did not permit the development of classification scale for water bodies
with mixed sediments. Revision of the boundary values is recommended after accumulation of
further data.
The default boundary values of M-AMBI have to be readjusted for each WBT after intercalibration
procedure for the whole Black Sea ecoregion takes place.
2.3 Mediterranean GIG
2.3.1 Intercalibration approach
Participation of countries in the Benthic Invertebrates subgroup:
Cyprus
France
Greece
Italy
Slovenia
Spain (Catalonia, Balearic Islands)
TYPOLOGY
In the Mediterranean 4 basic intercalibration types have been agreed. These are shown in table 2.3.1
below:
43
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0046.png
Table 2.3.1:
Mediterranean coastal common intercalibration types.
Type
CW - M1
CW - M2
CW - M3
CW - M4
Name of Type
Rocky shallow coast
Rocky deep coast
Sedimentary shallow coast
Sedimentary deep coast
Substratum (1)
rocky
rocky
sedimentary
sedimentary
Depth (2)
shallow
deep
shallow
deep
(1) In many cases different seabed substrata will occur within one water body type. The dominant substratum should be
defined.
(2) Depth division is based on 40 m depth at 1 mile distance from the coastline.
Benthic invertebrates experts defined that the four Mediterranean coastal IC types (see section 1
General part), primarily defined on substratum composition and depth profile, are not relevant to the
MED-GIG benthic invertebrates intercalibration.
It is important to note that the intercalibrated methods in this phase are habitat specific (e.g soft
bottom, hard bottom).
All methods are for use in soft sediment habitats.
Methods for other habitats
are still under development and cannot be assessed at this stage.
OPTION 3 (direct comparison of national methods) has been used.
All pressures are considered but the methods are particularly sensitive to organic enrichment.
2.3.2 National methods that were intercalibrated
Member State
Cyprus
France
Greece
Italy
Slovenia
Spain - Catalonia
Spain- Balearic is.
Method
Bentix
Multimetric approach (AMBI, Shannon
Diversity, BQI Trophic Index)
Bentix
AMBI, M-AMBI, Bentix
M-AMBI
MEDOCC
MEDOCC
Status
Finalized
Under development
Finalized
Under development
Finalized
Officially accepted
Officially accepted
Boundary
agreement
yes
no
yes
no
yes
yes
yes
2.3.3 Reference conditions and class boundary setting
As option 3 has been selected for the intercalibration of this quality element each Member State has
derived biological reference conditions and the boundaries according to their classification systems.
The criteria for establishment of biological reference conditions and the boundary setting procedure
are described below separately for each country.
44
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
GREECE and CYPRUS
Reference conditions
Species sensitivity
General reference conditions are based on the normative definition, which states, “All the
disturbance-sensitive taxa associated with undisturbed conditions should be present”.
Reference sites from Greek data were defined, presenting reference conditions for biological element
macroinvertebrates. The fauna is composed of mostly sensitive species (GI) and corresponding
mean Bentix values are amongst the highest: Bentix>5. In these cases the composition of the fauna
corresponds to sensitive species over 75 %. In special cases where muddy bottoms are encountered
within a reference site Bentix values are expected to reach values over 4 and sensitive species
percentage over 50 %.
Species ecology
Another aspect biological reference conditions setting is based on the autoecology of species.
Each species is designated with an ecological identity as extracted from scientific literature, so it
is possible to identify the species belonging to, or characterizing each type of community > habitat
> water body, and thus to establish reference conditions on an ecological basis. Species reference
lists are established for each kind of habitats-communities (Simboura & Zenetos, 2002) and the
link among community-habitat-water body type is given (Simboura et al., 2005, Sete presentation,
Simboura) following the EUNIS habitat classification scheme for European coasts.
Diversity indices
Other indices as the Shannon Diversity index and species Richness are expected to be among the
highest according to the given type of habitat, in the sites under reference conditions. For example
over a large set of data from Greek coastal areas and for a sample size of 0.1m2, Hmax=6.3
and Smax=110 for mixed sediments, while for muds the respective values were Hmax=4,6
and Smax=39. Generally for the above standard reference sample area and for mixed sediments
H values in reference sites are expected to be over 5 and S over 80 and for muddy bottoms H
over 4-4.5 and S over 30. However, discrepancies in the values of these indices may arise from
sampling methodology differences and habitat particularities. Another point to be considered is
that in transitional zones of disturbance (ecotone) Shannon diversity and species richness maybe
significantly high leading to misinterpretation of reference conditions; so diversity values should be
cross-checked with biotic indices.
Boundary Setting
The Bentix index (Simboura & Zenetos, 2002) is a newly developed biotic index based on the
relative percentage of ‘sensitive’ (GS) and ‘tolerant’ (GT) species in the fauna weighted analogously
to derive a single formula:
Bentix = (6 X %GS + 2 X %GT)/100
(http://www.hcmr.gr/english_site/services/env_aspects/bentix.html).
The formula was initially developed as Bentix = [(6 X %GI + 2 X ( %GII + %GIII)]/100 where
GI includes the sensitive and indifferent taxa, GII the tolerant and second order opportunistic and
GIII the first order opportunistic taxa. As the multiplying factor 2 is the same for groups GII and
GIII, these groups can be merged to simply represent all tolerant taxa (GT) versus all sensitive taxa
(GS). The resulting classification scheme and Ecological Quality Ratio (EQR) which is the universal
comparison scale for different metrics is given in table 2.3.2.
45
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0048.png
The Bentix index was developed in the Mediterranean ecoregion where the benthic fauna is usually
very diverse and evenly distributed with no one species naturally dominating over 10 %. Also under
slight disturbance conditions there might be the situation where a high diversity of sensitive and
opportunistic species may co-exist giving the impression of a high quality status. Given the naturally
even distribution of the fauna, the groups are weighted equally in the Bentix formula taking into
account simply only the ratio of the groups’ occurance in the fauna: the probability of one species
randomly picked up from the fauna to belong to a “tolerant” over a “sensitive” group is 3:1. Also the
indifferent group is counted with the sensitive. This approach combined with an equal scaling of the
ranges of good and moderate classes (2.5-3.5, 3.5-4.5 respectively), results to successfully designate
the community health in the Mediterranean benthic ecosystem where, naturally a high number of
species are evenly distributed over the benthic population.
Table 2.3.2:
The Bentix index classification scheme and EQR.
EcoQS
High
Good
Poor
Bad
Bentix values
4.5 < Bentix < 6
EQR
>0.75-1
Moderate
3.5 < Bentix < 4.5
2.5 < Bentix < 3.5
2.0 < Bentix < 2.5
0
>0.58-0.75
>0.42-0.58
>0.33-0.42
0
* For naturally stressed muddy habitats only, boundary limits 4.5 and 3.5 are reduced to 4 and 3 respectively.
** Also for comparison with the m-Ambi EQR, the EQR scale starts from value 2 corresponding with EQR=0, since
the area among 0 and 2 is a non-value area (as for Ambi=6 the EQR=0).
Class boundary values were set by checking community composition of tolerant/sensitive species
over the Bentix values for Greek and Cyprus IC data. This ratio is selected as paired metrics to
locate class boundaries.
Figure 2.3.1 shows the plotting of the percentage of tolerant and sensitive taxa along the gradient of
decreasing Bentix index. Within the range high to poor, changes in the percentage of sensitive and
tolerant taxa are gradual and indicated by the smoothing lines.
Based on this plot, it is evident that the cross-line point of the two lines corresponding to the value
of Bentix=4, represents the class center of “good” where literally the two ecological groups of
tolerant and sensitive share the fauna by 50 % each. This point corresponds with the ecotone point
of the transitional zone, middle of good class.
The points at equal distances (0.5) in each side of the crossline represent the high-good boundary
limit with value 4.5, and at the other side of the center the boundary between good/moderate with
value 3.5. The HG and GM boundaries and the center of good class are indicated by vertical lines.
At the high to good class boundary (Bentix=4.5), the percentage of the sensitive taxa drops to less
than 60 % of the fauna and the percentage of the tolerant taxa accounts for more than 40 %. At the
good to moderate class boundary (Bentix=3.5), the percentage of tolerant species becomes over 60
% (roughly 2/3 of the fauna) and the sensitive taxa less than 40 % (1/3 of the fauna).
46
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0049.png
Figure 2.3.1:
Bentix values
(from Greece and Cyprus IC data)
plotted against the percentages of the
two ecological groups. P: poor class,
M: moderate class, G: good class,
H: high class.
Compliance with Normative definitions
Figure 2.3.2 presents the degradation model of the percentage contribution of the three main
ecological groups treated initially in the formula: GI (sensitive + indifferent), GII (tolerant + second
order opportunistic), GIII (first order opportunistic) in the benthic fauna in relation with the values
of the BENTIX index. The sequence of quality classes and class boundaries are interpreted in
terms of shifts of ecological group percentages. At the border of good to high status (Bentix=4.5)
the sensitive group accounts roughly for more than 60 % or more than 2/3 of the fauna, while the
tolerant group as a whole (tolerant plus opportunists) accounts for less than 40 % or less than 1/3
of the fauna. It is important to stress here that for purely muddy habitats where the benthic fauna is
normally dominated by some tolerant species, and only in this class border among high and good, a
possible refinement of the boundary limit would change 4,5 to 4.
This is to some extent in accordance with the normative conditions, which states that “The level of
diversity and abundance of invertebrate taxa is slightly outside the range associated with undisturbed
conditions. Most of the sensitive taxa of the type specific communities are present”.At the border of
good to moderate status (bentix=3.5) the sensitive group accounts roughly for less than 40 % or
less than 1/3 of the fauna, while the tolerant group as a whole (tolerant plus opportunists) accounts
for more than 60 % or more than 2/3 of the fauna. This is to some extent in accordance with the
normative conditions, which states that “The level of diversity and abundance of invertebrate taxa is
moderately outside the range associated with undisturbed conditions. Most of the sensitive taxa of
the type specific communities are absent”.
47
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0050.png
Figure 2.3.2:
Degradation model of benthic fauna composition in terms of ecological groups percentages
(GI, GII, GIII) from heavily polluted to undisturbed communities in relation to the Bentix index tools.
Index response to the impact gradient
Plotting the Bentix results from Greek IC data with chemical parameters available, the Bentix index
showed a linear relationship with the pressure gradient, expressed by chemical pressure factors
(nutrients, oxygen and organic carbon content in sediment) with no discontinuities (Fig. 2.3.3).
Using Cyprus data were the pressure gradient is represented by the increasing distance from fish
farm areas the results of the Bentix EQS assessment shows a good response of the index in relation
with the gradient (Figure 2.3.4).
In the three case areas the EQS is improving (moderate-good or moderate-good-high) with
increasing distance. In series C the presence of Posidonia meadows at 300 and 1000 m distance
from the cages determine the high EQS status.
Figure 2.3.3:
Correlation of Bentix
with chemical parameters.
48
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0051.png
Series A
6
4. 5
3
1. 5
0
Μ
0
A-1
G
50
G
200
G
500
G
1800
G
5000
6
4.5
3
1.5
A-re f
0
Μ
0
B -1
G
Series B
G
250
H
1000
50
B-2
B -3
B -Ref
Increasing distance from cages
A-2
A-3
A-4
L -re f
Increasing distance from cages
Series C
6
4 .5
3
1 .5
Μ
0
CI
Μ
200
CI I
CIII Pos CV Pos
H
300
H
1000
G
40 00
CVI
Figure 2.3.4:
Variation of the Bentix index and resulted
EQS along a gradient of increasing distance from fish
farm cages in Cyprus (distances in meters are marked
below the EQS).
0
Increasing distance from cages
Results of the comparison
AMBI-BENTIX
The two indices were compared after revising the scores of both indices reaching an agreement
on certain ambiguous scores. A direct comparison among the two biotic indices EQRs (Figure
2.3.5) shows the relation among the two indices’ Ecological Quality Ratio (EQR) which is linear
and statistically significant (R2=61, p=0.000). The lines in the plot define the ranges of each class
set by the boundary limits of each method. The Ambi boundary limits were set according to WFD
ecological status classification (Muxica et al., 2005). The two indices are linearly correlated, but the
EQS assessment rendered by each method differs as shows a Box-and-Whisker plot comparison of
the EQS assessment versus the values of the EQRs of the two indices (Figure 2.3.6). This difference
is due to a) the differential weight each index puts in the various ecological groups and b) in the
different boundary limits of Ambi values and EQRs set among classes as fig. 2.3.5 shows.
In the good class range, the two indices give a common assessment. However, the Ambi
classification method gives a wider “good” class range compared with the Bentix, classifying most
moderate and often high class sites according to the Bentix to the “good” class. Generally, moderate
and high class ranges in the Ambi are more compressed compared to the Bentix scheme.
These differences are structural: in the Ambi method the indifferent group species are weighted
separately, while Bentix counts them with the sensitive. Also the Ambi gives different significance
to each “tolerant” group of species. In the Bentix method all tolerant species are weighted equally
versus the sensitive ones. Besides, the scaling of the distances among classes is different in the two
methods. The Bentix sets equal distances for the moderate (2.5-3.5) and good (3.5-4.5) classes,
while the Ambi renders a wider good class (1.2-3.3) compared to the moderate (3.3-4.3) and high
(0-1.2).
49
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0052.png
Comparison of AMBI-BENTIX EQRs
1.00
R
Comparison of AMBI-BENTIX EQRs
1.00
0.80
R
2
= 0.6126
2
= 0.6126
0.80
0.60
Ambi EQR
0.60
Ambi EQR
0.00
0.00
0.00
0.20
0.40
0.40
0.20
0.20
0.00
0.20
0.40
Bentix EQR
0.40
Bentix EQR
0.60
0.80
1.00
0.60
0.80
1.00
Figure 2.3.5:
Comparison among the AMBI and BENTIX EQRs and class boundary thresholds.
Box-and-Whisker Plot
1
0.8
0.6
0.4
0.2
0
AMBI EQR
Box-and-Whisker Plot
1
BENTIX EQR
0.8
0.6
0.4
0.2 G
0
M
EQS
P
B
1
0.8
0.6
0.4
0.2
0
BENTIX EQR
Box-and-Whisker Plot
Box-and-Whisker Plot
1
0.8
0.6
0.4
M
EQS
P
B
AMBI EQR
H
H 0.2 G
0
Figure 2.3.6:
Box-and-Whisker plot of EQS classification results using AMBI and BENTIX on Greek
M
data.
P
this
B
IC
For
H
G
M
P
B
H
G
comparison EQR=0 for Ambi=6 and Bentix=2.
EQS
EQS
m-AMBI - BENTIX
The Bentix was compared with the combined AMBI-Shannon Diversity-Species Richness
multivariate method. In this method a factorial analysis of the three combined indices, using high
and bad values for all indices determine the distance of each point-station from the two extremes
(high and bad). The threshold values for the EQS classification (EQR determination) were based
upon the REFCOND (non-intercalibrated values; Borja et al., 2007) and are referred below (Muxica
et al., 2006).
For this comparison two sets of habitats was decided to be treated separately. This is because the
indices of Shannon Diversity and species richness vary a lot depending on the type of habitat.
50
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0053.png
For this reason two main types of habitats were decided to be treated separately: the mixed and
vegetated sediments and the purely muddy sediments which is considered a special kind of naturally
stressed biotope where normally is inhabited by several tolerant species. Also the Bentix index
renders some modifications in the class boundaries for the muddy bottom habitats.
Various sets of “reference” or “highest values” were tested for the indices to run the factor analyis.
Finally it was decided to select the ones well above the maximum values encountered in real data.
However there is some speculation to use these highest values as reference values in general, as
these indices (diversity and species richness) depend much on specific conditions of the substrate,
sampling size and methodology, seasonal cycles etc.
In the comparison of Bentix and m-Ambi EQRs the lowest EQR=0 was equated
with AMBI=6 and Bentix=2 because of the lack of numeric values in the area 0-7 for Ambi and 0-2
for Bentix.
a) mixed sediments
Highest values for H (=6) and S (=120) were selected above the maximum values from real data
(reference sites) and refer to the standard unit of sampling surface (0.1 m
2
) and the mixed sediments
habitat. The variability explained by the data is high (69 %) and the agreement in the EQS
assessment is moderate (kappa value-0.52), but in the limit with very good (Figure 2.3.7).
1.0
0.9
0.8
0.7
EQR-Multivariate
0.6
0.5
0.4
0.3
0.2
0.1
0.0
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
EQR-Bentix
y = 0.9751x + 0.0959
R
2
= 0.6928
Figure 2.3.7:
Correlation among m-AMBI and BENTIX methods’ EQRs in the mixed
sediments habitat.
High and Bad values of the indices selected
for the factor analysis in the mixed sediments
analysis
HIGH
H
S
AMBI
6
120
0
BAD
0
0
6
Status Class Boundary
REFCOND (non intercalibrated values)
EQS
B/P
P/M
M/G
G/H
EQR
0.2
0.41
0.62
0.83
51
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0054.png
b) purely muddy bottoms
In this type of habitat are considered those stations sampled in muddy bottoms with silt and clay or
mud content of 85 % and over (according to Folk, 1954 classification scheme).
0.8
0.7
0.6
EQR-multivariate
0.5
0.4
0.3
0.2
0.1
0.0
0.0
0.1
0.2
0.3
0.4
EQR-Bentix
0.5
0.6
0.7
0.8
R
2
= 0.3778
Figure 2.3.8:
Correlation among m-AMBI and BENTIX methods’ EQRs in the muddy sediments habitat.
High and Bad values of the indices selected
for the factor analysis in the muddy sediments
analysis
HIGH
H
S
AMBI
5
40
0
BAD
0
0
6
Status Class Boundary
REFCOND (non intercalibrated values)
EQS
B/P
P/M
M/G
G/H
EQR
0.2
0.41
0.62
0.83
The agreement of m-AMBI and Bentix in this kind of habitat type is almost perfect (kappa
value=0.98). The explained variability is moderate (40 %) because of the high aggregation of data
in the moderate class. However, again care should be taken in the selection of high values for the
diversity and species richness, which could only be considered as a highest threshold for reference
conditions setting. Figure 9 shows an overall comparison of EcoQ classification results for all Greek
data and separately for muds where different reference values for m-Ambi are used. The agreement
between Bentix and m-Ambi is good, especially in the muddy habitats.
Figure 2.3.10 shows the comparison among indices in the Cyprus data. Ambi tends towards higher
classes, m-Ambi towards lower classes and the agreement is better among Bentix and m-Ambi.
52
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0055.png
100
90
80
70
60
50
100
40
90
30
80
20
70
10
60
0
50
40
30
20
10
0
H
G
Ambi
M
m-ambi
P
Bentix
B
Greek IC data comparison of indices
100
90
80
70
60
50
100
40
90
30
80
20
70
10
60
0
50
40
30
20
10
0
H
G
Ambi
M
m-ambi
P
Bentix
B
Habitat 2 (muds)
H
G
Ambi
M
m-ambi
P
Bentix
B
H
G
Ambi
M
m-ambi
P
Bentix
B
Figure 2.3.9:
Relative frequency distribution of EcoQ status of all Greek data and separately for muddy habitat
according to the various indices tested.
100
90
80
70
60
50
40
100
30
90
20
80
10
70
0
60
50
40
30
20
10
0
Figure 2.3.10:
Relative frequency distribution of EcoQ
status of Cyprus data.
H
G
Ambi
M
m-ambi
P
Bentix
B
FRANCE
The results here presented have to be considered only as
Data used for the
Bentix
intercalibration exercise was obtained using an identical methodology at all sampling stations:
5 samples per station
use of Van Venn or Smith Mac Intyre bucket
sample size 0.1 m²
sample depth between 20 and 35 metres (according to depth of Posidonia meadow lower limit).
The exercise was mainly based on datasets collected in Languedoc Roussillon by the ARAGO
laboratory, situated in Banyuls. Corsican data was supplied by the STARESO station. The company
CREOCEAN provided data acquired during numerous impact studies in Languedoc Roussillon.
Data collected by CREOCEAN at reference sites (Languedoc Roussillon and Provence-Alpes-Côte
d’Azur) during the 2005 cruise was also used (Fig. 2.3.11).
Indices tested in the framework of the intercalibration exercise
Each dataset was processed using H’, AMBI, BENTHIX and TI (Trophic Index) index formulas as
shown below:
A- Shannon-Weaver (H’) Index
The Shannon-Weaver index is a diversity index taking into account both the number of species in a
sample as well as their relative abundance, and is hence an efficient tool for assessing the ecological
balance of ecosystem populations.
53
H
G
M
an exercise.
Ambi
m-ambi
P
B
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0056.png
Ni = number of individus belonging to the species i; N = total number of individuals per m²
A maximum Shannon index (H’max) is achieved when all species are found in equal numbers. A
zero trend indicates that the population is dominated by a single species.
B- AZTI Marine Biotic Index (AMBI) & Biotic Index (BI)
The AMBI (AZTI Marine Biotic Index), also referred to as the benthic coefficient (BC), is based
on ecological successions. The 5 ecological groups (GI, GII, GIII, GIV and GV) are defined
according to species polluosensitivity. Group I comprises species which are the most sensitive to
hypertrophication; group V comprises opportunistic species inhabiting reduced sediment. The
reliability of this index also depends on the extent of sampling: if data is too limited, the same
average value will be achieved but with a high standard deviation. It has the advantage of being
applicable to all environments, as it is based on a predefined list of species.
%GI, %GII, %GIII, %GIV, %GV = relative abundance of various trophic groups versus total
number.
AMBI values range from 0 to 6. A 0 value indicates an unpolluted environment, whereas a 6 value
indicates major pollution and an azoïc environment.
1
3
2
1. Gulf of Lions : 214 stations
2. Reference sites : 7 stations
3. Corsica : 15 stations
Figure 2.3.11:
French Mediterranean coastal regions sampled during the intercalibration
exercise on soft-substrate benthic populations.
54
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0057.png
C- Biological Quality Index (BQI)
The BQI index is used to evaluate the composition of biological communities and hence pinpoint
environmental disturbances. BQI values are only valid for the target habitat and the maximum value
measured in that particular habitat. Moreover, the reliability of this index depends largely on the
extent of sampling. Extensive data is therefore needed to calculate the ES500,05
19
of each species
and perform a reliable analysis.
Ni = number of individus belonging to the species i; N = total number of individuals per m²; S:
Number of species
The BQI index is mainly used to detect physical disturbances to the environment.
D- Trophic Index (TI)
This index assesses the feeding habits of species present in a sample on the basis of the ecological
succession principle. Species are classified into 4 trophic groups (1: suspension feeders, 2: detritus
feeders, 3: deposit feeders, 4: anaerobic substrate species).
n1, n2, n3, n4 = number of species from trophic groups 1, 2, 3 and 4
N = total number of species
Index values range from 0 to 100. Values above 60 indicate a normal population, unaffected by
sediment enrichment with organic matter. Values between 30 and 60 indicate a population
imbalance , slightly affected by sediment enrichment with organic matter. Values below 30 highlight
a disturbed population impacted by sediment deterioration due to excessive enrichment with organic
matter. The Trophic Index is therefore used to pinpoint environmental disturbances of organic
origin.
Using the values calculated at each station with the Shannon, AMBI, BQI and Trophic indices,
we can establish ecological status class boundaries per index, as defined in the Water Framework
Directive (“High”, “Good”, “Moderate”, “Poor” and “Bad”) (Tab. 2.3.3).
Results of the comparison
LANGUEDOC – ROUSSILLON REGION
Data from 214 stations was analyzed. Thanks to in-depth knowledge of the sampling sites,
population counts and surveys conducted at each station, we were able to establish that 170 (79 %)
were undeteriorated.
19
ES50:
represents the theoretical number of species obtained if this sample contained only 50 individuals. On the basis
of this result, we can establish an ES50 distribution curve for this species according to the associated theoretical specific
number. The first 5 % of this distribution are referred to as
ES50
0.05
.
55
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0058.png
Table 2.3.3:
EcoQ values for the Shannon, AMBI, BQI and Trophic indices.
EcoQ
High
Good
Moderate
Poor
Bad
H’
H’ > 4
3 < H’ ≤ 4
2 < H’ ≤ 3
1 < H’ ≤ 2
H’ < 1
AMBI
AMBI ≤ 1.2
1.2 < AMBI ≤ 3.3
3.3 < AMBI ≤ 4.3
4.3 < AMBI ≤ 5.5
5.5 < AMBI ≤ 6
BQI
Depth. < 20m
BQI > 18.8
14.1 < BQI ≤ 18.8
9.4 < BQI ≤ 14.1
4.7 < BQI ≤ 9.4
BQI ≤ 4.7
Depth. > 20m
BQI > 26.4
19.8 < BQI ≤ 26.4
13.2 < BQI ≤ 19.8
6.6 < BQI ≤ 13.2
BQI ≤ 6.6
IT > 80
60 < IT ≤ 80
50 < IT ≤ 60
30 < IT ≤ 50
IT ≤ 30
The Shannon Index (H’) enabled an efficient discrimination of deteriorated stations, 89 % of which
obtained a “Moderate” “Poor” or “Bad” score. Moreover, the majority of undeteriorated stations
(86 %) achieved a good score.
The AMBI index did not enable discrimination and all stations were classified as “undeteriorated”.
We noted a drop in the index value at higher depths.
The BQI did not provide satisfactory discrimination between the various stations either. Although
all deteriorated stations scored badly, so did half (46 %) of the undeteriorated stations.
The Trophic Index (TI) failed to identify 16 % of the deteriorated stations. The index value dropped
at higher depths and 50 % of undeteriorated stations obtained a low score.
H’
degraded stations have bad indices.
Most not degraded stations have
good indices
AMBI
All the stations are considered
as good
Indices decreased with depth
BQI
degraded stations have bad indices.
Many not degraded stations have
bad indices
IT
None degraded stations have
bad indices.
Indices decreased with depth
Figure 2.3.12:
Comparison of Languedoc-Roussillon site classifications using the AMBI, Shannon-Weaver, BQI and
Trophic indices.
56
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0059.png
Out of the 214 stations sampled in Languedoc Roussillon, 79 % were defined as having a “Good” or
“High” status. Application of the various indices to the Languedoc Roussillon stations showed that
the H’ index provided the best characterization of deteriorated and undeteriorated sites (Tab. 2.3.3)
Table 2.3.3:
Comparison of scores obtained using each index versus actual impacts evaluated at Languedoc-Roussillon
sites.
Eco Q
Biotic Indices
Actual Impact
Shannon (H’)
BQI
AMBI
TI
“High” or “Good”
79 %
69 %
26 %
100 %
57 %
“Moderate”, “Poor” or “Bad”
21 %
31 %
64 %
0%
43 %
CORSICA
The dataset included 15 stations sampled for the last 5 years by STARESO.
Generally speaking, the sediment particles found around Corsica are fairly large, hence incurring
low species densities. This particular characteristic impedes index reliability, in particular that of the
Shannon (H)’ index which indicates a poor ecological status even in undeteriorated environments.
We could circumvent this difficulty by calculating the EQRs of each reference site using major
community-type sampling (Fig. 2.3.13).
Figure 2.3.13:
Application of H’, AMBI and TI indices for the classification of Corsican
coastal sites.
57
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0060.png
Reference sites
7 stations corresponding to reference sites in Languedoc Roussillon (2 stations) and Provence-
Alpes-Côte d’Azur (5 stations) were sampled during the 2005 cruise.
Overall, the sampled stations showed balanced populations, with relatively to very high biomasses
and densities. Data processed using the H’ and BQI indices gave “High” results at all stations; the
AMBI index gave “Good” to “High” results and the Trophic Index gave “Moderate” to “Good”
results (Fig. 2.3.14). Using BQI calculations, just 27 % of cruise target species were identified using
the data collected from over 200 stations. If results obtained with the H’ and BQI indices prove
similar, the Shannon index would hence appear to be more efficient than the BQI
Analysis of the various results demonstrated that no single index is capable of characterizing
environmental disturbances as a whole. None of the indices are exhaustive, and each has its pros and
cons. Some are harder to implement than others. However, comparatively speaking, the Shannon index
proved more efficient at highlighting environmental disturbances along French Mediterranean coasts.
The Shannon Index could usefully be associated with the AMBI and usual biological parameters
(specific numbers, density and biomass).
The BQI requires relatively complicated calculations and an extensive database. The database used
to obtain the BQI index (data from over 200 stations) resulted in the identification of just 157 species
Figure 2.3.14:
Classification of
reference sites using the H’, BQI,
AMBI and TI indices.
58
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0061.png
out of the 586 identified at the reference sites during the cruise. In addition, the BQI and H’ (Fig.
2.3.15) are strongly correlated and are both structural indices, although H’ is very easy to calculate.
Although the TI and AMBI are generally associated (both are organic pollution indicators), they
actually appear to differ in several ways due to the AMBI’s weaker range versus the TI, which is
more penalizing that the AMBI.
Moreover, the results highlight a close relationship between particle size and population composition,,
due to the fact that particle size defines the types of habitat available in the environment.
In view of the heterogeneous character of the sediment particles found in French Mediterranean
waters (Languedoc-Roussillon, Provence, Corsica), the experts who took part in this intercalibration
exercise recommended using stratified sampling per major community type (4 particle types) and per
reference station (undisturbed zone). This would give us a reference notwithstanding the type of sample
collected during monitoring and would also offer greater sampling flexibility (samples taken from the
predominant substrate of monitored water masses), as, to all intents and purposes, it is very difficult to
obtain samples with similar particle characteristics at all French Mediterranean coastal stations.
ITALY
Results presented here have to be considered only as an exercise which is based on the data
available till now for the Italian coasts. Data were collected in 50 stations along 11 coastal Regions
within the National Monitoring programme (2001-2006) of the Italian Ministry of Environment
(still under evaluation).
Major reasons of concerns about this data set and the validation of the ecological classification
obtained by their use are related to the following points:
– a high percentage of individuals not identify to species level is present in some samples;
– no chemical-physical parameters (oxygen, organic matter, nutrients, sediment grain size)
are available for most of the stations. The chemical-physical parameters are fundamental
to interpret the results and to understand the reliability of the quality indices tested.
Relation between H' and BQI
40
35
30
25
BQI
20
15
10
5
0
0
1
2
3
H'
4
5
6
y = 2.2693e
0.4857x
R = 0.7386
2
Figure 2.3.15:
Correlations between the Shannon (H’) and BQI indices.
59
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0062.png
The a-priori classification of the intercalibration sites is in some case questionable.
No real reference sites exist.
Having this in mind, the Multimetric AMBI method (M-AMBI) (Muxika et al., 2007) with several
reference conditions (RC) and boundaries, adapted to Italian data, were applied and tested. Three
different approaches that give similar reference values for AMBI, richness (R) and diversity (H),
and that provide the most reliable ecological status classification, are described below.
Reference conditions
The WFD identifies four options for determining reference conditions: 1) existing undisturbed sites,
2) historical data and information, 3) models, and 4) expert judgement. Option 3 and 4 were selected
to define different reference values.
1. RC1
In this case, samples with individuals in ecological group I >70 % (sensu Borja et al., 2000) were
selected. Then, among those only samples with R> 20, H> 2 were considered as reference. The high
percentage of sensitive species is here considered more important than high R and H. The median
value of AMBI, R and H in these samples, is used as reference . The mean or median value from the
distribution of reference site values are considered the most robust values to be used as reference
and relatively few data are needed for sufficient confidence in RC (Wallin et al., 2003).
RC1
AMBI
0.5
R
30
H
4
2. RC2
In this second attempt, the reference values were calculating on all the data set. For this reason a
higher percentile than the median (the 90th percentile) was used for R and H and a lower one (the
10th) for AMBI.
RC2
AMBI
0.4
R
42
H
4
3. RC3
In this case the Intercalibration sites, which were a-priori classified in High/Good ES and the
stations indicated by Regional Agencies as “control sites”, were considered as reference sites . In
order to establish the reference values, not being sure of the a-priori classification of these sites, we
have considered a higher (lower for AMBI) percentile than the median value: the 75th percentile for
R and H, and to the 25th for AMBI.
RC3
AMBI
0.8
R
33
H
3.3
Boundary Setting
Different boundaries were set on the basis of the different reference conditions considered (1, 2, 3).
1. The 10th percentile of EQR values obtained for the reference samples was selected as the “upper
anchor” (Wallin et al., 2003) and the class boundary between high and good. The width of the four
remaining classes was evenly spaced over the remaining interval. This has resulted in the following
class boundaries.
60
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0063.png
1
High status
Good status
Moderate status
Poor status
Bad status
≥ 0.81
≥ 0.61
≥ 0.41
≥ 0.20
<0.20
2. The 75th percentile of EQR values (obtained by applying M-AMBI in all the samples) was
selected as the “upper anchor” and the class boundary between high and good. The width of the four
remaining classes was evenly spaced over the remaining interval. This has resulted in the following
class boundaries.
High status
Good status
Moderate status
Poor status
Bad status
≥ 0.80
≥ 0.60
≥ 0.40
≥ 0.20
< 0.20
2
3. The 30th percentile of AMBI, R and H calculated in the reference sites has been selected as the
“upper anchor” and the class boundary between high and good. The width of the four remaining
classes has been evenly spaced over the remaining interval. This has resulted in the following class
boundaries.
High status
Good status
Moderate status
Poor status
Bad status
≥ 0.78
≥ 0.59
≥ 0.39
≥ 0.20
< 0.20
3
Results of the comparison
M-AMBI options 1, 2, 3
Results obtained by the application of M-AMBI with the three different RC and boundaries are
reported in the figure 2.3.15. Results of K-analyses are shown in table 2.3.4. M-AMBI applied
following option 2, gives the most severe classification of the Italian sites with 19 % of the samples
classified in moderate status; with option 3, only the 7 % of the sample results in moderate status.
Despite these differences, the K-analyses shows very good agreement between the three methods.
Table 2.3.4:
K values and agreement among Ecological
status classification obtained by applying M-AMBI with
different RC and boundaries (options 1, 2, 3).
1
1
2
3
0.82
Very good
0.76
Very good
0.63
Good
2
0.82
Very good
3
0.76
Very good
0.63
Good
61
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0064.png
70
1
60
50
Percentage %
40
30
20
10
0
2
3
H
G
M
ES
P
B
Figure 2.3.15:
Percentage of samples classified in High, Good, Moderate, Poor, and Bad
Ecological Status by applying M-AMBI with different RC and boundaries (options 1, 2, 3).
M-AMBI options 1, 2, 3/Bentix
Results obtained by applying M-AMBI with different RC and boundaries were compared with those
obtained by Bentix Index; the different boundaries are reported in table 2.3.5. The Bentix EQR
values for Italian data were calculated by Mika Simùboura.
Table 2.3.5:
Boundaries values for Bentix and for M-AMBI (options 1, 2, 3).
Boundaries
H/G
G/M
M/P
P/B
Bentix
0.75
0.58
0.42
0.33
M-AMBI 1
0.81
0.61
0.41
0.20
M-AMBI 2
0.80
0.60
0.40
0.20
M-AMBI 3
0.78
0.59
0.39
0.20
Figure 2.3.16 shows, as an example, the relationship between Bentix EQR and M-AMBI 1 EQR
(the M-AMBI option that gives results nearest to those obtained by Bentix). Only a low correlation
exists; the major difference is about the H/G classification, the 32 % of the sample classified in high
ecological status by Bentix result in Good ES for M-AMBI (Fig. 2.3.17).
Results of Kappa analysis (table 3) confirms low agreement between the two different Ecological
Status classifications.
62
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0065.png
1.00
0.80
0.60
1.00
0.40
0.80
0.20
0.60
0.00
0.400.00
0.20
0.00
0.00
Bentix
Bentix
0.20
0.40
0.60
0.80
1.00
M-AMBI 1
r=0.221 p<0.01 n=169
0.20
0.40
0.60
0.80
1.00
M-AMBI 1
r=0.221 p<0.01 n=169
Figure 2.3.16:
Correlation between bentix and M-AMBI 1. Vertical and horizontal lines
indicate the boundaries.
70
Bentix
60
50
Percentage %
70
40
60
30
50
20
40
10
30
0
20
H
G
M
P
B
ES
Bentix
M-AMBI 1
M-AMBI 1
Figure 2.3.17:
Percentage of samples classified in High, Good, Moderate, Poor, and Bad
10
Ecological Status by applying Bentix and M-AMBI 1.
0
H
G
M
P
B
ES
K values and agreement among Ecological status classification were obtained by applying M-AMBI
with different RC and boundaries (options 1, 2, 3) and Bentix index
1
Bentix
0.40
Low
2
0.30
Low
3
0.38
Low
SLOVENIA
Reference conditions
Slovenian reference conditions were set on expert judgement, since in the Slovenian sea and whole
Gulf of Trieste there aren’t any proper reference sites. Two sampling sites from the area with
minimal known human impact (SD_VT2_P1 and SD_VT2_P2) were taken into consideration when
setting the reference conditions. They were also compared with sites in the Italian part of the Gulf
of Trieste and found as suitable. Averaging values from those two stations and adding 15 % on them
have set the reference conditions.
63
Percentage %
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0066.png
SLOVENIA
Metric
AMBI
Reference condition
1,3/6
5,8
110
Shannon-Wiener (H’)
Richness(S)
Setting of Boundaries
Boundaries were determined according to expert judgement. This was the only possible method,
since the Slovenian data were not equally distributed over the whole range of EQR values and due
to the fact that more than one metrics has been used to calculate EQR.
SLOVENIAN BOUNDARIES
H/G
0,83
G/M
0,62
M/P
0,41
P/B
0,20
Boundary between High and Good ecological class was set according to the reference conditions and
natural variability. Natural variability defines width of High class. In our case natural variability is
presumed to be around 20 %, meaning that upper and lower limit of High class differ for 20 %. H/G
boundary (lower limit) was calculated by taking median from EQR values of the two stations used
in calculating reference conditions (SD_VT2_P1 and SD_VT2_P2) and subtracting additional 5 %
on this value. Subtracting of 5 % was done because median of our actual data lays 15 % from upper
limit, so to get the lower limit, which differs from upper for 20 % this subtraction must be done.
Other boundaries were set equidistantly from the H/G boundary (0,83): between G/M on 0,62,
between M/P on 0,41 and between P/B on 0,20.
Results of the comparison
M-AMBI/BENTIX
M-AMBI
1
0,9
0,8
0,7
M-AMBI
0,6
0,5
0,4
0,3
0,2
0,1
0
0
0,2
0,4
Bentix-EQR
0,6
0,8
1
y = 0,9142x + 0,1154
R = 0,7537
2
Figure 2.3.18:
Correlation among EQR_M-AMBI (SI) and EQR_BENTIX (GR, CY) using
Slovenian benthic invertebrate data.
64
1,00
0,90
0,80
0,70
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0067.png
M-AMBI
0,6
0,5
0,4
0,3
=
Slovenia (M-AMBI),
Table 2.3.5:
Agreement for ecological status analysis between the methods used
y
in
0,9142x + 0,1154
2
Greece/Cyprus (Bentix).
0,2
R = 0,7537
0,1
M-AMBI (SI)
0
0,2
BENTIX (GR, CY)
0,6
Bentix-EQR
M-AMBI (SI)
0
1,00
0,83
BENTIX (GR, CY)
0,4
1,00
0,8
1
1,00
0,90
0,80
0,70
EQR_M-AMBI (SI)
0,60
0,50
0,40
0,30
0,20
0,10
0,00
H i gh
G o od
EQS_M-AMBI (SI)
1.00
0.90
0.80
0.70
EQR_BENTIX (Gr,Cy)
0.60
0.50
0.40
0.30
0.20
0.10
0.00
High
Good
EQS
Bad
B ad
Figure 2.3.19 . Boxplot-and-Whisker plot (75 percentile) of EQS classification results using
M-AMBI (SI) and BENTIX (GR, CY) on Slovenian IC data.
65
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0068.png
M-AMBI/MEDOCC
1
0,9
0,8
EQR_M-AMBI (SI)
0,7
0,6
0,5
0,4
0,3
0,2
0,1
0
0,0
0,1
0,2
0,3
0,4
EQR-MEDOCC
0,5
0,6
0,7
0,8
y = 0,8819x + 0,2431
R = 0,583
2
Figure 2.3.20:
Correlation among EQR_M-AMBI (SI) and MEDOCC using Slovenian benthic
invertebrate data.
Table 2.3.6:
Agreement for ecological status analysis between the methods used in Slovenia (M-AMBI) and Spain
(MEDOCC) (Slovenian benthic invertebrate data).
M-AMBI (SI)
MEDOCC
1,00
0,90
0,80
0,70
EQR_M-AMBI (SI)
0,60
0,50
0,40
0,30
0,20
0,10
0,00
High
Good
EQS_M-AMBI (SI)
0,8
0,7
0,6
EQR_MEDOCC
0,5
0,4
0,3
0,2
0,1
0
G
M
EQS_MEDOCC
B
Bad
M-AMBI (SI)
1,00
0,75
MEDOCC
1,00
Figure 2.3.21:
Boxplot-and-
Whisker plot (75 percentile) of EQS
classification results using M-AMBI
(SI) and MEDOCC on Slovenian IC
data.
66
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
SPAIN (Catalonia and Balearic Islands)
The MEDOCC index is an adaptation to the western Mediterranean area of the AMBI index
developed for the Atlantic coast (Borja et al.,2000) based on sensitivity/tolerance of the species.
Species sensitivity
Species from the dataset have been assigned to one ecological group (EG) in relation to the
sensitivity/tolerance of the species mainly to organic enrichment.
Ecological groups have been divided into four categories: sensitive, indifferent, tolerant and
opportunistic species, based on available literature, AMBI list (http://www.azti.es), Simboura &
Zenetos, (2002); http://www.hcmr.gr/english_site/services/env_aspects/bentix.html), Pinedo (1998)
and expert judgment.
MEDOCC index
The MEDOCC index is able to detect organic enrichment following communities succession
described by Pearson & Rosenberg (1978) and posterior authors. The main differences with the
original AMBI method proposed by Borja et al. (2000) are the following:
Change in the categories of the ecological groups in some species. Although Borja has defended
the idea of assigning species to the same ecological group uniquely according to the species’
biogeographical latitudinal range, it can be considered that species react differently depending on
inter-species interaction and environmental conditions (Rosenberg et al., 2004; Dauvin, 2007). In
other way, assigning a species to different ecological groups according to the region could become
subjective related to the experience and expertise of the scientist (Dauvin, 2007).
During the intercalibration exercise it was attempted to get an unanimous species scores assignation
but we did not arrive to an agreement for some species; thus finally we consider the necessity of
changing ecological groups in some species that respond differently to organic enrichment in the
Mediterranean and the Atlantic. Carvalho et al. (2006) also suggest different EGs to those assigned
in AMBI list for some species. In the following table we include only those species that change EGs
comparing with AMBI list and species with no ecological group in AMBI list but assigned to one in
MEDOCC index.
67
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0070.png
Table 2.3.7:
Species changing the ecological group and species with no ecological group in AMBI list.
Species
Acanthocardia tubercula
Ampelisca typica
Ampharete sp.
MEDOCC
2
2
2
2
2
3
2
2
2
2
1
1
4
2
3
2
1
1
2
3
1
3
3
4
2
1
1
3
AMBI
1
1
1
1
2
1
3
1
1
2
-
-
-
Species
Gouldia minima
Loripes lacteus
MEDOCC
2
2
3
3
2
3
3
3
2
2
2
2
2
2
2
2
2
3
1
1
2
2
1
1
1
2
2
AMBI
1
1
1
2
1
2
2
3
1
1
1
4
1
3
1
4
2
2
2
1
3
3
2
2
3
1
-
Iphinoe trispinosa
Lumbrineris latreilli
Magelona mirabilis
Marphysa sp.
Mastobranchus sp.
Pariambus typicus
AMPHARETIDAE
Amphiura chiajei
Anthura gracilis
Amphipholis squamata
Apseudes latreillei
Aricidea suecica
Caecum sp.
Aricidea catherinae
Bodotria pulchella
Capitellethus sp.
Cardium sp.
Corbula gibba
Micronephthys maryae
Parvicardium ovale
Parvicardium exiguum
Parvicardium juvenil
Pectinaria koreni
Phtisica marina
3
4
3
2
2
1
1
2
2
2
3
2
2
-
-
Corophium runcicorne
Cyclichna cylindracea
Chone longiseta
Diastilis rugosa
Eurydice sp.
Phyllodoce mucosa
Prionospio caspersi
Retusa truncatula
Scolaricia typica
Plagiocardium papillosum
Pseudolirius kroyeri
Retusa umbilicata
Scolelepis squamata
Scoloplos armiger
Sigalion mathildae
Euclymene collaris
Exogone naidina
Exogone verugera
Fabriciola tonerella
Glans sp.
Galathowenia oculeata
Glycera oxycephala
Goniadella galaica
Sigalion squamosus
Tellina pulchella
Spiochaetopterus solitarius
Four ecological groups have been considered (instead of five in AMBI): sensitive (EGI), indifferent
(EGII), tolerant (EGIII) and opportunistic species (EGIV). The reduction of the five EGs to four is
proposed as a way to avoid errors in assignation of species to the first or second-order opportunists
EG. At the current level of knowledge, we are not able to distinguish between both order of response
to disturbance.
The calculation of the index is as following
MEDOCC=
[(0 x % EGI + 2 x %EGII + 4 x % EGIII +6 x % E GIV)]/100
Index response to the impact gradient
For the development of this methodology to asses the ecological status and to obtain the boundaries
required by the WFD we have analysed samples from fine to muddy fine sediments from Catalonia
and Balearic Islands. Although the communities’ composition in terms of ecological groups is
68
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0071.png
expected to respond to the extent of anthropogenic pressure (sensu lato), this method is developed to
detect mainly organic enrichment. Plotting the MEDOCC results from our data with organic matter
content in sediment, MEDOCC index shows a significant linear relationship with the pressure
gradient (Figure 2.3.22).
Reference conditions and class boundary setting
(Catalonia and Balearic Islands, SPAIN)
Reference conditions
One of the problems in deriving reference conditions arises from the absence of unimpacted areas
in some European regions. The WFD identifies four options for determinig reference conditions: 1)
existing undisturbed sites, 2) historical data and information, 3) models, and 4) expert judgement.
Option 3 and 4 have been selected to define reference sites for Catalonia and Balearic Islands,
thus selecting the best situation (sample) of our dataset where the most species belong to EGI
(sensitive species) and EGII (indiferent species). After choosing this “best of all” situation from
our dataset, we have excluded all tolerant (EGIII) and opportunistic species (EGIV) and created a
new theoretical situation where the fauna is composed of only sensitive (EGI: 90 %) and indifferent
species (EGII: 10 %).
Boundary Setting
Class boundary values are obtained by checking community composition of sensitive/tolerant
species over the MEDOCC index for the dataset (Catalonia and Balearic Islands dataset). Figure
2 shows the plotting of the percentage of all EGs considered along the gradient of increasing
MEDOCC index in Catalonia and Balearic Islands. The sequence of quality classes and class
boundaries are interpreted in terms of shifts of ecological group percentages.
The relation between MEDOCC values and the ecological status classification of the WFD is
based on the ecological theory describing the communities succession in an increasing disturbance
gradient followed by Glémarec & Hily (1981), Hily (1984), Majeed (1987), Grall & Glémarec
(1997), and Borja et al. (2000).
The five levels of ecological status in the WFD are described as below:
High status: “The level of diversity and abundance of invertebrate taxa is within the range
normally associated with undisturbed conditions. All the disturbance-sensitive taxa associated
r=0.362 p<0.001
6.00
5.00
4.00
MEDOCC
3.00
2.00
1.00
0.00
0.00
0.50
1.00
1.50
organic matter %
2.00
2.50
3.00
Figure 2.3.22:
Pearson correlation of MEDOCC index in Catalonia and Balearic Islands with organic matter ( %) in
sediments.
69
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0072.png
with undisturbed conditions are present”. This could be associated to normal or unpolluted benthic
community, dominated by the Ecological Group I. We propose the border between high and good
ecological status for MEDOCC equal to 1.6; that is, sensitive ecological group (EGI) accounting for
more than 40 % of total abundances.
Good status: “The level of diversity and abundance of invertebrate taxa is slightly outside the
range associated with the type-specific conditions. Most of the sensitive taxa of the type-specific
communities are present”. This could be associated to unbalanced or slightly polluted benthic
community, dominated by the Ecological Group III. The border between good and moderate
ecological status is 3.2 for MEDOCC index; in this case tolerant ecological group (EGIII) accounts
for 20-50 %, but sensitive taxa (EGI) are also present (10-40 %).
Moderate status: “The level of diversity and abundance of invertebrate taxa is moderately outside
the range associated with the type-specific conditions. Taxa indicative of pollution are present”. This
could be associated to a transitional benthic community, dominated by the Ecological Group III and
IV. We propose that tolerant ecological group (EGIII) accounting for nearly 50 % and opportunistic
taxa accounting less than 45 %. The border value between moderate and poor ES is 4.77 for
MEDOCC.
Poor status: “… the relevant biological communities deviate substantially from those normally
associated with the surface water body type under undisturbed conditions”. This could be associated
to polluted areas, dominated by the Ecological Group IV but still with some presence of species
belonging to the ESIII. We propose that the opportunistic ecological group (EGIV) accounts for
more than 45 %. The border between poor and bad ecological status is 5.5 for MEDOCC index.
Bad status: “… In which large portions of the relevant biological communities normally associated
with the surface water body type under undisturbed conditions are absent”. It includes areas heavily
polluted and azoic sediments. We propose that opportunistic ecological group (EGIV) accounting
for more than 80 %.
Thus, the ranges of the MEDOCC index and Boundaries obtained for Catalonia and Balearic Islands
data following the ecological theory are showed in Table 2.3.8. Boundaries are obtained rescaling
MEDOCC values (between 0-6) from 1 to 0.
Figure 2.3.23:
MEDOCC values relating to the percentages of the ecological groups for
the Catalonia and Balearic Islands dataset. Vertical lines show boundaries of the different
ecological status. H: High ecological status; G: good ES; M: moderate ES; P: poor ES; and
B: bad ES. I: Sensitive species; II: indifferent species; III: tolerant species; IV: opportunistic
species.
70
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0073.png
Table 2.3.8:
MEDOCC index and EQRs obtained for Catalonia and Balearic Islands.
Ecological Status
High
Good
Moderate
Poor
Bad
MEDOCC values (6-0)
(0<MEDOCC<1.6)
(1.6<MEDOCC<3.2)
(3.2<MEDOCC<4.77)
(4.77<MEDOCC<5.5)
(5.5<MEDOCC>6)
Boundaries (0-1)
0.73
0.47
0.20
0.08
0
It is important to note that the assessment of boundaries is an intrinsic part on the development of
the MEDOCC methodology. This method is aimed at describing and suggesting the way to establish
boundaries for ecological status classes based on the composition of the communities in terms of
ecological groups, instead of proposing fixed values for the boundaries between each ecological
status. In short, the important part of the method described above is the possibility of creating the
relative abundance curves of each EG along a increasing gradient of MEDOCC index adapted to
a particular dataset, and delimitate boundaries based on the dominance of EGs, as stated above.
In fact for the macroinvertebrates communities from Catalonia and Balearic Islands the distribution
of ecological groups over the MEDOCC index (Figure 2.3.22) can not be compared with those
obtained for Atlantic waters over the AMBI values (Borja et al., 2000).
The major differences are due to the distribution of species belonging to ecological group III
(tolerant species). In the case of Catalonia and Balearic Island tolerant species are more associated
to the presence of opportunistic species (EGIV) whereas in Atlantic waters these group of species
are more associated to the presence of sensitive taxa (EGI). That means that if we take into account
the boundaries considered for Atlantic communities, the “good” ecological status would be
dominated by EGIII and EGIV. Those conditions do not agree with the levels of ecological status
described in the WFD.
The same pattern of distribution of the ecological groups has been observed for other dataset in
the Mediterranean ecoregion (Italia, Slovenia) during the Intercalibration exercise. So, due to
differences in benthic assemblages in terms of ecological groups we consider that the boundaries
proposed by AMBI index are not applicable to the Catalonia and Balearic Island communities and
probably, to other Mediterranean areas.
The EQR values (Table 2) were calculated by rescaling MEDOCC values (between 0-6) from
1 (high) to 0 (bad) and taking into account the reference condition situation. That is, with a
reference condition corresponding to a situation where the faunistic assemblage is composed of 90
% of sensitive (EGI) and 10 % of indifferent species (EGII) we obtain a MEDOOC value equal
to 0.2. This means that after considering the reference condition, the MEDOCC values for each
station have to be corrected by this value. Therefore, after considering the reference condition the
MEDOCC values ranges from 0.2 to 6. (Example: MEDOCC value=3.2; MEDOCC corrected by
Reference conditions, 3.2-0.2=3.0; EQR (from 0 as High ES to 1 as Bad ES)=3.0/5.8=0.52; EQR
(from 0 as Bad ES to 1 as High ES, following WFD)=0.48; Ecological Status=Good).
Results of the comparison
We have run BENTIX and M-AMBI (combination of three indexes: AMBI-Shannon Diversity-
Species Richness) in our data applying boundaries established by each method. We then compare
the ecological status obtained with MEDOCC with those obtained from the other methods allowing
to us to know the percentage of agreement.
71
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0074.png
BENTIX/MEDOCC
The MEDOCC was compared with BENTIX. The EQR values for our data applying BENTIX index
were obtained by Mika Simboura. The BENTIX limits between ES were selected from the boundary
setting protocol elaborated by Greece (Table 2.3.9). Figure 2.3.24 shows a linear relationship
statistically significant (r=0.714; p<0.001) between both EQRs values. The lines in the plot define
the ranges of each ecological status by the boundary limits of each method.
Table 2.3.9:
Ecological Status boundaries for MEDOCC and BENTIX methods.
EQS
H-G
G-M
M-P
P-B
Boundary EQR_MEDOCC
0.73
0.47
0.20
0.08
Boundary EQR_BENTIx
0.75
0.58
0.42
0.33
Figure 2.3.24 shows the relationship between MEDOCC and BENTIX. The 52 % of the samples
show an agreement in the Ecological Status (shady areas in the Figure 3). The most significant
differences are related to the H/G Ecological status. The 33 % of the stations with “high” ES by one
method obtains “good” ES for the other. Anyway, the number of stations in critical situation, thus is
“moderate”, “poor”, and “bad” ES for one of the methods and “high” or “good” for the other is low:
12.4 %. The main problem is that using the “H-G” boundary of BENTIX (Table 2.3.9, EQR= 0.75) in
our dataset we are considering that “high” ES accounts for more than 80 % of sensitive and indifferent
species. A community with this dominance of species does not agree with the definition of “high” ES
according BENTIX (see BENTIX description in this final report) which is taking into account more
than 60 % of sensitive and indifferent species to classify a site as “high” ES. When applying BENTIX
index in Catalonia and Balearic Islands, if we want to consider “high” ES following BENTIX
considerations, the EQR value should be lower than 0.75. Changing the EQR boundary for H-G ES
would improve the agreement. Figure 2.3.25 shows the comparison of Ecological Status classification
results for Catalonia and Balearic Islands with MEDOCC and BENTIX index.
Figure 2.3.24:
Comparison between BENTIX and MEDOCC EQRs with the class boundaries.
Shady areas show the space where the two methods have the same status classification.
72
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0075.png
70
MEDOCC
60
BENTIX
50
Percentage (%)
40
30
20
10
0
H
G
M
P
B
Figure 2.3.25:
Relative frequency distribution of Ecological Status (High, Good, Moderate, Poor, and Bad) of Catalonia
and Balearic Islands data with BENTIX and MEDOCC indexes.
M-AMBI/MEDOCC
The MEDOCC was compared with M-AMBI. M-AMBI uses a factorial analysis to determine
the distance of each station from the two extreme situations (reference conditions bad and high).
Various sets of reference conditions were tested for highest values. Finally we decided to select the
maximum values encountered in real data, as it was chosen in MEDOCC index. The highest values
were: AMBI= 0.14; R=100; H’=5.54. Bad reference conditions were: AMBI=6; R=0; H’=0.
The M-AMBI limits between ES were selected from Borja et al. (2007) (Table 2.3.10). The two
indices were compared showing a linear relationship statistically significant (r=0.517; p<0.001) in
Figure 2.3.26. The lines in the plot define the ranges of each ecological status by the boundary limits
of each method.
Figure 2.3.26 shows the relationship between MEDOCC and M-AMBI. The 34 % of the samples
show an agreement in the Ecological Status (shady areas in the Figure 2.3.26). The most significant
differences are related again to the H/G Ecological status. The 59 % of the stations have “high” ES
Table 2.3.10:
Ecological Status boundaries for MEDOCC and
M-AMBI methods.
EQS
H-G
G-M
M-P
P-B
Boundary EQR_MEDOCC
0.73
0.47
0.20
0.08
Boundary
EQR_M-AMBI
0.85
0.55
0.39
0.20
73
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0076.png
with one index but “good” ES for the other. Anyway, the number of stations in critical situation,
thus is is “moderate”, “poor”, and “bad” ES for one of the methods and “high” or “good” for the
other is low: 6.7 %. Figure 2.3.27 shows the comparison of Ecological Status classification results
for Catalonia and Balearic Islands with MEDOCC and M-AMBI index.
We believe that the slight relationship between both methods could be related mainly to the
assignation of the boundaries, which, as we have explained above, they have to be obtained from
each dataset depending on the evolution of the ecological groups considered. This disagreement is
understandable if we observe the boundary for “H-G” ES considered for M-AMBI in Table 2.3.10.
Selecting an elevated EQR value for H-G ES (=0.85) enhance the high number of stations classified
as “good” for M-AMBI in the Figure 5; this result is the main responsible of the disagreement.
Anyway, we also think that the inclusion in the multivariate method of the diversity index and
species richness worsen the results, as those indexes change depending on other environmental
variables and they are not necessarily related to the degree of habitat disturbance. As an example,
Figure 2.3.24:
Comparison between MAMBI and MEDOCC EQRs with the class boundaries.
Shady areas show the space where the two methods have the same status classification.
90
80
70
60
Percentage (%)
50
40
30
20
10
0
H
G
M
P
B
MEDOCC
M-AMBI
Figure 2.3.27:
Relative frequency
distribution of Ecological Status
of Catalonia and Balearic Islands
data with M-AMBI and MEDOCC
indexes.
74
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0077.png
the station with “bad” ES classification using both methods shows similar values in H’ and species
richness to that of other stations classified as “high” ES.
2.3.4 Results of the comparison
Each country has tested the different methods with their own data applying boundaries established
by each method. Moreover, they compared the ecological status obtained with the different methods
in order to know the percentage of agreement. Results are described, for each country.
2.3.5 Results of the harmonisation – Boundary EQR values
Member
State
Cyprus
France
Greece
Italy
Slovenia
Spain -
Catalunya
Method
Bentix
Multimetric approach (AMBI, Shannon Diversity,
BQI Trophic Index)
Bentix
AMBI with factor analysis
Bentix
MEDOCC index
MEDOCC index
Boundary
High/Good (EQR)
0.75
Boundary
Good/Moderate (EQR)
0.58
0.75
0.58
M-AMBI (AMBI, Shannon Diversity, S)
0,83
0.73
0.73
0,62
0.47
0.47
Spain-
Balearic is.
2.3.6 Open issues and need for further work
For all: need to consider different sub-regions (eastern, western Med), and to develop analysing
tools (multimetric indices) for hard bottom substrates.
Still some comparison/harmonization between methods should be done.
For Italy: Application of MEDOCC methods to Italian data, in collaboration with Spanish colleagues.
In the attempt of taking into account the variability of Italian coastal areas, a classification of the
benthic invertebrates stations using the same method used by the MED GIG phytoplankton group for
typologies
will be tested. Coastal sites are characterized according to different freshwater influence:
Type 1, coastal sites highly influenced by freshwater inputs; Type 2, coastal sites not directly affected
by freshwater inputs; Type 3, coastal sites not affected by freshwater inputs. Development of RC for
each of these typologies will be assessed.
For Spain: Future work has to be focussed in the harmonization of boundaries. Although a
comparison of the results of the ecological status with the different methodologies has been
performed, the boundaries obtained by each member state have not been harmonized to get a better
agreement. The base of this exercise must to be focused on the agreement around the definition of the
conditions of the different Ecological Status.
2 countries (France and Italy) are working on the methods to be used as national: either adopting
one method used in other countries or developing their own.
75
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0078.png
2.4 NE Atlantic GIG
2.4.1 Intercalibration approach
In the NE Atlantic seven basic intercalibration types have been agreed. These are shown in table
2.4.1 below:
Table 2.4.1:
NEA GIG Intercalibration Types.
New Type ID
Name
Salinity
Tidal range (m)
Depth (m)
Current
velocity
(knots)
Medium
(1 - 3)
Medium
(1 - 3)
Low
(< 1)
Low
(< 1)
Exposure
Mixing
Residence
time
Days
Days
CW –NEA1/26
CW – NEA3/4
Exposed or sheltered,
euhaline, shallow
Polyhaline, exposed or
moderately exposed
(Wadden Sea type)
Deep, low current,
sheltered
Polyhaline, microtidal,
sheltered, shallow
(Skagerrak inner arc
type)
Fjord with a shallow
sill at the mouth with
a very deep maximum
depth in the central
basin with poor
deepwater exchange.
Polyhaline, microtidal
exposed, deep
(Skaggerak outer arc
type)
Transitional waters
Fully saline
(> 30)
Polyhaline
(18 - 30)
Fully saline
(> 30)
Polyhaline
(18 - 30)
Mesotidal
(1 - 5)
Mesotidal
(1 - 5)
Mesotidal
(1 - 5)
Microtidal
(< 1)
Shallow
(< 30)
Shallow
(< 30)
Deep
(> 30)
Shallow
(< 30)
Exposed or
sheltered
Exposed or
moderately
exposed
Sheltered
Sheltered
Fully mixed
Fully mixed
CW – NEA7
CW – NEA8
Fully mixed
Partially
Stratified
Days
Days-
Weeks
CW – NEA9
Polyhaline
(18 - 30)
Microtidal
(< 1)
Deep
(> 30)
Low
(< 1)
Sheltered
Permanently
Stratified
Weeks
CW – NEA10
Polyhaline
(18 - 30)
Microtidal
(< 1)
Deep
(> 30)
Low
(< 1)
Exposed
Permanently
Stratified
Days
TW – NEA11
Oligo-
Euhaline
(0 - 30)
Mesotidal
(1 – 5 )
Shallow
(< 30)
Medium
Sheltered or
moderately
Exposed
Partially- or
Permanently
Stratified
Days-
Weeks
The above types occur in Member States’ waters as detailed below in table 2.4.2:
Table 2.4.2:
Member States sharing types.
Type
CW – NEA1/26
CW – NEA3/4
CW – NEA7
CW – NEA8
CW – NEA9
CW – NEA10
TW – NEA11
X
X
X
X
X
X
x
BE
X
DE
X
x
DK
X
ES
X
FR
X
IE
X
NL
X
x
x
x
x
x
X
X
x
x
X
X
NO
X
PT
X
SE
UK
X
76
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0079.png
For benthic invertebrates, all classification schemes intercalibrated relate only to the soft sediment
infauna component. Member States use the same classification schemes in all their types.
Differences occur in the reference conditions for the types; these are specific for the habitat type,
and for some Member States (NL and DE), sometimes even specific for the water body. However,
the basic metrics in each country’s scheme remains the same. As Option 3 has been used, the
classification schemes are compared at the type level and boundaries set that are specific to
particular types. This process is described below.
2.4.2 National methods intercalibrated
The national methods that have been assessed in this intercalibration exercise are shown in table
2.4.3 below:
Table 2.4.3:
Member States national methods.
Member State
BE
DE
DK
ES
FR
IE
NL
NO
PT
SE
UK
Method
Benthic Ecosystem Quality Index (BEQI)
1
Multimetric Factorial Analysis (M-AMBI)
Danish Quality Index (DKI)
Multimetric Factorial Analysis (M-AMBI)
Multimetric Factorial Analysis (M-AMBI)
Infaunal Quality Index (IQI)
Benthic Ecosystem Quality Index (BEQI)
1
Norwegian Quality Index (NQI)
Portuguese Benthic Assessment Tool (P-BAT)
Biological Quality Index (BQI)
Infaunal Quality Index (IQI)
Status
Agreed
Agreed
Agreed
Agreed
Agreed
Agreed
Agreed
Agreed
Agreed
Agreed
Agreed
NB. BEQI
1
-For the intercalibration exercise, only the level 3 of BEQI (within habitat quality) is applied.
2.4.3 Reference conditions and class boundary setting
Type - NEA1/26 and NEA7
Reference Conditions
DE, DK, ES, IE, NO, PT, UK
As Option 3 has been selected for the intercalibration of this quality element each Member State has
derived biological reference conditions for the indices within their classification systems. Reference
conditions are specific for habitat sampled and sampling method. For this intercalibration exercise,
all Member States used reference conditions that relate to subtidal, mud/sand habitats and these
are described below in table 2.4.4. (If different habitats were to be assessed, alternative reference
conditions would need to be used.)
General physico-chemical and hydromorphological reference conditions for all types have not been
defined.
77
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0080.png
Table 2.4.4:
Reference conditions.
MS
Method
Sample area
(m
2
)
Component metrics - reference values
Taxa no.
DK
ES
IE
NO
PT
UK
DKI
M-AMBI
IQI
NQI
P-BAT
IQI
0.1
0.4
0.1
0.1
68
Shannon’s = 4.1
Margarlef = 5
Simpsons = 0.97
0
1-AMBI/7 = 0.96
42
68
Diversity
Shannon’s
(logbase2) = 5
Shannon’s = 4
Simpsons = 0.97
AMBI
0
1
1-AMBI/7 = 0.96
0.78
Combined
reference value
DE
Reference values used for the Intercalibration of the German M-AMBI classification methods are
specific for waterbody and habitat type. Because no regional reference sites or sample data were
available DE used reference values from NL database, which is based on similar habitats and shown
in table 2.4.5 below. Germany will generate own local reference data within Phase 2.
Table 2.4.5:
DE reference conditions.
Method
WaterBody
Component metrics - reference values
Taxa no.
M-AMBI
M-AMBI
Vortrapptief, (NEA 1 subtidal 18m), fine sand - sand
Hoher Weg (NEA 26), low, littoral sand)
31
17
Diversity
Shannon’s
AMBI
0.107
0.393
2.66
2.22
NL/BE
The BEQI method (used by NL and BE) requires a reference dataset for each habitat within a water
body. The requirement of a reference dataset is different from the other MS multimetrics, which
only require a (maximal) reference value for a specified habitat. The reference values are also
related to sampling surface (determined by a randomisation procedure) and defined for each WFD
boundary.
NL:
The reference values for reaching good status (good/moderate boundary, sampling surface of
1m
2
) for the fine muddy sand coastal habitat of the Dutch coast (Hollandse kust and Waddenkust)
are summarised in table 2.4.6 below.
Table 2.4.6:
NL reference conditions.
Habitat type
Dutch coast (fine
muddy sand Q1)
Sampling
surface
1 m
2
No. of species
60
Similarity
0.74
Density (ind/
m
2
)
2584 and 7975
Biomass
(gAFDW/ m
2
)
14.2 and 52.4
78
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0081.png
Reference conditions correspond to assessment class Good/Moderate. Reference values are based
on the habitat type Q1 cluster of stations within six nautical miles from the coast, using samples
between 1983 and 1990.
N.B.
The analysis of the benthic communities of Hollandse kust and Waddenkust did not reflect
differences related with the current WFD typology: Hollandse kust (NEA3) and Waddenkust (NEA
1). The typology differences between the areas only reflect the polyhaline versus euhaline water
characteristics. As such in the intercalibration exercise, the Hollandse kust and Waddenkust and
Eems-Dollard kust are taken together in the NEA 1 type intercalibration comparison.
BE:
The reference values for reaching good status (good/moderate boundary) for the fine muddy sand
coastal habitat (for the minimal sampling surface of 1.5 m²) are summarised below in table 2.4.7.
Table 2.4.7:
BE reference conditions.
Habitat type
Fine muddy sand
Abra alba
community
Sampling
surface
No. of species
60
Similarity
0.84
Density (ind/m
2
)
1516 and 5394
Biomass (gAFDW/ m
2
)
Not yet determined
1.5 m
2
Too few data were available in the Belgian WFD benthos dataset to determine the reference
conditions for the other habitats; this could only be done for the
Abra alba
community. In fact,
the habitat typology of the Belgian continental shelf is based on four different macrobenthic
communities with typical species associations in between. The reference values for the three
other communities will have to be determined in future research on historical data or monitoring
in the future. The communities are each characterised by different habitat specifications as briefly
illustrated below.
Macrobenthic communities (habitats):
– Abra alba
community: shallow muddy sand
– Nephtys cirrosa
community: well-sorted mobile sands
– Ophelia limacina – Glycera lapidum
community: medium to coarse sand
– Macoma balthica
community: shallow sandy mud
The macrobenthos status classification will have to be performed for these four communities
separately, based on community-oriented monitoring, to come to an overall macrobenthos
assessment for the Belgian coastal waters.
Some important issues and difficulties to be highlighted:
The Belgian reference data for the different habitats is gathered in the period 1994 – 2000 and
comes not from a `natural, undisturbed` reference period, as required by the WFD. However, this
reference dataset aims to give a reflection of the spatial and temporal variability within the habitats
Using the BEQI method, reference values for species richness, species composition and density
have been given for one community. Reference biomass data is not yet available.
Class Boundary Setting Procedure
As Option 3 has been used, Member States have derived their own initial boundaries for the
National classification methodology.
Two examples of boundary setting procedures are as follows for i) the IQI (sample level procedure)
and ii) the BEQI (waterbody level procedure incorporating biomass):
i) Infaunal Quality Index (IQI)
79
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0082.png
The Infaunal Quality Index (IQI) operates over a scale range from 0 (impacted) to 1 (non-impacted).
Initially, class boundaries were set at equidistant points along the scale (0.2, 0.4, 0.6, and 0.8). These
boundaries were then modified to ensure that the communities conform to the status characteristics
as defined in the Normative Definitions. Figure 2.4.1 illustrates the process followed in setting and
testing the ecological status boundaries.
Set equidistant boundaries
within the IQI range for
each status class.
STATUS
IQI range
High
>0.8
Good
0.6-0.8
Moderate
0.4-0.6
Poor
0.2-0.4
Bad
<0.2
Benthic Invertebrate Abundance Dataset
Calculate IQI (Infaunal Quality
Index) for each sample:
((
0.38
×
AMBI
)
+
(
0.08
×
1
λ'
)
+
(
0.54
×
S
IQI
IQI
IQI 0.1
Intercalibration technical report
– Part 3 Coastal/Transitional,
Section 2 Benthic Invertebates
)
0.4
)
/0.6
In each status class assess*
AMBI ecological group
proportions
Taxa abundances using
SIMPER analysis
Breakdown of Phyla, Class,
Order, Family and Genus
using SIMPER analysis and
reference to MNCR habitat
characterising taxa
No
Does the taxa
composition in each
status class fit with
Normative
Definitions?
Yes
Continue to test
status class
ranges on
further datasets
of the same
habitat type
I
II
III
IV
V
Calculate the 1st and
3rd quartiles for each
status class
Assess the
quartiles as in *
above
%
40
20
AMBI group proportions at the upper and lower quartiles of each status class
100
80
60
Adjust the status
class boundaries
towards the 1st
and/or 3rd
quartile so taxa
composition fits
with Normative
Definitions
0
P
PQ
3
M
Q
1
M
Q
3
H
B
BQ
3
PQ
1
BQ
1
Q
1
G
3
1
M
G
Status class (Quartile 1 and 3 indicated)
Figure 2.4.1:
Flow diagram of stepwise process followed to establish boundaries for each status class which reflect the
Normative Definitions.
80
G
H
Q
H
Q
Q
3
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0083.png
The class boundaries were set using samples from an organic enrichment gradient; the sewage
sludge disposal site at Garroch Head (data courtesy of the Fisheries Research Services, Scotland).
Benthic invertebrate abundance data had been obtained across the pressure gradient annually over
a 20-year period (total of 186 samples). The samples were taken from habitats classified as type
A5.3 (marine sublittoral muds) under the European Nature Information System (EUNIS) habitat
classification scheme.
The IQI, which is comprised of three indicators of benthic ecosystem health (Simpson’s evenness,
taxa number, AZTI Marine Biotic Index (AMBI)), was calculated for all samples.
The IQI is calculated as:
IQI =
((0.38× AMBI
IQI
)+ (0.08×1 − λ’
IQI
)+ (0.54×S
IQI 0.1
)− 0.4)/0.6
Where:
AMBI
IQI
= Observed value of (1-AMBI/7) divided by maximum expected value under reference
conditions for a given habitat type.
1-λ’
IQI
S
IQI
= Observed value of Simpsons evenness index (1-λ’) divided by the maximum expected
value under reference conditions for a given habitat type.
= Observed number of taxa divided by the maximum expected value under reference
conditions for a given habitat type.
AMBI assigns taxa a sensitivity score (ecological group, EG) as follows: EG I – sensitive taxa,
EG II – indifferent taxa, EG III – tolerant, EG IV – opportunistic, EG V – indicator taxa. The
composition of an assemblage in terms of ecological groups is expected to respond to the extent of
anthropogenic pressure (e.g. Figure 2.4.2). The proportions of sensitive taxa, tolerant taxa and taxa
indicative of pollution were used to establish class boundaries.
BIOTIC COEFFICIENT
0
100
90
PERCENTAGE OF GROUPS
80
70
60
50
40
30
20
10
0
0
1
2
3
BIOTIC INDEX
4
6
5
7
II
IV
I
III
AZOIC SEDIMENT
BAD
STATUS
V
1
2
3
4
5
6
POLLUTION
WFD
UNPOLLUTED
HIGH
STATUS
SLIGHTLY POLLUTED
GOOD
STATUS
MEANLY POLLUTED
MODERATE
STATUS
POOR
STATUS
EXTREM.
HEAVILY
`POLLUTED POLLUTED
INCREASING POLLUTION
Figure 2.4.2:
The AMBI biotic coefficient, relating the ecological groups present in a sample to an assessment of the
benthic invertebrate community (Borja et al, 2003).
81
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0084.png
PERC
20
10
0
0
1
2
II
Using equidistant class boundaries initially, AMBI ecological group proportions in the gradient
dataset were assessed as
UNPOLLUTED
the proportions of sensitive
MEANLY POLLUTED
to whether
taxa, tolerant taxa
HEAVILY EXTREM.
and taxa
POLLUTION
SLIGHTLY POLLUTED
`POLLUTED POLLUTED
indicative of pollution were in line with the Normative Definition for each status class (Figure
HIGH
GOOD
MODERATE
POOR
BAD
2.4.3).
WFD
STATUS
STATUS
STATUS
STATUS
STATUS
3
BIOTIC INDEX
4
6
5
7
The boundaries were then adjusted to ensure they captured the sampling occasions that provided
INCREASING POLLUTION
the highest agreement with the Normative Definitions. This was achieved by plotting the ecological
group proportions within the first and third quartiles of each status (Figure 2.4.4) and adjusting the
boundaries until agreement with the Normative Definitions was maximised.
The validity of the adjusted boundaries was then assessed by analysis of the composition of the
taxa in each status class. SIMPER analysis (PRIMER©) was carried out, assessing the top 90 %
100
EG I
80
EG II
EG III
60
%
EG IV
EG V
40
20
0
B
P
M
Status
G
H
Figure 2.4.3:
The proportions of AMBI ecological groups for each ecological status class (B
= Bad, P = Poor, M = Moderate, G = Good, H = High) as determined by an equidistant split of
the IQI.
100
EG I
EG II
EG III
EG IV
EG V
80
60
%
40
20
0
BQ1
B
BQ3
PQ1
P
PQ3
MQ1
M
MQ3
GQ1
G
GQ3
HQ1
H
HQ3
Status class (Quartile 1 and 3 indicated)
Figure 2.4.4:
AMBI group proportions of status classes (B = Bad, P = Poor, M = Moderate, G = Good, H = High) and
their upper (Q3) and lower (Q1) quartiles using preliminary equidistant divisions of the IQI.
82
AZ
IV
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0085.png
of contributing taxa (family level taxonomic discrimination). Changes in composition between the
status classes were evaluated to determine whether the adjusted ranges mirrored the Normative
Definitions in terms of contributing taxa. In addition, the Marine Nature Conservation Review
(MNCR) Habitat Directive types were used to ensure an understanding of the characterising taxa
and their composition expected for a particular habitat type.
The resulting IQI boundaries and are shown in table 2.4.8 and AMBI group proportions in Figure
2.4.5 respectively.
ii) Benthic Ecosystem Quality Index (BEQI)
The BEQI statistically integrates the risk of misclassification at water body level related to sample
size in the methodology and boundary setting. Based on permutation calculations, reference values
are determined for each component metric and class boundary.
The reference values are calculated per habitat over increasing sampling surfaces. This allows
for the estimation of the reference value for any given sampling surface. The reference for a 1m
2
sampling surface is based on a set of 2000 artificial random samples out of the reference dataset.
Out of the randomisation procedure, each component metric (indicators: density, biomass, species
richness, species composition changes), a 5
th
percentile value is selected as the value that has to be
reached to achieve good status (the value of the good/moderate boundary) (Figure 2.4.6).
Table 2.4.8:
Class Status boundaries – IQI assessment
(prior to optimisation through the Intercalibration exercise).
Status Class Boundary
High/ Good
Good/ Moderate
Moderate/ Poor
Poor/ Bad
IQI
0.80
0.65
0.43
0.20
100
EG I
EG II
EG III
EG IV
EG V
80
60
%
40
20
0
B
P
M
Status
G
H
Figure 2.4.5:
AMBI group proportions within each status based on the Garroch Head data
using the following boundaries: High/Good =0.8; Good/Moderate = 0.65; Moderate/Poor =
0.43; Poor/Bad = 0.2.
83
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0086.png
poor
moderate
50
good
high
High status (1 – 0.8) :
– 0.8) :
High status (1
median
median
40
Good status (0.6 - 0.8): 5
-
Good status (0.6
0.8): 5
the
the percentile
percentile
number of species
30
Moderate status (0.4 – 0.6): 2/3
– 0.6):
boundary
good
Moderate status (0.4
of good
2/3 of
value
boundary value
value
20
Poor status (0.2 –
(0.2
1/3
0.4): 1/3 of good boundary
Poor status
0.4):
of good boundary value
10
Bad status (0 – 0.2)
Bad status (0
– 0.2)
0
0.02
0.25
0.47
0.7
1.6
1.82
2.5
3.4
4.3
4.52
4.75
4.75
0.92
1.15
3.17
3.62
3.85
sampling surface (m
2
)
Figure 2.4.6:
The boundary settings for number of species in relation to sampling surface. A similar protocol is applied
for similarity (species composition changes).
For the parameters density and biomass, a two side deviation from the reference values is scored
(Figure 2.4.7). The other boundary values were adopted from this value (equal intervals), except the
high/good reference value, which is also directly extracted from the randomisation procedure.
poor min
poor min
mod min
mod min
good min
good min
high min
high min
median
median
high max
high max
good max
good max
mod max
mod max
poor max
poor max
14
Bad status (0 – 0.2): < 2 x good
good boundary value
Bad status (0 – 0.2): < 2 x
boundary value
Poor status (0.2 – 0.4): < 5/3 of good boundary value
Poor status (0.2 – 0.4): < 5/3 of good boundary value
Moderate status (0.4 – 0.6): < 4/3 of good
good boundary value
Moderate status (0.4 – 0.6): < 4/3 of
boundary value
the percentile
Good status (0.6 - 0.8): < 97.5
97.5
the
percentile
Good status (0.6 - 0.8): <
12
10
ind./m
2
(fourth root)
8
ste and
75
ste
percentile
High status (1 – 0.8)
0.8) : between 25
ste
and
75
ste
percentile
High status (1 –
: between 25
the
percentile
Good status (0.6 - 0.8): > 2.5
2.5
the
percentile
Good status (0.6 - 0.8): >
Moderate status (0.4
(0.4 – 0.6): > 2/3 of
boundary value
Moderate status
– 0.6): > 2/3 of good
good boundary value
6
4
Poor status (0.2 – 0.4): > 1/3 of good boundary value
Poor status (0.2 – 0.4): > 1/3 of good boundary value
Bad status (0 – 0.2): < 1/3 of good
good boundary value
Bad status (0 – 0.2): < 1/3 of
boundary value
2
0
0.25
0.92
1.15
1.37
2.27
3.40
3.85
4.07
4.30
0.02
0.47
0.70
1.60
1.82
2.05
2.50
2.72
2.95
3.17
3.62
4.52
4.97
Sampling surface (m
2
)
Figure 2.4.7:
The boundary settings for density in relation to sampling surface. A similar protocol is applied for
biomass.
84
4.97
1.37
2.05
2.27
2.72
2.95
4.07
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0087.png
The boundaries selected from the randomisation procedure, with the corresponding WFD class
boundaries are summarised below in table 2.4.9.
Table 2.4.9:
NL boundaries.
Boundary
High/Good: 0.8
Good/Moderate: 0.6
Moderate/Poor: 0.4
Poor/Bad: 0.2
No. species and species composition change
Median
5
th
percentile
2/3 of Good/Mod value
1/3 of Good/Mod value
Density & biomass
25 and 75 percentile
2.5
th
and 97.5
th
percentile
2/3 and 4/3 of Good/Mod value
1/3 and 5/3 of Good/Mod value
The boundaries set by the BEQI method are therefore statistical significance levels related to a fixed
WFD Class boundary value.
Type - NEA3/4
Reference Conditions
Intercalibration was carried out between the BEQI (Netherlands) and the M-AMBI (Germany).
Reference boundary values for four habitats of major importance in NEA3/4 types are shown below
in table 2.4.10 (for G/M status).
Table 2.4.10:
NL and DE reference conditions.
BEQI
Habitat
High Littoral Mud
Middle Littoral
Muddy Sand
Low Littoral Sand
Brackish Sub
Littoral
M-AMBI
High Littoral Mud
Middle Littoral
Muddy Sand
Low Littoral Sand
Brackish Sub-
Littoral
Sample
surface (m
2
)
3
3
3
3
No. species
13
17
13
26
No. species
18
23
17
16
Similarity
0.68
0.7
0.6
0.82
AMBI
2.7
0.947
0.393
1.541
Density (ind/m
2
)
448 - 7643
269 - 12063
106 - 7384
1810 - 103353
Shannon
2.16
2.34
2.22
2.178
Biomass
(gAFDW/ m
2
)
4.1 - 20.6
18.4 - 58.9
4.3 - 24.3
18.7 - 88.8
85
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
Class Boundary Setting Procedure
Netherlands (BEQI):
see section above for procedure described under NEA1/26.
Germany (M-AMBI):
The class boundaries to classify the results of the M-AMBI were set after
comparison of the assessment results using the boundaries given by Muxika
et al.
(2007) and Spain
(Intercalibration Report, 2007) with a classification by expert judgement. The boundaries were
then modified to ensure that the communities conform to the status characteristics as defined in the
Normative Definitions of the WFD.
The High/Good boundary (0.85) was in agreement with the boundary given by Muxika et al.
(2007) but higher than the value of Borja (0.77) (Intercalibration report 2007). The Good/Moderate
boundary was set at 0.7, which is higher than the boundaries given by Muxika et al. (2007) (0.55)
or Borja (Intercalibration report 2007) (0.53). The reference values chosen for the M-AMBI
calculations were determined by NL (van Hoey et al. 2007) on a data basis including the years 1969
to 1983 for the littoral Wadden Sea and the years 1988 to 1990 for the Wadden- and Eemskust.
These reference datasets derived therefore from times, in which human impacts on the North Sea
and the Wadden Sea were already obvious and first changes in benthic species composition and
biomass were detected (e.g. Beukema 1991, Kröncke 1995). Therefore the ‘reference data do not
represent an unaffected status’ and the boundaries needed adjustment. Further calculations should
be carried out with references derived from own data sets of the relevant stations. The boundary
G/M was set as 0.7, because for all of the chosen intercalibration stations and for many additional
stations (Heyer 2007) the M-AMBI results corresponded with the expert judgement.
Type - NEA8/9/10
Reference Conditions
Denmark, Sweden and Norway have these types.
In Sweden when deriving the type-specific class boundaries, the greatest emphasis was placed on
the good–moderate boundary. This boundary was primarily determined with the aid of comparative
data from high–good areas. Comparative data was chosen for each national type from regions
lacking local discharges; in practice areas with the highest mean BQI values existing for that type.
Sequential tests identifying the level of BQI (20th percentile) where a water body significantly
differs from the comparison material were made to assist in the setting of good-moderate boundary.
In those types where the amount of data was insufficient, expert assessments were conducted based
on existing data and data from nearby types with similar properties. Once the good–moderate
boundary was determined, it was deemed acceptable to consider the area from the good–moderate
boundary up to the highest observed index value in the existing type as mainly constituting a status
of good. Two-thirds (2/3) of the span exceeding the good–moderate boundary was assigned a
status of good, while the upper third was reserved for the status of high. The area below the good–
moderate boundary was divided into three equal intervals for the remaining boundaries. It was not
possible to describe reference conditions due to a lack of unaffected areas today and historical data
were not collected with the same methodology as in use today. As shown above the best available
data from unimpacted areas has been used as a reference in setting the boundaries. This data
assumed having a status of at least good.
In Denmark the Good-Moderate boundary was determined by the 5
th
percentile of samples/sites
classified as being in at least Good status. The High-Good boundary was set midway between the
GM value and the theoretical highest index value of 1. The reference state using this approach is the
highest diversity (Hmax) and the lowest
86
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
AMBI found in the material having at least Good status from the type in question. In accordance
with the normative definition concerning sensitive species the reference state is characterized by
high dominance of sensitive species (Ecological group I according to the AMBI system), followed
by group II. Group III and the opportunists, groups IV and V are subdominant and virtually absent
respectively.
In Norway the 90-percentile (the border value between the 90 % lower and 10 % higher values
among all the grab samples from the reference stations) was used to quantify the reference value for
the metric.
Class Boundary Setting Procedure
Again as Option 3 has been used Member States have derived their own initial boundaries for the
national classification methodology. An example of this from Sweden is as follows:
The performance of Swedish, Danish and Norwegian national methods of assessing status have been
compared by Sweden on Swedish national and regional monitoring data from coastal deeper parts (>
20 m depth) of the Kattegat/Skagerrak area. Three steps have been carried out in this intercalibration
exercise, comparison of species sensitivity classifications, comparison of index values in different
environments and comparison of status classification on water body level.
Species classifications
A major part of the Swedish (BQI), Danish (DKI, Borja et al. 2006) and Norwegian (NQI) method
for classification is based on species sensitivity classifications. In the Swedish method an objective
method was used to classify species sensitivity to environmental degradation. In general, few species
appear in degraded areas. Thus, a species that frequently appears in species-poor environments are
given a low index value compared with species that only appear in species-rich environments. We
have used Hurlberts diversity index as a measure of the number of species in a sample. The species
sensitivity value has been calculated as the 5th percentile of that species diversity-value frequency
distribution, where each observation in the frequency distribution represents an individual’s
diversity value from the sample in which that individual appeared. We have used a large amount of
both Swedish and Danish data when calculating the sensitivity values.
The Danish and Norwegian method uses AMBI species classifications (see Borja et al 2000). AMBI
classifies species into five ecological groups, I = Sensitive, II = Indifferent, III = Tolerant, IV =
Second order opportunists and V = First order opportunists.
The correspondences between the two classifications are not that good. There are only small
differences between distributions of sensitivity values in the ecological groups I, II, III and IV.
Index correlations
To see how the indices correlates the BQI, DKI and NQI was calculated in each grab sample (~0.1
m
2
) and the results where plotted against each other. The Swedish good and moderate boundary
lies around a BQI value of 12 for this type. Up to this value there is an almost linear relationship
but over this value DKI and NQI tends to level off while BQI continues to increase. Still there is
a rather good correlation between BQI, DKI and NQI. The relationship between NQI and DKI is
nearly perfect, not unexpectedly because they both use the same components i.e. AMBI.
87
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0090.png
28
26
24
22
20
18
16
14
12
10
8
6
4
2
0
Ecological group I Sensitive
Rare
Common
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Sensitivity value
28
26
24
22
20
18
16
14
12
10
8
6
4
2
0
Ecological group II Indifferent
Rare
Common
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Sensitivity value
28
26
24
22
20
18
16
14
12
10
8
6
4
2
0
Ecological group III Tolerant
Rare
Common
No of taxa
No of taxa
No of taxa
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Sensitivity value
28
26
24
22
20
18
16
14
12
10
8
6
4
2
0
Ecological group IV Second order opportunistic
Rare
Common
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Sensitivity value
28
26
24
22
20
18
16
14
12
10
8
6
4
2
0
Ecological group V First order opportunistic
Rare
Common
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Sensitivity value
28
26
24
22
20
18
16
14
12
10
8
6
4
2
0
Ecological group not assigned
Rare
Common
No of taxa
No of taxa
No of taxa
y = 0,0002x
3
- 0,0077x
2
+ 0,1272x
R
2
= 0,7532
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Sensitivity value
Figure 2.4.8:
Swedish sensitivity values in different ecological groups. Only species found in the Swedish coastal part
of deeper (> 20 m depth) areas in the Kattegat/Skagerrak are included in the comparison. Common taxa, found in more
than 500 samples, are indicated with purple color.
1
y = 6E-
3
- 0,0043x
2
+
R
2
= 0,7872
1
1
y = 0,8204x + 0,1515
R
2
= 0,9214
0,8
0,8
0,8
0,6
DKI
0,6
NQI
0,6
NQI
0,4
0,4
0,4
0,2
0,2
0,2
0
0
5
10
BQI
15
20
0
0
5
10
15
20
0
0
0,2
0,4
0,6
0,8
1
B QI
D KI
Figure 2.4.9:
Correlation between BQI, DKI and NQI.
To compare how the indices behaves in different disturbed situations plots were made on data from
the deepest part of the Gullmar fjord where low oxygen levels have affected the fauna at several
occasions and data from the Saltkälle fjord where organic loading have decreased and fauna
recolonized after closure of a sulphite mill factory in 1966.
In both of these examples the BQI, DKI and NQI behaves similarly, only minor differences can be
seen.
88
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0091.png
0,8
BQI
DKI
NQI
14
12
10
8
6
4
2
BQI
0,8
L9
0,6
0,4
BQI
DKI
NQI
1965
1970
1975
1980
1985
1990
1995
0,6
DKI and NQI
0,4
0,2
0,0
1980
1985
1990
1995
2000
2005
0
Figure 2.4.10:
Comparison of index response to low oxygen content at Alsbäck 118 m depth in Gullmar fjord on the
Swedish west coast. The indices are shown with standard error of the mean. Data from monitoring programmes and
Agrenius pers comm.
16
14
12
BQI
10
8
6
4
2
0
1965
1970
1975
1980
1985
1990
BQI
DKI
NQI
1995
L6
0,8
0,6
DKI and NQI
16
14
12
BQI
10
8
6
4
2
0
0,4
0,2
0,0
2000
0,2
0,0
2000
0,8
0,6
0,4
16
14
12
BQI
0,8
L11
0,6
DKI and NQI
16
14 L12
12
10
8
6
4
2
0
1965
1970
1975
1980
1985
1990
0,4
BQI
DKI
NQI
1995
8
6
4
2
0
BQI
DKI
NQI
1965
1970
1975
1980
1985
1990
1995
0,2
0,0
2000
0,8
0,6
0,4
DKI and NQI
0,2
0,0
2000
18
16 L18
14
12
10
8
6
4
2
0
1965
BQI
BQI
DKI
NQI
1970
1975
1980
1985
1990
1995
0,2
0,0
2000
Figure 2.4.11:
Comparison of index response to decreased organic load after closure of sulphite mill factory in 1966.
Stations L6, L9, L11, L12 and L18 are shown with increasing depth (25 – 44 m) and increasing distance to discharge
from river Örekilsälven in the inner part of the Saltkällefjord on the Swedish west coast. Station L9 had a low value in
1988 due to a temporary drop in oxygen concentration. The indices are shown with standard error of the mean. Data
from Bagge 1969, Leppäkoski 1975, Rosenberg 1972, 1973, 1976 and monitoring programmes.
89
DKI and NQI
10
BQI
DKI and NQI
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0092.png
Water body status
To compare the assessment of status of water bodies’ national indices and boundaries were applied
on data from a selection of Swedish water bodies with at least 5 stations sampled in a year. Denmark
and Norway has not yet presented any method for how to go from sample to water body level so
for comparison the mean value of DKI and NQI were compared with the 20
th
percentile of BQI
according to the Swedish method.
Comparisons on time series from three water bodies have been made. The results of these
comparisons are shown. The agreement between the three methods are good, only in a few instances
does the status differ slightly between them.
Type - NEA11
This is the generic single transitional water type covering all transitional waters in eight countries.
As such it is very wide ranging and demands methods that cover all the salinity ranges in the
different transitional waters.
Reference Conditions
Reference conditions derived for coastal waters may be used in the euhaline parts of transitional
waters assuming equivalent habitats (sediment types) are assessed. The reference conditions from
the coastal water type into which the transitional water drains may therefore be applicable. For the
N m Hallands kustvatten (vd)
16
14
12
10
8
6
4
2
0
1970
N m Hallands kustvatten (vd)
1,0
0,8
NQI
0,6
0,4
0,2
1975
1980
Saltkällefjorden
1,0
0,8
DKI
NQI
0,6
0,4
0,2
1971
1973
1975
1977 1979
0,0
1967 1969
1,0
0,8
DKI
0,6
0,4
0,2
1984
1989
1994
1999
0,0
1979
1984
1989
1994
1999
NQI
1971
1973
1975
1977 1979
1985
1990
0,0
1970
1,0
0,8
0,6
0,4
0,2
0,0
1967 1969 1971
1973
1975
1977 1979
1975
1980
Saltkällefjorden
1985
1990
N m Hallands kustvatten (vd)
1,0
0,8
DKI
0,6
0,4
0,2
BQI
1975
1980
Saltkällefjorden
1985
1990
0,0
1970
16
14
12
10
8
6
4
2
0
1967 1969
BQI
Gullmarn centralbassäng
16
14
12
10
8
6
4
2
0
1979
Gullmarn centralbassäng
1,0
0,8
0,6
0,4
0,2
0,0
1979
Gullmarn centralbassäng
BQI
1984
1989
1994
1999
Figure 2.4.12:
Comparison of water body status assessed with Swedish (left column), Danish (middle column) and
Norwegian (right column) methods. Error bars in Swedish graphs denotes 20
th
and 80
th
percentile. Note that the
Swedish method uses the 20th percentile for assessment. The agreements between the methods are good. Data from
Bagge 1969, Leppäkoski 1975, Rosenberg 1972, 1973, 1976 and monitoring programmes.
90
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
polyhaline and mesohaline stretches within transitional waters different reference criteria may be
required for use in the specified classification scheme. At this time, these reference conditions are still
under development. All required reference conditions are not expected to be fully developed before
June 2007 therefore postponing intercalibration in transitional waters until Phase II, 2006 to 2009.
Class Boundary Setting Procedure
If assessments are made in the euhaline areas of transitional waters (or with a similar salinity regime
to the adjoining coastal waterbody) then the classification tool for coastal waters would be used in
the same habitats to which that tool applies. As such the boundary setting procedure would be the
same as described above for the coastal waters types. The same principles would be used in setting
boundaries in other areas of transitional waters with different salinity regimes but the classification
would be anchored by different reference condition specific to those areas. At this time, there are
no classification systems developed that allow a comparison between Member States, therefore
postponing intercalibration in transitional waters until Phase II, 2006 to 2009.
Any future approach should build on the lessons learnt from the coastal water intercalibration
exercise. It has been suggested that as many transitional waters have been identified as ‘heavily
modified water bodies, it may complicate the simple aggregation of present metrics adapted to cover
salinity ranges and differences in habitats, for some classification assessments.
2.4.4 Results of the comparison
Type - NEA1/26 and NEA7
DK, ES, IE, NO, PT, UK
Data have been exchanged by Member States and a standardised Intercalibration dataset collated.
This dataset has been run through each proposed national classification scheme to determine the
matches in status obtained when assessed by the different methods. For example, Danish data were
assessed using the UK classification scheme, and vice versa, and the matches in status assessed.
Five classification schemes (M-AMBI, IQI, NQI, DKI and P-BAT) were available for this
assessment. The assessment indicated the agreement of sites that have been assigned similar status
(High/Good or Moderate/Poor/Bad) by the different assessment methods. The EQR boundaries
in the ecological status assessment were then altered, to achieve the highest agreement between
Member States (i.e. optimisation of boundaries).
Following optimisation of the boundaries, agreement between boundaries was assessed according to
guidelines issued by Ecostat to compare the agreement in classes both for all five classes and only
three, high, good and moderate or worse. In addition further evaluation was made using harmonised
boundaries and allowing a 0.05 EQR range of deviation around the boundaries. These results are
presented in the table 2.4.11 below.
91
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0094.png
Table 2.4.11:
Percentage agreement between classification methods. (Following optimisation of boundaries).
Agreement
(5 classes)
Average A
(DKI –DK)
Average B
(M-AMBI –
FR/ES)
Average C
(NQI – NO)
Average D
(PBAT – PT)
Average E
(IQI – UK/IE)
Overall average
ABCDE
77.098
71.77
72.635
70.995
71.578
72.815
Agreement
(3 classes)
80.303
75.01
76.298
74.448
75.275
76.267
Agreement
(5 cl.+0.05 EQR)
97.918
95.098
94.033
94.958
96.55
95.711
Agreement (3
cl. +0.05 EQR)
98.13
95.273
94.173
95.205
96.69
95.894
Average class
difference
0.035
0.013
0.01
0.018
-0.035
0.008
Absolute
average class
difference
0.068
0.078
0.085
0.085
0.08
0.079
Table 2.4.12:
Overall summary data for this comparison.
EQR deviation
Number of classes
No deviation
76.3
3
No deviation
72.8
5
0.05 EQR
95.9
3
0.05 EQR
95.7
5
Percentage of
classification agreement
In the case of three classes the percentage of classifications that differ by 2 classes is 0.479 for no
deviation allowance and 0.028 for the 0.05 EQR deviation allowance situations.
Table 2.4.13:
Average and absolute class differences.
Average of class difference
Absolute average of class difference
0.008
0.079
This is a high level of agreement bearing in mind that the data was collected by participating
countries independently, in many cases before intercalibration really started and the inherent
variability of biological data. This is deemed by the GIG to be an acceptable level of agreement and
validates the national boundaries.
DE
In the intercalibration report of Heyer (2007) a comparison of the methods (BEQI, M-Ambi)
showed following results:
Comparison of the BEQI, M-AMBI, and IQI were carried out for two sites in NEA1/26. (This
differs from the above exercise where a standardised Intercalibration dataset was used by all the
Member States involved.)
Table 2.4.14 shows the compilation of the assessment results for different assessment methods for
the German intercalibration stations: blue =’high status’(‘I’), green = ’good status’(‘II’), yellow =
’moderate status’(‘III’) and orange = ’poor status’(‘IV’).
92
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0095.png
Table 2.4.14:
DE comparisons.
Waterbody Station
Type
Sediment
Expert
judgement
M-AMBI
BEQI
Assessment,
Assessment,
DE boundaries NL boundaries
0.77
0.46
0.52
0.36
0,53
0,79
IQI Assessment,
UK boundaries
(pre-harmonisation)
0.78
0,57
0,68
NEA1
NS2, Vortrapptief,
Fine sand
mean 2002 to
to sand
2004
Hoher Weg 1988
Hoher Weg 2004
Low
littoral
sand
‘II’ to ‘III’
‘II’ to ‘III’
NEA26
Type NEA1:
The assessments of the ecological status of station ‘NS2’ by three different methods
showed following results: ‘high status’ assessed by the ‘IQI, ‘good status’ assessed by the
‘M-AMBI’ and ‘poor status’ assessed by the ‘BEQI’. The assessment by the M-Ambi reflected the
status given by expert judgement better than the other methods (status ‘good’ to ‘moderate’)
Type NEA26:
For the year 1988, the ecological status of the stations was assessed as ‘moderate’
by the three methods. For 2004, the status of the stations was assessed as ‘good’ by the IQI and
the BEQI, while the M-AMBI led again to the assessment ‘moderate’. The results of the different
methods were in good agreement, also with the expert judgement (status ‘good’ to ‘moderate’).
NL
This intercalibration/comparison of methods is on a habitat level, as the BEQI evaluates directly at
a habitat level within a waterbody. This approach to intercalibration at a water body level may be
completed in Phase II Intercalibration.
Comparison of the BEQI, M-AMBI, IQI, DKI and NQI on Dutch data was done for the habitat Q1
(fine muddy sand) in two water bodies in the North Sea coastal zone, Waddenkust and Hollandse
Kust. DKI, IQI and M-AMBI scores were calculated with reference settings suggested by the
member states for their defined habitats in the earlier exercise (fixed reference) and with reference
values determined from the Q1 reference dataset of the North Sea coastal zone (Table 2.4.15). The
reference setting is the same for Hollandse Kust and Waddenkust. NQI does not allow for use of
local reference values.
The assessment period is 3 years (2002-2004) with a total of 15 samples per habitat in the two water
bodies. Except for the BEQI method the average of assessment station scores within a habitat was
taken as the overall habitat status.
Assessment results were very similar for both water bodies (Figure 2.4.13). DKI, IQI and NQI
are hardly different between water bodies. BEQI and M-AMBI show more variation between
Hollandse Kust and Waddenkust. In all cases, the outcomes were higher when the local determined
reference settings were used, in three out of six cases use of the local reference resulted in a shift
Table 2.4.15:
Local reference values used for index calculation. Reference values based on the Q1 cluster of stations
within six nautical miles from the coast, sampled between 1983 and 1990.
Bad
Waddenkust
Hollandse kust
6
6
AMBI
High
0.107
0.246
Bad
0
0
Shannon-Wiener
2.66
2.83
High
Bad
0
0
# Species
High
31
31
Simpson
Max
0.91
0.92
93
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0096.png
Figure 2.4.13:
Bar graph of five metrics calculated for two water bodies of the Dutch coast,
Hollandse kust and Waddenkust. Calculations are done using a fixed reference (in red)
suggested by the member states and a local reference (blue) determined from a reference data
set (Q1 cluster). Boundaries of the indices are indicated with horizontal lines.
to a higher quality status. Of the five compared methods the BEQI method assigned both water
bodies to a lower quality class than the other methods, regardless of the reference used (fixed
or local). Differences in assessment ranged as much as between poor and high status. These
differences are caused by differences in the applied methodology. In the first place, total biomass
of the macrobenthic community is included in the BEQI but not in the other assessment methods.
Biomass has increased considerably in the North Sea coastal zone since the period of reference
setting, causing very low scores for the sub-metric biomass and subsequent low overall scores. Also
the measure of similarity between reference and assessment was rated not more than moderate at
the Hollandse Kust and Waddenkust, as it detects the changes in species composition caused by an
invasive species.
Besides the use of different input and sub-metrics, the BEQI relates assessment and reference
conditions differently. The BEQI explicitly takes spatial and temporal variability into account. The
class boundaries are determined by the variance in the reference data.
To increase the number of data for comparison, index scores are calculated for several time intervals
(four periods: 91-95, 96-98, 99-01 and 02-04). BEQI responds strongest to the temporal and spatial
differences with a coefficient of variation (cv) of 0.12. M-AMBI showed less variation in time and
space with a cv of 0.09. The other indices DKI IQI and NQI responded least with cvs of 0.06, 0.05
and 0.05 respectively. M-AMBI, DKI, IQI and NQI are all positively correlated. This is not really
surprising as the indices rely on the same or similar sub metrics. The BEQI correlates poorly (not
significant) with the other methods.
Based on this very limited comparison between the BEQI and classification methods, the BEQI
seems to rate water bodies at a lower ecological class than the other methods. Further comparisons
with a larger dataset with a larger variation in conditions are needed to further evaluate the
performance of BEQI relative to the other indices.
94
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0097.png
Figure 2.4.14:
Scatterplot matrix of the index scores
of the different methods calculated for two water
bodies and four time periods in the coastal zone of the
Netherlands, Waddenkust and Hollandse Kust. Sample
ellipses are plotted together with the data.
Type – NEA3/4
NL
At present Germany applies the M-AMBI to the Wadden Sea, which is assigned to Intercalibration
type cluster NEA 3/4. A comparison is made between the results of the BEQI method for the
different habitats of the Dutch Wadden Sea, and M-AMBI values calculated for the same assessment
data. Additionally the BEQI and M-AMBI are calculated for benthos datasets from Norderney and
the Leybucht, supplied by Germany for the intercalibration exercise. Norderney data cover a five-
year period from 1998 to 2002. For Leybucht, data from a longer period are available and there
the six most recent years from 1998 to 2003, are selected for the BEQI and M-AMBI calculation.
Information on biomass of the macrobenthic species is not available, so only three of the four sub-
metrics are calculated.
Results of the BEQI calculations are presented in Table 2.4.16. For Norderney reference conditions
of Middle Littoral Muddy Sand are used. Leybucht results are relative to the reference for High
Littoral Mud. Total densities of the macro fauna are higher than the reference. Number of species
is very large, in both areas more species are found than described in the reference data. Similarity
is rated poor in both areas. The overall score is good for Norderney and moderate for the Leybucht.
This difference is due to the larger deviance of total density in the Leybucht. This very strong
divergence from the reference conditions is caused by high numbers of
Hydrobia ulvae, Tubificoides
benedeni
and
Corophium
sp. This could indicate a disturbed situation, however it is questionable if
the reference description is appropriate in this case. No
T. benedeni
or other Oligochaetes are present
in the reference set, which might suggest that this group was neglected in the Dutch monitoring, or
alternatively this ecotype was not well covered. However, excluding
T. benedeni
in the calculations
still leads to a poor density status.
M-AMBI scores are calculated for the same six assessment samples as are used for the BEQI
scores, four habitats of the Dutch Wadden Sea are presented and the two German sites discussed
above. M-AMBI scores are calculated per station and then averaged per habitat. Reference values
are outlined in Section 2.3.3 (reference conditions NEA 3/4).
95
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0098.png
Table 2.4.16:
BEQI results of three sub-metrics for Norderney
(years 1998 to 2002) and Leybucht (years 1998-2003)
in the German Wadden.
Table 2.4.17:
BEQI and M-AMBI assessment (Spanish pre-optimisation boundaries) results for the Dutch and German
Wadden Sea habitats. The samples in the Dutch Wadden Sea consist of stations in several areas.
A comparison between the BEQI and M-AMBI scores for the Dutch and German Wadden Sea
habitats is made in Table 2.4.17. In three out of six cases the status classification is the same for both
metrics. In the other three cases BEQI gives a lower quality status, twice moderate versus good and
once moderate versus high for the Leybucht. In the Leybucht the difference is mainly caused by the
very high abundances of three species with much stronger effect on BEQI than M-AMBI. Moderate
BEQI scores in the Dutch Brackish Sub-Littoral and High Littoral Mud habitats are caused by
large biomass values relative to the reference and moderate similarities. This partly explains the
difference in status score.
A more detailed analysis of the comparison between BEQI and M-AMBI results for the Wadden
Sea does not seem adequate until the number of cases is substantially larger. In the next section
data from the Coastal zone of the North Sea, the Wadden Sea and the Eems-Dollard are combined
to increase sample size and variation in assessment conditions for a comparison between the two
indices. This generates a larger dataset with a larger range of variation in conditions reflected in the
indices. M-AMBI scores are calculated per assessment station and averaged per habitat. M-AMBI
96
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0099.png
reference values are selected from the same reference datasets as are used for the BEQI reference
description. Calculation of BEQI scores and results are presented in previous sections of this report.
The comparison was made based on 20 observations from the combined water bodies. In Figure
2.4.15 the AMBI scores are plotted against the BEQI scores. The range of M-AMBI scores is 31
% and the range of BEQI 43 % of the scale from 0 to 1. In this limited comparison most M-AMBI
assessments fall within a single category, good and a few in high. The BEQI status assessments
mainly fall within two categories moderate and good. Overall the two indices are positively
correlated with a Pearson correlation coefficient of 0.45 (P<0.05).
Within water bodies scores are
not related. In five out of twenty cases the status is judged similar by BEQI and M-AMBI, in the
fifteen remaining cases the M-AMBI status was higher than the BEQI status.
BEQI and M-AMBI are indices with several differences in methodology. This leads to differences in
the outcomes when applied to the same data. On a large scale there is agreement in the direction of
the response of the two methods, shown as a positive correlation. However BEQI is more sensitive
to changes in the assessment conditions relative to the reference than M-AMBI.
DE
The assessments for the monitoring stations ‘T1’ to ‘T8’ by the BEQI and M-AMBI were in good
agreement: the ecological status of these stations was assessed by both methods as ‘good’ and these
assessments are also in agreement with expert judgement.
The assessments for the stations of the ‘Leybucht’ from the BEQI and M-AMBI were different:
the ecological status of these stations was assessed as ‘moderate’ by the BEQI and as ‘good’ and
‘moderate’ by the M-AMBI. The BEQI method was in full agreement with expert judgement, which
assessed the area as ‘moderate’ ecological status.
1.0
0.8
M-AMBI
0.6
0.4
poor
mod.
good
high
REGION
Eems Dollard
bad
0.2
North Sea
0.0
0.0
bad
0.2
poor
0.4
mod.
0.6
good
0.8
high
1.0
Wadden Sea
BEQI
Figure 2.4.15:
M-AMBI scores (Spanish pre-optimisation boundaries) plotted against
BEQI scores from different water bodies. Quality boundaries are included. Points in
the shaded regions have the same status for both indices. The black line is the x = y
relationship. The blue line is the linear relationship between M-AMBI and BEQI.
97
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0100.png
Compilation of the assessment results for different assessment methods for the German
intercalibration stations, blue = ’high status’(‘I’), green = ’good status’(‘II’), yellow = ’moderate
status’(‘III’) and orange = ’poor status’(‘IV’) ecological as shown in table 2.4.18 below.
The boundaries (H/G and G/M), that were proposed by Borja et al. (2007) would result in much
too high ecological assessments for the German intercalibration sites, since in comparison to
historical papers (eg. Hagmeier & Kändler 1927, Hagmeier 1925, Linke 1939) - at all stations
more or less changes in species composition and numbers were detected. Therefore the boundaries
(H/G and G/M) had to be adjusted, which was done by expert judgement under consideration of the
Normative Definitions of the WFD. The assessment results with the adjusted German boundaries
resulted in assessments, which were, at all stations, in better agreement with expert judgement and
in most cases in agreement with the other assessment methods.
Table 2.4.18:
DE assessments in type 3/4.
Waterbody type Station
Sediment
Expert
judgement
M-AMBI
Assessment,
DE boundaries
0.75
BEQI
Assessment,
NL boundaries
0.63
NEA ¾
Norderney,
Mean T1 to T8,
1998 to 2003
Middle littoral
muddy sand
‘II’
NEA ¾
Leybucht,
mean autumn
High littoral mud
2003, mud stations
‘III’
0.71
0.51
BE
The comparability between the BEQI (Benthic Ecosystem Quality Index) and the other benthic
evaluation methods is tested on a temporal dataset (1995-2003) at a slightly organically enriched
station in the Belgian coastal waters. The BEQI evaluates the benthic status on habitat/water body
level (by directly pooling the samples), whereas for the other benthic evaluation methods, the
average of the EQR (Ecologocial Quality Ratio) of the samples within the habitat/water body is
used as the habitat/water body assessment. This way of habitat/water body level assessment has
been accepted at the NEA-GIG benthos workshop in Lisbon (February 2007) in anticipation of
the final acceptance of the habitat/water body level assessment methods, which are currently in
development in other European countries. For an appropriate comparison, only the third level of the
BEQI method is applied. The parameter biomass could not be included due to lack of sufficient data.
This analysis delivers 22 assessment cases (based on 200 samples) for testing the comparability
between the different methods at habitat/water body level. Through the use of the average of the
EQR’s of the samples as habitat/water body assessment, the comparison best takes into account a
standard deviation, in order not to consider variability of the result within sample-based methods
as a real difference with another methodology. Therefore, the average of the standard deviations
of the sample EQR’s of each method has been taken into account. A small deviation from the class
boundaries less than 0.05 (IQI, NQI) or 0.06 (DKI, m-AMBI, PT) EQR units, is not considered as a
real misclassification. The same principle is applied in other GIG’s (Alpine and central-Baltic GIG)
for the intercalibration, but they arbitrarily selected 0.05 EQR units. The results of the comparison
between the BEQI and the other methods are shown in table 2.4.19.
Agreement between boundaries of the BEQI and the other methods was further assessed according
to guidelines issued by ECOSTAT to compare the agreement in classes, both for all five classes
98
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0101.png
and for only three classes (being (1) high, (2) good and (3) moderate or worse). In addition, a
comparison was also made between allowing a 0.05 EQR range of deviation around the boundaries
and not allowing any deviation. The results of the comparison between the BEQI index (whole
water body assessment) used in BE and NL and the other national methods are presented in figures
2.4.16, 2.4.17 and tables 2.4.20, 2.4.21, 2.4.22.
Table 2.4.19:
The percentage of cases in which a certain class difference is found between the BEQI and the other
methods (a deviation of 0,05 taken into account).
class difference
-4
-3
-2
-1
0
1
2
3
4
equal or 1 class
IQI
0
4,5
22,7
50,0
22,7
0
0
0
0
72,7
DKI
0
18,2
40,9
31,8
9,1
0
0
0
0
40,9
m-AMBI
0
0
9,1
50,0
40,9
0
0
0
0
90,9
NKI
0
0
9,1
45,5
45,5
0
0
0
0
90,9
PT
0
0
4,5
45,5
50,0
0
0
0
0
95,5
Total
0
4,5
17,3
44,5
33,6
0
0
0
0
78,2
1,0
0,9
0,8
0,7
0,6
0,5
0,4
0,3
0,2
0,1
0,0
1995/2
1995/3
1995/4
1995/1
1996/1
1996/2
1996/3
1997/1
1997/2
1997/3
1997/4
1999/1
1999/2
1999/3
1999/4
2002/1
2002/2
2002/3
2002/4
2003/1
2003/2
2003/3
High
Good
Moderate
Poor
Bad
BEQI
IQI
DKI
m-AMBI
NKI
PT
Figure 2.4.16:
Comparison of the results from the BEQI and the other national methods (no standard deviation of 0,05
EQR taken into account) for 22 assessments of one site that is subject to some organic enrichment in the Belgian coastal
waters. One assessment is a cluster of a series of samples per season (indicated by number 1 to 4).
1,0
0,9
99
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0102.png
1,0
0,9
0,8
0,7
0,6
0,5
0,4
0,3
0,2
0,1
0,0
1995/2
1995/3
1995/4
1995/1
1996/1
1996/2
1996/3
1997/1
1997/2
1997/3
1997/4
1999/1
1999/2
1999/3
1999/4
2002/1
2002/2
2002/3
2002/4
2003/1
2003/2
2003/3
High
Good
Moderate
Poor
Bad
BEQI
IQI
DKI
m-AMBI
NKI
PT
Figure 2.4.17:
Comparison of the results from the BEQI and the other national methods (standard deviation of 0,05
EQR taken into account) for 22 assessments of one site that is subject to some organic enrichment in the Belgian coastal
waters. One assessment is a cluster of a series of samples per season (indicated by number 1 to 4).
Table 2.4.20:
Percentage of classification agreement between the BEQI and the other benthic methods for the four
different comparison options.
3 classes
no deviation
0 (40,9)
0 (86,4)
31,8 (4,5)
36,4 (22,7)
54,5 (0,0)
24,5 (30,9)
5 classes
no deviation
0,0
0,0
13,6
18,2
18,2
10,0
3 classes
0,05 EQR deviation
36,4 (13,6)
9,1 (45,5)
63,6 (0,0)
68,2 (0,0)
77,3 (0,0)
50,9 (11,8)
5 classes
0,05 EQR deviation
22,7
9,1
40,9
45,5
50,0
33,6
IQI
DKI*
m-AMBI
NKI
PT
Total
*DKI comparability results were very low; more investigation is needed to clarify this and meanwhile caution is needed
by the interpretation of this result.
For the option of using three classes, the percentage of classifications that differs 2 classes is given between brackets.
Table 2.4.21:
Average of class difference between the BEQI and the other benthic methods.
3 classes
no deviation
-1,4
-1,9
-0,7
-0,9
-0,5
-1,1
5 classes
no deviation
-1,8
-2,3
-1,1
-1,3
-0,9
-1,5
3 classes
0,05 EQR deviation
-0,8
-1,4
-0,4
-0,3
-0,2
-0,6
5 classes
0,05 EQR deviation
-1,1
-1,7
-0,7
-0,6
-0,5
-0,9
IQI
DKI*
m-AMBI
NKI
PT
Total
100
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0103.png
Table 2.4.22:
Absolute average of class difference between BEQI and other benthic methods.
3 classes
no deviation
1,4
1,9
0,9
5 classes
no deviation
1,8
2,3
1,3
0,9
3 classes
0,05 EQR deviation
0,8
1,4
0,3
5 classes
0,05 EQR deviation
1,1
1,7
0,7
0,6
m-AMBI
NKI
PT
DKI*
IQI
0,7
1,1
0,4
Total
1,1
0,5
1,5
0,6
0,2
0,5
0,9
Figures 2.4.16 and 2.4.17 illustrate that all the methods show similar trends, but still significant
differences remain. The conclusion is that, for the current state-of-the-art, the BEQI is generally
one class more stringent than the other methods. The PT shows the highest comparability with the
BEQI, followed by the NKI and the m-AMBI. The lowest comparability is observed with the IQI
and especially the DKI, with which a classification difference of about 2 classes is observed.
The reasons for the deviations can be described as follows:
• different approaches are applied within the BEQI and the other methods concerning
statistical confidence (requirements sufficiently sampled area) and water body level
assessment versus sample level assessment
• the multimetric set-up of the BEQI includes community species composition, which is more
sensitive than a classification based on assigning species to specific categories (sensitive,
tolerant etc.)
• the multipressure set-up of the BEQI detects the impact of a combination of pressures
on the benthos, not only eutrophication or enrichment with organic matter or specific
substances, but also physical disturbances or impacts of invasive species
• significantly higher densities in relation to the reference are considered as a negative impact
within the BEQI and contribute to an assessment result with a lower classification
• the amount of reference data used within the BEQI is still very limited and not ideal yet.
This has an influence on the reliable estimation of the natural reference situation and the
derived statistical boundaries. Further elaboration of the reference dataset and improvement
of the method is necessary and will enhance comparability. However, this cannot be done
on the very short time and will be performed as part of the monitoring contracts.
The comparability at the level of the habitat/water body will be improved in the second
intercalibration round, when every country has its definitive assessment method at habitat/water
body level.
Class Boundaries – Member States
(in summary)
Type NEA1/26
DE, DK, ES, FR, IE, NO, PT, UK
Despite differences in index construction, classifications (M-AMBI, DKI, IQI, NQI and P-BAT) for
DK, ES, IE, NO, PT, and UK behave in a similar way. Class boundaries were harmonised through
the Intercalibration exercise and status classifications are in good agreement between the Member
States.
101
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
NL/BE
As the BEQI method evaluates directly at the water body level, and in the current intercalibration
exercise no water body intercalibration was done, an adjustment of class boundaries is not
considered appropriate. Adjustment of the class boundaries in the harmonisation of ecological
status classification of the water body will affect the statistical significance levels for the individual
parameters.
Status classification between Belgium and the Netherlands is in full agreement as the BEQI is used
by both Member States.
The assessment results comparing the BEQI (BE, NL) with M-AMBI, DKI, IQI, and NQI (DK, ES,
FR, IE, NO, UK) clearly result from the differences in methods (as described earlier).
Type - NEA3/4
Adjustment of the class boundaries for the German m-AMBI, as a result of the comparison with the
BEQI, is fully consistent for larger (5 year) datasets and sufficient using a one year comparison and
expert judgement.
Status classification between DE and NL is in good agreement.
Type - NEA8/9/10
Despite differences in index construction and species classifications indices behave in a similar way
and status classifications are in good agreement between Sweden, Denmark and Norway.
Type - NEA11
As described above the boundaries agreed for coastal waters could be used in transitional waters in
areas with a similar salinity regime. However, as NEA11 intercalibration will be postponed until
Phase II, class boundaries for habitats within varying salinity regimes are not available at this time.
2.4.5 Results of the harmonisation – Boundary EQR values
Type - NEA1/26 and NEA7 (EQRs) (Site level assessment, soft sediment habitat)
Good/Moderate
Denmark
0.53
France
0.53
Germany
0.70
Ireland
0.64
Spain
0.53
Portugal
0.58
Norway
0.81
United Kingdom
0.64
High/Good
0.67
0.77
0.85
0.75
0.77
0.79
0.92
0.75
Type - NEA1/26 (EQRs) (Waterbody level assessment, all habitats)
Belgium
Netherlands
102
0.60
0.60
0.80
0.80
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
Type - NEA3/4 (EQRs)
Germany
Netherlands
Good/Moderate
0.70
0.60
High/Good
0.85
0.80
Type - NEA8/9/10 (EQRs)
Denmark
Norway
Sweden
Good/Moderate
0.63
0.81
0.68
High/Good
0.82
0.92
0.89
Type - NEA11 (EQRs)
Boundaries cannot be agreed by June 2007. Boundaries are to be completed in Phase II of the
Intercalibration process.
2.4.6 Open issues and need for further work
There are a high number of issues remaining to be resolved in Phase II so clear prioritization must
take place.
The intercalibration process to date has concentrated on the assessment of classification methods for
the subtidal soft sediment infauna within waterbodies. Classifications relate to general disturbance
pressures (hazardous substances, organic enrichment, eutrophication, etc) but not necessarily to
physical disturbance (method dependent). (Data used by Member States, however, may relate to
multi-pressure environments where physical disturbance may play a role.) Hydromorphological
pressures may require a different assessment methodology which takes more account of spatial
variation (e.g. habitat extent tools). May also need to revisit the coastal waters types to see if any
subdivision is necessary.
The key work area in the next phase will be to develop intercalibration of assessment methods that
are applicable in the polyhaline, mesohaline and oligohaline habitats of transitional waters. Both
intertidal and subtidal habitats need to be considered.
The majority of developed classifications for the coastal soft sediment infauna, deal with
samples grouped within specifically defined habitat types. There is now a general need to assess
classifications at a water body level, incorporating sampling design and risk of misclassification.
The Dutch/Belgian methodology (BEQI) has made a start with such an approach.
There is also a need to further investigate the intercalibration of methods for other habitats (e.g.
maerl, mobile sands, zostera beds, gravels) where they will be sampled by Member States.
Reference conditions and methods are still under development and would need to be assessed in a
further stage of the Intercalibration process.
The development of reference species lists for all habitats to be considered and agreement on
the designation of sensitive species, the role of invasive species in classification, and community
changes in habitats where there is little occurrence of “sensitive” species is required.
103
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0106.png
Methods applicable for hard substratum and subtidal epibenthos assessments are yet to be
developed.
Results of ongoing FP6 research could not be taken fully into account in this phase of
Intercalibration but should be addressed in starting a second phase of Intercalibration. Better
communication is required between the scientists involved in FP7 research related to WFD and the
people involved in the application of the science (according to the CIS process). This requires new
ways of interactive bridging, which should be made transparent in programming both processes.
All future work is dependent on adequate funding being made available by countries within the
GIG. We should also consider using the FP7 framework if possible.
3 Discussion
3.1 Comparability between GIGs
Table 2.4.23:
Summary of intercalibration topics per GIGs.
NEA GIG
INDICES/Metrics
N
O
. TYPES
OPTION
BEQI, M-AMBI, DKI,
IQI, NQI, P-BAT
7
3
MED GIG
BENTIX, Multimetric
approach, AMBI, M-AMBI,
MEDOCC
4 not relevant
3
BAL GIG
Shannon, AMBI,
M-AMBI
1
3
percentage from
average reference
values; ecological status
boundaries equally scaled
(80 % for good, 60 % for
moderate, 40 % for poor
and 20 % for bad status),
H/G, G/M.
Expert judgement,
unaffected areas or good
status sites
Refine typology
BOUNDARY
SETTING
PROCEDURE
Equal classes revised by
expert judgement
Expert judgement on
indices values,
Percentage from reference
sites
REFERENCE
CONDITION
Expert judgement on
indices values,
Expert judgement on
indices values,
Percentage from reference
sites
consider different sub-
regions, introduce new
multimetric indices
OPEN ISSUES
Refine typology, relate to
pressure,
Metrics
Intercalibration for the benthic invertebrates quality element has been successfully performed by
all of the four Coastal Waters GIGs. AMBI and its multivariate version (M-AMBI) results to be the
favourite experts choice (or at least the more common), together with Shannon index (in some cases
aggregated with other indices) and a few highly specific national methods.
104
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
Types
Most of the types result to be covered by the Intercalibration results, even though there is a common
view (e.g. MEDGIG) stating that the a-priori types are definitely not relevant for the quality element
ecological assessment.
Option
Option 3 has been chosen by all GIG, thus Member States have developed their own classification
schemes, these have been assessed against each other through the exchange and evaluation of data
and the boundaries in each scheme harmonized to give an acceptable level of agreement.
Boundary Setting Procedure
All Coastal Waters GIGs are providing results for the H/G and G/M boundaries (at least), applying
similar approaches based on the expert supervised selection of appropriate percentiles as boundaries
best fitting with evident changes in ecological status.
Reference Condition
The expert judgment has been selected as the favorite option in setting Reference Condition.
3.2 Open issues and need for further work
To refine typology going down to a level of details that include the sub-ecoregion in the process has
been highlighted as a major task for the next phase.
4 References
Anger, K. 1975. On the influence of sewage pollution on inshore benthic communities in the south of Kiel Bay.
Part 1. Qualitative studies on indicator species and communities. Merentutkimuslait. Julk./Havsforskningsinst.
Skr. 239:116-122.
Anger, K. 1977. Benthic invertebrates as indicators of organic pollution in the Western Baltic Sea. Int. Revue
ges. Hydrobiol. 62:245-254.
Bäck, S., Kauppila, P., Kangas, P., Ruuskanen, A., Westberg, V., Perus, J.
& Räike, A. 2006. A biological
monitoring programme for the coastal waters of Finland according to the EU Water Framework Directive.
Environmental research, engineering and management 4 (38): 6-11 (ISSN 1392-1649)
Blomqvist, M., Cederwall, H., Leonardsson, K. & Rosenberg, R. 2007. Bedömningsgrunder för kust och
hav. Bentiska evertebrater 2006. Rapport till Naturvårdsverket 2007-04-11. 70 pp. (in Swedish with English
summary)
Borja A., Muxika I., Franco J. (2003). The application of a Marine Biotic Index to different impact sources
affecting soft-bottom benthic communities along European coasts.
Mar. Pollut. Bull.
46 : 835 – 845
Borja, A., A. B. Josefson, A. Miles, I. Muxika,, F. Olsgard, G. Phillips, J. G. Rodríguez and B. Rygg. 2007.
An approach to the intercalibration of benthic ecological status assessment in the North Atlantic ecoregion,
according to the European Water Framework Directive.
Mar. Poll. Bull.,
55: 42-52.
Borja, A., Franco, J., Muxika, I., 2003. Classification tools for marine ecological quality assessment: the
usefulness of macrobenthic communities in an area affected by a submarine outfall. ICES CM 2003/Session
J-02, Tallinn, Estonia, 24–28 September.
Borja, A., J. Franco & V. Pérez, 2000. A marine biotic index to establish the ecological quality of soft bottom
benthos within European estuarine and coastal environments, Marine Pollution Bulletin, 40(12): 1100-1114.
105
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
Carvalho, S., M. Barata, F. Pereira, M. B. Gaspar, L. C. da Fonseca, and P. Pousão-Ferreira. 2006. Distribution
patterns of macrobenthic species in relation to organic enrichment within aquaculture earthen ponds.
Mar. Poll.
Bull.,
52: 1573-1584.
Clarke, K. R., R. M. Warwick, 1994. Change in marine communities: an approach to statistical analysis and
interpretation. Plymouth, Plymouth Marine Laboratory, 144 pp.
Common implementation strategy for the Water Framework Directive (2000/60/EC). Guidance Document
№ 13, Overall Approach to the Classification of Ecological Status and Ecological Potential, 2005, EC, 53 pp.
Common implementation strategy for the Water Framework Directive (2000/60/EC). Guidance Document
№ 5, Transitional and Coastal Waters Typology, Reference Conditions and Classification Systems, 2003, EC,
116 pp.
Dauvin, J-C. 2007. Paradox of estuarine quality: Benthic indicators and indices, consensus or debate for the
future.
Mar. Poll. Bull.,
55: 271-281.
Diaz, R. J. & Rosenberg, R. 1995. Marine benthic hypoxia: a review of its ecological effects and the behaviour
responses of benthic macrofauna. Oceanogr. Mar. Biol. Ann. Rev. 33: 245-303.
Directive 2000/60/EC of the European Parliament and of the Council of 23 October 2000 establishing a
framework for Community action in the field of water policy, 72 pp.
Fauvel, P., 1923. Polychètes errantes. Faune de France, Vol. 5, Paris.
Fauvel, P., 1927. Polychètes sédenteires. Faune de France, Vol. 16, Paris.
Glémarec, M. and C. Hily. 1981. Perturbations apportées à la macrofaune benthique de la baie de Concarneau.
Acta Œcologica,
2: 139-150.
Grall, J. and M. Glémarec. 1997. Using biotic indices to estimate macrobenthic community perturbations in
the Bay of Brest.
Estuar. Coast. Shelf Sci.,
44: 43-53.
Gray, J. S., Shiu-Sun Wu, R. & Ying Or, Y. 2002. Effects of hypoxia and organic enrichment on the coastal
marine environment. Mar. Ecol. Prog. Ser. 238: 249-279.
Hansson, M. & Håkansson, B. 2004. Indelning av Svenska Övergångs- & Kustvatten i typer enligt Ramdirektivet
för Vatten. Dnr: 2002/1796/1933 SMHI (http://www.vattenportalen.se/
Helawell, J.M. 1986. Biological Indicators of Freshwater Pollution and Environmental Management. Pollution
Monitoring Series. Elsevier Appl. Sci. Publ., London.
Heyer, K. 2007; German Intercalibration Report for the NEA-GIG (May 2007) ‘ Assessment of German
Coastal Waters (NEA1/26, NEA3/4) by benthic invertebrates with the ‘M-AMBI’ assessment method’ for the
Intercalibration Report- Coastal GIGs for the 6th milestone report for CIS
Hily, C. 1984. Variabilité de la macrofaune benthique dans les milieux hypertrophiques de la Rade de Brest.
Ph
D thesis,
Úniversité Bretagne Occidentale. Vol 1, 359 pp; Vol 2, 337 pp.
Järvekülg, A. 1970. Bottom fauna as an indicator of pollution of the marine benthal in the vicinity of Tallinn.
Estonian Contribution to the International Biological Programme 1: 158-193.
Kangas, P., Byholm, L., Stigzelius, J. 2001: Changes in zoobenthos communities. In: Kauppila, P. & Bäck, S.
(eds.). The state of Finnish coastal waters in the 1990s. Helsinki, Finnish Environment Institute. The Finnish
Environment ; 472: 79-88.
Kangas, P., Bäck, S. & Kauppila, P.(eds) 2003. Suggestions for a Typology of Coastal Waters for the Finnish
Coast According to the European Union Water Framework Directive (2000/60/EC). Mimeograph series of
Finnish Environment Institute 284, 51 p. (in Finnish with an English summary)
Karlson, K., Rosenberg, R. & Bonsdorff, E. 2002. Temporal and spatial large-scale effects of eutrophication
and oxygen efficiency on benthic fauna in Scandinavian and Baltic waters – a review. Oceanogr. Mar. Biol.
Ann. Rev. 40: 427–489.
Labrune C., 2006. Utilisation de la macrofaune benthique en tant qu’indicateur des changements
environnementaux. PhD, Université de Perpignan: 1-250 + annexes.
Labrune C., Amouroux J.M., Sarda R., Dutrieux E., Thorin S., Rosenberg R., Gremare A., 2006. Characterization
of the ecological quality of the coastal Gulf of Lions (NW Mediterranean). A comparative approach based on
three biotic indices.
Mar. Pollut. Bull.
52: 34-47.
106
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
Landner, L., Nilsson, K. och Rosenberg, R. 1977. Assessment of industrial pollution by means of benthic
macrofauna surveys along the Swedish Baltic coast. Vatten 3:324-379.
Leppäkoski, E. 1975. Assessment of degree of pollution on the basis of macrozoobenthos in marine brackish-
water environments. Acta Academiae Aboensis. 35: 1-96.
Licari M.L, 1998. Mise au point d’un système d’aide à l’interprétation des données benthiques en milieu marin
et lagunaire. PhD, Ecole Pratique des hautes Etudes : 1-280 + annexes.
Lipej, L., P. Mozetič, M. Orlando Bonaca, B. Mavrič, M. Šiško & N. Bettoso (2006). Opredelitev ekološkega
stanja morja v skladu z Vodno direktivo (Water Framework Directive, 2000/60/ EC). Poročila 85. Morska
Biološka Postaja, Nacionalni Inštitut za Biologijo, Piran, 180 str.
Majeed, S. A. 1987. Organic matter and biotic indices on the beaches of North Brittany.
Mar. Poll. Bull.,
18:
490-495.
Mandaville, S.M. 2002. Benhic Macroinvertebrates in Freshwaters – Taxa Tolerance Values, Metrics, and
Protocols. Project H-1, Soil and Water Conservation Society of Metro Halifax. http://chebucto.ca/Science/
SWCS/SWCS.html, 48 s.
Marinov, T., 1977. Fauna of Bulgaria, 6. Polychaeta. Sofia, Publishing house of the Bulgarian Academy of
Sciences, 258 pp. (In Bulgarian).
Morduhay-Boltovskoy, M. D., (Ed.), 1968. A key to Black Sea and Azov Sea fauna, vol. I, Kiev, Naukova
Dumka, 437 pp. (In Russian).
Morduhay-Boltovskoy, M. D., (Ed.), 1969. A key to Black Sea and Azov Sea fauna, vol. III, Kiev, Naukova
Dumka, 340 pp. (In Russian).
Morduhay-Boltovskoy, M. D., (Ed.), 1972. A key to Black Sea and Azov Sea fauna, vol. II, Kiev, Naukova
Dumka, 536 pp. (In Russian).
Muxika I., Borja A., Bald J., 2007. Using historical data, expert judgement and multivariate analysis in assessing
reference conditions and benthic ecological status, according to the European water framework Directive.
Mar.
Poll. Bull.,
55: 16-29.
Muxika, I., Borja, Á., Bonne, W., 2005. The suitability of the marine biotic index (AMBI) to new impact
sources along European coasts. Ecological Indicators, 5: 19-31.
NFS 2006:11 http://www.naturvardsverket.se/sv/Lagar-och-andra-styrmedel/Lag-och-ratt/Foreskrifter-och-
allmanna-rad/Foreskrifter-utgivningsordning/#2006
Pearson, T.H. and Rosenberg, R., 1978. Macrobenthic succession in relation to organic enrichment and pollution
of the marine environment. Oceanogr. Mar. Biol. Ann. Rev. 16: 229–311.
Perus, J., Bäck, S., Lax, H-G., Westberg, V., Kauppila, P. & Bonsdorff, E. 2004. Coastal marine zoobenthos as
an ecological quality element: a test of environmental typology and the European Water Framework Directive.
Coastline Reports 4: 27-38.
Perus, J., Bonsdorff, E., Bäck, S., Lax, H-G., Westberg, V. Villnäs, A. 2007. Zoobenthos as indicator of ecological
status in coastal brackish waters: A comparative study from the Baltic Sea. Ambio 36 (2-3): 250-256.
Pinedo, S. 1998. Structure and dynamics of Western Mediterranean soft-bottom communities along a
disturbance gradient. Natural and man-induced variability in the Bay of Blanes.
Ph D thesis,
Universidad de
Barcelona, Spain. 177 pp.
Rosenberg, R., Blomqvist, M., Nilsson, C.H., Cederwall, H., Dimming, A., 2004. Marine quality assessment
by use of benthic species-abundance distributions; a proposed new protocol within the European Union Water
Framework Directive. Mar. Poll. Bull. 49, 728-73
Shannon, C. E. & W. Weaver, 1963. The mathematical theory of communication, University Illinois Press,
Urbana, 117 pp.
Simboura N, Panayotidis P, Papathanassiou E., 2005. A synthesis of the Biological Quality Elements for the
implementation of the European Water Framework Directive in the Mediterranean Ecoregion: the case of
Saronikos Gulf.
Ecol. Indicators,
5: 253-266.
Simboura, N. and A. Zenetos. 2002. Benthic indicators to use in ecological quality classification of Mediterranean
soft bottom marine ecosystems, including a new Biotic index.
Med. Mar. Science,
3(2):77-111.
107
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
Simboura, N., E. Papathanassiou & D. Sakellariou, 2007.The use of a biotic index (Bentix) in assessing long
term effects of dumping coarse metalliferous waste on soft bottom benthic communities. /Ecological Indicators,
7(1): 164-180.
Todorova V. & Konsulova T., 2005. Manual for collection and treatment of soft bottom macrozoobenthos
samples. Online: http://bsc.ath.cx/documents/ExpertNetwork/docs/
Expert %20Network %20- %20Zoobenthos/Todorova %20Manual_zoobenthos.doc
Vuori, K-M., Bäck, S., Helssten, S., Karjalainen, S. M., Kauppila, P., Lax, H-G., Lepistö, L., Londesborough, S.,
Mitikka, S., Niemelä, P., Niemi, J., Perus, J., Pietiläinen, O-P., Pilke, A., Riihimäki, J., Rissanen, J., Tammi, J.,
Tolonen, K., Vehanen, T., Vuoristo, H., Westberg, V. 2006. The basis for typology and ecological classification
of water bodies in Finland. The Finnish Environment. 807. 151 pp. (in Finnish with English summary)
Wallin M., Wiederholm T., Johnson R.K., 2003. Guidance on establishing reference condition and ecological
class boundaries for inland surface waters. Produced by CIS Working Group 2.3 – REFCOND: 86 pp.
Wiederholm, T. 1973. Bottom fauna as an indicator of water quality in Sweden´s large lakes. AMBIO 2:107-
110.Warwick R (1986) A new method for detecting pollution effects on marine macrobenthic communities.
Mar.Biol. 92: 557-562.
5 Annexes
Annexes can be downloaded from the following address:
http://circa.europa.eu/Public/irc/jrc/jrc_eewai/library?l=/intercalibration_2&vm=detailed&sb=Title
Section 2 – Benthic Invertebrates – Overview of Annexes
A – Baltic GIG
Annex 2.1: Comparison of the Finnish and Swedish macroinvertebrate classification systems using Finnish coastal data
from the Baltic Sea
108
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0111.png
Section 3 – Phytoplankton
1 Introduction
All four coastal water GIGs have been able to produce results for the phytoplankton quality element.
NE Atlantic, Baltic and Mediterranean GIG have intercalibrated single metrics based on chlorophyll
a
concentration, while Romania and Bulgaria in the Black Sea GIG agreed on the phytoplankton
boundaries based on seasonal average biomass (mg/m
3
). Only NEA GIG intercalibrated also other
metrics (besides chlorophyll a) indicative for phytoplankton composition. Those were Indicator
Taxa (Frequency of Phaeocystis Cell counts) and Taxa Cell Counts (Frequency of phytoplankton
taxa cells counts).
All the GIGs have chosen option 2 or a hybrid between option 2 and 3 in this phase of the
intercalibration process. Member States have developed metrics as part of their own classification
schemes. All Selected metrics have been chosen by each GIG for assessment and agreement of
boundaries in this phase of the intercalibration process. At this stage not all the metrics that make up
Member States’ schemes can be intercalibrated, so it is not possible to produce EQRs for the whole
quality element. Boundaries have been agreed for the selected metrics.
2 Methodology and results
2.1 Baltic GIG
2.1.1 Intercalibration approach
The Baltic Sea Geographical Intercalibration Group (GIG) includes the whole or parts of the
coastline of the following countries: Germany, Denmark, Estonia, Finland, Latvia, Lithuania, Poland
and Sweden (Table 2.1.1).
The common coastal water types are characterised by the descriptors of the System B typology.
The typology factors are based on the common typology framework presented in the guidance
on the typology for the coastal and transitional waters
20
. In the Baltic Sea GIG, the common
intercalibration types were characterized using basic salinity and exposure with further delineation
based on depth and number of ice cover days (Table 2.1.1). One transitional water type (TW B 13)
was identified. All countries agreed to focus the intercalibration on the quality elements that are
sensitive to eutrophication pressures.
The common intercalibration types were characterized by the following descriptors:
– Salinity (using practical salinity scale): low (0,5-3) and high (3-6) oligohaline, mesohaline (6-22)
– Depth: all shallow (<30 m)
20
Guidance document No. 5 ‘Transitional and Coastal Waters - Typology, Reference conditions, and Classification
systems’. Common Implementation Strategy of the Water Framework Directive, Available at: http://forum.europa.eu.int/
Public/irc/env/wfd/library
109
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0112.png
– Exposure (using agreed Pan-European scale
21
):
exposed, sheltered and very sheltered
– Duration of ice cover: >150 days/ year, 90-150 days/ year, no or very short ice cover
For phytoplankton a hybrid between option 2 and 3 was used. A common metric – summer mean
of chlorophyll
a
from May/June to September
22
- was agreed. Intercalibration was performed by
comparison of the results of the national assessments. Within each type member states essentially
agreed on common reference and classifications values for the whole type or for sub areas in the type.
Details of the national methods, reference condition and boundary setting, as well as references to
earlier literature and reports within each country, are available in Annex 3.1-3.7.
Table 2.1.1:
Description of Baltic Sea Common intercalibration types that have included in the intercalibration exercise.
Type
CW B0
CW B2
CW B3 a
CW B3 b
CW
B12 a
Eastern
Baltic Sea
CW
B12 b
Western
Baltic Sea
CW B13
Salinity psu
0.5- 3
3-6
3-6
3-6
5-8
Exposure
Sheltered
Sheltered
Sheltered
Exposed
Sheltered
Depth
Shallow
Shallow
Shallow
Shallow
Shallow
Ice days
> 150
> 150
90-150
90 -150
-
Other Characteristics
Sites in Botnian Bay (Northern Quark)
Sites in Bothnian Sea
Sites in the area extending from the southern
Bothinian Sea to the Archipelago Sea and
the western Gulf of Finland
Sites in the Gulf of Riga
8 - 22
Sheltered
Shallow
-
Sites at the Southern Swedish coast and the
South western Baltic Sea open coast along
Denmark and Germany
Sites along the coast of the Estonia, Latvia
and Lithuania, the Polish coast and the
Danish island “Bornholm”
Lagoons
Transitional water. Sites along the coast of
Lithuania and Poland
6-22
Exposed
Shallow
-
CW B 14
TW B 13
6-22
6-22
Sheltered
Exposed
Shallow
Shallow
-
Countries sharing types that have been intercalibrated:
Types CWB0, CWB2, CWB3a, CWB3b:
Finland, Sweden.
Type CWB12a:
Estonia
Type CWB12b:
Germany, Denmark, Sweden.
Type CWB13:
Denmark, Estonia, Lithuania, Latvia, Poland.
Type CWB14:
Denmark, Poland
Type TWB13:
Lithuania, Poland.
21
22
According to the definitions of the common European exposure categories; Guidance document No. 5
The definition of the length of the summer period depends on the latitude and area of the Baltic Sea.
110
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
2.1.2 National methods that were intercalibrated
In the Baltic Sea GIG average summer chlorophyll
a
concentration was chosen as the metric for
the intercalibration of the phytoplankton element. All countries used the monitoring data available
in their national data bases to set the reference conditions for their national types. At the time of
the intercalibration, final national metrics were not yet available. Chlorophyll
a
values for summer
growth season were averaged to obtain reference conditions and boundary values for seasonally
defined phytoplankton metric. The period of aggregation varied depending of the latitude. Finland
and Sweden used average chlorophyll
a
values from data sampled between June and August for
the types in the Northern Baltic Sea. For the types in the Baltic Proper (B13) and Gulf of Riga
(B12) Estonia used growth period between June-September to calculate average values. In the
Southern Baltic Sea, Denmark and Poland used values averaged over May-September. In Germany
chlorophyll
a
values were originally averaged for the period between March-October in their
national reference conditions setting.
Further information on national methods and assessments can be found in Annexes 3.1-3.7.
2.1.3 Reference conditions and boundary setting:
Denmark
Two methods have been used to establish reference conditions for phytoplankton biomass expressed
as chlorophyll
a
in Danish waters:
1. Development of reference conditions using historical Secchi depth measurements and
relationships between Secchi depth and chlorophyll
a
obtained from recent monitoring
data from Danish coastal waters (more detailed presentation of the method and procedure
is presented in the Annex 3.1) Reference conditions for chlorophyll
a
are calculated as
predictions from the relationships corresponding to an average secchi depth around the
beginning of the 20th century.
2. Reference conditions were estimated from a combination of 1) hind-casted nutrient inputs
(loading of total nitrogen) to the Danish straits based on estimates of the nitrogen surplus
from Danish agriculture and estimated changes in point sources, 2) characterization (expert
judgment) of reference loading using the hind-casted estimates, and 3) historical nitrogen
inputs projected into total nitrogen (TN) levels and related to chlorophyll
a
levels in coastal
waters (see Annex 3.1 for detailed presentation)
Two different approaches were used to set the chlorophyll
a
boundaries between good and moderate
ecological status for the selected Danish intercalibration sites (described in detail in Annex 3.2):
1. Historical secchi depth observations are compared to chlorophyll
a
-secchi depth relationships
established from recent data (as described above). Boundaries between good and moderate
status for chlorophyll
a
are then defined as reference conditions plus 50 % in accordance with
the HELCOM Eutro approach.
2. Relationships between nitrogen loading and total nitrogen (TN) as well as between TN and
chlorophyll
a
are established using recent monitoring data. Site-specific boundaries for TN
and chlorophyll
a
were predicted from modelled time series of nutrient inputs to the Danish
straits. The boundary values of nutrient inputs for different time periods were been selected
using expert judgement.
Comparing the results from the two approaches indicated that the use of historical nitrogen inputs
(method 2) generally resulted in lower reference conditions and boundaries than the use of historical
111
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0114.png
secchi depths (method 1). Confidence intervals for the historical secchi depth method are also wider,
suggesting that the use of historical nitrogen input is giving more precise and unbiased estimates of
reference conditions (Table 2.1.2).
Table 2.1.2:
Comparison of good - moderate boundary values of the summer mean chlorophyll
a
(µg l
-1
) from a number
of Danish coastal water bodies, based on two different approaches: 1) Historical secchi depths and chlorophyll
a
-
secchi relationships, and 2) historical nitrogen inputs projected into total nitrogen levels and subsequently to chlorophyll
a
levels. In the second approach median values are shown in order to obtain a comparable value to the geometric mean
of the first approach.
Area
Northern Kattegat
Fakse Bay
Hirtshals
Hjelm Bay
Århus Bay
Geometric
mean
2.28
2.06
2.32
2.42
3.66
3.76
2.79
2.55
2.40
1) Historical secchi depths
Lower
confidence
1.72
1.51
1.75
1.65
2.57
2.82
2.42
2.11
Upper
confidence
3.01
2.81
3.06
2.73
5.23
5.01
3.22
3.54
2) Historical nitrogen inputs
Median
1.69
1.62
1.53
1.84
1.31
6.13
2.25
1.55
1.31
6.10
2.84
Lower
confidence
1.45
1.04
0.86
1.60
1.01
5.57
1.97
1.05
0.65
5.25
2.62
Upper
confidence
1.96
2.34
2.37
2.09
1.68
6.74
2.56
2.17
2.55
7.11
3.07
West of Bornholm
North of Funen
Dybsø Fjord
Outer Wadden Sea
Northern Sound
Inner Wadden Sea
The recommended boundaries for the Danish intercalibration sites, based on the relationship
between total nitrogen and chlorophyll
a
(as method 2), are presented in Table 2.1.3. The good-
moderate boundaries are calculated as the range between the median value and the upper confidence
value from the Table 2.1.2. The high-good boundary and the reference conditions are established
by expert judgement based on historical nitrogen loading and corresponding TN and chlorophyll
a
relationships.
Table 2.1.3:
Recommended chlorophyll
a
boundaries (
arithmetic mean for summer values between May
–September,
as µg l
-1
) for the Danish intercalibration sites in the Baltic Sea.
Intercalibration sites
Fakse Bay
Hjelm Bay
Bornholm west
Dybsø Fjord
Reference conditions
1.4
1.2
0.9
High/good
boundary
1.5
1.3
1.1
Good/moderate
Boundary
1.7 – 2.6
1.5 – 1.9
1.6 – 3.1
Estonia
In Estonian waters reference conditions were obtained by setting empirical relationships between
nutrient concentrations, water transparency and phytoplankton parameters. The time series data
from the period 1993-2005 was used to assess average concentrations levels and trends for summer
112
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0115.png
period (between June to September) in each national coastal type. Reference conditions were set by
using frequency distributions and assuming that 20 % deviation from the lowest concentrations as
reference conditions. For the Secchi depth transparency, 20 % deviation from the maximum values
was proposed as reference conditions, respectively.
Expert judgement based on analysis of the monitoring data from the period 1993-2005 was used in
the boundary setting. The areas belonging to the Baltic GIG common type B13 have been monitored
only in 2005. The time series data were considered to assess average concentrations levels and
trends in each coastal type. The high-good and good-moderate boundary values for chlorophyll
a
were derived as 20, and 50 % deviation from the reference conditions (Table 2.1.4).
Table 2.1.4:
Recommendations for the reference conditions and classification boundaries for the Estonian coastal
waters belonging to the common intercalibration type B13. The boundary value for the high status class represents 20 %
deviation and the good-moderate boundary 50 % deviation from the reference conditions. The two lower boundaries are
derived as 70 % and 90 % deviation from the reference conditions.
Parameter
Total nitrogen
(annual mean)
Total phosphorus
(annual mean)
Secchi depth
Chlorophyll
a
Biomass
Reference
value
10.6
0.28
8
1.1
0.29
unit
µM
µM
m
mg m
-3
mg l
-1
% deviation from reference conditions
20 %
12.27
0.336
7.20
1.32
0.348
50 %
15.90
0.42
6.00
1.65
0.43
70 %
18.02
0.476
4.40
1.87
0.493
90 %
20.14
0.532
2.40
2.09
0.551
Finland
In Finnish coastal waters, the reference values of chlorophyll
a
were estimated based on three
methods: (1) empirical modeling using the relationships between chlorophyll
a
and Secchi depth,
(2) statistical analyses on frequency distribution data by using 5 % deviation from the lowest
values as the boundary between high and good status and (3) trend analyses of chlorophyll
a
in the
intercalibration sites or in the vicinity of them (detailed presentation in Annex 3.4).
Type specific reference conditions for phytoplankton have been estimated using empirical
relationships between concentrations of chlorophyll
a
and Secchi depth of the present monitoring
programs and the historical Secchi values of the early 1900s (Annex 3.4). The reference chlorophyll
a
concentrations represent mean values which were estimated from the empirical equations using
the historical Secchi values covering the Finnish intercalibration types. Following the Swedish
model, the 95 % confidence limits of variation were estimated to represent the boundary between
high and good at its maximum.
Scientific grounds are required to justify the boundaries between good and moderate status.
Therefore, changes in eutrophication, revealed by other ecological variables, were used to find out
if there is an ecologically reasonable base to estimate the chlorophyll boundary between good and
moderate status in the Finnish coastal types. The variables included were the depth limit of
Fucus
vesiculosus
and the occurrence of cyanobacteria in phytoplankton assemblages in relation to Secchi
depth and chlorophyll
a
concentrations.
113
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0116.png
Germany
In the German coastal waters, modelling was used to calculate loading of total nitrogen (TN) from
rivers (see Annex 3.7). The marine value of 10 µM TN was taken from the literature was considered
to be consistent with pristine coastal data. This value was also taken for the Baltic Sea due to the
long time period between pristine and recent conditions. References for all salinities were estimated
using mixing diagrams. For local calculations of references, recent mean salinity data and the
freshwater reference of the next eastern located river were used, assuming a residual current moving
westward along the surface of the German Baltic coastline. The reference chlorophyll
a
values were
calculated using regression between total nitrogen (TN) and chlorophyll
a
concentrations during the
growth season (months March- October).
Boundary values were calculated from summer values of recent monitoring data. The analysis
is based on 438 data sets sampled during 1978 and 2004 on 8 outer coastal sites along a salinity
gradient from 7 to 17 PSU, which represent the extended intercalibration sites. A significant
correlation between TN and chlorophyll
a
could not be found for these intercalibration sites.
Therefore the boundary setting was performed by linear division along the actual gradient of
chlorophyll
a
values. The 80 %-percentile (3.4 µg chlorophyll
a
/l) marked the boundary poor-bad.
The boundaries between good and moderate status as well as between moderate and poor status
represent the three equal parts between high and poor state of the water type.
Poland
Reference conditions were calculated on the basis of empirical relationships obtained from
contemporary monitoring data. Only data obtained in summer period (May-September) in
1999-2005 were taken into consideration – in total: 61 chlorophyll
a
concentration values with
corresponding water transparency values (SD) and TN concentration values (detailed description in
Annex 3.5).
Assuming that 6 m is a water transparency reference value for summer – estimated from historical
data, corresponding chlorophyll
a
concentrations values were calculated by means of linear
regression. Chlorophyll
a
reference values were also estimated on the basis of total nitrogen (TN)
concentration reference values determined for outer Puck Bay, using the method of extrapolation of
temporary trends. Results of reference value calculations obtained by different methods are shown
in the Table 2.1.5.
In the calculations of the relationship between chlorophyll
a
and TN, only summer nitrogen
concentrations were used because data from the winter period was incomplete. However, the
relationship between chlorophyll
a
and transparency was better, and therefore the chlorophyll
a
reference concentration for Kępa Redłowska intercalibration site was estimated by using the
chlorophyll
a
and water transparency regression. The obtained reference values (3.7 mg·m
-3
) was
Table 2.1.5:
Chlorophyll
a
(chl-a) reference values for Kępa Redłowska area assessed on the basis of relationship with
water transparency and total nitrogen (TN) concentrations.
Method of reference values determination
Linear correlation chl-a/ transparency
Exponential relationship chl-a/ transparency
Linear correlation chl-a/ TN
Exponential relationship chl-a/ TN
Chlorophyll
a
reference value
3.7
3.7
3.0
3.1
114
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0117.png
much higher than the reference value (1.2 mg·m
-3
) obtained for the Swedish coastal waters type
for the CWB13. However, the this value reflects the strong impact of the Vistula River – one of the
biggest rivers in Europe. Therefore a significant part of the Gulf of Gdańsk (inner Gulf of Gdańsk)
was assigned to be transitional waters having thus higher reference values for that region than for
other coastal waters belonging to the type B13.
Discontinuities, which could indicate class boundaries, in the chlorophyll
a
concentration/ water
transparency relationship were not found. Consequently, the ecological status classification was
based on statistic methods.
Values 25 % and 50 % were agreed to be the acceptable deviation from the reference conditions
(following recommendation of the HELCOM EUTRO project), that is, if the chlorophyll
a
concentration increases by 25 % or 50 % in relation to reference value, ecological state is still good.
With regard to considerable scatter of chlorophyll
a
concentration values, which indicates significant
natural variability of environmental conditions, the higher acceptable deviation from reference value –
50 % was adopted. In this way the boundary between good ecological status and moderate ecological
status (3.7 + 50 %) was obtained and it equals chlorophyll
a
concentration value – 5.6 mg·m
-3
.
The boundary between moderate and poor ecological status was established on the basis of statistical
analyses on the frequency distribution of data. This value was 7.3 mg·m
-3
, which is a median in
value range 6.0<chl-a<9.0 mg·m
-3
. Proposal of ecological status classes for Kępa Redłowska
intercallibration site is presented in Table 2.1.6.
Table 2.1.6:
Ecological classification for Kępa Redłowska intercallibration area.
High/
reference value
<3.7
Good
3.8 – 5.6
Moderate
5.6 – 7.3
Poor
7.3 - ?
Bad
?
Sweden
The reference values for total nitrogen (TN) were estimated using historical data on Secchi depth,
and by using empirical relations between nutrients and Secchi depth based on current data. The
modern relationships between nutrients and phytoplankton biovolume and chlorophyll
a
were used
to estimate reference conditions for these variables. Historical secchi depth data indicate the Secchi
depth has decreased in the Baltic proper since the beginning of the 20
th
century from ca 10 m to 7
m. We have used 10 m as reference value for summer Secchi depth in the Baltic proper.
From the empirical relationships between Secchi depth and TN, a reference value for total nitrogen
of 15.3 µM (214 µg/l) was estimated for the open coastal waters of the Baltic proper (See Annex
3.6 for details). Using the TN reference value of 15.3 µM and the empirical relationships found
between chlorophyll
a
and TN, a reference value for chlorophyll of 1.2 µg/l was estimated for the
open coastal waters of the Baltic proper
Outside major freshwater outflows, salinity gradients cause conditions in the surface water bodies to
vary greatly. Therefore a simple mixing model was used to calculate individual surface water body
reference conditions corrected for background concentrations in freshwater total nitrogen discharges
according to salinity (Annex 3.6).
115
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0118.png
The reference value indicates the average pristine condition. There is natural variation around this
value. The boundary between high and good status was set at a level where this natural variation is
judged to be exceeded. The 95 % confidence limit of variation was used as guidance.
Classifications for the Baltic proper coastal areas are based on dose-effect relations with total
nitrogen which was used to calibrate classes to assure a similar outcome of classification
exercises. In the Baltic proper the relation between phytoplankton and TN was stronger than
between phytoplankton and TP and consequently only TN was used to develop classifications
for phytoplankton. The empirical relationship between chlorophyll and TN was used to set
corresponding boundaries for chlorophyll. The southern Gulf of Bothnia is influenced by the Baltic
proper and a similar classification has been used in both areas. Further north in the Gulf of Bothnia
the influence of humic substances from freshwater increases which motivate an upward adjustment
of the classification limits for Secchi depth, chlorophyll
a
and phytoplankton
The classification of the general condition in open coastal areas as moderate means that the
ecological status of many coastal waters is not mainly determined by the local discharges, but rather
the general conditions of the Baltic Sea (Fig. 2.1.1). This calls for methods to separate the effect on
coastal waters into two parts: one that is directly manageable by local measures and one part that
is manageable by efforts by all countries together. An empirical model was developed to provide
quantitative estimates of the possible improvements in Secchi depth and reductions in chlorophyll
a
and phytoplankton biomass resulting from local measures (Annex 3.6).
Estimated chlorophyll
a
values for reference conditions and classification boundaries, as well as
the subsequent Ecological Quality Ratios for Swedish coastal waters corresponding the common
intercalibration types for the Baltic Sea are presented in Table 2.1.7.
Mixing model
Tot-N
Ex. measured tot-N
R.v. fresh
FW corrected
+OS corrected
Measured in open
water
OS corr.
FW correction
R.v. in open
sea
Salinity
Measured
salinity
Salinity
in open sea
Fig. 2.1.1:
Mixing model used to correct for natural background of
fresh water inputs of total nitrogen
(FW correction).
The fresh water corrected value forms the basis for classification of ecological status. The model also can be used to
estimate the influence of the
open sea on total nitrogen
(OS correction). The open sea correction is used to estimate the
degree to which local measures will have an effect on local ecological status (R.v. = Reference values).
116
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0119.png
Table 2.1.7:
Reference values (Rv), class boundaries (HG=high-good, GM= good-moderate, MP= moderate-poor, PB=
poor-bad) and corresponding EQR for summer levels of chlorophyll
a
(µg/l). Grey marking means that values should be
corrected according to salinity and estimated natural background load of nitrogen in respective area before comparison with
class boundaries (see text). Intercalibration sites correspond to the the common intercalibration types of the Baltic GIG.
2.1.4 Results of the comparison and harmonisation
Type B0:
Salinity 0.5-3, sheltered, shallow, >150 ice days
Sites
Preliminary
Status
1
-
H/G
G/M
H/G
Reference
values
1.3
1.4
1.1
1.2
Calculated value
1995 – 2000
2.2
Data missing
1.6
Calculated
value
2000 – 2004
2.0
1.4
Time period
for ref values
1925-1932
1900-1920
1900-1920
Finland
Sweden
Classification Classification
G
H
H
G
H
H
Bergö (FI)
Gaviksfjarden (SE)
Holmöarne, N
Kvarken (SE)
The Quark –
Orefjärden (SE)
1
Borderline
in the Intercalibration network register
Classification table
Type B 0
Sweden: Gaviksfj.
Holmöarna
Örefjärden
Finland
Boundary high/
good
1.8
1.5
1.8
1.8
EQR H/G
0.78
0.73
0.67
0.72
Boundary good/
moderate
2.3
2.0
2.3
2.7
EQR G/M
0.61
0.55
0.52
0.48
Period
June – August
July – August
117
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0120.png
Type B2:
Salinity 3-6, sheltered, shallow, >150 ice days
Sites
Prelimin.
Status
1
H/G
G/M
G/M
G/M
H/G
H/G
G/M
G/M
Reference
values
1.4
1.4
1.4
1.4
1.4
1.4
1.4
1.4
Calculated
value
1995 – 1999
-
3.8
2.4
3.3
-
-
-
-
Calculated
value
2000 – 2004
1.4
3.2
4.2
2.1
Time period
for ref values
1925-1934
1925-1934
1925-1934
1925-1934
1900-1920
Finland
Classifikation
Data 2000-2004
H
M or lower
M or lower
G
Sweden
classifikation
H
M or lower
M or lower
G
Domarkobban (FI)
Järviluoto (FI)
Pjelaxfjarden (FI)
Rounakari (FI)
Harkskars and Trodjefjard (SE)
Langvinds- and Skarsfjarden (SE)
The Bay of Gavle, Outer (SE)
Ljusnan and Voxnan (SE)
1
Borderline
in the Intercalibration network register
Classification table
Type B 2
Sweden
Finland
Boundary high/
good
1,8
1,8
EQR H/G
0,78
0,78
Boundary good/
moderate
2.3
2,6
EQR G/M
0,61
0,54
Period
June – August
July – August
Type B3:
Salinity 3-6, sheltered, shallow, 90-150 ice days
Sites
Prelimin. Reference
Status
1
values
G/M
H/G
H/G
G/M
G/M
G/M
1.8
1.8
1,5
1.5
1.2
1.3
Calculated Calculated
value
value
1995 – 1999 2000 – 2004
4.9
3.7
2.1
2.7
2.7*
6.3
6.6
2.4
4.2
2.7**
3.3***
Time
period for
ref values
1925- 1934
c.1930 –
1950
Finland
Sweden
Classification Classification
M or lower
M or lower
H
M
G
G or M
M or lower
M or lower
M or lower
M or lower
M or lower
M or lower
Bågaskär I (FI)
Långskär I (FI)
Putsaari II (FI)
Seili II (FI)
The Askoe area
(SE)
The Yxla area
(SE)
1
Borderline
in the Intercalibration network register
* 0 m, Jun-Aug 1995-1999, 1 station, ca 6 samplings/year during Jun-Aug
** 0 m, Jun-Aug 2000-2003, 1 station, ca 6 samplings/year during Jun-Aug
*** 0 m, Aug 2001, 2004, 2005, 6 stations, one sampling per year.
Classification table
Type B 3
Sweden: Askoe area
(More exposed)
Finland
I: Archipelago Sea
II: Western Gulf of
Finland
Boundary high/good EQR H/G Boundary good/moderate EQR G/M
1.5
1.6
a
2.2
2.6
0.80
0.81
0.82
0.58
1.8
1.9
a
2.9
4.0
0.67
0.69
0.62
0.38
Period
June – August 0 m
June - August 0 m
July – August
Sweden: Yxla area
a: Dynamic reference values (and boundaries) estimated from salinity during sampling.
118
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0121.png
Type B 12b:
Salinity 6-22, sheltered, shallow
Sites
Borderline in Reference Calculated Calculated Time period
Intercalibration
value
value
value
for reference
network
1995-1999 2000-2004
values
database
G/M
G/M
G/M
G/M
G/M
G/M
1,4
1,4
<1,1
<1,1
1,2
1,8
1,9
1,6
1,8
2,7
2.6**
1904-1911
1903-1912
German
classification
Swedish
classification
Danish
classification
Faxe Bay, DK
Hjelm Bay, DK
Darss-Zingst DE
Geltinger Birk, DE
Arc. Torhamn, SE
Koiguste laht, EE
1,8
4,6
2,7*
G/M
G
G
M or lower
M or lower
G/M
G
G
M or lower
M or lower
G/M
G
G
M or lower
M or lower
* June - Aug 1996-1999, discrete samples between 0-4m, 1 station, 1 sampling per year (station K19).
** June - Aug 2000-2004, discrete samples between 0-4m, 1 station, 1 sampling per year (station K19).
Classification table
Type B 12
Denmark
Sweden
Germany, Geltinger Birk
1)
Germany, Darss-Zingst
Estonia
1)
Site
Boundary high/good
1.5
1,5
1,1
1,1
2.2
EQR H/G
0.93
0.80
-
-
0.82
Boundary good/moderate
1,9
1,9
1,9
1,9
2.7
EQR G/M
0.74
0.63
< 0.58
< 0.58
0.66
Period
May – September
June – August
June – September
June – September
June – September
at the mouth of Flensburger fjord.
Type B 13:
Salinity 6-22, exposed, shallow
Sites
Borderline in Reference
Intercalibration
value
network
database
G/M
H/G
G/M
G/M
H/G
H/G
G/M
1,2
1,1
3.2
3.85
1.1
<1,5
< 3.7
Calculated
value
1995-000
Calculated Time period Estonia/Latvia
value
for ref
classification
2000-2004
values
1,4
4,5
8,6
?
?
1958-59
?
?
?
?
G
Danish
classification
G
Bornholm vest (DK)
Kudema Bay (EE)
Open coast 6 (LT)
Baltic Sea 3 (TW) (LT)
LI (LV)
1
Rowy (PL) *
Kepa Redlowska (TW) (PL)
Classification table
Type B 13
Denmark
Estonia and Latvia
1
Boundary high/good
1,3
1,3
EQR H/G
0.85
0.92
Boundary good/moderate EQR G/M Period
1.65
1,6
0.66
0.75
May – September
June – September
Data below are not comparable with the information from other countries
Lithuania coast
Poland Rowy*
Lithuania outside lagoon (TW) 4.8
Poland Kepa Redlowska (TW)
1.5
3.7
4
5
6
3
5.6
?
May – September
June - September
May – September
* Polish coastal water type B 13 – Rowy is situated too close to river mouth so it is not representative for type or water body.
1
LV has adopted the assessment made by Estonia. The sites in Estonia and in Latvia are very similar (expert
judgement). LV has
not developed their own assessment system and has no data from the intercalibration site.
119
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0122.png
Type B 14:
Salinity 6-22, sheltered, shallow lagoons
Sites
Borderline in
Reference Calculated
Intercalibration
value
value
network database
1995-2000
0,9
5
4
2,3
Calculated
value
2000-2004
1,3
Time period
for ref values
Polish
Danish
Classification Classification
G/M
Dybsø Fjord (DK)
G/M
1
* H/G
Mielizna Borzynska (PL)
Piaski Dziewicze (PL)
1
*
G/M
Classification table
Type B 14
Denmark
Boundary high/good EQR H/G
1,1
0.82
Boundary good/moderate EQR G/M
1,6
10
7,5
0.56
Period
May – September
?
?
Data below not comparable with the information from other countries
Mielizna Borzynska (PL)* 5
Piaski Dziewicze (PL)*
4
* Transitional waters. Data not comparable with the information from other countries
120
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0123.png
2.1.5 Results of the harmonization – Boundary EQR values
Summery table with the results of the phytoplankton intercalibration in the Baltic Sea area.
Reference
conditions
1.1 – 1.4
Boundary
High/good
1.5 – 1.8
EQR
High/Good
0.67 – 0.78
Boundary
Good/ moderate
2.0 – 2.3
EQR
Good/ Moderate
0.52 – 0.61
Type B 0
Sweden
Finland
Sweden
Type B 2
1.3
1.4
1.8
1.8
0.72
0.78
2.7
2.3
0.48
0.61
Finland
Type B 3
1.4
1.8
0.78
2.6
0.54
Sweden:
Askoe area (More
exposed)
1.2
1.3
1.5
1.6
0.80
0.81
1.8
1.9
0.67
0.69
Sweden: Yxla area
Finland
I: Archipelago Sea
II: Western Gulf of
Finland
Type B 12
Denmark
1.8
1.5
1.4
2.2
2.6
1.5
0.82
0.58
0.93
2.9
4.0
1.9
0.62
0.38
0.74
Sweden
Germany,
Geltinger Birk
Germany,
Darss-Zingst
Estonia
1.2
1.5
1.1
1.1
2.2
0.80
-
-
0.82
1.9
0.63
< 1.1
< 1.1
1.8
1.2
1.9
1.9
2.7
1.6
< 0.58
< 0.58
Type B13 CW
Denmark
0.67
0.75
Estonia and Latvia
Type B13 TW
Poland
1.1
< 3.7
1.3
3.7
1.3
0.85
?
0.92
1.65
5.6
0.67
0.67
Lithuania
Type B 14
Denmark
3.85
0.9
4,8
1.1
0.80
0.82
6,0
1.6
0.64
0.56
Phytoplankton: parameter indicative of biomass (Chlorophyll a)
Results: Ecological quality ratios and parameter values
The following results refer to summer mean May/June – September
121
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0124.png
Type and country
Ecological Quality Ratios for the
national classification systems
High-Good
boundary
0.76
0.78
0.71
0.81
CW B 12 a
Eastern Baltic Sea
Salinity 5-8 psu
All countries sharing the type
CW B3 b
Exposed
All countries sharing the type
CW B3 a
Sheltered
All countries sharing the type
CW B2
All countries sharing the type
CW B0
All countries sharing the type
Good-Moderate
boundary
0.56
0.56
0.49
0.68
High/-Good
boundary
Parameter values/ranges
Chlorophyll a µg/l
Good-Moderate
boundary
2.3 (2.0 -2.7)
2.5 (2.3 -2.6)
3.5 (2.9 – 4.0)
1.9 (1.8 – 1.9)
1.7 (1.5 – 1.8)
1.8
2.4 (2.2 - 2.6)
1.6 (1.5 – 1.6)
0.82
0.66
2.2
2.7
CW B 12 b
Western Baltic Sea
Salinity 8 -22 psu
All countries sharing the type
CW B 13
Denmark, Estonia and Latvia
CW B 14
Denmark
0.92
0.92
0.82
0.90
0.63
0.75
0.56
0.66
1.3 (1.1 – 1.5)
1.3
1.1
4.2
1.9
1.6
1.6
5.8
TW B 13
All countries sharing the type
Countries sharing types that have been intercalibrated:
Types CWB0, CWB2, CWB3a, CWB3b: Finland, Sweden.
Type CWB12a: Estonia
Type CWB12b: Germany, Denmark, Sweden.
Type CWB13: Denmark, Estonia, Lithuania, Latvia, Poland.
Type CWB14: Denmark, Poland
Type TWB13: Lithuania, Poland.
2.1.6 Open issue and need for further work
All member states were participating in the intercalibration of the chlorophyll a metrics and setting
of reference conditions and boundaries as well as EQR values for chlorophyll
a
(Latvia agreed
on the boundaries for type CWB13 developed by other member states). Further development and
intercalibration is needed for other phytoplankton parameters. Assessment methods were not yet
ready for phytoplankton species composition and phytoplankton blooms in any of the Baltic GIG
member states.
122
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0125.png
2.2 Black Sea GIG
2.2.1 Intercalibration approach
Typology
In the Black Sea was identified only one common type, as is shown in the table below:
Tab 2.2.1:
Black Sea GIG Coastal Water Types included in the intercalibration.
Type
CW – BL1
Salinity
mezohaline
Tidal range
microtidal
Depth
shallow
Substratum
mixed
The countries participating at this intercalibration exercise into the Black Sea region are Bulgaria
and Romani
2.2.2 Reference conditions and class boundary setting
a. BULGARIA
Dr. Assoc. Prof. Snejana Moncheva, Junior Res. Natalia Slabakova, Marine Biology and Ecology
Department, INSTITUTE OF OCEANOLOGY, BAS, Varna
Identification of type-specific biological reference conditions for the coastal waters
After the critical review of the possible phytoplankton descriptors and the outlined constrains and
uncertainties for the identification of type-specific biological reference conditions for the coastal
waters the following metrics have been selected:
• Phytoplankton biomass [mg/m
3
]
• Chlorophyll a [mg/m
3
] and
• Phytoplankton blooms (need of further development)
Approach
The approach for the identification of reference conditions is based on:
• historical data from the relatively pristine period of the Bulgarian Black Sea coast (1954-1970
– published seasonal data);
• 10
th
percentile (lower quartile) of a long-term data set (period 1983-2005) to test applicability
for definition of reference values and for the selection of “bad values” (from the period of
intensive anthropogenic eutrophication – the 80-ies) (Heiskanen et al., 2005). This step was
crucial for the identification of reference values for chlorophyll a. Due to lack of chlorophyll a
measurements from the reference period and before 1990, the 10
th
percentile from a long-term
data set (1990-2006) was determined and its applicability tested against the results from the
analysis of the phytoplankton data set .
• expert judgment
• As initially phytoplankton is an indicator of eutrophication mainly, the typology groups were
considered in relation to the level of exposure and depth. In addition due to the influence of
the Danube river nutrient enriched waters through the main Black Sea current along the Black
Sea coast the moderately-exposed-intermediate and exposed-shallow types were combined
123
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0126.png
into one group. Thus the classification tool was developed for the following 2 typology
groups: moderately exposed - shallow (CW602330; CW602310) and moderately-exposed-
intermediate/exposed-shallow types (CW602210; CW602220; CW602230; CW602321).
• Due to the high seasonal variability of phytoplankton communities composition and abundance
the average annual values are not adequate and the classification tool should be developed on
seasonal bases
In order to illustrate the approach the Galata/Varna Bay sites were selected on the basis of the best
(only) available long-term data set, over the period 1954-2006 (phytoplankton) and 1990-2006
(chlorophyll a).
The data sets of the two sites were merged into one data array based on the following assumptions:
1) the intensive hydrophysical interaction between the two sites that to a great extent determine
the spatial distribution and dynamic of phytoplankton communities thus reflecting the synergy
between the local land base sources and the Danube influence through the main Black Sea
currents.
2) Principally the analytical procedure of phytoplankton processing (biomass estimation) are
comparable for the data array 1954-1998, assuming 25 % analytical error inherent in the
method itself. The difference of sampling frequency (seasonal sampling for the historical data
versus monthly /bimonthly sampling after 1980) should be underlined.
3) Originally phytoplankton is the biological quality element most sensitive to external forcing
(nutrient over-enrichment (eutrophication), light, currents etc.).
4) the values estimated for the period 1954-1970 correspond to High status, and those from the
period 1983-1998 to Bad status, and they represent the worst High and the worst Bad along
the Bulgarian Black Sea coast (Moncheva et al, 1995, 2001, Moncheva, Kamburska, 2002).
Results
Table 2.2.2:
Statistical summary of phytoplankton biomass B[mg/m3] by seasons (Galata-Varna Bay).
winter
years
n
Median*
Max*
1954-
1970
15
1214
6800
winter
1983-
1997
18
7763
3088
42260
spring
1954-
1970
14
3995
2700
12290
spring
1983-
1997
45
36892
8055
583139
summer
1954-
1970
11
526
534
900
summer
1983-
1997
39
39035
23196
157789
autumn
1954-
1970
15
783
489
2483
autumn
1983-
1997
18
31639
854
296675
annual
1954-
1970
55
1845
1234
12290
annual
1983-
1997
120
28832
8798
583139
Average* 2075
*the values within the 95 % confidence interval after the statistical processing of the data array
As apparent from the data presented the average annual gives no idea about the alterations in the
seasonal pattern of phytoplankton dynamics, the severe increase in summer biomass in particular
(related hypoxia/anoxia conditions and ecosystem deterioration as one of the most important
ecological concerns) bearing implications for the proper environmental management policy and
monitoring design.
Principally the historical median was selected as the seasonal reference value, and the “average”
for autumn only. The threshold “bad” was based on the 90 percentile of the long-term data set and
expert judgment.
As apparent from Fig. 1. the lower quartile of the long-term data set for spring (A) is very close to
the historical reference, which does not hold for summer (B).
124
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0127.png
A)
B)
Fig. 2.2.1:
Cumulative distribution of spring and summer phytoplankton biomass [mg/m
3
].
125
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0128.png
Table 2.2.3:
Inventory of chlorophyll a [mg/m3] data by sites.
Site
Kaliakra
Varna Bay
Galata
Burgas Bay
Winter
period
1995-2005
1990-2005
1990-2005
1987-2005
No
Samples
9
39
9
32
Spring
period
1990-2006
1991-2005
1990-2006
1984-2006
No
Samples
22
101
16
51
Summer
period
1990-2006
1990-2006
1990-2006
1987-2006
No
Samples
28
93
30
53
Autumn
period
1990-2006
1990-2006
1990-2006
1982-2006
Total
No
No
Samples Samples
23
11
46
82
88
321
66
182
The chlorophyll a reference and threshold values were determined based on the 10 (or 25) percentile
and the 90 percentile (bad) of the long-term data set – Fig.2., Table.3
The proposed classification and QRs based on phytoplankton biomass and chlorophyll a to be
applied for the Bulgarian Black Sea coastal waters is presented in Tab.4 and 5.
Fig. 2.2.2:
Cumulative distribution of long-term chlorophyll a data for winter and summer.
126
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0129.png
Table 2.2.4:
Seasonal “high” values as inferred by the cumulative distribution.
Site
Kaliakra
Galata
V. Bay
Winter
Spring
High – chl. a [mg/m
3
]
25th %
0.97
2.2
2.1
10th %
0.65
1.2
0.8
Summer
Autumn
10th %
0.77
1.4
1.05
25th %
2.30
1.7
1.7
10th %
0.79
1.3
1.4
25th %
1.12
2.1
1.2
10th %
1.32
0.8
0.7
25th %
1.82
1.3
1.2
Burgas Bay
Table 2.2.5:
Classification and EQR based on phytoplankton biomass [mg/m
3
] by seasons.
Type
CW602330
CW602310
CW602210
CW602220
CW602230
CW602321
Type
CW602330
CW602310
CW602210
CW602220
CW602230
CW602321
Type
CW602330
CW602310
CW602210
CW602220
CW602230
CW602321
Type
CW602330
CW602310
CW602210
CW602220
CW602230
CW602321
Winter
Index
PhB [mg/m3]
High
(1200) 2070
Good
4150
Moderate
7600
Poor
11600
Bad
>15000
PhB [mg/m3]
EQR
Index
PhB [mg/m3]
(1200)1700
0.85
High
(2700) 3700
3200
0.7
Spring
Good
5200
6000
0.45
Moderate
9500
9100
0.16
Poor
14500
>12000
Bad
>15000
PhB [mg/m3]
EQR
Index
PhB [mg/m3]
(2500)2900
0.85
High
(600-900) 1160
4100
0.7
Summer
Good
2750
7400
0.45
Moderate
5500
11600
0.16
Poor
7600
>15000
Bad
>10000
PhB [mg/m3]
EQR
Index
PhB [mg/m3]
(750) 1100
0.85
High
(1000) 1650
2150
0.7
Autumn
Good
3300
4000
0.45
Moderate
6100
6100
0.16
Poor
9200
>8000
Bad
>12000
PhB [mg/m3]
EQR
(1000) 1350
0.85
2700
0.7
4950
0.45
7500
0.16
>10000
127
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0130.png
Table 2.2.6:
Classification and EQR based on chlorophyll a data [mg/m
3
] by seasons.
128
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0131.png
In addition based on the correlation chl. α/Secchi depth as inferred from the available data for the
north-western Black sea the regression equation was used (Fig.3) to suggest classification thresholds
for the Secchi depth presented on – Table 6.
At this stage it is not possible to suggest classification based on phytoplankton blooms for the
following reasons.
The reference data (1954-1970) are seasonal, as well as the recent data which makes quantification
and the application of the 6 years monthly frequency records as suggested by some expert groups
(WFD ITR, 2007) impossible. The records for the species that were the most frequent drivers of
ecosystem dysfunction (Prorocentrum minimum and the related hypoxia events in summer during
the eutrophication period – Moncheva et al., 1995, 2001) are of little use due to the shifts in the
phytoplankton taxonomic composition after 2000 (Moncheva et al, 2006) – Table 7.
Figure 2.2.3:
Correlation between chlorophyll a
[mg/m3] and Secchi disk depth [m] – log scale
(after V. Doncheva. 2007).
Table. 2.2.7:
Classification based on TRIX and Secchi depth.
All types and seasons
Index
TRIx
Secchi depth[m]
EQR
High
2-4.5
>4.5
0.9
Good
4.5-5.4
4.5-3.5
0.7
Moderate
5.5-6.4
3.5-2.5
0.5
Poor
6.5-7.9
2.5-1.5
0.3
Bad
>8
<1.5
129
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0132.png
Table 2.2.8:
Summary of phytoplankton blooms (Galata-Varna Bay).
Descriptor
Winter
1954-1970
Bacillariophyceae
Skeletonema costatum
Pseudonitzschia seriata
Detonula confervaceae
Winter
1983-1998
Bacillariophyceae
Skeletonema costatum
Pseudonitzschia seriata
Dinophyceae
Heterocapsa triquetra
Prorocentrum minimum
13
31
43.47
Spring
1983-1998
Bacillariophyceae
Skeletonema costatum
P. delicatissima
Dinophyceae
Heterocapsa triquetra
Prorocentrum minimum
24
63
220
Summer
1983-1998
Bacillariophyceae
Cerataulina pelagica
Pseudonitzschia delicatissima
Nitzschia tenuirostris
Pseudinitzschia seriata
Rhizosolenia fragilissima
Dinophyceae
Gymnodinium uberimum
Prorocentrum minimum
Alexandrium monilatum
Lingulodinium polyedrum
Haptophyceae
Phaeocystis pouhettii
21
73
481
Autumn
1983-1998
Skeletonema costatum
Detonula confervaceae
Leptocylindrus minimus
Dinophyceae
Prorocentrum minimum
Alexandrium monilatum
Lingulodinium polyedrum
Oxyphisis oxytoxoides
15
33
60.17
Most typical species
No species
No blooms
Max N [1x10
6
cells/l]
Descriptor
3
6
8.2
Spring
1954-1970
Bacillariophyceae
Pseudonitzschia delicatissima
Pseudinitzschia seriata
Dinophyceae
Prorocentrum minimum
Most typical species
No species
No blooms
Max N [1x10
6
cells/l]
Descriptor
Most typical species
7
16
13.5
Summer
1954-1970
Bacillariophyceae
C.caspia
Dinophyceae
Prorocentrum minimum
No species
No blooms
Max N [1x10
6
cells/l]
Descriptor
2
4
2.1
Autumn
1954-1970
Bacillariophyceae
Skeletonema costatum
Detonula confervaceae
Leptocylindrus minimus
Dinophyceae
Most typical species
No species
No blooms
Max N [1x10
6
cells/l]
2
2
5.7
130
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0133.png
b. ROMANIA
The site selected was Constanta/Mamaia Bay were the phytoplankton samples have been collected
twice per week. The long-term data set is available for the period 1960-1970 and 1986-1997 for this
site.
Phytoplankton samples have been collected from the surface layer (enabling more then 500 ml of
water per sample) and immediately treated with a formaldehyde solution of 4 % concentration,
followed by further processing based on sedimentation method (Morozova-Vodyanitzkaya, 1954;
Bodeanu, 1987-1988). The taxonomic determination and cells counting by species of respective
samples has been carried out by means of plankton inverted microscopes, equipped with 40x
magnification lens, in the case of small-sized cells (less then 15-20 µm) or with 16.3x / 20x
magnification lens, in the case of larger-sized cells. Afterwards numerical densities (cell×l
-1
) and
biomass (mg/m
3
) were calculated for each specific component of each taxonomic group and for total
phytoplankton.
In order to identify the biological reference condition for the Romanian coastal waters the
phytoplankton biomass (mg/m
3
) has been selected. The identification of reference condition is based
on historical data from pristine period of the Romanian Black Sea coast (1960-1970), from period
of intensive eutrophication (’80 years).
In order to identify the EQRs for the common body waters, we found the values estimated for the
period 1960-1970 corresponding to high status and those from the period 1986-1997 to bad status.
The limits between these classes, for good, moderate and poor have been decided based on expert
judgment.
Table 2.2.9:
Summary of phytoplankton biomass (mg / m
3
) by season.
1960-1970
Average
Max
Winter
2692
Spring
2583
Summer
690
Autumn
540
921
6603
Annual
1627
1986-1997
Average
Max
5725
1255
19165
5064
1035
10664
46809
19165
6747
5465
34647
61195
31195
Table 2.2.10:
Classification based on phytoplankton biomass (mg/m
3
).
High
2500
EQR
3000
EQR
1000
EQR
1500
EQR
Good
3600
5700
2500
3500
Moderate
Winter
5500
Spring
9000
Summer
6000
Autumn
7200
Poor
8000
14000
7500
9500
Bad
> 10000
> 15000
> 12000
> 15000
131
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0134.png
2.2.3 Results of the harmonization – Boundary EQR values
Table 2.2.11:
Phytoplankton biomass.
Ecological status
Winter
Range
EQR
High
< 2000
> 0.85
Good
2000 - 3000
0.7
Moderate
3001 - 5500
0.45
Poor
5501 - 12000
0.16
Bad
> 12000
< 0.16
Spring
Range
EQR
< 3000
> 0.85
3000 - 5000
0.7
5001 - 9000
0.45
9001 - 20000
0.16
> 20000
< 0.16
Summer
Range
EQR
< 1000
> 0.85
1000 - 2500
0.7
2501 - 5000
0.45
5001 - 10000
0.16
> 10000
< 0.16
Autumn
Range
EQR
< 2000
> 0.85
2000 - 3500
0.7
3501 - 6500
0.45
6501 - 15000
0.16
> 15000
< 0.16
It was agreed that the assessment system for phytoplankton have to be applied for surface layer
samples, with a monthly frequency sampling, at least between May and September; the assessment
should be made based on seasonal values and for annual assessment will be made using the “one
out, all out” principle; harmonization of EQR values resulted with separation of the most northern
water body of the common water body type in Bulgaria (CW602230) from the rest of the bodies at
the same type, and eventually separation of a new type.
The rest of the metrics have not been harmonized because Romanian part was not ready with their
reference values and EQRs.
2.2.4 Open issues and need for further work
The future work plan will focus on preparation of assessment system for bloom concentrations,
bloom species and chl α, according to the methods provide in the Black Sea Commission Manual.
Other parameters will be considered at a later stage.
2.3 Mediterranean GIG
2.3.1 Intercalibration approach
Participation
of countries in the Phytoplankton subgroup:
Cyprus
France
Greece
Italy
Slovenia
Spain (Valencia, Catalonia, Balearic Islands)
+ Croatia (Accession Country)
132
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0135.png
Data availability
The examination of the datasets provided by each MedGIG Member States highlighted a huge data
heterogeneity, mainly due to different monitoring schemes. Table 1 summarized the main features
of the datasets, highlighting the availability of key features (sites, parameters, vertical profile and
amount of data) needed for the typological identification of water bodies.
Data used are referred to Chl
a
concentrations measured in “coastal waters”, as defined in the
Directive:
“within the 1 nautical mile distance from the coastline”,
A common decision is to recommend the use of nearshore data only (500-1500 m from the coast),
(while for the IC all “coastal” data have been considered, as reported above), when not differently
advised by MS
23
.
Table 2.3.1:
Data features for MED GIG MSs datasets.
France
Italy
Slovenia
Cyprus
Spain
MS
N° Sites
11
8
N° Records
1250
2541
1997-2006
2001-2004
2005
1997-2004
1991-2006
2000-2004
Period
Freq (d)
15
7
not available
available
available
Profile
available
available
Temp Sal data
Croatia
Greece
117
48
19
2
332
5
1784
1109
158
30
1225
2000-2004
regional
120
30
60
not available
available
not available
available
available
available
not available
available
available
Typology
Phytoplankton experts defined that the 4 Mediterranean Coastal IC types, (see Section 1 General
Part), based primarily on the substratum composition and the depth profile, cannot be applied to
the IC for the current BQE, within the Mediterranean basin: the classification criterion is based
mainly on the morphological features of the bottom and therefore it is not so meaningful in a
“phytoplankton perspective”.
A new typology has been developed, mainly focused on hydrological parameters characterizing
water bodies’ dynamics and circulation. The typological approach is based on the introduction of
the static stability parameter (derived from temperature and salinity values in the water column):
such a parameter, having a robust numerical basis, can describe the dynamic behaviour of a coastal
system. The group agreed then to adopt surface density as a proxy indicator for static stability as
both Temperature and Salinity are relevant in the dynamic behavior of a coastal marine system: both
are involved in circulation and mixing dynamics and all information is then nested in the surface
density parameter (Russo et al., 2006, submitted).
On the basis of surface (density (σ t) values three major water types have been defined:
Table 2.3.2:
density thresholds defining new coastal water types.
σ t (density) (kg m
-3
)
Type I
<25
25<d<27
Type II
Type III
>27
23
NOTE from Spain: From 90s in Spain there has been an important effort to sample Chl-a stations located at inshore
(over the line coast). In the main of cases these data are the base of the methods used along the country and the base of
the part of the results Spain shares in this intercalibration exercise. Although no other MSs work with this kind of data, it is
important to stress that all the reference conditions and boundaries proposed in this intercalibration exercise are referring
to the nearshore.
133
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0136.png
The same three water types are defined below as salinity classes, since for Spain (Catalonia)
variability due to seasonal fluctuation is marginal: using relationships between density and salinity,
Spain calculated the following table (at 18 °C)
Table 2.3.3:
salinity thresholds defining new coastal water types.
Annual Mean Salinity (psu)
Type I
<34.5
34.5<d<37.5
Type II
Type III
>37.5
This type subdivision based only on salinity, is perfectly comparable with the previous ones, based
on density, agreed by the rest of Mediterranean MSs. For more details about the use of salinity
instead of density see the Annex 3.8: Spain MS report on phytoplankton element.
The three different water types, in an ecological perspective, can be described as follows:
Type 1 coastal sites highly influenced by freshwater inputs
Type 2 coastal sites not directly affected by freshwater inputs
Type 3 coastal sites not affected by freshwater inputs
A further distinction has been suggested and approved by the MSs, regarding the splitting of the
coastal water type 3 in two different sub basins, the Western and the Eastern Mediterranean one,
according to the different trophic conditions
24
:
Type 3 WM
Type 3 EM
Furthermore, Spain proposed the subdivision of type II, which include marine waters with
intermediate salinities) in two subtypes: type II-A and type II-B
25
.
Types presence in the different MSs was reviewed, and finally defined, as shown in the following
Table 2.3.4.
24
The Levantine Basin of eastern Mediterranean is characterized as nutrient-deficient and therefore ultra-oligotrophic
in comparison to the Atlantic Ocean (Berman et al., 1984). Furthermore, eastern Mediterranean is more P-limiting to the
growth of phytoplankton, in contrast to the general dogma that N is the more limiting nutrient in marine systems (Krom
et al., 1991). Recent studies made on phytoplankton biomass in the deeper waters of eastern Mediterranean reveal that
prevailing oligotrophic conditions result in low chlorophyll-a concentrations ranging from 0.1 to 0.2 µg L-1 (Krom et al.,
1992). It has also been shown that chlorophyll-a concentrations off the coast of Cyprus are among the lowest in the region
and ranged from 10 to 90 ng L-1 (Bianchi et al., 1996). Recent studies along the coastal waters of Cyprus confirmed its
oligotrophic status (Argyrou, 2005, 2006).
The coastal waters of Cyprus are classified as Type III (no freshwater input – density greater of 27), due to their hydro-
graphical features and the prevailed physicochemical characteristics; in fact mean salinity of coastal waters of Cyprus is
39,1 psu. The annual mean of Chl a for the years 2004 to 2006 ranged from 0,07 to 0,11 µg L-1 while, the calculated 90th
Percentile ranged from 0,09 to 0,2 respectively. The overall average level of Chl a for the entire period, 2004 to 2006, was
0,086 and the respective 90th Percentile was 0.188. These values were used for the assessment of the ecological status
of the coastal waters of Cyprus according to the Eutrophication Scale, which was developed by Ignatiades et al. (1992)
and Karydis (1999), and further modified by Siokou & Pagou, 2000; Pagou, 2000) based on nutrient and phytoplankton
data collected from several coastal and marine areas from Greece.
25
The South of Spain (the main part of Andalusian coast) is clearly affected by the influence of the Atlantic waters, so the
natural salinity, nutrients and Chl-a concentrations do not correspond with type III. Moreover the lower salinities of before
defined type II were explained by freshwater inputs, coming mainly from the continent. It should be emphasized that in
the vicinities of Gibraltar Strait there are also lower salinities that come from the Atlantic, and that is why this subdivision
in Type II-A (the original one) and Type II-B (affected by atlantic influence) was proposed by Spain. For more details con-
sult the
Annex: Spain MS report on phytoplanckton element.
It should be also considered the relationship between the
reference conditions and boundaries defined by the NEA GIG for the atlantic waters in the western part of the Gibraltar
Strait.
134
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0137.png
Table 2.3.4:
Types occurrence per MSs in Coast MED GIG.
Types description
Type I
Highly influenced by
freshwater input
A
B
W
E
Moderately influenced
by freshwater input
(continent influence)
Influenced by Atlantic
waters
Continental coast , Not
influenced by freshwater
input
Not influenced by
freshwater input
X
France
X
X
X
X
X
X
X
X
X
Spain
Italy
X
X
X
Slovenia
Croatia
Greece
Cyprus
Type II
Type III
Based on this new Typology and on the available data for the different types in the Member States,
Intercalibration was performed as follows:
for
Type I
France is not able at the moment to provide enough data to be compared with the Italian
data; therefore no Intercalibration is performed for this typology at the present phase
for
Type II
Spain had to make distinction within the same type (as specified below (
3
)). As this
subdivision is not found in the other Mediterranean MS, IC performed on Type II does not include
the Spanish Type II B (influenced by Atlantic waters)
for
Type III
the distinction between Type III Western Mediterranean and Type III Eastern
Mediterranean was already agreed for evident ecological differences within the 2 Mediterranean
basins. Only Greece and Cyprus belong to the Eastern Mediterranean basin, therefore Type III E
Intercalibration was performed only between these 2 countries
2.3.2 National methods that were intercalibrated
Since we did not elaborate a methodology based on a common data set, which would be then
adopted at national level of each MS, we compared boundaries (Chl-a concentrations and EQRs)
derived from national methods having different status at present (under development, finalized,
officially accepted). Available methodologies descriptions are detailed or cited in the Annex 3.8:
National methods included in the intercalibration.
Only one parameter of the BQE phytoplankton was considered for this phase of the Intercalibration
process: Chlorophyll
a
concentration as parameter/indicator for biomass.
Common statistical analysis on Chl-a, nutrients and physico-chemical data, and some multivariate
techniques have been performed in order to facilitate the reaching of a wide agreement for the
intercalibration process. Since a methodology based on a common data set, which would be then
adopted at national level of each MS, was not elaborated, different metrics of this parameter and
different statistical approaches for setting the boundaries (derived from national methods, where
defined) were analysed and compared. Boundaries are in terms of Chl-a
concentrations and EQRs
that have different status of implementation/finalization, at present in the Mediterranean Member
State (see Table below). Different metrics of this parameter and different statistical approaches for
setting the boundaries were compared.
135
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0138.png
The finally agreed approach that has been followed for the intercalibration, is an the hybrid option, as
described in the Intercalibration Guidance as follows:
“Boundary values are first established with national classification assessment methods (as in Option
3). The subsequent comparison of the boundary values could then be done with the help of a
common metrics method (as in Option 2).”
National methods adopted, mostly, three kinds of metrics: percentile90
th
, annual geometric mean
(geomean) and average. Depending on the MS the metrics were calculated using only surface data
or water-column integrated data, covering different period (e.g. one year in case of geomean and 5
to 6 years when using percentile 90
th
. The metrics used by MSs are shown in the table below.
Table 2.3.5:
Metrics adopted by MSs.
MS
France
Spain
Italy
Slovenia
Greece
Cyprus
X
X
Percentile90
th
X
X
X
G
A
A
A
Average (A-arithmetic,
G-geometric)
Based on: Raw data (R)/
Geomean (G)
R
R
G
R
2.3.3 Reference conditions and class boundary setting
For the 3 intercalibrated types every MS defined reference conditions, boundaries and EQRs
applying their own methodologies. These methodologies are detailed or cited in the 2.3. 1.
Reference conditions
Reference conditions will be different according to different water types.
Each MS proposed its own reference conditions based on their phytoplankton experts’ knowledge.
All of them based their calculations of reference conditions by selecting High status stations from
their monitoring programmes. For more details of the methods see Annex 3.8: National MSs reports
on phytoplankton element.
Type-specific reference conditions, as suggested by MSs, are listed in table 2.3.6, below.
Table 2.3.6:
Type-specific Reference conditions, expressed as surface Chl a concentration (μg /L)
RC
France
Type II-A
1.9
<2
Type III WM
1.10
0.4
*
*
*
<1
Type III EM
*
*
Percentile90th
Average
Metric
Italy
Spain
Spain
Greece
Slovenia
0.77
*
*
0.98
0.99
0.46
*
*
Percentile90th
Cyprus
* Not applicable
0.08
0.08
*
Percentile 90
th
(Geomean)
Annual (geomean)
Annual average
Annual average
136
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0139.png
Setting of Boundaries
In the following table the MSs boundaries, as chl
a
concentration,
(μg /L)
, for each water type are
listed:
Table 2.3.7:
Type-specific MSs boundaries.
Reference
1,8
<2
1,9
0,98
0,77
0,99
<1
1,1
0,46
0,4
0,08
0,08
H/G
2,4
2
2,3
1,15
1
1,28
1
1,3
0,54
0,51
0,1
0,1
EQR
H/G
0,75
0,83
0,85
0,77
0,78
0,85
0,85
0,78
0,8
0,8
G/M
3,5
4
3,5
1,72
1,24
1,62
2
1,8
0,7
0,64
0,4
0,4
EQR
G/M
0,51
0,54
0,57
0,62
0,61
0,61
0,66
0,63
0,2
0,2
Metric
Percentile 90th
Geometric mean
Percentile 90th
Percentile 90th
Mean
Geométric mean
Geometric mean
Percentile 90th
Percentile 90th
Mean
Geométric mean
Percentile 90th
Percentile 90th
TRUE VALUES
TRUE VALUES
T1
T2 - A
France
Italy
France
Spain
Spain
Italy
Slovenia
T3 – Western Med France
Spain
Spain
Italy
T3 – Eastern Med Cyprus
Greece
TRUE VALUES
2.3.4 Results of the comparison and harmonization
Phytoplankton experts from MSs decided to adopt a final agreement based on 90th percentile on
raw data and, at least, monthly sampling frequency. In order to make MSs values more comparable
Slovenia and Italy decided to translate their own values in new boundaries calculated using the same
metric as the other MSs (90
th
percentile). For details on MS calculations see Annex 3.8.
Using the same metric the proposed boundaries by each country were very similar, thus the group
came to the agreement reported in the following paragraph.
Harmonization of boundaries and EQR values
The boundary values are expressed for the metric 90
th
percentile, assuming that at least 5years data
are available, with monthly sampling, in the surface layer.
Type I:
not intercalibrated, only Italy has enough data
Type II
Table 2.3.8:
Harmonized boundaries for Type II - A.
Type
T2 - A
MS
Slovenia
France
Spain
Italy
REFERENCE
1.9
H/G
2.4
EQR
0.80
G/M
3.6
EQR
0.53
137
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0140.png
Type III
Table 2.3.9:
Harmonized boundaries for Type III.
Type
T3 – Western Med
T3 – Eastern Med
MS
France
Spain
Italy
Cyprus
Greece
REFERENCE
0.9
0.08
H/G
1.1
0.1
EQR
0.80
0.80
G/M
1.8
0.40
EQR
0.50
0.20
2.3.5 Open issues and need for further work
A few suggestions have been pointed out by MS for an eventual further intercalibration activity:
1. Include species composition analysis or blooms frequency analysis for a better
understanding of the system’s behavior and efficiency/status Initially common indicator
metrics will be considered (option 1 or 2), but it may be possible to move to option 3 when
member states will have developed their full phytoplankton classification systems.
2. Improve dose/response analysis correlating pressures (nutrients) with trophic conditions.
2.4 NE Atlantic GIG
2.4.1 Intercalibration approach
Option 2 has been used in the intercalibration process for phytoplankton. Three metrics have been
selected: chlorophyll, Indicator Taxa (Frequency of
Phaeocystis
Cell counts) and Taxa Cell Counts
(Frequency of phytoplankton taxa cells counts). Boundaries have been agreed for all countries for
the measurement of chlorophyll a using a 90
th
percentile metric. Total cell counts and
Phaeocystis
metric are not relevant in all countries’ waters, but common thresholds have been agreed where
applicable.
Generic typologies for NEAGIG types
In the NE Atlantic six basic intercalibration types have been agreed. These are shown in the Table 1.
Types were distinguished by salinity, tidal range, depth, velocity, exposure, mixing and residence
time.
The above types occur in Member State’s waters as detailed in Table 2.
138
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0141.png
Table 2.4.1:
Broad geographical types within the North East Atlantic GIG.
New Type ID
CW
–NEA1/26
a,b,c,d,e
CW –
NEA3/4
CW – NEA7
CW – NEA8
Name
Exposed or sheltered,
euhaline, shallow
Polyhaline, exposed or
moderately exposed
(Wadden Sea type)
Deep, low current,
sheltered
Polyhaline, microtidal,
sheltered, shallow
(Skagerrak inner arc
type)
Fjord with a shallow
sill at the mouth with
a very deep maximum
depth in the central
basin with poor
deepwater exchange.
Polyhaline, microtidal
exposed, deep
(Skaggerak outer arc
type)
Transitional waters
Salinity (PSU)
Fully saline
(> 30)
Polyhaline (18
- 30)
Fully saline
(> 30)
Polyhaline (18
- 30)
Tidal range (m)
Mesotidal
(1 - 5)
Mesotidal
(1 - 5)
Mesotidal
(1 - 5)
Microtidal
(< 1)
Depth (m)
Shallow
(< 30)
Shallow
(< 30)
Deep
(> 30)
Shallow
(< 30)
Current
velocity
Medium
(1 - 3 knots)
Medium
(1 - 3 knots)
low
(< 1 knot)
low
(< 1 knot)
Exposure
Exposed or
sheltered
Exposed or
moderately
exposed
Sheltered
Sheltered
Mixing
Fully mixed
Residence time
Days
Fully mixed
Days
Fully mixed
Partially
Stratified
Days
Days-Weeks
CW – NEA9
Polyhaline (18
- 30)
Microtidal
(< 1)
Deep
(> 30)
low
(< 1 knot)
Sheltered
Permanently
Stratified
Weeks
CW – NEA10
Polyhaline (18
- 30)
Microtidal
(< 1)
Deep
(> 30)
low
(< 1 knot)
Exposed
Permanently
Stratified
Days
TW – NEA11
(will be split
into sub-
types)
Oligo-Euhaline
(0 - 35)
Mesotidal
(1 – 5 )
Shallow
(< 30)
Medium
Sheltered or
moderately
Exposed
Partially- or
Permanently
Stratified
Days-Weeks
Table 2.4.2:
The occurrence of the broad geographical types and sub-geographical types within each member state.
NB.
Type
CW – NEA1/26 has been split down into 5 sub-types due to regional geographic differences within the North
East Atlantic Geographic Intercalibration Group.
Type
CW - NEA1/26a
CW - NEA1/26b
CW- NEA1/26c
CW- NEA1/26d
CW- NEA1/26e
CW – NEA3/4
CW – NEA7
CW – NEA8
CW – NEA9
CW – NEA10
TW – NEA11
X
X
X
X
X
x
x
x
x
x
x
x
X
X
X
x
x
X
x
X
X
X
X
X
BE
DK
FR
X
X
X
DE
IE
X
X
NL
NO
X
PT
ES
X
SE
UK
X
X
139
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
2.4.2 National methods that were intercalibrated
For the phytoplankton element, three metrics were chosen:
Chlorophyll a
Elevated cell counts
Phaeocystis
blooms
National methods from each country have all used a variation of these metrics within their
own assessment protocols. All countries have used chlorophyll-a as a primary assessment
tool for phytoplankton, though typically chlorophyll is used within a suite of phytoplankton
tools and calculated as an EQR. All MS have agreed to use chlorophyll-a as a metric within the
intercalibration process. Elevated cell counts and
Phaeocystis
are not all used uniformly across
member states in the national methods and as such, have not been adopted as an intercalibration
tool for all MS. Five countries are using the elevated count tool and four counties are using the
Phaeocystis
tool. National methods for each MS are listed below and further discussed in Table 3.
UK
Biomass
Blooms/Counts
SPAIN
Biomass
Blooms/Counts
- 90
th
percentile of chlorophyll
a
data
- Frequency (%) of blooms of any single phytoplankton taxa
- Seasonal succession of functional groups
- 90
th
percentile of chlorophyll
a
data
- Frequency (%) of blooms of any single phytoplankton taxa.
REPUBLIC OF IRELAND
Biomass
- 90
th
percentile and median of chlorophyll-a data
Blooms/Counts
- Frequency (%) of blooms of any single phytoplankton taxa
- Seasonal succession of functional groups (under development)
NETHERLANDS
Biomass
Blooms
FRANCE
Biomass
Blooms/Counts
Composition
PORTUGAL
Biomass
Blooms/Counts
SWEDEN
Biomass
Blooms/Counts
- 90
th
percentile of chlorophyll
a
data
-
Phaeocystis
counts
- 90
th
percentile of chlorophyll
a
data
- Frequency (%) of blooms of any single phytoplankton taxa.
Blooms of all species
- Blooms of harmful species
- 90
th
percentile of chlorophyll
a
data
- Frequency (%) of blooms of any single phytoplankton taxa.
Blooms of all species
- 90
th
percentile and median of chlorophyll-a data, (original
Swedish metric: mean of chlorophyll-a data)
- Frequency (%) of phytoplankton, total cells of all taxa, divided
into two size fractions, >20 µm and 2-20 µm (>20 µm includes common
species, e.g.
Skeletonema costatum
and many
Chaetoceros
spp.)]
140
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0143.png
DENMARK
Biomass
BELGIUM
Biomass
Blooms/Counts
Composition
GERMANY
Biomass
Blooms/Counts
NORWAY
Biomass
Blooms/Counts
- 90
th
percentile of chlorophyll a data
- 90
th
percentile of chlorophyll a data
-
Phaeocystis
counts
- Blooms of harmful species (under development, thus not intercalibrated yet)
- 90
th
percentile of chlorophyll-a data
-
Phaeocystis
counts
- 90
th
percentile of chlorophyll-a data
- 90
th
percentile of biovolume/cellcarbon (under development)
- Frequency (%) of blooms of any single phytoplankton taxa >5 µm
(Need further development)
- Seasonal succession of functional groups (under development)
Table 2.4.3:
Description of national methods using a chlorophyll a metric.
Description of national method/
assessment of phytoplankton biomass
using the plant pigment chlorophyll
UK
Metric: 90
th
percentile of the
chlorophyll-a data set (µg L
-1
).
• Sampling frequency: monthly (*)
• Season: growth period, from March
to October (*)
• Period: 6-year (*)
Boundary classes agreed for national
assessment
Atlantic Coast - Exposed West Coast
of England & Northern Ireland
(NEA 1/26a)
High
-90
th
% <= 5
Good
-90
th
% <= 10
Mod
-90
th
% <=20
Poor
-90
th
% <= 40
Bad
-90
th
% > 40
North Sea/Eastern Irish Sea
(NEA 1/26b)
High
-90
th
% <= 10
Good
-90
th
% <= 15
Mod
-90
th
% <=20
Poor
-90
th
% <= 40
Bad
-90
th
% > 40
Spain
Metric: 90
th
percentile of the
chlorophyll-a data set (µg L
-1
).
• Sampling frequency: monthly (*)
• Season: growth period, from February
to October (*)
• Period: 6-year (*)
High
90
th
percentile < the first threshold
Good
90
th
percentile ≥ the first threshold and
< the second threshold
Moderate
90
th
percentile ≥ the second threshold
Note: some difference in
sampling strategy may occur
along the NEA Spanish Coast
The Spanish Coast has been
split into 2 NEA GIG Types
1/26a and 1/26e, and 5 regional
areas, boundary values for
H/G and G/M are given in the
‘Results of Harmonization’
section of this chapter.
Variation in methods
Chlorophyll is also investigated
in the elevated count tool.
141
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0144.png
ROI
Metric: 90
th
percentile and median of the
chlorophyll-a data set (µg L
-1
).
• Sampling frequency: monthly (*)
• Season: growth period, from March
to October (*)
• Period: 5-year (*)
High
Median and 90
th
below H/G boundary
Good
Median and 90
th
below G/M boundary
Moderate
Median and/or 90
th
above G/M
boundary
H/G boundary = 5 (µg L
-1
)
a
G/M boundary = 10 (µg L
-1
)
a
It differs from the proposed
metric in having a second
statistic based on the median
value.
use both acetone and
hot methanol for chlorophyll
extraction, hence the need to
develop two sets of reference
conditions relevant to the
analysis (see Table 5).
When reporting chlorophyll-a
data, based on the hot methanol
method, the data is not
corrected by subtracting for
Phaeopigment and is therefore
reported as Total Pigment.
a
RoI
France
Metric: 90
th
percentile of the
chlorophyll-a data set (µg L
-1
).
• Sampling period: March to October.
• Sampling frequency: once a fortnight
or once a month.
• Period: 6-year
• Concern coastal waterbodies
For the Channel and Atlantic coast
except North of France
(type NEA 1/26a)
High
90
th
% <= 5
Good
90
th
% > 5 and <= 10
Mod.
90
th
% > 10 and <= 20
Poor
90
th
% > 20 and <= 40
Bad
90
th
% > 40
For the French Northern coast
(type NEA 1/26b)
High
90
th
% <= 10
Good
90
th
% > 10 and <= 15
Mod.
90
th
% > 15 and <= 20
Poor
90
th
% > 20 and <= 40
Bad
90
th
% > 40
Metric: mean summer (March –
September) chlorophyll-a concentration
(µg/l) and on the maximum abundance
of Phaeocystis (10
6
cells/l).
The class boundaries for the 90 %ile
summer chlorophyll-a concentrations are
water-type specific.
For Phaeocystis the boundaries high/
good and good/moderate are set to
10
6
cells/l and 10
7
cells/l, respectively.
These boundaries are identical for all
waterbodies.
NEA 1/26b)
High
90
th
% <= 10
Good
90
th
% <= 15
Mod.
90
th
% <= 30
Poor
90
th
% <= 60
Bad 90
th
% > 60
142
Netherlands
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0145.png
Metric: 90
th
percentile of the
chlorophyll-a data set (µg L
-1
).
• Sampling frequency: once a fortnight
or once a month.
• Season: growth period, from February
to October
• Period: 6-year
For the North Sea and Norwegian
Sea (NEA 1/26a) and Skagerrak
(NEA 9)
High
90
th
% <= 2.5
Good
90
th
% > 2.5-5.0
Mod.
90
th
% > 5.0
For Skagerrak (NEA 8)
High
90
th
% <= 1.5
Good
90
th
% > 1.5-3.0
Mod.
90
th
% > 3.0
For Skagerrak (NEA 10)
Chlorophyll-a: Norway is
using methanol for chlorophyll
a extraction according to
Norwegian Standard 4767.
Norway
High
90
th
% <= 3.0
Good
90
th
% > 3.0-6.0
Mod.
90
th
% > 6.0
Metric: 90
th
percentile of the
chlorophyll-a data set (µg L
-1
).
NEA 1/26c – North Sea – German
Bight & Jutland
High/Good: 5 µg/L
Good/Moderate: 7.5 µg/L
NEA 1/26d – Denmark Skagerrak
High/Good: 3 µg/L
Good/Moderate: 4 µg/L
NEA 8 – (Kattegat + small area of
Skagerrak)
High/Good: 1.5 µg/L
Good/Moderate: 3 µg/L
Metric: 90
th
percentile of the
chlorophyll-a data set (µg L
-1
).
• Sampling frequency: monthly
• Season: growth period, March -
September
• Period: 6-year
High
Median and 90
th
below H/G boundary
Good
Median and 90
th
below G/M boundary
Moderate
Median and/or 90
th
above G/M
boundary
H/G boundary
NEA8 (Kattegat + small area of
Skagerrak): 1.5 µg/L
NEA9 (Skagerrak): 2.5 µg/L
NEA10 (Skagerrak): 3 µg/L
G/M boundary
NEA8 (Kattegat + small area of
Skagerrak): 3 µg/L
NEA9 (Skagerrak): 5 µg/L
NEA10 (Skagerrak): 6 µg/L
Metric: 90
th
percentile of the
chlorophyll-a data set (µg L
-1
).
• Sampling frequency: all available data
• Season: all available data
• Period: all available data Sa
Type NEA 1/26e
High
90
th
percentile <8 µg/L
Good
90
th
percentile ≥ 8 - <12
Moderate
90
th
percentile ≥12
Note: some difference in
sampling strategy may occur
along the NEA Portuguese
Coast. Data was sparse so all
available data was used
Chlorophyll-a The Swedish
classification tool differs by
using mean values during June-
August, resulting in different
reference values and borders
between the classes.
Portugal
Sweden
Denmark
143
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0146.png
Metric: 90
th
percentile of the chlorophyll
“a” data set (µg L
-1
).
• Sampling frequency: monthly or once
a fortnight from 2008 onwards
• Season: growth period, from March
to October
• Period: 5-year, in 2007 with lower
frequency
Metric: 90
th
percentile of the chlorophyll
“a” data set (µg L
-1
).
Belgium
NEA 1/26b
High
90
th
% <= 10
Good
90
th
% > 10 and <= 15
Mod.
90
th
% > 15 and <= 30
Poor
90
th
% > 30 and <= 45
Bad
90
th
% > 45
Chlorophyll is measured
with the same frequency
as the
Phaeocystis
counts
measurements (once a month
or once a fortnight, from 2008
onwards).
144
Germany
NEA GIG Type 1/26c – North Sea -
German Bight & Jutland –
High/Good: 5 µg L
-1
Good/Moderate:7.5 µg L
-1
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0147.png
Table 2.4.4:
Description of national methods using phytoplankton counts and/or composition.
Description of national method
1 Frequency ( %) of blooms of any
single phytoplankton taxa.
2 Frequency ( %) Phaeocystis above
10
6
Cells/l)
UK
Elevated counts of phytoplankton
groups are recorded and compared
against thresholds, and calculated as
a percentage of the total number of
sampling occasions.
There are 4 sub--metrics of
phytoplankton, including
Total cell count,
individual species count,
Phaeocystis
count and
Chlorophyll concentrations.
Final assessment is on the mean of the
four phytoplankton metrics
Boundary classes agreed for
national assessment
Variation in methods/
Comments on tools
Single (individual) taxa cell
counts:
Threshold: 250,000 cells per liter
(large phytoplankton)
Percentage exceeding threshold:
High – Count ( %) < 20 %
Good – Count ( %) < 20 %-39 %
Mod – Count ( %) < 40 %-69 %
Poor – Count ( %) < 70 %-90 %
Bad – Count ( %) > 90 %
Phaeocystis
cell count metric:
Threshold: 1,000,000 cells per liter
Percentage exceeding threshold:
High – Count ( %) < 10 %
Good – Count ( %) 10 %-17 %
Mod – Count ( %) > 17 %
Poor – Count ( %) > 35 %
Bad – Count ( %) > 80 %
All geographical areas except
Iberian upwelling coast and
Western Cantabrian coast
NEA 1/26a
High -Frequency
of samples
exceeding the threshold < 20 %
Good-Frequency
of samples
exceeding the threshold ≥ 20 % and
< 40 %
Moderate-Frequency
of samples
exceeding the threshold ≥ 40 %.
Iberian Coast and Western
Cantabrian Coast NEA 1/26e
High -Frequency
of samples
exceeding the threshold < 30 %
Good-Frequency
of samples
exceeding the threshold ≥ 30 % and
< 50 %
Moderate-Frequency
of samples
exceeding the threshold ≥ 50 %.
NEA Type 1/26a
High -Frequency
of samples
exceeding the threshold < 20 %
Good -Frequency
of samples
exceeding the threshold ≥ 20 % and
< 40 %
Moderate -Frequency
of samples
exceeding the threshold ≥ 40 %.
Elevated counts of single
species are taken as a sub-
metric within a tool, and final
assessment is calculated as
a combination of the 4 sub-
metrics
Other tools are used in the
national assessment including
Seasonal succession of
functional groups
Community indicator species
(under development)
Boundaries apply to UK NEA
Type 1/26a & NEA Type 1/26b
Spain
The phytoplankton taxa included in
the analysis are the species routinely
monitored in coastal waters.
Three different approaches are
followed depending on the degree
of expertise in the identification and
counting step
Only diatoms, dinoflagellates and
euglenophyceans are considered.
Small flagellates and coccoids are not
included.
Small flagellates and coccoids are
included and the phytoplankton is
classified by two size categories:
small (< 20
µm)
and large
phytoplankton (> 20
µm).
Small flagellates and coccoids are
included, but size categories are not
established.
Sampling frequency: monthly (*)
Season: whole year, from January to
December (*)
Period: 6-year (*)
The use of two thresholds
(large and small phytoplankton)
is not convenient for the
national methodology. But, for
intercalibration purposes in the
coast of the País Vasco (Eastern
Cantabrian coast) thresholds
have been calculated for
large (75,000 cells/L) and
Small phytoplankton
(750,000 cells/L).
Note: some difference in
sampling strategy may occur
along the NEA Spanish Coast
NB: Cells/litre thresholds
for the other Spanish regions
and types are documented, in
the reference conditions and
‘Results of Harmonisation’
sections of this chapter.
ROI
Percentage of samples where counts of
individual phytoplankton taxa (> 20
µm)
are above a threshold of 250,000 cells/l.
Sampling frequency: monthly between
January and December.
Assessment period: 5-6 years
For Irish coastal waters it will
be necessary to distinguish
between phytoplankton
populations that are known to
originate offshore and show
no response to local nutrient
conditions and those resident
populations that do. Blooms
that are considered to have
originated offshore (based
on existing knowledge of
the region’s oceanography)
rather than in response to in
situ growth conditions will be
omitted from the analysis.
145
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0148.png
France
Metric: percentage of samples where a
single taxa count is above the exceeding
threshold.
Definition of the exceeding threshold:
100 000 cells per liter for large cells,
and 250 000 cells per liter for small
cells.
Sampling period : all the year
Sampling frequency: once a fortnight
or once a month.
Concern coastal waterbodies
Period: 6-year
Blooms of all species
Types NEA 1/26a & 1/26b
High
- Count ( %) < 20
Good
- Count ( %) >= 20 and < 40
Mod –
Count( %) >= 40 and < 70
Poor –
Count( %) >= 70 and < 90
Bad –
Count ( %) >= 90
Portugal
Phytoplankton counts are carried
out. When elevated chlorophyll
concentrations are detected (>12µg L
-1
G/M threshold).
Elevated counts of phytoplankton
single taxon groups are recorded
and compared against the cell count
threshold, (microphytoplankton 100,000
cells /L, micro+nanophytoplankton
1,000,000 cells /ml). The occurrences
of single taxon blooms are calculated
as a percentage of the total number of
sampling occasions.
The metrics are
total cell count,
single taxon counts
Iberian Upwelling Coast
NEA 1/26e
High -Frequency
of samples
exceeding the threshold < 30 %
Good-Frequency
of samples
exceeding the threshold ≥ 30 % and
< 50 %
Moderate-Frequency
of samples
exceeding the threshold ≥ 50 %.
The
Phaeocystis
metric
is
not applicable for the Iberian
Upwelling Coast
Denmark
Norway
N/A
Percentage of samples where a single
taxa count is above the exceeding
threshold.
Definition of the exceeding threshold:
500 000 cells per liter.
Sampling period : all the year
Sampling frequency: once a fortnight
or once a month.
Period: 6-year
N/A
Blooms of all species, frequency of
samples – TBC
High
- Count ( %) < 20
Good
- Count ( %) >= 20 and < 40
Mod. –
Count( %) >= 40 and < 70
Poor –
Count( %) >= 70 and < 90
Bad –
Count ( %) >= 90
N/A
The metric need further
development because among
other factors the cell size is not
taken into consideration.
Postponed to phase II
intercalibration.
Sweden
Percentage of samples with Total cells of
phytoplankton above a set threshold.
Microplankton (>20 µm):
Tentative – no threshold (cells/L) at
present
Nanoplankton (2-20 µm):
Tentative – no threshold (cells/L) at
present
Sampling frequency: monthly
Season: growth period,
March - September
Period: 6-year
Total cell counts, frequency of
samples - TBC
High
– < 20 %
Good
– 20 % - 39 %
Moderate –
40 % - 69 %
Poor –
70 % - 90 %
Bad –
> 90 %
This metric is not fully
developed and evaluated for
the Swedish NEA GIG 8, 9, 10,
but needs further work, if to be
used at all.
In Sweden phytoplankton
biovolume mean values during
June-August have been used,
resulting in different reference
values and borders between the
classes.
Postponed to phase II
intercalibration.
146
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0149.png
Belgium
Metric
Phaeocystis
counts
Sampling frequency: monthly or once
a fortnight from 2008 onwards
Period: 5-year
Cell Threshold: 1,000,000 Cells/l
% Phaeocystis samples exceeding
threshold (NEA Type 1/26b)
HIGH - <10 %
GOOD - 10 – 17 %
MODERATE - >17 %
POOR- >35 %
BAD- >80 %
% Phaeocystis samples exceeding
threshold (NEA Type 1/26b)
HIGH - <10 %
GOOD - 10 – 17 %
MODERATE - >17 %
POOR- >35 %
BAD- >80 %
NEA GIG Type 1/26c – North Sea
- German Bight & Jutland & NEA
Type 3/4 – North Se - Ems, Weser,
Elb Coast
HIGH - <10 %
GOOD - 10 – 17 %
MODERATE - >17 %
POOR- >35 %
BAD- >80 %
A metric evaluating
elevated counts of harmful
phytoplankton taxa is under
development.
Netherlands
Metric
Phaeocystis
counts
Sampling frequency: monthly during
growing season (March-Sept (incl.))
Period: 6-year
Cell Threshold: 1,000,000 Cells/l
Sampling takes place in the
growing season only. The
frequency is calculated under
the assumption that there are no
Phaeocystis
blooms outside the
growing season.
Germany
Metric
Phaeocystis
counts
Cell Threshold: 1,000,000 Cells/l
2.4.3 Reference conditions and class boundary setting
Initially reference conditions for the phytoplankton quality element were put forward by each MS,
using established background criteria from unimpacted sites (assumed to be in reference condition),
modelling and regression from impacted site data, marine conventions, scientific literature and
through expert agreement in the GIG. The diversity of measures required for setting reference
conditions reflects the extremely diverse and localised ecohydrodynamics of the NEA GIG causing
gross and subtle modification to phytoplankton responses to nutrients and climate.
The following outlines the reference setting protocols established by each Member state, and the
reference thresholds that were derived for each of the different types.
United Kingdom (NEA GIG TYPE 1/26A and 1/26B)
• Atlantic Coast – Exposed West Coast of England and Northern Ireland (NEA 1/26a)
• North Sea, Eastern Irish Sea and Channel (NEA 1/26B)
Chlorophyll-a metric
An appropriate standard for assessing chlorophyll
a
concentration can be derived from the
background nutrient concentrations by making some reasonable assumptions about nutrient
conversion to plant biomass. From practical experience the UK has adopted 10 µg L
-1
chlorophyll
a
as a guide for assessment in North Sea, Channel and Eastern Irish Sea Waters. It is therefore
proposed that for UK Type 1/26b offshore waters that 10 µg L
-1
chlorophyll
a
is adopted as the 50
% elevated value (implying a background value of 6.7 µg L
-1
and a reasonable C : Chl factor of
0.012). For further information on conversion of C : Chl see Malcolm
et al
(2002).
For Atlantic Coastal Waters which include the Exposed West Coast of England and Northern Ireland,
the UK has adopted 5 µg L
-1
chlorophyll
a
as a guide for assessment. It is therefore proposed that
for Atlantic Coast offshore waters 5 µg L
-1
chlorophyll
a
is adopted as the 50 % elevated value
(implying a background value of 3.33 µg L
-1
).
147
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0150.png
Proposed Chl-a reference values for UK waters
• NEA GIG Type 1/26A – Atlantic Coast – 3.33 µg L
-1
• NEA GIG Type 1/26B – North Sea, Channel, Eastern Irish Sea – 6.7 µg L
-1
Taxa count metric
Proposed reference values for the taxa count metric – UK
• NEA GIG Type 1/26a – Atlantic Coast & NEA 1/26b – North Sea, Channel, Eastern Irish Sea
Cell Threshold: 250,000 cells/L (large phytoplankton)
• Reference Percentage exceedent threshold: 16.7 % (2 samples, associated with natural spring
and autum blooms. Assumes monthly sampling)
• Percentage exceedent thresholds: H/G 20 %, G/M 40 %
Phaeocystis metric
Proposed reference values for the
Phaeocystis
metric – UK
• NEA GIG Type 1/26a – Atlantic Coast & NEA 1/26b – North Sea, Channel, Eastern Irish Sea
Cell Threshold: 1,000,000 cells/L.
• Reference percentage exceedent threshold: 8.30 % (1 sample in 12 months as part of natural
spring bloom. Assumes monthly sampling)
• Percentage exceedent thresholds: H/G 10 %, G/M 17 %
Spain (NEA GIG 1/26a and 1/26e)
The NEA coast of Spain has been divided into five different sub-areas that reflect the natural
variation in abiotic factors driving the phytoplankton communities: Eastern Cantabrian Coast,
Western Cantabrian Coast, Western Iberian Coast (upwelling), Canary Islands and Atlantic Coast
(Southern Spain). A summary of the boundary setting protocol in each of these areas is provided
below.
Eastern Cantabrian Coast (NEA 1/26a)
At the North of Spain, it extends between the border with France in País Vasco (43º 22´ N, 1º 46´
W) and Cape Peñas in Asturias (43º 39´ N, 5º 50´ W).
Chlorophyll-a metric
The 90
th
percentile of the chlorophyll-a in reference stations ranged between 0.8 µg L
-1
(País
Vasco) and 2.8 µg L
-1
(Asturias). If these values are compared to those observed in other Atlantic
coastal waters (e. g. France, Portugal) the phytoplankton biomass in the Eastern Cantabrian coast is
considerably lower. However the specific hydrographical conditions of the Eastern Cantabrian coast
in relation to other Atlantic areas can explain the differences in chlorophyll “a” concentration. The
Cantabrian shelf is narrower compared to the French shelf. Also, nutrient fluxes to the Cantabrian
shelf are distributed among several small rivers along the coast and no large plumes are formed
(Diez et al. 2000). The intensity of the Iberian upwelling system decreases eastward along the
Cantabrian shelf and it only slightly affects the Eastern Cantabrian Coast (Valencia et al. 2004;
Mason et al. 2005; Lavin et al. 2006).
148
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
Taxa count metric
Phytoplankton counts were conducted since 2002 in the coastal waters of País Vasco. The
anthropogenic eutrophication impact in the studied stations is considered negligible due to the
strong hydrodynamic forces that act on this coast and, therefore, the phytoplankton is assumed to
be in High or Good status. Low chlorophyll concentrations (see above) support this assumption.
Water samples were collected at surface every 6 months. Sampling seasons corresponded to spring
(May) and summer (August). Standard methods were used for field and laboratory: Van-Dorn
bottle sampling, inverted microscopy and Uthermöhl. Based on the work of Borja et al. (2004), a
bloom frequency of 20 % was considered the boundary between High and Good status and 40 %
the boundary between Good and Moderate status. In order to calculate the abundance threshold that
defines a bloom,
three different approaches were followed:
1.
Only diatoms, dinoflagellates and euglenophyceans were included in the data analysis. Small
flagellates and coccoids were not included. By this method,
500,000 cells L
-1
was the most
suitable threshold to define a bloom.
2.
Small flagellates and coccoids were included in the data analysis and the phytoplankton
community was split into two size categories: small (< 20
µm)
and large phytoplankton (>
20
µm).
By using a criteria based on the Equivalent Spherical Diameter (ESD), 102 taxa
were classify as large phytoplankton and 126 taxa as small phytoplankton. The thresholds
that best fitted with this method were:
75,000 cells L
-1
for large phytoplankton and
750,000 cells L
-1
for small phytoplankton.
3.
Small flagellates and coccoids were included in the data analysis, but size categories were
not established. The resulting threshold to be applied to any single phytoplankton taxa
(regardless of its cellular size) was
750,000 cells L
-1
.
The most accurate method was based on the use of two different thresholds, for large and small
phytoplankton respectively (approach 2). However, the size-fractionation method was very time-
consuming as it required obtaining and processing the information of the ESD for the phytoplankton
taxa. Therefore, it was found not suitable as a routine phytoplankton monitoring method. Other
alternative methods were based on the use of a single threshold (approach 1 and 3) and the results
obtained by them in the Basque coast (País Vasco) were acceptable and very similar. At present, in
the Eastern Cantabrian Coast the approach 3 and the 750,000 cells L
-1
threshold have been chosen
because the monitoring programs provide accurate counts of the small phytoplankton forms.
However, it is important to note that the abundance threshold to define a phytoplankton bloom
depends on the methodology used (the counting expertise and the type of data analysis employed).
This is evidenced by the different thresholds obtained with similar samples when using the
approach 1 (only some taxonomic groups), the approach 2 (size-fractionation) or the approach 3 (all
phytoplankton taxa).
Western Cantrabrian Coast (NEA 1/26e)
At the North of Spain, it extends between Cape Peñas in Asturias (43º 39’ N, 5º 50’ W) and Cape
Estaca de Bares in Galicia (43º 47´ N, 7º 41´ W).
Chlorophyll -a metric
The chlorophyll “a” data were obtained in the coast of Asturias (43º 36´ N, 6º 08´ W). The 90
th
percentile value in the reference station was 2.2 µg L
-1
. In addition to the time series of field data
from Asturias, the influence and regional variation of the upwelling process over the Western
Cantabrian coast needs to be taken into consideration to establish sensible limits for the whole
region. The review of published data by Bode et al. (1996) showed geographic differences in the
upwelling regime, so that the mean chlorophyll “a” concentration increases westward along the
149
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
N-NW Spanish coast in most oceanographic stages. These geographic differences are lower during
unproductive phases (summer stratification and winter) and the highest during the upwelling
stage, when the mean integrated chlorophyll “a” concentration increases twofold from the Western
Cantabrian Coast to the Western Iberian Coast (Southern Galicia). In turn, this geographic gradient
in the mean chlorophyll “a” concentration should reflect on a gradient for the limits of the 90
th
percentile chlorophyll “a” along the N-NW coast of the Iberian Peninsula.
Taxa count metric
By expert judgement, in accordance with the westward gradient of increasing upwelling influence
along the N-NW Spanish coast, 750,000 cells L
-1
is proposed as the most suitable threshold for
defining a phytoplankton bloom in the Western Cantabrian Coast and, similarly to the Iberian
upwelling coast (see below), 30 % and 50 % as the boundaries between classes (High/Good, Good/
Moderate). As with phytoplankton biomass, it is important to stress that the current paucity of data
available for intercalibration, together with the high spatial and temporal variability associated to
the wind-driven upwelling processes, may require a thorough revision of the proposed thresholds,
once suitable databases become available.
Western Iberian Coast (upwelling) (NEA 1/26e)
The Iberian upwelling coast in Spain extends along the coast of Galicia, from the Cape Estaca de
Bares (43º 47’ N, 7º 41’ W) to the border with Portugal.
Hydrographic changes associated with upwelling and their influence on nutrient regimes and
phytoplankton cycles in the Galician coast are well-documented. Many studies have focused on the
hydrographical and biological features of the Iberian upwelling system with particular emphasis
on the enhancement of biological productivity (Fraga 1976; 1981; Álvarez-Salgado et al. 1996;
Moncoiffé et al. 2000; Cermeño et al 2006). The Galician coast is also characterised by the presence
of wide coastal embayments or rías. These are deep estuaries (ca. 40 m depth near the mouth),
whose intense exchange of water, and hence materials, with the ocean leads to their consideration as
coastal intrusions of the shelf. The fertilizing process in the rías leads to higher primary production
rates than in other estuarine or coastal systems (Boynton et al. 1982). For example in the Ría de
Vigo mean annual primary production rates are about 800 mg C m
-2
d
-1
(Prego 1994) and the
maximum values from spring to autumn range between 700 and 1800 mg C m
-2
d
-1
(Vives and
Fraga 1961, in Nogueira et al. 1997). In terms of biomass and primary production there are not
significant differences between the upwelling season and the spring bloom. Although seasonally
variable in frequency and intensity, the upwelling is an episodic process driven by the wind. The
close relationship between wind regime, nutrient inputs and chlorophyll concentration has been
reported in several papers (Moncoiffé et al. 2000, Cermeño et al. 2006). This close relationship with
the wind regime reflects on a high frequency of phytoplankton blooms and very high short-scale
variability in sea surface chlorophyll “a” concentration.
Chlorophyll-a metric
The data were obtained at the Ría de A Coruña (43° 21.8’ N, 8° 22.2’ W) and in the middle zone
of the Ría de Vigo (42º 14.5’ N, 8º 45.8’ W). The calculated values of the 90
th
percentile were 4.73
µg L
-1
at the A Coruña station (monthly sampled, from 1992 to 2006) and 10.55 µg L
-1
at the Vigo
station (two samples per week, from 1987 to 1996). A previous analysis by the INTECMAR Centre
in Galicia (Phytoplankton
tools for NEA type 1 Galician waters),
calculated the 90
th
percentile from
37 stations distributed in the southern part of the Spanish upwelling coast (between the border with
Portugal and Cape Finisterre) and 13 stations in the northern part (between Cape Finisterre and
Cape Estaca de Bares). All the stations were weekly sampled from 1992 to 2004 and the values
obtained for the 90
th
percentile were 4 and 5 µg L
-1
in the southern and northern part of the Spanish
upwelling coast, respectively.
150
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
Taxa count metric
Phytoplankton cell counts in the Ría de A Coruña (43° 21.8’ N, 8° 22.2’ W) were carried out from
1992 to 1998. Samples were collected monthly, preserved with Lugol’s solution and counted under
an inverted microscope following the technique described by Uthermöl (1958). The Ría de A
Coruña can be considered a bay with important oceanic influence, whose status is considered High
or Good. According to the data from the Ría de A Coruña, a value of 10
6
cells L
-1
is proposed as the
threshold for a phytoplankton bloom. This coincides with the value proposed by the INTECMAR
Centre (Phytoplankton
tools for NEA type 1 Galician waters)
after a study of data from 40 stations
weekly sampled along the Galician coast from 1999 to 2004. According to the data analysis and the
knowledge of the high variability in sea surface phytoplankton in the area (see above), the High/
Good and Good/Moderate boundaries are set at 30 % and 50 % bloom frequencies, respectively.
Canary Islands (NEA 1/26a)
Chlorophyll-a metric
The Canary Islands (~28° N) are located in a transition zone between the oligotrophic waters of the
open ocean and the NW Africa upwelling system (Barton et al. 1998). Although there is a slight
increasing gradient in chlorophyll-a towards the Eastern islands, the mean values are low during
most of the year all around the Archipelago and maxima are typically < 1 µg L
-1
(Aristegui 1990)
.
Taxa count metric
Taking into account the oligotrophic character of the Canary Islands in comparison to other Spanish
NEA coastal waters, 500,000 cells L
-1
is proposed as the threshold for defining a phytoplankton
bloom and 20 % and 40 % as the boundaries between classes (High/Good, Good/Moderate) in this
coastal area.
Atlantic Coast (NEA 1/26a)
The Atlantic coast in the South of Spain extends from the mouth of the Guadiana River at the
Portuguese border to the mouth of the Guadalmesí River at the Straight of Gibraltar. This coastline
belong to the area of the Gulf of Cadiz in Andalucía and it can be distinguished from the coastline
open to the ocean (from Cape San Vicente to the North of Spain) as it represents an area of transition
between the Mediterranean Sea and the Atlantic Ocean (OSPAR 2000).
Chlorophyll-a metric
The Atlantic coast in the South of Spain presents a large continental platform and it receives the
runoff from two important rivers (Guadalquivir and Guadiana Rivers).
In addition, some upwelling influence has been described in the Gulf of Cadiz (Garcia et al.
2002). Also, the continuity with the Portuguese coast, westward of the Guadiana mouth, must be
considered and the thresholds for phytoplankton biomass should not differ importantly between
them. Extensive data series for chlorophyll in reference monitoring stations are not available.
Therefore, on the basis of expert judgment, the thresholds proposed for the 90
th
percentile of
chlorophyll “a” in the Atlantic coast (S Spain) are 5 µg L
-1
(between high and good status) and 10
µg L
-1
(between good and moderated status). Nevertheless, in the future, chlorophyll data that are
being acquired now could lead to consider a new definition of these thresholds for the Atlantic coast
in the South of Spain.
Taxa count metric
For the determination of the cell abundance that indicates a bloom event in the Gulf of Cadiz, data
series corresponding to 6-year periods have been studied. The minimum sampling frequency was
monthly and, in some cases, weekly or fortnightly. Small phytoplankton forms (mainly flagellates)
151
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0154.png
were not recorded. Different abundance values to define a phytoplankton bloom were studied both,
in reference stations and, in other areas considered with a slightly higher level of eutrophication.
Attending to the results, the threshold of 500,000 cells L
-1
was set to define a bloom event, with
20 % and 40 % as the boundaries between classes (High/Good, Good/Moderate).
Proposed reference values for the chlorophyll “a” (90
th
percentile) metric – Spain
• Eastern Cantabrian Coast (NEA Type 1/26a) 1 - 3 µg L
-1
• Western Cantabrian Coast (NEA Type 1/26e) 2 – 5 µg L
-1
• Western Iberian Coast upwelling (NEA Type 1/26e) 4 -11 µg L
-1
• Canary Islands (NEA Type 1/26a) < 1 µg L
-1
• Atlantic Coast (NEA Type 1/26a) < 5 µg L
-1
Proposed reference values for the taxa count metric – Spain
• Eastern Cantabrian Coast (NEA1/26a)-Threshold 750,000 cells/L- % Frequency Exceeding cell
threshold: Reference 16.70 %, H/G 20 %- G/M 40 %
• Western Cantabrian Coast (NEA1/26e)- Threshold 750,000 cells/L- % Frequency Exceeding
cell threshold: Reference 25 %, H/G 30 %- G/M 50 %
• Western Iberian Coast (upwelling, NEA1/26e)- Threshold 1,000,000 cells/L- % Frequency
Exceeding cell threshold: Reference 25 %, H/G 30 %- G/M 50 %
• Canary Islands (NEA1/26a)- Threshold 500,000 cells/L- % Frequency Exceeding cell threshold:
Reference 16.70 %, H/G 20 %- G /M 40 %
• Atlantic Coast (NEA1/26a)- Threshold 500,000 cells/L- % Frequency Exceeding cell threshold:
Reference 16.70 %, H/G 20 %- G/M 40 %
Portugal
NEA GIG TYPE 1/26e Western Iberian Coast (upwelling)
Chlorophyll-a metric
The coast of Portugal extends north from Cape St Vincent to the border with northern Spain at
the Minho. This coastline has a narrow continental shelf and is subject high hydrodynamics and
upwelling.
The south coast of Portugal extends west from Cape St Vincent to the border with southern Spain at
the Guadiana. Although less exposed than the west coast it also has a narrow continental shelf and is
subjected to upwelling, (Garcia
et al.
2002).
The continuity with the Andalucian coast, east of the Guadiana mouth, and the northern border
with Spain must be considered, and the thresholds for phytoplankton biomass should not differ
significantly between Portugal and Spain. The 90
th
percentile of chlorophyll “a” for Portugal is 5 µg
L
-1
(boundary 8 µg L
-1
between high and good status) and 10 µg L
-1
(boundary 12 µg L
-1
between
good and moderate status). Nevertheless, these may be revised in the future, as more chlorophyll
data becomes available.
Proposed Reference chl-a values for Portuguese waters
• NEA GIG TYPE 1/26e Western Iberian Coast (upwelling) -
• Reference condition <4 µg L
-1
, High/ Good- 8 µg L
-1
, Good/Moderate - 12 µg L
-1
152
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0155.png
Taxa count metric
Proposed Reference taxa counts values for Portuguese waters
• NEA GIG TYPE 1/26e Western Iberian Coast (upwelling) Cell Threshold: 100,000 cells/L (large
phytoplankton threshold), 1,000,000 cells/L (small phytoplankton threshold)
• Percentage exceedent thresholds: Reference = 25 % (3 blooms expected per year, spring,
summer and autumn, with monthly sampling), H/G <30 %, G/M <50 %
Republic of Ireland
NEA GIG TYPE 1/26a (Atlantic Coast and Western Irish Sea)
Chlorophyll-a metric
Reference criteria were based on contemporaneous data collected from undisturbed waterbodies
that were considered to be close to or at reference as determined by expert judgment, the Article
5 Characterisation report and supporting environmental data. The proposed class boundaries for
excessive cell counts were based on the analysis of phytoplankton data collected from coastal
waterbodies that are considered to be close to or at reference. Reference criteria were provided for 2
of the 3 intercalibration common metrics (i.e. chlorophyll biomass and excessive cell counts).
The metric based on
Phaeocystis
was not considered to be relevant for Irish coastal waters and was
therefore not included in Ireland’s intercalibration exercise.
A 100 % agreement was obtained in a comparison of the intercalibration common metric against
the national assessment method (see Appendix x). It should be noted that the chlorophyll data used
by the Republic of Ireland for the intercalibration exercise was derived using 2 different extraction
methods that are known to have different extractant efficiencies. Intercomparisons of the two
extractants have shown that chlorophyll levels extracted using the hot methanol method can be, on
average 25 - 40 % higher than cold acetone. The difference can vary considerably depending on
phytoplankton species composition at time of sampling. As a consequence of these differences it
was deemed necessary to develop separate class boundary criteria for both extraction methods.
The boundary setting criteria for chlorophyll in Irish coastal waters, for both extraction techniques,
is summarised in Table 5. The difference in magnitude between both sets of criteria, with the hot
methanol criteria being twice that of the cold acetone, appears to be consistent with observations,
first, that methanol is the more efficient extractant and second, that the hot methanol technique is
likely to overestimate chlorophyll values because it does not correct for degradation products.
Table 2.4.5:
Median and 90
th
percentile-a (µg L
-1
) boundary criteria for different pigment extraction techniques.
Extractant
Median
Boundary
high/good
good/moderate
2.5
5.0
5.0
10.0
5.0
10.0
10.0
20.0
Cold Acetone
90 %ile
Median
Hot Methanol
90 %ile
153
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0156.png
Proposed reference chl-a values for ROI waters
• NEA GIG Type 1/26A – Irish Sea (Acetone Method) – 3.7 µg L
-1
• NEA GIG Type 1/26A – Atlantic Sea (Acetone Method) – 3.4 µg L
-1
Taxa count metric
Proposed reference taxa count values for ROI waters
• NEA GIG Type 1/26A – Irish Sea & Atlantic Sea Cell Threshold: 250,000 cells/L (Large
phytoplankton)
• Reference Percentage exceedent threshold: 16.7 % (2 samples, associated with natural spring
and autumn blooms. Assumes monthly sampling)
• Percentage exceedent thresholds: H/G 20 %, G/M 40 %
Netherlands
NEA Type 3, 4 & NEA Type 1/26b
• Holland and North Delta Coast (NEA Type 3)
• Dutch Wadden Sea (NEA Type 4)
• Zeeland Coast and Wadden Coast (NEA Type 1, 1/26b)
Chlorophyll-a metric
Originally all coastal waters of UK (east) and NL were classified as one type (NEA1) and therefore
had the same reference values. However, with regard to the mean salinity values the Holland Coast
more properly fell in the NEA3 type. The difference between the UK discharges and the Dutch
discharges is a factor 1.7, and between the German discharges and the Dutch discharges is a factor
2.3. Assuming that the concentrations of nutrients in the pristine situation were identical in the
British and the continental rivers, the nutrient load on the Holland Coast can be expected to be 1.7
times higher than that in the British waters and 2.1 times higher than in the German waters. New
class boundaries of NEA3 have been calculated based on freshwater discharges. Therefore the
elevated values for the 90-percentile of the chlorophyll-a concentrations in the Holland Coast will
be in the range of 11 to 17 µg L
-1
i.e. at least a factor 2.1 higher than the values of the German
Coast and at most 1.7 higher than the values for the English coast due to the higher run off of the
Netherlands/Holland Coast typology.
According to Cadée and Hegeman (2002) the annual average chlorophyll-a concentrations in
the 70’s of the last century have been between 2 and 5 µg L
-1
. With the relationships between
annual mean and summer mean, the related range of the summer mean could be calculated and by
multiplying the numbers of this range with the factor 2 (relation mean – 90-percentile) the range of
the summer 90-percentile has been derived, as being 5 – 13 µg L
-1 1
. Assuming that the chlorophyll
concentrations in the 70’s were elevated in relation to the undisturbed background concentrations,
but still below 1.5 times the background, the value 14 μg/l deduced from the freshwater discharges
agrees very well with the value deduced from Cadée & Hegeman (2002). In the framework of the
“Watersysteemverkenning” (Water System Exploration) in the Netherlands, so called reference values
(which were in fact the boundaries between High and Good) for a number of functional groups and
individual species have been calculated (Baptist & Jagtman, 1997). For chlorophyll the “reference”
value has been expressed as the 90-percentile value of the concentration. The calculation of this value
has been based on model simulations and expert judgement and is calculated as 14.3 μg/l for NEA3 in
Dutch coastal waters. Taking 14 µg/l as boundary between High and Good gives a reference value of
9.3 µg L
-1
(2/3 * 14) and a boundary between Good and Moderate of 21 µg L
-1
(1.5 *14).
154
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0157.png
The waterbodies Zeeland Coast and Wadden Coast are belonging to the euhaline water type, NEA1.
Although the influence of freshwater in the Zeeland Coast is slightly higher than in UK coastal
waters the class boundaries for the 90-percentile summer chlorophyll-a concentrations for both
areas could be set on 10 and 15 µg L
-1
, with a reference values of 6.7 µg L
-1
Class boundaries for NEA4 for the Dutch Wadden Sea are set equal to NEA3 (14 and 21 µg L
-1
),
based on comparable influence of Rhine discharges.
Proposed reference chl-a values for Netherlands Waters
• NEA GIG Type 4 – Dutch Wadden Sea – 9.3 µg L
-1
• NEA GIG Type 3 – Holland and North Delta Coast – 9.3 µg L
-1
• NEA GIG Type 1 (1/26b) – Zeeland Coast and Wadden Coast – 6.7µg L
-1
Phaeocystis metric
Proposed reference
Phaeocystis
values for Netherlands waters
• NEA GIG Type 1 (1/26b) - Zeeland Coast and Wadden Coast, NEA Type 3 - Holland and NEA
Type 4 - North Delta Coast, Dutch Wadden Sea - Cell threshold: 1,000,000 cells/L
• Reference percentage exceedent threshold: 8.30 % (1 sample in 12 months as part of natural
spring bloom. Assumes monthly sampling)
• Percentage exceedent thresholds: H/G 10 %, G/M 17 %
France
Chlorophyll-a metric
French Channel and Atlantic coast (except Northern part) / type NEA 1/26a
Reference values are based on data collected from undisturbed sites that are considered to be
close to or at reference as determined by expert judgment. The 90
th
percentile calculated on these
reference sites shows that the reference values may be different between sites:
site
Cherbourg Port / 2004-2005
Donville / 2002-2006
Chausey / 2001-2006
Bréhat / 1992-2006
Ouessant - cale de Porz Arlan / 2006
Groix nord / 1996-2006
Port Joinville / 2006
Bouée 7 / 2003-2006
90
th
percentile
4.40
2.58
2.29
1.71
1.30
3.99
2.20
2.85
This is considered as reflecting the diversity of the French coast, since the values are consistent
between large geographical regions: the lowest values concern the Northern and Western parts
of Brittany, while the highest ones concern primarily Normandy. The reference value is therefore
proposed as a range between the lowest (1.30) and the highest value (4.4) of these reference sites.
155
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0158.png
Northern part of the French coast (North Sea) / type NEA 1/26b
No reference site could be found in this region. However, data collected in two offshore sites (site 4
SRN Dunkerque and site 3 SRN Boulogne) show that the values of 90
th
percentile chlorophyll-a on
the period 1992-2006 are respectively 9.9 and 6.2 µg/L. It may be considered by expert judgment
that these values are consistent with the reference value calculated for the type NEA 1/26b. So it is
decided that, in the absence of reference site for this part of French coast, the calculated value of
6.7 µg/L should be the reference value.
Proposed reference chl-a values for French Waters
• NEA GIG Type 1/26a – French Channel and Atlantic Sea – 1.3 – 4.4 µg L
-1
• NEA GIG Type 1/26b – French North Sea – 6.7 µg L
-1
Taxa count metric
Proposed reference taxa count values for French waters
• NEA GIG Type 1/26a – French Channel and Atlantic Sea – 100,000 cells/L (Large
phytoplankton threshold, 250,000 cells/L (Small phytoplankton threshold)
• Reference Percentage exceedent threshold: 16.7 % (2 samples, associated with natural spring
and autumn blooms. Assumes monthly sampling)
• Percentage exceedent thresholds: H/G 20 %, G/M 40 %
Sweden
NEA GIG TYPE 8, 9 and 10
• Öresund and Kattegat (8)
• Skagerrak (9)
• Skagerrak (10)
Chlorophyll-a metric
Sweden is using the method outlined in the Combine manual of HELCOM (www.helcom.fi).
In short this means; fluorometric measurement, calibrated against spectrophotometric method,
ethanol as solvent. Sweden reports Chlorophyll-a, without subtraction of Phaeopigment.
Along the Swedish west coast (Skagerrak, Kattegat and Öresund) reference values of chlorophyll
are based on off shore data. Temporal changes of Secchi depth and Total Nitrogen were used to
establish “unaffected” (background) values. The relations between Secchi depth, Total Nitrogen,
salinity and chlorophyll-a were then used to establish the Chlorophyll a reference value for each
Type Area. Boundary classes were calculated using the relationships between Secchi depth, Total
Nitrogen and chlorophyll-a. The boundary between GOOD and MODERATE was set with the
assumption that the Skagerrak and Kattegat are not unaffected by man. Other class boundaries were
set by a 50-75 % increase between the classes.
Proposed reference chl-a values for Swedish waters
• NEA GIG Type 8 – Öresund and Kattegat – 1.0 µg L
-1
• NEA GIG Type 9 – Skagerrak – 1.7 µg L
-1
• NEA GIG Type 10 – Skagerrak – 2.0 µg L
-1
156
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0159.png
Abundance of phytoplankton as a metric has not been used in Sweden and reference values and
boundaries can therefore not be presented at present. However Sweden is developing phytoplankton
abundance as a metric, which can be used as a parallel to the Swedish phytoplankton metric
BIOVOLUME of phytoplankton.
Norway
Chlorophyll-a metric
The method for measuring chlorophyll-a in Norway is in accordance to Norwegian Standard 4767
using methanol for pigment extraction. There is no correction for pheopigments.
For establishing reference values and boundary criteria “Classification of environmental quality
in fjord and coastal waters. A guide” (Norwegian Pollution Control Authority 1997) and expert
judgment has been used. Lack of long term series from monitoring programs from relevant areas
have prevented use of historical data.
Proposed reference chl-a values for Norwegian waters
• NEA GIG Type 1/26a – North Sea, Norwegian Sea – 1.7 µg L
-1
• NEA GIG Type 8 – Skagerrak (see map) – 1.0 µg L
-1
• NEA GIG Type 9 – Skagerrak (see map) – 1.7 µg L
-1
• NEA GIG Type 10 – Skagerrak (see map) – 2.0 µg L
-1
Norway has not used abundance of phytoplankton as a metric, as reference values and boundaries
can not be presented at this stage. However, phytoplankton as a metric is under development as well
as cell carbon and biovolume.
Belgium
NEA GIG TYPE 1/26b
Chlorophyll-a metric
For Chl-a, a background Chl-a concentration was determined for the application of phytoplankton
assessment criteria for the Belgian Continental Shelf. Eutrophication in the BCS occurs in the form
of massive, ephemeral
Phaeocystis
colony blooms during spring (April-May).
Phaeocystis
colonies
generally occur simultaneously with diatom (mainly
Guinardia
spp.) blooms that considerably
affect the Chl-a concentrations. Analysis of historical data (ASMO 98/3/Info.1-F) suggests
that foam accumulation would occur from a
Phaeocystis
cell concentration of 10
7
l
-1
. This cell
concentration can be converted into a Chl-a concentration of 5 μg Chl-a l
-1
(C : Chla = 29) using the
experimentally determined factor of 0.5 pg Chl-a /Phaeocystis cell (Rousseau
et al.,
1990).
The contribution of diatoms to the spring bloom has been established from the analysis of a 13
year time series at a monitoring station (330) located in the centre of the Belgian Continental
Shelf. Seasonal evolution of Chl-a concentrations measured on a weekly basis indicates that the
average Chl-a
concentration of 9.2 μg l
-1
corresponds to the
Phaeocystis
pre-bloom situation, itself
determined as a
Phaeocystis
concentration higher than 1x10
6
cells l
-1
(Cadée & Hegeman, 1991).
This allows us to consider a threshold of 15 μg Chl-a
l
-1
from which nutrient over-enrichment would
result in ecosystem disturbance.
Proposed reference chl-a values for Belgium waters
(Calculated reference condition, based on 50 % deviation of reference from the high/good boundary)
NEA GIG Type 1/26b – North Sea – 6.7 µg L
-1
157
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0160.png
Phaeocystis metric
Proposed reference
Phaeocystis
values for Belgium waters
NEA GIG Type 1/26b – North Sea – Cell threshold: 1,000,000 cell/L
Reference percentage exceedent threshold: 8.30 % (1 sample in 12 months as part of natural
spring bloom. Assumes monthly sampling)
Percentage exceedent thresholds: H/G 10 %, G/M 17 %
Denmark
• NEA GIG Type 1/26c – North Sea (German Bright & Jutland)
• NEA GIG Type 1/26d – Denmark Skagerrak
• NEA GIG Type 8 – Kattegat + Small area of Skagerrak
Chlorophyll-a metric
Proposed reference chl-a values for Denmark waters
(Calculated reference condition, based on 50 % deviation of reference from the high/good
boundary)
• NEA GIG Type 1/26c – North Sea (German Bright & Jutland) – 3.3 µg L
-1
• NEA GIG Type 1/26d – Denmark Skagerrak – 2.0 µg L
-1
• NEA GIG Type 8 – Kattegat + Small area of Skagerrak – 1.0 µg L
-1
Germany
Chlorophyll-a metric
Proposed reference chl-a values for German waters
• NEA GIG Type 1/26c – North Sea - German Bright & Jutland – 3.3 µg L
-1
• NEA GIG Type 3/4 – North Se - Ems, Weser, Elb Coast – 3.3 µg L
-1
Phaeocystis metric
Proposed reference
Phaeocystis
values for German waters
• NEA GIG Type 1/26c – North Sea - German Bright & Jutland – Cell threshold: 1,000,000 cell/L
• NEA GIG Type 3 4 – North Sea - Ems, Weser, Elb Coast – Cell threshold: 1,000,000 cell/L
• Reference percentage exceedent threshold: 8.30 % (1 sample in 12 months as part of natural
spring bloom. Assumes monthly sampling)
• Percentage exceedent thresholds: H/G 10 %, G/M 17 %
Class Boundary Setting Procedures
The reference values put forward by each MS for the typologies were specific to each type, and
vary geographically. The boundary setting protocols investigated conditions that influenced the
geographical difference between the NEA GIG types. Examination of the conditions influencing the
variation between NEA types supported the delineation of the geographical areas, and the reasoning
behind the differences in reference conditions.
158
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0161.png
The NEA GIG covers a very large geographic area from Southwest Spain (including the Canary
Islands, 28
o
N) to Northern Norway (71
o
N). These areas will have different non-anthropogenic
pressures, which would affect the onset, rate of growth and overall productivity of phytoplankton
populations in identical typologies, consequently different thresholds need to be set. The main non-
anthropogenic drivers are:
• Isolation
• Freshwater discharge & catchment geology
• Upwellings
• Continental shelf width / break
• Coastal geomorphology & Sediment loads
• Tidal height & coastal currents
• Basic salinity regime
The OSPAR Quality Status Review (2000) recognizes some of these hydromorphic features in its
approach to regional seas (Figure 2.4.1).
However for the purposes of setting chlorophyll boundaries and thresholds the local nature of all the
factors above also needs to be taken into account.
Major influences on sub-types
The first layer of types was determined primarily on freshwater inflow. Details of freshwater flow
characteristics are found in Table 6. On the basis of freshwater flow, an initial boundary setting
procedure differentiates 9 broad coastal types as shown in Table 7
Higher freshwater flows within the Dutch coastal zone compared to other NEA-GIG coasts, change
the typology of the Holland Coast from North Sea euhaline (NEA1) into a polyhaline (NEA3).
However, freshwater flow does not account for all the variability between GIG types and further
delineation is explained by both upwelling and localised influences.
Within each of the broad typologies, further delineation was required to account for upwelling and
localized and regional variations within the typologies.
Figure 2.4.1:
Hydromorphic areas of the NEA GIG Atlantic area.
159
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0162.png
Table 2.4.6:
Freshwater flow characteristics of each region associated with reference conditions of chlorophyll.
Zone
Western Iberian coast upwelling
Cantabrian coast
Atlantic Coast
Atlantic Coast
North Sea
Irish Sea
German Bight
Scandinavian Coast
Member States
Portugal / Spain West coast
North Spanish Coast
French Atlantic coast
Atlantic Coast (UK and IE)
North Sea (UK, NL, Belgium)
Irish Sea (UK & IE)
Germany & Denmark
Denmark, Sweden & Norway
Freshwater Discharge/ Runoff
(km
3
/year)
63
16
46.5
40
122
80
22
21.5
For example, because of the extreme variation in latitude within the type NEA1/26 it was
not possible to agree a single range of chlorophyll thresholds across the entire area as there are
significant differences in how phytoplankton reacts to natural and anthropogenic nutrient
enrichment. There are clear differences between the open clearer Atlantic waters and the more
enclosed sea areas and special cases e.g. where natural upwelling occurs, particularly on the Western
Iberian coast.
It was therefore agreed by the GIG that the only way to produce meaningful results was to split type
NEA1/26 into five sub-types, NEA1/26a, b, c, d, e. The sea areas grouped under this new typology
are described in Table 7.
Type 1/26a generally reflects open clearer Atlantic Waters and Type1/26b generally reflects more
enclosed sea areas. Spain and Portugal are both affected by the Iberian Upwelling, which reacts
very differently from other regions in the NEA GIG, with naturally nutrient enriched coastal waters
and more natural phytoplankton blooms expected through out the year. This upwelling area has been
designated Type 1/26e.
Kattegat and Skagerrak also behave differently from the other areas. It is a complex system with low
salinity (surface values from salinity 10-12 in south Kattegat increasing up 25-30 in the Skagerrak)
and the areas are greatly influenced by inflow of water from the Baltic Sea, the German Bight, the
North sea and the Atlantic. Stratification, sometimes three layers, occurs in most of the area in the
growing season due to the different salinity in the different inflowing water types and the relatively
shallow water. Due to these complex factors, the Scandinavian coast is divided into type 8, 9, 10
with additional separation of small areas into 1/26a, 1/26c and 1/26d.
There is also a change of chlorophyll thresholds across the Spain/ France border. The differences
in morphological and hydrographical features between the Spanish and French shelves have been
taken into account to justify lower thresholds of the Eastern Cantabrian coast (Spain’NEA1/26a).
The Eastern Cantabrian coast has almost no coastal plain and the rivers are small, whereas the French
shelf presents a wide coastal plain and much higher river runoff.
160
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0163.png
Table 2.4.7: Description of NEA GIG types and associated areas.
New Type
NEA1/26a
NEA1/26b
NEA1/26c
NEA1/26d
NEA 1/26e
NEA8
NEA9
NEA10
Areas
Eastern Cantabrian Coast, Canary Islands. Atlantic Coast, Western Irish Sea, Scandinavian Coast:
North Sea Norway + Norwegian Sea
Eastern Irish Sea, North Sea (UK East Coast, Channel, NL Wadden Coast, Zeeland Coast,
Belgium Coast)
German Bight & Jutland
Scandinavian Coast: Denmark Skagerrak
Western Iberian Coast upwelling, Western Cantabrian Coast
Denmark, Norway, Sweden: Kattegat + small area of Skagerrak
Sweden, Norway: Skagerrak
Sweden, Norway: Skagerrak
Thus if we take into account, the variation between freshwater flow, upwelling and localised
regional differences, the NEA GIG can be separated into 9 distinct coastal areas. The nine coastal
types will be intercalibrated in Phase I. Intercalibration on NEA Type 11 (Transitional waters) will
be taken up in Phase 2 Intercalibration. The final delineation of NEA GIG Coastal Types for Phase I
intercalibration is illustrated in Figure 2.
NB.
The map in Figure 2 shows the regions and areas and the coastal types that apply, Germany &
Netherlands appears to have two types overlying each other, this is due to off shore coastal islands
that respond to North Sea influences defined by type NEA 1/26b and an inshore area (Wadden Sea
Coast) which response to influences defined by type NEA3/4.
2.4.4 Results of the comparison
As option 2 has been used, the boundaries have been agreed by experts representing all countries
in the GIG phytoplankton sub-group. Therefore the final agreed results presented below are the
results of comparing expert views on what the boundaries should be related to type differences and
reference levels.
2.4.5 Results of the harmonization – Boundary thresholds and EQR values
Metric-Chlorophyll-a concentration (90
th
Percentile) (Boundary Setting Procedure and Results
(All Coastal Water Types: NEA 1/26a,b,c,d,e, NEA3/4, NEA7, NEA8, NEA9, NEA10))
This tool is based on the location of the 90
th
percentile of chlorophyll concentration against
boundary thresholds set for the High/Good and Good/Moderate boundaries. The tool assumes that
6 years of chlorophyll data are available with monthly sampling during the growing season. The
assessment is made against the phytoplankton agreed growing season which varies with latitude.
Recommendations from Quasimene on chlorophyll analysis show an accepted 12 % error in
sampling and analysis. The inherent sampling error can then account for the small differences in
reference values within the same type and same boundary conditions (i.e. Portugal and Spain –
Iberian Upwelling coast).
161
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0164.png
Figure 2.4.2:
NEA Coastal Typology
Delineation, Phase I Intercalibration.
The same level of deviation from reference values to the High/Good and Good/Moderate boundaries
have been agreed within these new (and existing) types and numeric 90
th
percentile thresholds
set for each of the geographic areas described above. There have been three different levels of
deviations within types set for chlorophyll, which translates into type specific EQRs for the three
different deviations (see Table 8). The three levels of difference between high/good and good/
moderate thresholds are set at 100 % for type 1/26a, 8, 9, 10, 50 % for type 1/26b, 1/26e, 3 and 4
and 33 % for type 1/26c and 1/26d. These typologies reflect the different way the phytoplankton
reacts in these waterbodies. In the clear open Atlantic waters, the phytoplankton respond quickly
to small changes in nutrients while those areas of more enclosed seas and upwelling’s tend to have
higher turbidity and nutrients, so may respond more slowly initially but bigger natural blooms can
be maintained.
NEA Types 1/26a, 8, 9, 10 are all relatively clear waters with relatively low natural inputs of
nutrients (therefore fast initial reaction but may not be sustained). For these types with low
reference thresholds, it was decided that in the absence of good empirical data, 100 % deviation
was considered to give a reasonable indication of a slight to moderate deviation from the high/good
162
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0165.png
boundary. ROI, France, Spain and Scandinavian countries all have the same percentage of change
between boundaries in areas where reference values are lower than 5 µg/L.
NEA Type 1/26b, 1/26e, 3, 4 are a combination of relatively high amounts of natural nutrients
(either from run off or upwelling) and / or relatively high turbidity (slower reaction which can
sustain bigger and longer blooms). In these areas where the reference values are higher, a 50 %
change has been agreed on in keeping with the OSPAR recommendation that 50 % is adequate to
describe a slight to moderate deviation from the reference level.
In NEA Types 1/26c and 1/26d, the inflow of deep water, relatively rich in nutrients but initially
unavailable to algae enters the Kattegat bottom waters and eventually mixing to the surface water
and re-exports to the Skagerrak frontal area.
Table 8 shows both the numeric chlorophyll thresholds and the associated EQRs. Table 9 shows a
summary of the type specific EQRs only. Extent of boundary conditions is illustrated in Figure 3.
Table 2.4.8:
Final boundary thresholds grouped by NEA GIG geographic zones and types for chlorophyll a metric.
Geographical Zone
Member States
New Type
Reference (calculated
from High/Good)
0.67
5.33
4.00
2.33
3.33
90 %ile
High /
Good
1
8
6
3.5
5
90 %ile
Good /
Mod
2
12
9
7
10
H/G
EQR
0.67
0.67
0.67
0.67
0.67
G/M
EQR
0.33
0.44
0.44
0.33
0.33
Canary Islands
ES
NEA1/26a
NEA1/26e
NEA1/26e
NEA1/26a
NEA1/26a
Western Iberian Coast PT, ES
upwelling
Western Cantabrian
Coast
Eastern Cantabrian
Coast
Atlantic Coast
ES
ES
FR , UK (exposed
West Coast & NI.),
ROI, ES (Gulf of
Cadiz, Sweden
UK (NI), ROI
Western Irish Sea
Eastern Irish Sea
North Sea
NEA1/26a
3.33
6.67
6.67
5
10
10
10
15
15
0.67
0.67
0.67
0.33
0.44
0.44
UK (England, Wales, NEA1/26b
Scotland)
NL (Wadden coast),
BE, UK (East Coast,
Channel), FR
NEA1/26b
North Sea - , “German DE, DK
Bight”, & Jutland
Scandinavian Coast
Scandinavian Coast
Scandinavian Coast
Scandinavian Coast
DK
DK,NO,SE
NO,SE
SE,NO
NEA1/26c
NEA 1/26d
NEA8
NEA9
NEA1/26a
NEA10
3.33
2.00
1.00
1.67
2.00
5
3
1.5
2.5
3
7.5
4
3
5
6
0.67
0.67
0.67
0.67
0.67
0.44
0.50
0.33
0.33
0.33
163
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0166.png
Table 2.4.9:
Summary of EQR ratios of Chlorophyll-a for the NEA GIG types and areas.
Type
NEA1/26a
NEA1/26b
NEA1/26c
NEA1/26d
NEA1/26e
NEA8
NEA9
NEA10
Areas
Eastern Cantabrian Coast, Canary Islands. Atlantic Coast, Western
Irish Sea, Scandinavian Coast: North Sea Norway + Norwegian Sea
Eastern Irish Sea, North Sea (UK East Coast, Channel, NL Wadden
Coast, Zeeland Coast, Belgium Coast)
German Bight and Jutland
Scandinavian Coast: Denmark Skagerrak
Western Iberian Coast upwelling, Western Cantabrian Coast
Denmark, Norway, Sweden: Kattegat + small area of Skagerrak
Sweden, Norway: Skagerrak (see map)
Sweden, Norway: Skagerrak (see map)
H/G EQR
0.67
0.67
0.67
0.67
0.67
0.67
0.67
0.67
G/M EQR
0.33
0.44
0.44
0.50
0.44
0.33
0.33
0.33
Chlorophyll
boundaries
H/G (ug/L)
Figure 2.4.3:
Chlorophyll boundary
conditions for NEA GIG types.
(Needs updating in early Phase II).
164
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
Metric - Taxa Cell Counts (“frequency of phytoplankton cell counts above a predefined
threshold, ranging approximately from 10
5
to 10
6
cells/l”) (Boundary Setting Procedure and
Results (All Coastal Water Types: NEA 1/26a,b,c,d,e, NEA3/4, NEA7, NEA8, NEA9, NEA10))
Frequent occurrences of phytoplankton taxa above bloom levels over the growing season are an
indication of ecosystem disturbance. Natural blooms in the spring and autumn are to be expected but
more frequent occurrences throughout the rest of the year are considered indicative of a degraded
system.
The sub metric is the frequency of blooms of any routinely identifiable phytoplankton taxa over a
defined threshold from monthly sampling is calculated over a 6 year period.
The definition of the index is based on:
Number of samples where a single taxa count is over a predefined threshold
Six years of routine monitoring data should be used from the whole year
A minimum of 12 sampling occasions (monthly) per year is recommended (but not essential).
CLASSIFICATION is based on the number of sampling times where the single taxa counts
exceed a threshold and is calculated as a percentage of the total number of samples collected
within one water type for the six year period.
As the chlorophyll thresholds are region-specific, the cell-count threshold has been adapted for
regions that present different trophic richness due to their natural (hydromorphological) features.
Therefore, different cell-count thresholds will be used within the different types and regions. For
example, the upwelling coast of NW Spain will have a higher threshold, as blooms there are of high
magnitude due to natural processes of nutrient enrichment. As the upwelling intensity decreases
gradually eastward, the Cantabrian Coast would present lower thresholds. The Canary Archipelago,
the most oligotrophic area in Spain, would present the lowest threshold. As for the frequency of
blooms, it should be noted that the Iberian upwelling zone is subject to a higher frequency of
blooms than expected in other coastal waters due to the occurrence of blooms during the summer,
which is a season of high upwelling intensity in that area. Therefore, the boundary classification
should be adapted for this special region where blooms are not always related to anthropogenic
eutrophication.
It was accepted at GIG meeting in Mieres that member states will use different phytoplankton
analytical levels. Two groups of routinely identified phytoplankton were determined. These are
loosely termed “large” (mainly microphytoplankton, chain formers and other readily identified
groups) and “small” (which includes, nanoplankton, microflagellates and the more difficult taxa).
There tend to be fewer large taxa then small, so two thresholds of cell count are required. Where
a member state identifies and counts phytoplankton from both groups the assessment requires a
primary assessment against the “large” threshold, if the sample passes then it is assessed against the
small threshold. The reference level is the frequency these local numbers are exceeded. These are
presented in the table below:
165
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0168.png
Table 2.4.10:
Summary of final intercalibrated boundary thresholds grouped by NEA GIG geographic zones and types
for taxa count metric. N/P = not participating.
Member
state
Typology
Area
Reference ( % Single
Taxa counts exceeding
thresholds)
Large
Phytoplankton
Threshold
Small
Phytoplankton
Threshold
Spain
Spain
Spain
Spain
Spain
Portugal
France
France
NEA1/26e
NEA1/26e
NEA1/26a
NEA1/26a
NEA1/26a
NEA1/26e
NEA1/26a
NEA1/26b
*W. Cantabrian
Coast
E. Cantabrian
Coast
*Western Iberian
Coast
Reference
= number of times a single taxa count exceeds the
threshold. If we expect a spring and autumn bloom and only
sample monthly then 2/12 = 16.7 %. High <20 %
*Upwelling Area Reference = 3 blooms per year spring, summer
and autumn expected, if sample monthly then 3/12 = 25 %,
High <30 % in Upwelling regions,
This symbol indicates the
assessment is made on all phytoplankton regardless of size
25
25
16.70
16.70
16.70
25
16.70
N/P
N/P
16.70
N/P
N/P
N/P
N/P
N/P
100,000
100,000
N/P
N/P
100,000
1,000,000
750,000
750,000
500,000
500,000
1000,000
250,000
N/P
250,000
N/P
Canary Islands
Atlantic Coast
*Western
Iberian Coast
Belgium
Netherlands
NEA1/26b
NEA3
Netherlands
Germany
NEA1/26b
NEA4
Netherlands
Germany
N/P
NEA1/26c
NEA3
NEA4
N/P
N/P
N/P
N/P
N/P
N/P
N/P
N/P
N/P
N/P
N/P
Germany
N/P
N/P
N/P
Denmark
Denmark
Sweden
NEA1/26c
NEA8
N/P
N/P
N/P
N/P
N/P
Denmark
Sweden
NEA1/26d
NEA8
N/P
N/P
N/P
N/P
N/P
N/P
Sweden
NEA9
Norway
NEA10
NEA8
N/P
Norway
NEA1/26a
NEA9
N/P
N/P
N/P
N/P
N/P
N/P
N/P
N/P
N/P
N/P
N/P
N/P
Norway
RoI
UK
UK
N/P
N/P
N/P
Norway
NEA10
NEA1/26a
NEA1/26a
16.70
N/P
N/P
N/P
N/P
NEA1/26b
16.70
250,000
16.70
250,000
250,000
N/P
N/P
166
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0169.png
Not all member states are participating in this metric. Participating countries include Portugal,
Spain, France, Republic of Ireland, and UK. The associated typologies are: NEA 1/26a, 1/26b,
1/26e, 3, and 4.
All single taxa counts above the threshold are calculated as a percentage of the overall sampling
count (over the six-year period). Percentages are calculated and scored as per Table 11(Borja et al,
2004). The value of the percentage decides the boundary classification.
Table 2.4.11:
Agreed Percentage Boundaries NEA1/26a, b, NEA3/4.
Phytoplankton Boundary Percentages
<20 % single taxa counts exceeding threshold
20 - 39 % single taxa counts exceeding threshold
40-69 % single taxa counts exceeding threshold
70-90 % single taxa counts exceeding threshold
>90 % single taxa counts exceeding threshold
Boundary classification
HIGH
GOOD
MODERATE
POOR
BAD
These boundaries equate to H/G and G/M EQRs of 0.84 and 0.43 calculated by dividing e.g. reference %, 16.7 by
boundary % e.g. 20 = 0.84 at the High/Good boundary etc.
Table 2.4.12:
Agreed Percentage Boundaries NEA1/26e (Spain & Portugal) for coastal waters affected by upwelling.
Phytoplankton Boundary Percentages
< 30 % single taxa counts exceeding the threshold
30 - 49 % single taxa counts exceeding the threshold
50 - 69 % single taxa counts exceeding the threshold
70 - 90 % single taxa counts exceeding the threshold
>90 % single taxa counts exceeding the threshold
Boundary classification
HIGH
GOOD
MODERATE
POOR
BAD
These boundaries equate to H/G and G/M EQRs of 0.83 and 0.51 calculated by dividing e.g. reference %, 25 by
boundary % e.g. 30 = 0.83 at the High/Good boundary etc.
Metric - Indicator Taxa (Frequency of
Phaeocystis
Cell counts above 10
6
Cells/l) (Boundary
Setting Procedure and Results (All Coastal Water Types: NEA 1/26a,b,c,d,e, NEA3/4, NEA7,
NEA8, NEA9, NEA10))
Classification is based upon percentage of months with one or more occasions where the counts of
Phaeocystis
exceed the predetermined threshold. The definition of the metric is:
Number of months with samples where the
Phaeocystis
cell count is over a predefined
threshold
Six years of routine monitoring data should be used from the whole year
A minimum of 12 sampling occasions (monthly) per year is recommended (but not essential).
CLASSIFICATION is based on the number of months with one or more sampling times where
the
Phaeocystis
counts exceed a threshold and is calculated as a percentage of the total number
of months collected within one water type for the six year period.
167
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0170.png
Table 2.4.13:
Agreed Thresholds and References for Phaeocystis threshold.
Member
state
Typology
Reference (%
Phaeocystis
counts exceeding thresholds).
N/P = not participating.
Reference
= number of times a single taxa count exceeds the
threshold. If we expect only a spring bloom and only sample
monthly then 1/12 = 8.33 %. High /Good=10 %,G/M=17 %
N/P
N/P
N/P
N/P
8.30
N/P
Phaeocystis
Threshold
Cells/l
N/P
N/P
N/P
N/P
1000000
1000000
1000000
1000000
1000000
1000000
1000000
N/P
N/P
N/P
N/P
N/P
N/P
N/P
N/P
N/P
N/P
1000000
1000000
N/P
N/P
Spain
Spain
Portugal
France
France
NEA1/26a
NEA1/26e
NEA1/26e
NEA1/26a
NEA1/26b
NEA1/26b
NEA3
NEA4
NEA3
NEA4
NEA1/26b
Belgium
Netherlands
Netherlands
Netherlands
Germany
Germany
Germany
Denmark
8.30
8.30
8.30
8.30
8.30
8.30
N/P
N/P
N/P
N/P
N/P
N/P
N/P
N/P
N/P
N/P
N/P
NEA1/26c
Denmark
Denmark
Sweden
Sweden
Sweden
Norway
Norway
Norway
Norway
RoI
UK
UK
NEA1/26d
NEA8
NEA8
NEA9
NEA10
NEA8
NEA9
NEA1/26c
NEA1/26a
NEA10
NEA1/26a
NEA1/26a
NEA1/26b
8.30
8.30
Only the Netherlands, Germany, UK and Belgium, are participating in this tool. Classification is
taken from the final percentage count of exceedances over the total sampling number. The present
threshold percentages related to boundary classifications are presented in Table 14. There is still
some discussion on the boundaries for moderate, poor and bad and further work is required on
these boundaries. As all participating countries have agreed the same method in terms of reference
conditions, boundary thresholds and EQRs within types then this intercalibration fulfils the criteria
set out under Option 1 in the intercalibration process.
168
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0171.png
Table 2.4.14:
Phaeocystis boundary thresholds.
Phaeocystis
Percentage exceedance boundaries
<10 %
Phaeocystis
exceeding threshold
10 – 17 %
Phaeocystis
exceeding threshold
>17 %
Phaeocystis
exceeding threshold
>35 %
Phaeocystis
exceeding threshold
>80 %
Phaeocystis
exceeding threshold
Boundary classification
HIGH
GOOD
MODERATE
Poor
Bad
These boundaries equate to H/G and G/M EQRs of 0.92 and 0.49 calculated by dividing e.g. reference %, 8.3 by
boundary % e.g. 10 = 0.83 at the High/Good boundary etc.
2.4.6 Open issues and need for further work
Transitional Waters – all metrics need agreement. Need to normalize data to salinity values.
At present there is only one ECO TW type, this will be a problem that we need to resolve.
Flow regime also affects diversity; this is another issue that would require normalization for all
countries in NEA GIG.
Sampling frequency – how can this affect the boundary thresholds.
Phytoplankton composition/community – requirement for further investigation.
Toxic species?
Cell carbon and biovolume methods for intercalibration
UK and ROI are possibly looking to modify chl-a 90 %ile tool, cell counts and functional
groups metrics to be applicable in transitional waters. NL and other countries are looking at
chl-a in transitional waters as well. Turbidity/light will be another normalizing issue.
• Risk of misclassification – share ideas.
Primary production – develop new, cheaper and more reliable methods.
Standardize size classes that phytoplankton taxa belong to. Start by country and construct a
standard size class list at MS level.
• A literature search needs to be carried out to find minimum and maximum size ranges for
phytoplankton.
Indicator species for eutrophication – OSPAR list exists but the relation between
eutrophication is not for all the species on the list. It is suggested we refine and develop a
better list.
Diversity and Composition metric needs to be linked to nutrient enrichment and other
pressures.
• Rules for combining elements for classification need to be agreed (Possible output from CIS
June Workshop in Paris).
Methods of chl a measurement need to be agreed/ intercalibrated
169
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0172.png
3 Discussion
3.1 Comparability between GIGs
Table 3.1.1:
Summary of intercalibration topics per GIGs.
NEA GIG
METRICS
N°. TYPES
OPTION
BOUNDARY
SETTING
PROCEDURE
Chl α, Indicator
Taxa, Taxa Cell
Count
8
2
Experts, 90iles,
H/G, G/M
MED GIG
Chl α
3
BAL GIG
Chl α
8
BLA GIG
Biomass
1
2
Quantiles,
All boundaries
2/3
Percentiles,
H/G, G/M
2/3
Percentiles,
H/G, G/M
Different
approaches,
empirical
relationships
between chl α and
Secchi, mixed
models, …
REFERENCE
CONDITION
MS-specific, 90ile
High status sites
Historical data
10th percentile
Expert judgement
OPEN ISSUES
Improve dose-
response
relationships
Missing metrics,
species composition
Missing metrics
Metrics
Intercalibration for the phytoplankton quality element has been successfully performed by all of the
four Coastal Waters GIGs, using Chlorophyll-α as metric. Other metrics have been considered and
tested by some GIGs (Indicator Taxa, Biomass).
Types
Not all types for each GIG have been covered, mainly because there were not shared by all MSs.
Option
Option 2 (MS agreed on the common metrics, created data sets relating MSs assessment methods
to the common metrics, agreed on High/Good and Good/Moderate class boundaries and established
relationships between common and national metrics) and a hybrid Option 2 and 3 (preliminary
national boundaries were compared and harmonized with those obtained by using a common dataset
for setting the final boundaries) approaches have been adopted for the intercalibration.
Boundary Setting Procedure
All Coastal Waters GIGs are providing results for the H/G and G/M boundaries (at least), applying
similar approaches based on the expert supervised selection of appropriate percentiles as boundaries
best fitting with evident changes in ecological status (90
th
percentile is the favorite choice).
Reference Condition
There’s no clear common view among GIGs (but also among MSs) about how to fix Reference
Conditions: in some cases the expert judgment has been the selected option, in some others a
mixture of modeling, statistics on historical data and empirical relationships with other metrics (e.g.
Secchi depth, see BALTIC GIG) has been applied.
170
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
3.2 Open issues and need for further work
The main task for the next intercalibration round is almost unanimously to include other metrics
(blooms, species composition, toxic species, indicator species) in order to complete the analysis. It
would be useful to focus on (strengthen) dose-response relationships in order to correlate pressures
and ecological status.
4 References
Agència Catalana de l’Aigua (ACA), 2005. Caracterisation of surface water bodies and analysis of the risk of
failing the objectives of the water framework directive (2000/60/CE) in the catalonian basins of the Ebro, the
Garona and the Senia, and revision of the internal basins of Catalonia. IMPRESS document. Departament de
Medi Ambient i Habitatge. (http://mediambient.gencat.net/aca/ca//planificacio/directiva/impress.jsp).
Álvarez–Salgado X. A., Rosón G., Pérez F. F. and Y. Pazos, 1993. Hydrographic variability off the Rías Baixas
(NW Spain) during the upwelling season. Journal of Geophysical Research 98(C8): 14447-14455.
Álvarez–Salgado X. A., Rosón G., Pérez F. F., Figueiras F. G. and Y. Pazos, 1996. Nitrogen cycling in an
estuarine upwelling system, the Ría de Arousa (NW Spain). I. Short–time–scale patterns of hydrodynamic and
biogeochemical circulation. Marine Ecology Progress Series 135: 259-273.
Argyrou, M. 2005. Report of the National Monitoring Programme of Cyprus for the Year 2004. Programme
for the Assessment and Control of Pollution in the Mediterranean Region (MEDPOL /UNEP). Department of
Fisheries and Marine Research, pp. 32.
Argyrou, M. 2006. Report of the National Monitoring Programme of Cyprus for the Year 2005. Programme
for the Assessment and Control of Pollution in the Mediterranean Region (MEDPOL /UNEP). Department of
Fisheries and Marine Research, pp. 41.
Aristegui J. 1990. Distribution of chlorophyll “a” in the Canary Island waters. Boletin del Instituto Español de
Oceanografia 6(2): 61-72.
Arístegui J., Hernández–León S., Montero M.F. and M. Gómez, 2001. The seasonal planktonic cycle in coastal
waters of the Canary Islands. Sci. Mar., 65 (Suppl. 1): 51-58.
Artigas L. F. 1998. Seasonal variability in microplanktonic biomasses in the Gironde dilution plume (Bay of
Biscay): Relative importance of bacteria. Oceanologica Acta 21: 563-580.
Bald J. 2005. Proposal for the physico-chemical quality status assessment within the estuarine and coastal
waters of the Basque Country (Northern Spain). PhD Thesis. University of Navarra (Spain) 262 p.
Bald J. A., Borja I., Muxika J., Franco and V. Valencia 2005. Assessing reference conditions and physico-
chemical status according to the European Water Framework Directive: a case-study from the Basque Country
(Northern Spain). Marine Pollution Bulletin, 50: 1508-1522.
Barton E. D., Aristegui J., Tett P, Canton M., Garcia-Braun J., Hernandez-Leon S., Nykjaer L., Almeida C.,
Almunia J., Ballesteros S., Basterretxea G., Escanez J., Garcia-Weill L., Hernandez-Guerra A., et al. 1998. The
transition zone of the Canary Current upwelling region. Progress in Oceanography 41: 455-504.
Berman, T., D.W. Townsend, S.Z. El-Sayed, G.C. Trees and Y. Azov. 1984. Optical transparency, chlorophyll
and primary productivity in the Eastern Mediterranean near the Israeli Coast. Oceanol. Acta, 7: 367-372.
Bianchi, T.S., A. Demetropoulos, M. Hadjichristophorou, M. Argyrou, M. Baskaran and C. Lambert. 1996.
Plant Pigments as Biomarkers of Organic Matter Sources in Sediments and Coastal Waters of Cyprus (eastern
Mediterranean).
Estuarine, Coastal and Shelf Science,
42: 103-115.
Bode A. and M. Varela, 1998. Primary production and phytoplankton in three Rias Altas (NW Spain): seasonal
and spatial variability. Scientia Marina 62: 319-330.
Bode A., Casas B,. Fernández E., Marañon E., Serret P. and M. Varela, 1996. Phytoplankton biomass and
production in shelf waters off NW Spain: spatial and seasonal variability in relation to upwelling. Hydrobiologia
341: 225-234.
171
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
Borja A., Franco J., Valencia V., Bald J., Muxika I., Belzunce M. J., and O. Solaun. 2004. Implementation of the
European water framework directive from the Basque country (northern Spain): a methodological approach.
Marine Pollution Bulletin 48: 209-218.
Botas J. A., Bode A., Fernández E. y R. Anadón. 1988. Descripción de una intrusión de agua de elevada
salinidades el Cantábrico Central: distribución de los nutrientes inorgánicos y su relación con el fitoplancton.
Investigación Pesquera 52: 561-573.
Botas J. A., Fernández E. and R. Anadón. 1990. A persistent upwelling off the central Cantabrian coast (Bay of
Biscay). Estuarine, Coastal & Shelf Science 30: 185-189.
Boynton W.R., Kemp W. M. and C.W. Keefe, 1982. A comparative analysis of nutrients and other factors
influencing estuarine phytoplankton production. In: Estuarine Comparisons. V.S. Kennedy (ed). Academic
Press, New York, 69-90.
Cadée, G.C. and J. Hegeman. 1991. Historical phytoplankton data of the Marsdiep. Hydrobiological Bulletin
24: 111-188.
Cermeño P., Marañón E., Pérez V., Serret P., Fernández E., and C. G. Castro, 2006. Phytoplankton size structure
and primary production in a highly dynamic coastal ecosystem (Ría de Vigo NW-Spain): seasonal and short-
time scale variability. Estuarine Coastal and Shelf Science 67: 251-266.
Diez I., Secilla A., Santolaria A., and Gorostiaga J. M. 2000. The north coast of Spain. In: Seas at the Millennium.
An environmental evaluation. C. Sheppard (ed.). Pergamon, Amsterdam, Volume I, 135-150.
European Environmental Agency. 2005. Source apportionment of nitrogen and phosphorus inputs into the
aquatic environment. EEA Report No 7/2005.
Figueiras F. G,. Niel F. X. and C. Mouriño, 1986. Nutrientes y oxígeno en la Ría de Pontevedra (NO de
España). Investigación Pesquera 50: 97-115.
Fraga F. 1976. Fotosíntesis en la Ría de Vigo. Investigación Pesquera 40 (1):151-167.
Fraga F. 1981. Upwelling of the Galician coast. Northwest Spain. In: Coastal Upwelling. F. A. Richards (ed).
American Geophysical Union, 176-182.
Garcia C. M., Prieto L., Vargas M., Echevarria F., García-Lafuente J., Ruiz J., and J.P. Rubín. 2002.
Hydrodynamics and the spatial distribution of plankton and TEP in the Gulf of Cadiz (SW Iberian Peninsula).
Journal of Plankton Research 24: 817-833.
Gil J. 2003. Changes in the pattern of water masses resulting from a poleward slope current in the Cantabrian
Sea (Bay of Biscay). Estuarine, Coastal & Shelf Science 57: 1139-1149.
Harding Jr., L. W. & E. S. Perry.1997. Long-term increase of phytoplankton biomass in Chesapeake Bay, 1950-
1994. Mar. Ecol. Prog. Ser., 157: 39-52.
Harris G. P. 1986. Phytoplankton Ecology: structure function and fluctuation. Chapman and Hall, London,
384 pp.
Ignatiades, l., Vounatsou, P., Karydis, M. (1992). A possible method for evaluating oligotrophy and eutrophication
based on nutrient concentration scales.
Mar. Poll. Bull.,
24: 238-243.
Karydis, Μ. (1999).
Evaluation report on the eutrophication level in coastal greek areas.
Univ. of Aegean,
Mytilini, February 1999 (in greek).
Krom, M.D., Kress, N., Brenner, S. Gordon, L.I. (1991). Phosphorus limitation of primary production in the
eastern Mediterranean Sea.
Limnol. Oceanogr.,
36: 424-432.
Krom, M.D., S. Brenner, N. Kress, A. Neori and L.I. Gordon. 1992. Nutrient dynamics and new production in
a warm-core eddy from the Eastern Mediterranean Sea. Deep-Sea Research, 39: 467-480.
Krom,D. M., N. Kress & S. Brenner. 1991. Phosphorus limitation of primary productivity in the eastern
Mediterranean Sea. Limnol. Oceanogr., 36(3): 424-432.
Lavin A., Valdes L., Sanchez F., Abaunza P., Forest A., Boucher J., Lazure P. and A. M. Jegou. 2006. The Bay
of Biscay: The encountering of the ocean and the shelf (18 b, E). In: The Sea, Volume 14. A.R. Robinson &
K.H. Brink (ed.). Harvard University Press, 933-1001.
172
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
Lipej L., P. Mozetič, M. Orlando Bonaca, B. Mavrič, M. Šiško & N. Bettoso. 2006. Evaluation of the Ecological
Status of Coastal Waters in accordance with the European Water Framework Directive (Water Framework
Directive, 2000/60/EC). Final national report. Marine Biology Station Piran, National Institute of Biology,
October 2006, 180 pp. (in Slov.)
Llope M. 2005. Variabilidad hidrográfica en un área costera del Mar Cantábrico. Modelado de plancton y
efectos de cambio climático. Tesis Doctoral. Universidad de Oviedo. Ph D. dissertation.
Lunven M., Guillaud J.F., Youénou A., Crassous M. P., Berric R., Le Gall E., Kérouel R., Labry C., and A.
Aminot. 2005. Nutrient and phytoplankton distribution in the Loire River plume (Bay of Biscay, France)
resolved by a new Fine Scale Sampler. Estuarine, Coastal & Shelf Science 65: 94-108.
Margalef R. 1974. Ecología. Ediciones Omega, Barcelona, 951 pp.
Margalef R., Durán M. and F. Sáiz. 1955. El fitoplancton en la Ría de Vigo de Enero de 1953 a Marzo de 1954.
Investigación Pesquera 2: 85-129.
Mason E., Coombs S., and P. B. Oliveira. 2005. An overview of the literature concerning the oceanography of
the Eastern North Atlantic region. Relat. Cient. Téc. IPIMAR, Série digital (http://ipimar-iniap.ipimar.pt) nº
33, 58 pp.
Moncoiffe G., Álvarez–Salgado X. A., Figueiras F. G. and G. Savidge. 2000. Seasonal and short–time–scale
dynamics of microplankton community production and respiration in an inshore upwelling system. Marine
Ecology Progress Series 196: 111-126.
Mozetič, P., J. Francé, M. Šiško & O. Bajt. 2005. Spatial and temporal patterns of phytoplankton assemblages
in a shallow coastal sea (Gulf of Trieste). In: Wassmann, P. & B. Ćosović (eds.): Eutrophication in the coastal
zone of the eastern Adriatic Sea. South-eastern Europe Programme Symposium. April 27-May 1, 2005, Hvar,
Croatia.
Nogueira E., Pérez F. F. and A. F. Ríos. 1997. Seasonal patterns and long–term trends in an estuarine upwelling
ecosystem (Ría de Vigo NW Spain). Estuarine Coastal and Shelf Science 44: 285-300.
OJEC (Official Journal of the European Communities). 2000. Directive 2000/60/EC of the European Parliament
and of the Council. L327/1 - L327/72.
Ojeda A. 2004. Biomasa fitoplanctónica y clorofila–a en las Islas Canarias occidentales (Mayo 1986).
Universidad de Las Palmas de Gran Canaria. Biblioteca digital.
Orive E., Franco J., Madariaga I., and M. Revilla. 2004. Bacterioplankton and phytoplankton communities.
In: Oceanography and Marine Environment of the Basque Country. A. Borja and M. Collins (ed.). Elsevier
Oceanography Series nº 70, Elsevier, Amsterdam, 367-393.
OSPAR Commission. 2000. Quality Status Report 2000. Region IV- Bay of Biscay and Iberian coast. OSPAR
Commission, London. 134 + xiii pp.
OSPAR Commission. 2003. OSPAR Integrated Report 2003 on the Eutrophication Status of the OSPAR
Maritime Area Based Upon the First Application of the Comprehensive Procedure.
Pagou, K. (2000).
Assessment of the trophic conditions in the Inner Thermaikos Gulf.
Technical Report for the
Ministry of Environment, Planning and Public Works, NCMR, Athens, December 2000, 11pp.
Prego R. 1994. Nitrogen interchanges generated by biogeochemical processes in a Galician Ría. Marine
Chemistry 45: 167-176.
Prego R. and F. Fraga, 1992. A simple model to calculate the residual flows in a Spanish Ria. Hydrographic
consequences in the Ría of Vigo. Estuarine Coastal and Shelf Science 34: 603-615.
Prego R., Barciela M. C. and M. Varela, 1999. Nutrient dynamics in the Galician coastal area (Northwestern
Iberian Peninsula): do the Rias Bajas receive more nutrient salts than the Rias Altas? Continental Self Researc
19: 317-334.
Rousseau, V., Mathot, S. and C. Lancelot. 1990. Calculating carbon biomass of Phaeocystis sp. from microscopic
observations. Marine Biology 107: 305-314.
Siokou-Frangou, I. & Pagou, K. (2000).
Assessment of the trophic conditions and ecological status in the
Inner Saronikos Gulf.
Technical Report for the Ministry of Environment, Planning and Public Works, NCMR,
Athens, March 2000,43pp. (in greek and english edition).
173
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
Smayda T. J., 1980. Phytoplankton species succession. In: The physiological ecology of phytoplankton.
I. Morris (ed.). Blackwell Science Publications, Oxford, 489 pp.
Uthermöhl H., 1958. Zur vervollkommnung der quantitative phytoplankton methodik. Mitteilungen
Internationale Vereinigung fuer theoretische und angewandtelimnologie. 91-38.
Valencia V., Franco J., Borja A., and A. Fontán. 2004. Hydrography of the Southeastern Bay of Biscay. In:
Oceanography and Marine Environment of the Basque Country. A. Borja and M. Collins (ed.). Elsevier
Oceanography Series nº 70, Elsevier, Amsterdam, 159-194.
Varela M., Prego R., Belzunce M. J. and F. M. Salas. 2001. Inshore–offshore differences in seasonal variations
of phytoplankton assemblages: the case of a Galician Ría Alta (Ría de A Coruña) and its adjacent shelf (NW of
Spain). Continental Shelf Research 21: 1815-1838.
Vergara et al. 2005. (Eds. Vergara Martín, J.M., Haroun Tabraue, R. y González Henríquez, N.) Evaluación
de Impacto Ambiental de Acuicultura en Jaulas en Canarias. Oceanográfica, Telde. ISBN: 84-609-4073-X.
Depósito Legal: GC 422-2005.
Vives F. and F. Fraga. 1961. Producción básica en la Ría de Vigo. Investigación Pesquera 19: 129-137.
Vollenweider,R. A., A. Rinaldi & G. Montanari. 1992. Eutrophication, structure and dynamics of a marine
coastal system: results of ten-year monitoring along the Emilia-Romagna coast (Northwest Adriatic Sea). Sci.
Tot. Environ. [Vollenweider, R. A., R. Marchetti & R. Viviani (eds.)],. Suppl.: 63-106.
WFD CIS Guidance Document No. 14: Guidance on the intercalibration process 2004-2006
Wiltshire, K. H. & C. D. Durselen. 2004. Revision and quality analyses of the Helgoland Reede long-term
phytoplankton data archive. Helogol. Mar. Res., 58(4): 252-268.
5 Annexes
Annexes can be downloaded from the following address:
http://circa.europa.eu/Public/irc/jrc/jrc_eewai/library?l=/intercalibration_2&vm=detailed&sb=Title
Section 3 – Phytoplankton – Overview of Annexes
A – Baltic GIG
Annex 3.1: Development of chlorophyll a reference conditions in Danish coastal waters
Annex 3.2: Boundary setting for chlorophyll a in Danish coastal waters
Annex 3.3: Chlorophyll a reference values and boundary setting in Estonian coastal waters
Annex 3.4: Chlorophyll a reference values and boundary setting in Finnish coastal waters
Annex 3.5: Chlorophyll a reference values and boundary setting in Polish coastal waters
Annex 3.6: Chlorophyll a reference values and boundary setting in Swedish coastal waters
Annex 3.7: Chlorophyll a reference values and boundary setting in German coastal waters
B – Mediterranean GIG
Annex 3.8 – National methods included in the intercalibration
174
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0177.png
Section 4 – Macroalgae
1 Introduction
Two of the four coastal water GIGs have been able to produce results for the macroalgae quality
element (Mediterranean and North-East Atlantic). Each GIG has several typologies that are found
in these waters. Not all countries within each GIG have all the types within its borders. Information
about the types and countries with each type in each GIG is described in the sections below.
Both GIGs have chosen Option 2 in this phase of the intercalibration process. Member States have
developed metrics as part of their own classification schemes. Selected metrics have been chosen by
each GIG for assessment and agreement of boundaries in this phase of the intercalibration process.
At this stage not all the metrics that make up Member States’ schemes can be intercalibrated, so it is
not possible to produce EQRs for the whole quality element. Boundaries have been agreed for the
selected metrics.
2 Methodology and results
2.1 Mediterranean GIG
2.1.1 Intercalibration Approach
Two classification methods were developed by two countries: EEI by Greece and CARLIT by Spain.
Intercalibration Option2 (ECOSTAT WG Guidance) was adopted. The InterCalibration Common
Metric (ICCM) used was the method “BENTHOS” developed by Spain.
Six different Mediterranean countries participated in the subgroup of macroalgae:
Cyprus
France
Greece
Italy
Slovenia
Spain
Macroalgae intercalibration has been carried out between Spain, Greece, Slovenia. Other countries
that could not provide data for the IC exercise or do not have data yet (Italy, France and Malta) or
provided data but shortly before the finalization of the report (Cyprus), approved the results.
Typology
During the early stages of the CIS the Mediterranean working group agreed in using only 2
parameters to distinguish water types, namely substrate composition and depth. Most of other
geomorphological parameters, described in Directive Annex II (1.2.4), were not relevant (i.e. tidal
regime) to distinguish different water types in relation to their ecological “significance” for the
175
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0178.png
Mediterranean Sea. Four main Types were then defined (Table 2.1.1). However, throughout the CIS,
following data analyses for the different BQEs, these types did not actually proved to be relevant for
the IC exercise, for some BQEs, as Mediterranean ecosystem is quite homogeneous in comparison
to Northern Seas (some ecological differences do exist but within the Mediterranean scale).
For Macroalgae Intercalibration the methods used are applied to macroalgal communities (species
composition and abundance) of the upper infralittoral zone (3.5 to 0.2 m depth) in rocky coasts,
with no types distinction.
Table 2.1.1:
Main water body types of coastal waters within the Mediterranean Sea.
Type
CW - M1
CW - M2
CW - M3
CW - M4
Name of Type
Rocky shallow coast
Rocky deep coast
Sedimentary shallow coast
Sedimentary deep coast
Substratum
1
Depth
2
rocky
rocky
sedimentary
sedimentary
shallow
deep
shallow
deep
1) Since in many cases different seabed substrata will occur within one water body type, the dominant substratum
should be taken under consideration.
2) Depth division is based on 40 m. depth, at 1 mile distance from the coastline.
2.1.2 National methods that were intercalibrated
The methods intercalibrated were: EEI and CARLIT. EEI was elaborated by Greece, while CARLIT
by Spain. Another method, BENTHOS, that was adopted as Intercalibration Common Metric
(ICCM), has also been developed by Spain in parallel with CARLIT. All these methods share a
common view for reference conditions.
The other Mediterranean countries that participated in the Macroalgae subgroup have decided
which methods they already applied or are on the way to apply on their country data on macroalgae
(Table 2.1.2).
Table 2.1.2:
Countries decisions for macroalgal methodologies.
Cyprus
France
Italy
Greece
Slovenia
Spain
QE 3: Macroalgae
Assessment Method
CARLIT
CARLIT
CARLIT, BENTHOS
EEI
EEI
EEI
Officially accepted
Officially accepted
Under consideration
Officially accepted
Officially accepted
Officially accepted
Status
176
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0179.png
Metric: Ecological Evaluation Index (EEI)
Ecological Evaluation Index (EEI) was designed to estimate the ecological status classes (ESC) of
transitional and coastal waters using benthic macrophyte communities as bioindicator
26
(Orfanidis et
al., 2001, 2003; Panayotidis et al., 2004; Orfanidis, 2007). It is based on the well-known pattern that
anthropogenic stress, e.g. eutrophication, heavy metal pollution, shifts the ecosystem from pristine
state, where late-successional species is dominant, to degraded state, where opportunistic species
through rapid growth and recruitment is dominant. This pattern can be explained from the species
competition abilities under abundant and limiting nutrient conditions and is in accordance to r- and
K-selection theory.
In order to evaluate the spatial scale-dependent ESC of the studied coast (EEI
EGR
), the area-weighted
value was calculated. For this purpose, the score of each site was multiplied by the percentage of the
coast area for which is considered to be representative and the products were summed. EEI values
higher than 0.5 indicate sustainable ecosystems of good or high ESC, whereas EEI values lower
than 0.5 indicate that the ecosystems should be restored to a higher ESC (Table 2.1.3). These values
should be regarded as provisional that have to be further verified in the future.
Figure 2.1.1:
Ecological State Groups (ESG) I, II of marine benthic macrophytes (macrophyte graphs are based on
diverse sources).
A readily measured component or metric of the biota that are used to provide long-term ecologically relevant
information about the state and trends of ecosystem. Such an approach effectively distinguishes responses of
human impact from natural variability, when supported by predictive modelling and sound ecological theory.
26
177
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0180.png
Table 2.1.3:
Classification scheme of transitional and coastal waters based on the Ecological Evaluation Index (EEI).
Metrics: BENTHOS, CARLIT
Both methodologies have been published recently and described in detail. The references of the
papers where methodologies have been described are the following:
BENTHOS:
Pinedo, S. García, M., Satta, M.P., de Torres, M. and Ballesteros, E., 2007. Rocky-
shore communities as indicators of water quality: A case study in the Northwestern Mediterranean.
Marine Pollution Bulletin. 55, 126-135.
CARLIT:
Ballesteros, E., Torras, X., Pinedo, S., García, M., Mangialajo and L., de Torres, M.,
2007. A new methodology based on littoral community cartography dominated by macroalgae for
the implementation of the European Water Framework Directive. Marine Pollution Bulletin. 55,
172-180.
2.1.3 Reference conditions and class boundary setting
Reference condition is the expression of high quality structure and function of aquatic ecosystems,
that should have not suffered any impact on their natural state because of human activities and
there is none or only very minor evidence of disturbance on each of the general physico-chemical,
hydromorphological and biological quality elements. Reference sites have been identified according
to the low pressures and impacts they receive in accordance with Annex V of WFD. In all methods
(EEI, BENTHOS, CARLIT) the reference sites are real sites (existing) and this allows the
application of the tested methodologies in these places.
Ecological Evaluation Index (EEI)
The reference conditions have to be presented by values of relevant biological quality elements
indicating high ecological status. For the description of macroalgal community of the rocky upper
infralittoral zone reference conditions in Greek coastal waters 62 samples from 26 putatively
pristine Aegean sites (Figure 2.1.2) dominated by
Cystoseira
cf.
crinita
community as part of the
Hellenic “NATURA 2000” data-base (see Panayotidis
et al.,
2001) in combination with the biotic
index Ecological Evaluation-EEI Index (Orfanidis et al., 2001; 2003) were used. The aim was (1)
to develop an objective and statistically valid “virtual” list of the most common algal species in the
Aegean under undisturbed conditions, and (2) to test the conceptual model and the EEI recently
developed by Orfanidis et al. (2001, 2003) for the implementation of Water Framework Directive
(2000/60/EC) in Greek coasts.
In total 113 taxa (73 Rhodophyceae, 25 Phaeophyceae, 15 Chlorophyceae) were identified in
Cystoseira
cf.
crinita
community of the Aegean Sea (Panayotidis et al., 2007). Nine (9) major taxa
178
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0181.png
(except
C.
cf.
crinita)
contributed cumulatively by 90 % in the community:
Haliptilon virgatum,
Cystoseira compressa, Jania rubens, Padina pavonica, Herposiphonia secunda, Corallina
elongata, Cladophora
spp.,
Sphacelaria cirrosa
and
Titanoderma cystoseirae.
Moreover, 34 taxa
contributed cumulatively by 99 %. Under-storey layer considerably dominated to the community
with most common representatives the red coralligenous algae
Haliptilon virgatum, Corallina
elongata
and
Jania rubens,
and the brown alga
Padina pavonica.
It was followed by
C. crinita
epiphytes distinguished in: 1) filamentous green (Cladophora
spp.), brown (Sphacelaria
cirrosa)
and red (Herposiphonia
secunda)
algae, and 2) in encrusting red algae (Titanoderma
cystoseirae
and
Hydrolithon
spp.).
Cystoseira compressa
contributed significantly (23.08 %) to
C. crinita
community indicating that these species share common habitat resources in the Aegean Sea.
Within the common
Cystoseira
cf.
crinita
taxa 21 (62 %) belong to opportunistic ESG II, whereas
13 (38 %) taxa belong to late-successional ESG I (Table 2.1.1). By contrast the ESG I taxa
dominated quantitatively (111 %; including
C.
cf.
crinita)
the ESG II (21.9 %) taxa in the
C.
cf.
crinita
community (Figure 2.1.5). This result did not change when naturally eutrophic North Aegean
sites (ESG I=128 %, ESG II=21 %) were differentiated form naturally oligotrophic sites of South
Aegean (ESG I=101 %, ESG II=22 %). This result is in accordance both: a) to the conceptual model
of Orfanidis et al. (2001, 2003) that “in less anthropogenic stressed coastal areas (pristine) the late-
successional species dominate” and b) to the basic assumption of the EEI that ESG I, II average
contribution in undisturbed areas is higher than 60 % and less than 30 %, respectively. Data from
tentatively pristine sites of Slovenian (Orlando-Bonaca et al., 2008) and Cyprus (pers. comm. M.
Argyrou) coasts as well as from less anthropogenic impacted sites of Catalan coasts (Arévalo et al.,
2007) have also verified the above hypotheses.
Since macroalgae and especially long lived genera like Fucales also follow long-term periodicity,
it is important to notice that their absence from a site should be regarded as indicative of
environmental degradation only if it is correlated with key abiotic parameters, e.g. water and
sediment nutrient concentrations, light attenuation.
Figure 2.1.2:
Map of putatively
pristine sites in the Aegean Sea,
Greece.
179
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0182.png
150
Aegean Sea
(n=62)
North Aegean Sea
(n=23)
South Aegean Sea
(n=39)
120
Coverage (%)
90
60
30
Mean
Mean±SE
Mean±1,96*SE
ESG I
ESG II
ESG I
ESG II
ESG I
ESG II
0
Figure 2.1.3:
Average coverage ( %) of ESG I, II in undisturbed Aegean Sites.
BENTHOS, CARLIT
Based on expert judgment and on historical data and information we used as RC for BENTHOS
and CARLIT existing undisturbed sites or with only very minor disturbance. A problem in deriving
reference conditions in the Catalan coast sites is the absence of unimpacted areas. Therefore, we
have chosen three reference zones outside Catalonia: Façade maritime du Parc Naturel Régional
de Corse (France), Parc Natural de Ses Salines (Balearic Islands, Spain) and Reserva Marina del
Nord de Menorca (Balearic Islands, Spain) (Figure 2.1.3). All these places have in common a
very low human influence with physico-chemical and hydrogeomorphological conditions similar
to the Catalan coast. We have also historical data for benthic communities in the Catalan coast
before 1980’s (Gibert, 1918; Seoane-Camba, 1975; Polo, 1978) and in the adjacent Albères coast
(Sauvageau, 1912; Feldmann, 1937; Gros, 1978; Thibaut et al., 2005) showing that previous littoral
vegetation in the Catalan coast was very similar to that currently observed in the selected reference
areas.
These sites, in rocky shores exposed to high irradiance levels, are characterized by dense
communities of several
Cystoseira
species:
C. mediterranea/amentacea
var.
stricta, C. crinita, C.
brachycarpa var. balearica, C. foeniculacea/barbata/spinosa var. tenuior/compressa var. pustulata.
Alternatively, in shadow zones (steep vertical cliffs, high hydrodynamic conditions)
Lithophyllum
byssoides
develops, forming important organogenic structures (trottoir).
In order to obtain reference conditions’ data for Benthos and Carlit methods we have sampled using
both methodologies at the same period of time, from May to June (2001). For Benthos, reference
conditions were assessed by sampling 11 stations situated within reference sites (4 in Corsica, 4 in
Menorca and 3 in Freus) and to obtain CARLIT’s reference conditions the whole coastline in these
three reference sites was cartographied.
180
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0183.png
Figure 2.1.3:
Location of study area and sites (Northwestern Mediterranean Sea). Reference sites are included in the map.
Common biological criteria of EEI, BENTHOS, CARLIT approaches for reference
conditions
EEI, BENTHOS, CARLIT approaches share common views for reference conditions within
Mediterranean Sea summarized as follows.
1. Macroalgal communities of high diversity should be dominated quantitatively by brown
algae mainly of the order Fucales in high irradiance sites and red algal Corallinales in
vertical cliffs.
2. Dense well-developed macroalgal communities thriving in the upper infralittoral zone with
most characteristic species belonging to the genera
Cystoseira, Sargassum, Lithophyllum,
Peyssonnelia, Corallina
and
Padina.
Other common species belong to the genera
Halopteris, Stypocaulon, Dictyota, Dictyopteris, Laurencia, Cladophora
and
Jania.
3. In shadow zones (exposed steep vertical cliffs)
Lithophyllum byssoides
develops, forming
important organogenic structures (trottoir). In marine caves with scarce light conditions a
sciaphilic vegetation of red and green algae dominant.
4. Spatio-temporal variability of the community’s composition and abundance affected by hard
substrata availability, intense and frequency of natural disturbances, e.g. hydrodynamism,
grazing, by seasonal cycle of light period and intense, and by limiting factors like nutrients.
Boundary setting (national level)
Boundaries are set according to biotic index and/or combined with the results of or multivariate
analysis. No statistical analysis exclusively to set boundaries. No discontinuities. Continuum of
possibilities with gradual disappearance/appearance of different indicator species.
Ecological Evaluation Index (EEI)
The conceptual model proposed by Orfanidis et al. (2001, 2003) used the functional differences of
benthic macrophytes related to their life-cycle strategy (r-,
K-selected
species) as a tool to evaluate
the ecological status of transitional and coastal waters (Figure 2.1.4). This scheme evidences the
181
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0184.png
existence of a gradient between two states, which represent the pristine (high-good) and the degraded
(low-bad) conditions. The dominance of the late-successional species of the genera
Cystoseira
form
communities indicative of pristine state, which is characterized, for example, by low nutrient and
clear water conditions, whilst the dominance of opportunistic seaweeds as
Ulva
and
Gracilaria,
and Cyanobacteria form communities indicative of degraded state, which is characterized by
high nutrients, heavy metals and turbid conditions. The coexistence of the late-successional like
Cystoseira, Sargassum, Corallina
species with opportunistic like
Ulva, Cladophora, Gracilaria,
Cyanobacteria species form communities that are indicative intermediate (moderate) conditions.
In this successional model the ecological role of
Corallina elongata
and its community should be
discussed in more details.
Corallina
is a slow growing calcified perennial red alga (ESG I) can be
found in the infralittoral zone of most rocky Mediterranean coasts subjected to different kinds of
natural, e.g. grazing, low irradiance, hydrodynamism or anthropogenic disturbance, e.g. pollution.
Therefore,
C. elongata
alone could not be relevant metric of ecosystem state and trends. For EEI,
which investigates anthropogenic stress at community level,
C. elongata
community is regarded as
indicative of both high to good or of moderate to bad conditions depending on high quantitative
contribution of ESG I or ESG II species, respectively. To estimate accurately community’s
contribution of small sized and delicately constructed ESG II species being very sensitive to adverse
hydrological conditions which predominate in open coasts, especially during the winter season,
seasonal and destructive sampling protocols are needed.
High status (10
EEI > 8). The composition of macroalgal taxa is consistent with undisturbed
conditions. There are no detectable changes in macroalgal cover due to anthropogenic activities.
This condition corresponds with unpolluted sites, where the late successionals taxa, especially
species of
Cystoseira
genus, represented by the ecological group ESG I account for more than 60
% of the mean macroalgae abundance-coverage and the early successionals taxa represented by the
groups ESG II account for 0-30 % of the macroalgae coverage.
Good status (8
EEI > 6). There are slight changes in the composition and abundance of macroalgal
taxa compared to the type-specific communities. Such changes do not indicate any accelerated
growth of phytobenthos or higher forms of plant life resulting in undesirable disturbance to the
balance of organisms present in the water body or to the physicochemical quality of the water. This
condition corresponds with slightly polluted sites (unbalanced). At the good status as is indicated
by the EEI, the ESG I group may range from 30 to 60 % while the ESG II from 0 to 30 % of the
macroalgae coverage, or the combination may thus that ESG I accounts for over 60 % and ESG II
between 30 and 60 % of the total macroalgae coverage.
Moderate status (6
EEI > 4). The composition of macroalgae taxa differs moderately from type-
specific conditions and is significantly more distorted from a good quality. Moderate changes in
the coverage macroalgal disturbance are evident and may be such as to result in an undesirable
disturbance to the balance of organisms present in the water body. This condition corresponds with
moderate polluted sites. At the moderate status as is indicated by the EEI, the two groups may
equally share the macroalgae coverage accounting for equally low, moderate or high percentages.
Poor status (4
EEI > 2). At the poor status as is indicated by the EEI, the late successional group
ESG I may account for 0-30 % and the early successional group ESG II for 30-60 % or the late
successional group may represent a coverage among 30 % and 60 % while the early successional
group may account for over 60 %.
Bad status (EEI=2): At the bad status the sensitive group ESG I accounts for 0-30 % and the early
successional group ESG II represents over 60 % of the total macroalgae coverage.
EEI application was also tested across nutrient gradients of different Mediterranean coasts of
Greece (Orfanidis et al., 2001; Panayotidis et al., 2004; Orfanidis and Panayotidis, 2005) and
Spain (Arévalo et al., 2007; see also Orfanidis, 2007) with available consistent seasonal data
(Figure 2.1.5). The result seems to be very promising indicating clear gradients between degraded
182
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0185.png
Figure 2.1.4:
Conceptual model of two alternative stable states of macroalgal communities
across an ecological status gradient in coastal waters. A conventional (A) and dynamic (B)
view of successional changes.
and pristine sites. A disagreement appeared in Site 2 of Tossa de Mar coasts located very close
to sewage outfall (Arévalo et al., 2007). In that site ESG I species, mainly
Corallina elongata,
dominated (ESG I=251 % mean annual coverage) to ESG II (12.2 %) species indicating a high
ESC. The very low coverage of ESG II species of green and red algae, as well as the absence of
Cyanobacteria (X. Torras correspondence) seems to be an interesting case that needs further
consideration. Figure 2.1.6 shows an exponential relationship between total dissolved inorganic
nitrogen (TDIN) and phosphorus mean concentrations and EEI values.
CARLIT/BENTHOS
Species of the genus
Cystoseira
(Fucales, Cystoseiraceae) dominate Mediterranean upper
infralittoral communities (Feldmann, 1937; Boudouresque, 1971) and are particularly sensitive to
any natural (Gros, 1978; Verlaque, 1987) or anthropogenic stress (Bellan-Santini, 1966; Ballesteros
et al., 1984; Hoffmann et al., 1988; Soltan et al., 2001) and, therefore, have experienced profound
changes and decline over extensive areas (Thibaut et al., 2005). The highly structured and productive
Cystoseira mediterranea/stricta/crinita
communities are observed in hydrodynamic environments
and non-polluted waters along the Northwestern Mediterranean coasts (Boudouresque, 1969;
Ballesteros, 1988). Increasing concentrations of organic matter and nutrients drives
Cystoseira-
dominated communities to be replaced by the red alga
Corallina elongata
(Bellan-Santini, 1965,
1968; Ballesteros et al., 1984; Giaccone, 1993) and the mussel
Mytilus galloprovincialis
(Bellan-
Santini, 1965, 1968; Bellan and Bellan-Santini, 1972). Green ephemeral algae begin to dominate
in highly disturbed environments and near freshwater discharges:
Ulva
(Golubic, 1970; Bellan
and Bellan-Santini, 1972; Rodríguez-Prieto and Polo, 1996),
Cladophora
(Belsher, 1977) or
Enteromorpha
(Ballesteros et al., 1984; Kadari- Meziane, 1994). Finally, the dominance of blue-
green algae (Oscillatoria,
Lyngbya, Phormidium)
indicates very degraded environments (Golubic,
1970; Littler and Murray, 1975; Murray and Littler, 1978).
183
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0186.png
Figure 2.1.5:
Implementation of
EEI across water quality gradients
in different Mediterranean coastal
waters: Kavala and Maliakos Gulfs
(Orfanidis and Panayotidis 2005),
Saronikos Gulf (Panayotidis et al.,
2004), Tossa de Mar coasts (Arévalo
et al., 2007).
Based on these ecological changes along pollution gradients the boundary between High and Good
conditions when
Cystoseira
communities occur in patches and do not make extensive-continuous
assemblages and the
Lithophyllum byssoides
belt displays symptoms of degradation. Samples
of
Cystoseira
assemblages indicate lower biomass of
Cystoseira
spp. and the substitution of the
sciaphilic species inhabiting the underlayer dense
Cystoseira
assemblages by
Corallina elongata
or
Mytilus galloprovincialis.
The disappearance of these sensitive species and its replacement by stress
tolerant species such as
C. elongata
and
M. galloprovincialis
defines the boundary between Good
and Moderate situations.
In Italy, Carlit method has been tested at Regional scale, in the Ligurian Sea, applying it in a
moderate urban gradient (figure 2.1.7) and in four Marine Protected Areas (MPAs), proposed as
hypothetical reference sites at a regional scale (Mangialajo et al., 2007). This study shows that Carlit
index is suitable to detect different kinds of anthropogenic pressures obtaining a good correlation
with different water column variables.
184
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0187.png
35
TDIN=37.07*exp
(-3.2*EEI)
, r=-0.9, p<0.000
P-PO4=12.03*exp
(-4.5*EEI)
, r=0.9, p<0.000
12
30
10
25
8
N-NO
3
( µ mol/l)
20
6
15
4
10
2
5
0
2
4
6
8
10
0
Ecological Evaluation Index (EEI)
Figure 2.1.6:
Total dissolved inorganic nitrogen (TDIN) and Phosphorus mean
concentrations plotted against EEI values. Data from different Mediterranean coastal
waters.
Figure 2.1.7:
Carlit EQR values along a moderate urban gradient in the Ligurian sea.
P-PO 4 ( µ mol/l)
185
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0188.png
2.1.4 Intercalibration
Material & Methods
Data sampled along Greek (Figure 2.1.8), Catalan (Figure 2.1.9), and Slovenian locations were
exchanged for intercalibration.
Data from Greek coasts were sampled from 11 sites (96 samples in total): Saronikos (6 sites, 6
sampling periods, 36 samples in total) and Maliakos (5 sites, 4 sampling periods, 60 samples in total)
Gulfs. Fifty one sites (1 sampling period, 151 samples in total) were sampled along the Slovenian
coasts (Lipej et al., 2006). All samples were destructive (1 to 3 random sample from a permanent
site-square 5 x 5 m per sampling period) on a quadrate 20cm x 20cm, which is considered to be the
minimal sampling area in the case of the Mediterranean infralittoral communities. In the laboratory
each sample was sorted carefully and the surface covered by each species in vertical projection
was quantified as % of cover (4 cm
2
= 1 % sampling surface). The total coverage usually exceeded
100 % due to the presence of different layers in the vegetation (canopy, under storey layer, crusts
and epiphytes). A nominal coverage value (0.1 %) was allocated to species showing insignificant
abundance. Taxonomically difficult taxa were consistently sub summarized to genus level as spp.
Data from Spain coasts were sampled from 48 sites (49 samples in total) Tossa-St. Feliu coast
(6 sites, High-Good Water Body) and Hospitalet-Ametlla coast (7 sites, Good-Moderate WB).
Furthermore, 36 extra sites from different Catalan places with Moderate to Bad ES were included
in the intercalibration exercise (Figure 2.1.9). The samples were collected in springtime in the upper
Figure 2.1.8:
Map of sampling sites of Maliakos and
Saronikos Gulfs. Colours correspond to Ecological Status
Classes in accordance to EEI.
186
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0189.png
!!
!
!
!
!!
!
!
!
!!
!
!!
!! !
!
!!
High-Good Intercalibration site
!
!
!!
! !
! !!
!
!!
!
!!
!!
!!
!
Good-Moderate Intercalibration site
!!
!
Figure 2.1.9:
Map of Catalonia sampling sites. Colours correspond to Ecological Status
Classes in accordance to Benthos methodology.
sublittoral zone on sub-horizontal rocky shores facing east, south or west. For each sample, the whole
community was collected from a 15 x 15 cm surface, using a hammer and a chisel. This surface is
large enough to quantitatively represent the littoral communities in the Northwestern Mediterranean
(Coppejans, 1980; Verlaque, 1987; Ballesteros, 1992). Samples were preserved in formalin: sea-
water at 4 % and were subsequently sorted in the laboratory. Algae and invertebrates were identified
to species level and quantified in terms of coverage (horizontal surface; Ballesteros, 1992).
2.1.5 Results of the comparison
EEI and BENTHOS were compared on 62 sites in Greece (11 sites) and Slovenia (51 sites), while
BENTHOS and CARLIT on 48 sites in Spain.
The relation between variable EQR-EEI and variable EQR-ICM (Benthos) is described by a linear
equation:
(1) EQR-ICM = 0.594 EQR-EEI + 0.3425, coefficient of determination R² = 0.85 (Figure 2.1.10).
The relation between variable EQR-ICM and variable EQR-CARLIT is also described by a linear
equation.
(2) EQR-ICM = 0.8604 EQR-CARLIT + 0.0709, coefficient of determination R ² = 0.77
(Figure 2.1.11).
187
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0190.png
0,8
0,6
1
0,4
0,8
0,2
EQR-ICM
0,6
0
0,4
0,2
0
0
0,2
0
0,2
R
2
= 0,8458
EQR-ICM
y = 0,594x + 0,3425
2
R = 0,8458
0,4
0,6
EQR-EEI
0,8
1
0,4
0,6
EQR-EEI
0,8
1
Figure 2.1.10:
Relation between EQR-ICM and EQR-EEI.
1
0,9
0,8
0,7
EQR-ICM
0,6
0,5
1
0,4
0,9
0,3
0,8
0,2
0,7
0,1
0,6
0
0,5
0,4
0,3
0,1
0
0,2
y = 0,8604x + 0,0709
2
R = 0,7739
EQR-ICM
y = 0,8604x + 0,0709
2
R = 0,7739
0,4
0,6
EQR -CARLIT
0,8
1
Figure 2.1.11:
Relation between EQR-ICM and EQR-CARLIT.
0,2
0
From the equations (1) and (2) we calculated new values of
0,8
class boundaries while replacing in the
0
0,2
0,4
0,6
1
equation (1) EQR-EEI by the boundary values of EQR defined for method EEI (ICM-boundaries
EQR -CARLIT
EEI) and the same for the boundary values of EQR defined for CARLIT (ICM-boundaries CARLIT)
by using the equation (2) (Tables 4 and 5). CARLIT and BENTHOS (ICM) have the same values of
class boundaries.
Table 2.1.4:
EQR of the class boundaries used for methods EEI and CARLIT.
Boundaries
H/G
G/M
EEI-Boundaries EQR
0.75
0.50
CARLIT-Boundaries
EQR
0.75
0.60
188
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0191.png
For each class boundary the new mean values of class boundaries between ICM-boundaries
CARLIT and ICM-boundaries EEI were calculated as well as their acceptable range (average ±
0.05; Table 5), defined as comparability acceptable range by the ECOSTAT group.
Table 2.1.5:
New EQR of the class boundaries calculated as mean values and their acceptable range (average + 0,05).
Boundaries
H/G
G/M
ICM-boundaries
(CARLIT)
0.72
0.59
ICM-boundaries (EEI)
0.79
0.64
Average
0.75
0.61
Acceptable range (average ± 0,05)
0.70 - 0.80
0.56 - 0.66
ICM-boundaries (CARLIT) and ICM-boundaries (EEI) are in the acceptable range of variability,
for their comparison, defined during the intercalibration process, in the ECOSTAT Working Group
(Figure 2.1.12).
1
0,9
0,8
0,7
0,6
ICM-EQR
0,5
0,4
0,3
0,2
0,1
0
SP
GR-SI-CY
Figure 2.1.12:
Intercalibrated EQR within the confidence interval.
2.1.6 Results of the harmonisation – Boundary EQR values
The harmonization process between the methods indicated that some differences, in respect to
the ecological meaning of certain species and communities, exist in the methods and, as a result,
each method has its own EQR values; nevertheless values are very close in all countries in which
methods have been applied.
189
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
2.1.7 Open issues and need for further work
Based on above described similarities and differences between the methods following topics were
emerged that needs to be searched in future:
1. To minimize differences between methods deeper knowledge in different aspects is needed:
(a) clarification of scale based ecological role of species like
Corallina elongata, Cystoseira
compressa
and their communities within the Mediterranean Sea, (b) method sensitivity to
environmental stress.
2.
Study of species-specific sensitivity and tolerance of dominant macroalgal species to
different disturbances.
3.
Development of type-specific reference conditions within rocky coasts, e.g. rocky shallow,
rocky deep, using new or existing literature data.
4. Application of common actions to describe different ecological status classes and test
different method effectiveness. Such an action is planned for spring/summer 2008 in
Slovenian coasts, where Slovenian and French experts will intercalibrate on a similar spatio-
temporal scale the CARLIT and EEI indices. The campaign will promote, if needed, the
development of local scales of sensitivity levels of species (CARLIT) or possible adjustment
of ecological state groups (EEI). It will also provide information/data on the ecological value
of the taxa under debate.
5. Italy, after having tested the CARLIT method at Regional scale, has presently introduced
its use, at the national level, in the new National Monitoring Programme (2008-2011) for
coastal waters. A first training course for the Regional Agencies, that operate in the fields,
has already been carried out.
190
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0193.png
2.2 NE Atlantic GIG
2.2.1 Intercalibration approach
In the NE Atlantic seven basic intercalibration types have been agreed. These are shown in table
2.2.1 below:
Table 2.2.1:
NEA GIG Intercalibration types.
New Type ID
CW
–NEA1/26a,b,c,d,e
CW – NEA3/4
Name
Exposed or
sheltered,
euhaline,
shallow
Polyhaline,
exposed or
moderately
exposed
(Wadden Sea
type)
Deep, low
current,
sheltered
Polyhaline,
microtidal,
sheltered,
shallow
(Skagerrak
inner arc
type)
Fjord with
a shallow
sill at the
mouth with
a very deep
maximum
depth in the
central basin
with poor
deepwater
exchange.
Polyhaline,
microtidal
exposed,
deep
(Skaggerak
outer arc
type)
Transitional
waters
Salinity
(PSU)
Fully saline
(> 30)
Polyhaline
(18 - 30)
Tidal range
(m)
Mesotidal
(1 - 5)
Mesotidal
(1 - 5)
Depth (m)
Shallow
(< 30)
Shallow
(< 30)
Current
velocity
Medium
(1 - 3 knots)
Medium
(1 - 3 knots)
Exposure
Exposed or
sheltered
Exposed or
moderately
exposed
Mixing
Fully mixed
Residence
time
Days
Fully mixed
Days
CW – NEA7
CW – NEA8
Fully saline
(> 30)
Polyhaline
(18 - 30)
Mesotidal
(1 - 5)
Microtidal
(< 1)
Deep
(> 30)
Shallow
(< 30)
low
(< 1 knot)
low
(< 1 knot)
Sheltered
Sheltered
Fully mixed
Partially
Stratified
Days
Days-Weeks
CW – NEA9
Polyhaline
(18 - 30)
Microtidal
(< 1)
Deep
(> 30)
low
(< 1 knot)
Sheltered
Permanently
Stratified
Weeks
CW – NEA10
Polyhaline
(18 - 30)
Microtidal
(< 1)
Deep
(> 30)
low
(< 1 knot)
Exposed
Permanently
Stratified
Days
TW – NEA11
(will be split into
sub-types)
Oligo-
Euhaline
(0 - 35)
Mesotidal
(1 – 5 )
Shallow
(< 30)
Medium
Sheltered
or
moderately
Exposed
Partially- or
Permanently
Stratified
Days-Weeks
191
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0194.png
The above types occur in Member State’s waters as detailed below in table 2.2.2:
Table 2.2.2:
Member States sharing the types.
Type
CW -
NEA1/26a
CW -
NEA1/26b
CW- NEA1/26c
CW- NEA1/26d
CW-
NEA1/26e
CW – NEA3/4
CW – NEA7
CW – NEA8
CW – NEA9
CW – NEA10
TW – NEA11
x
X
X
X
X
x
x
x
x
x
x
x
X
X
X
x
x
X
x
X
X
X
X
X
BE
DK
FR
X
X
X
DE
IE
X
X
NL
NO
X
X
PT
ES
X
SE
UK
X
X
Option 2 has been used in this phase of the intercalibration process. Three metrics have been
selected, Perennial Intertidal Algae (in type NEA1/26a,b,e), Opportunistic Macroalgae (in types
NEA1/26a,b,c/11) and Perennial Subtidal Algae (in types NEA8/9/10). Boundaries have been
agreed for some countries for each metric where applicable. This is described below.
2.2.2 National methods that were intercalibrated
National methods are summarised in table 2.2.3 below. Further details can be found in the following
sections and appendices
Table 2.2.3:
National Methods.
Member State
BE
DK
FR
DE
IE
NL
NO
PT
ES
SE
UK
Yes
Yes
Yes
Yes
Yes
Yes
Only in TW’s
Yes
Yes
Yes
Yes
Yes
RSL
CFR
P-MarMAT
MAB
Subtidal algae
192
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
RSL (Reduced Species list) – Rocky shore tool
(Perennial Intertidal Algae)
A tool for ecological quality assessment of intertidal rocky shore macroalgae based on the species
richness from a reduced species list (RSL tool) of between 68 and 70 species had been developed
over the last 4 years by UK and has been adopted with slight local variation by Republic of Ireland
and Norway. The metric is based on five components:
Numerical species richness which incorporates a deshoring factor to correct for the overall
nature of the shore and its suitability for algal attachment.
The proportion of Rhodophyta species within the overall sample
The proportion of Chlorophyta species within the overall sample
The proportion of opportunistic species within the overall sample which includes the species
Blidingia, Chaetomorpha, Enteromorpha, Ulva, Ectocarpus, Pilayella and Porphyra.
The ratio of perennial forms (ESG I) to annual or ephemeral forms (ESG II)
CFR – Quality of Rocky Bottoms
(Perennial Intertidal Algae & Opportunistic macroalgae)
Composed of three indicators: percentage cover of characteristic macroalgae (Cover), characteristic
macroalgae population richness (Richness) and percentage cover of opportunistic species
(Opportunists) relative to the total vegetated surface.
P-MarMAT – Portuguese Marine Macroalgae Assessment Tool
(Perennial Intertidal Algae & Opportunistic macroalgae)
This includes all components from the RSL (species richness, proportion of greens, proportion of
reds, ESG ratio, proportion of opportunists and shore description) but in addition also includes
coverage of opportunists (%).
MAB – Macroalgae Blooming
(Opportunistic macroalgae)
This tool is based on 5 components:
% cover of algae over the available intertidal area (or the whole intertidal area for Germany)
calculated as an average
The total extent of the algal blooms measured in hectares
The biomass of the algae taken as an average per squared meter over the available intertidal area
The Biomass of the algae taken as an average per squared meter over the affected bloom area only
Presence of entrained algae calculated as a % of the total quadrats in which it is recorded.
Subtidal Algae
(Perennial Subtidal Algae)
This tool incorporates three metrics
Metric 1 describes the depth extension of selected perennial macroalgal species in
Scandinavian coastal waters. The species selected for each type are all 1) perennial, 2)
commonly occurring in the type and 3) easy to determine.
193
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
Metric 2 describes the cover of the macroalgal community along depth gradients. This is still
under development.
1) ‘Total cover’: overall cover of the community representing values up to 100 %, or
2) ‘Cumulated algal cover’: summed cover of all individual species, representing values up to
several 100 % as the algae may grow in several stories.
Metric 3 describes the macroalgal composition in rocky littoral and sublittoral zones. The
metric includes species richness (or total number of species recorded), proportions of red,
green and opportunist, ratio of ephemeral against perennial algae, ratio between Ecological
Status Groups I and II (ESG ratio), lower depth extension of a few selected species and their
abundance as well as the physical properties of the location. Again this metric is still under
development.
2.2.3 Reference conditions and class boundary setting
Metrics -RSL (Reduced Species List) -
Rocky Shore Tool,
CFR & P-MarMAT
CW - NEA1/26a,b,e
Member states involved in intercalibration of rocky shore tools:
Reduced Species List (RSL): (applied only to intertidal areas).
NB: the RSL doesn’t incorporate cover:
• UK
• Republic of Ireland
• Norway
CFR: (Applied either for intertidal or subtidal areas).
NB: the CFR incorporates cover:
• Spain
• Portugal
P-MarMAT:
NB: the P-MarMAT incorporates coverage of opportunists:
• Portugal
• Spain
France is yet to decide which metric will be adopted, but this means they will ‘automatically’
intercalibrate with whichever member state’s tool they adopt.
Reference Conditions
General reference conditions are based on the normative definition, which states, “All disturbance-
sensitive macroalgal and angiosperm taxa associated with undisturbed conditions should be present
and the levels of macroalgal cover and angiosperm abundance should be consistent with undisturbed
conditions”. The reference values indicated in the nominative definition are calculated as the
maximum values possible for the selected indices comprising the macroalgae metric.
The RSL (Reduced Species List) macroalgae metric indices include: Total numerical species
richness based on a reduced species list which includes a correction factor based on the shore
description, the proportion of red species within the sample, the proportion of green species,
the proportion of opportunist species listed as
Blidingia, Chaetomorpha, Enteromorpha, Ulva,
Ectocarpus, Pilayella
and
Porphyra,
and the ratio of ESG I/ESG II (ESG 1 – late successionals or
perennials and ESG 2 – ephemerals or annuals)
194
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0197.png
This translates into the following definition (table 2.2.4):
Table 2.2.4:
RSL reference values.
Biological element
Marine benthic
macroalgae
Optimal Metric
Scores
Scotland
England/Wales/RoI
NI
Norway
Reference value
Diverse community of red, green and brown seaweeds with high levels of species richness.
Cover variable depending on local physical conditions but species richness relatively constant
temporally. Red species present as richest group along with a high proportion of long-lived
spp. Opportunist and green species should constitute a lower proportion of the algal present.
Species Richness
≥35
≥35
≥34
≥33
% Green Species
≤12
≤15
≤20
≤20
% Red Species
≥55
≥55
≥45
≥40
ESG Ratio
≥1.0
≥1.0
≥0.8
≥0.8
% Opportunists
≤10
≤10
≤15
≤15
The reduced species list has a maximum of 70 Species that should be present for UK& IE shores,
and 68 for Norway. The actual species within these lists vary to represent regional differences in
species composition. These are the reference species numbers for the ‘reduced species list’ from
which species richness is derived.
In terms of the indices agreed for this metric then a reference site should score the maximum
possible i.e. 24 points, giving an EQR of 1.0 (or 20points if no shore description is available). The
tool now works on an EQR sliding scale and no longer equates to a point system as previously
described, the new method is much more accurate and is easier to see when a site is sitting on, or
close to, a class boundary.
For UK, IE, and Norway bodies reference values were established using historic data, historic
reports and publications and expert judgement. Expert judgement was then used to refine
recommended levels of macroalgae species richness to correspond to desired reference conditions
of maximum species richness and desirable community composition within natural and undisturbed
water bodies.
Spain and Portugal have tested the RSL and found that it was not appropriate for their coasts.
Spain has developed the CFR (Quality of Rocky Bottoms) Index (Juanes
et al.,
2008), which is
composed of three indicators: percentage cover of characteristic macroalgae (Cover), characteristic
macroalgae population richness (Richness) and percentage cover of opportunistic species
(Opportunists) relative to the total vegetated surface. The reference conditions for this metric are as
follows (table 2.2.5):
Table 2.2.5:
CFR reference values
Metric
Ref. Cond.
Characteristic Macroalgae Cover
Populations Richness
Relative Coverage of Opportunists
70 %
6
9%
195
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0198.png
Portugal has developed de P-MarMAT (Portuguese Marine Macroalgae Assessment Tool). Which
includes all metrics from RSL (species richness, proportion of greens, proportion of reds, ESG ratio,
proportion of opportunists and shore description) and coverage of opportunists ( %). The reference
conditions for this metric are as follows (table 2.2.6):
Table 2.2.6:
P-MarMAT reference values.
Metric
Species richness
Proportion of greens
Proportion of reds
ESG ratio
Proportion of opportunists
Shore description
Coverage of opportunists
Ref. Cond.
25
10
70
2.5
10
7
10
France is yet to decide which metric will be adopted, but this means they will ‘automatically’
intercalibrate with whichever member state’s tool they adopt.
Boundary Setting Protocol
UK, Ireland, Norway:
The Reduced Species List Index has been tested using UK, Irish, Norwegian, Portuguese and
Spanish Data. Boundaries have been agreed through data testing and common consensus (expert
judgement). As no direct relationship with a pressure gradient could be established, step 8 in the
boundary setting procedure was invoked (see Annex C of Milestone 6 report).
The full boundary setting protocol is laid out in the Milestone 6 report Annex C. At present only
UK, Ireland, and Norway are fully intercalibrated and have accepted the boundaries agreed for the
RSL index.
Spain:
The CFR index was developed in Cantabria (Juanes
et al.,
2008), region where it has been
extensively applied to intertidal and subtidal macroalgae communities (Guinda et al, 2008).
This method has been tested as an alternative method to that of the UK, in order to test a more
quantitative approach that may reflect, in a homogeneous way, the ecological condition of our
hard substrate habitats all through the extent of the water bodies (intertidal+subtidal). Based on
independent assessments of the status at different bathymetric levels (intertidal/ depth ranges), the
index provides a tool for the consideration of all the possible “coastal reef habitats” that, according
to the normative (e.g. Habitat Directive) or the expert judgements, are important in each water body.
Furthermore, the initial formulation of the index (indicators, boundary values) has been intensively
discussed by experts before the last version has been adopted.
The last version of the CFR index provides a quantitative approach for reflecting, in a homogeneous
way, the ecological condition of hard substrate habitats throughout the extent of the water bodies
in the North coast of Spain, based on independent assessment of the quality status at different
bathymetric levels (intertidal/subtidal). This metric integrates those features suggested by the WFD
196
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0199.png
for the assessment of seaweed communities, including the composition (Richness, R; presence
of Opportunistic species, O) and the abundance (Cover, C). The Richness value (R) evaluates the
number of different “characteristic macroalgae” populations that present a significant coverage
(ca. >1 %), according to a previously established list for each biogeographic region. Moreover, the
Cover score (C) assesses the extent that is occupied by those assemblages, considered all together.
The third indicator quantifies the abundance of Opportunistic species (O) in relation to the total
vegetated surface, as one of the first symptoms of several anthropogenic disturbances, mainly
related to nutrient enrichment.
Further details about theoretical basis and application procedures may be obtained from the
previous version of this index (Juanes
et al.,
2008), which included additional assessments of the
physiological status.
Different quality thresholds were considered for diverse degrees of exposure (slope) of intertidal sites
and depth ranges of subtidal stations. Table 2.2.7 shows the boundaries adopted for each metric.
Table 2.2.7:
Value boundaries and scores assigned to each of the three metrics of the CFR index, for its
application at different intertidal and subtidal zones.
CFR: Quality of Rocky Bottoms Index
Score
45
35
20
10
0
Semiexp. Int.
70-100%
40-69%
20-39%
10-19%
<10%
Exposed Int.
50-100%
30-49%
10-29%
5-9%
< 5%
Cover
(1)
5 - 15 m
70-100%
40-69%
20-39%
10-19%
<10%
15 - 25 m
50-100%
30-49%
10-29%
5-9%
< 5%
(1)
% Cover of Characteristic Macroalgae (CM).
Score
20
15
10
5
0
Semiexp. Int.
4-5
2-3
1
0
>5
Populations Richness
Exposed Int.
>3
3
2
(2)
5 - 15 m
4-5
2-3
1
0
>5
15 - 25 m
4-5
2-3
1
0
>5
1
0
(2)
Characteristic macroalgae populations richness.
Score
35
25
15
5
0
Opportunistic species
Intertidal
10-19%
20-29%
30-69%
<10%
(3)
5 - 15 m
5-9%
<5%
15 - 25 m
5-9%
<5%
70-100%
20-49%
50-100%
10-19%
20-49%
50-100%
10-19%
(3)
Relative cover of opportunistic or pollution indicator species respect to the total vegetated
surface.
197
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0200.png
The final CFR value is calculated by adding the scores obtained for each of the three metrics
(C+R+O). To obtain the corresponding EQR value ranging between 0 and 1, the CFR/100 value is
calculated and the corresponding status is assigned according to the ranges of the next table 2.2.8
(based on REFCOND 2003):
Table 2.2.8:
CFR ranges.
CFR Scores
83-100
62-82
41-61
20-40
0-19
EQR
0.83-1
0.62-0.82
0.41-0.61
0.2-0.4
0-0.19
Status
High
Good
Moderate
Poor
Bad
Portugal:
The metrics selected for the Portuguese marine macroalgae assessment tool (P-MarMAT) include
the ones from the RSL methodology proposed by UK, Norway and RoI (species richness, proportion
of greens, proportion of reds, ESG ratio, proportion of opportunists and shore description), and
the coverage of opportunists. A reduced species list was developed to Portugal (keeping the same
number of algae from green -8-, red -20- and brown -24- groups initially proposed by other MS),
and a new metric was included on the assessment method (coverage of opportunists). The scores
for the species richness, proportion of greens, proportion of reds, ESG ratio, and the proportion of
opportunists were calculated from the RSL adapted to Portugal (PT-RSL), and the shore description
followed the same procedures of the ones proposed earlier by the above mentioned MSs. The
coverage of opportunists represents the percentage cover of opportunists (defined from the PT-RSL)
on the total area covered by marine macroalgae in the assessed shore. The metrics species richness
and % coverage of opportunists have a factor of 2 on the contribution to the total score. The sum of
scores is converted in a 0 to 1 scale (EQR according to the WFD definitions), that afterwards allows
the determination of the EQS of the shore (according to WFD normative definitions). Table 2.2.9
shows the candidate metrics, the boundaries adopted for each of them to Portugal on this stage of
the IC exercise, and provides the way to go from the sum of scores to the EQS.
Table 2.2.9:
Candidate metrics comprising the P-MarMAT, their boundaries, and conversion of sum of scores in EQR
and EQS classes.
Species Richness (*)
Proportion of Reds
ESG Ratio
Quality
0-5
Bad
Poor
30 - 40
1 - 1.5
5-8
Moderate
20 - 30
1.5 - 2
9 - 16
17 - 24
10 - 20
2 - 2.5
8 - 11
55 - 70
10 - 20
10 - 20
0.6 - 0.8
Good
22 - 28
Good
High
0 - 10
> 2.5
1-7
> 24
Proportion of Greens
40 - 100
0 - 30
0-1
-
30 - 45
30 - 40
45 - 55
20 - 30
70 - 100
0 - 10
0 - 10
29 - 36
0.8 - 1
High
Proportion of Opportunists
Shore Descriptions
Coverage of Opportunists
(%) (*)
Sum of scores
EQR
EQS
40 - 100
70 - 100
0 - 0.2
Bad
0-7
15 - 18
30 - 70
12 - 14
20 - 30
0.2 - 0.4
Poor
8 - 14
Moderate
0.4 - 0.6
15 - 21
(*) factor of 2
198
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0201.png
Metric - MAB (Macroalgal Blooms) -
Opportunistic Macroalgae
(NEA1/26/11)
Reference Conditions
General reference conditions should fulfil the criteria for high status as defined by the Water
Framework Directive: There should be no or only very minor anthropogenic alterations to the values
of the physico-chemical and hydromorphological quality element relative to those associated with
undisturbed conditions. Specifically reference sites should also fulfil the criteria for high status of
the quality element ‘macroalgae and angiosperms’ as defined by the WFD: All disturbance-sensitive
macroalgal and angiosperm taxa associated with undisturbed conditions should be present and
the levels of macroalgal cover and angiosperm abundance should be consistent with undisturbed
conditions. The proposed metrics/indices are % cover and total area of opportunist macroalgae over
the available intertidal habitat (AIH – UK/IE), or % cover over the total intertidal (DE), biomass of
opportunist macroalgae and presence of entrained algae within the sediment definitions of which are
given below in table 2.2.10 (UK/IE):
Table 2.2.10:
UK/IE reference values.
Measurement
% cover of AIH
Total area (ha)
Biomass (g m2) of AIH
Biomass (g m2) of affected area
Presence of entrained algae
(% of Quadrats)
% cover of total intertidal
Definition
The % cover of algae is estimated as an average over the whole of the available
intertidal habitat for the whole waterbody
The total extent of algal bloom, measured in hectares and based on the external
perimeter of the bloom
This is the average biomass of algae per metre squared over the whole of the
available intertidal habitat
This is the average biomass of algae per metre squared over the affected area
only
This is where algae is seen to be growing > 3cm into the underlying sediment
indicating the likelihood of regeneration
The percentage cover of algae is estimated as an average over the whole
intertidal for the waterbody
This translates into the following definition (table 2.2.11):
Table 2.2.11:
UK/IE reference value definitions.
Biological element
Marine benthic macroalgae
Reference value
Opportunistic macroalgal blooms of anthropogenic origin should be absent
or if present should cover less than 5 % of the available intertidal habitat and
total biomass of macroalgae per square meter should be less than 100.
Total area coverage of opportunist macroalgae should be less than 100
hectares with no effect on quality status.
Entrained algae should only present in less than 5 % of quadrats.
DE – Opportunistic macroalgal blooms of anthropogenic origin should be
absent or if present should cover less than 1 % of the total intertidal of the
waterbody.
Generally directed at intertidal sedimentary shores in both transitional and
coastal waters.
199
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
Agreed Indices, Scoring Criteria and Use in Classification
The assessment system uses two metrics, firstly, % cover of opportunist macroalgae
which
responds directly to increased nutrients, turbidity and reduced salinity and refers to the total area
covered by opportunist macroalgae calculated as a percent of the total available intertidal habitat
consisting of sedimentary substrate (UK/IE) or to the total intertidal (DE). This metric responds
to anthropogenic disturbance by an increased % cover of algae. Secondly, the extent of cover of
opportunist macroalgae, which responds as above but is required to account for the total size of the
macroalgae bloom with the water body. This metric responds to anthropogenic disturbance by an
increased extent of opportunist macroalgae cover.
As nutrient input increases there is often a corresponding increase in turbidity and resulting decrease
in light attenuation subsequently providing a habitable environment for those species tolerant
to such conditions, these opportunist species take advantage of the condition and dominate the
community often to the detriment of other species. This is where macroalgae blooms occur. This
biological metric does seem to respond over the whole potential gradient of impact whereby each
quality status has defined levels of % cover and total extent of opportunist macroalgae, which
correspond directly to the changes in quality. Each quality status has well defined and static class
boundary and there is no overlap between classes.
The above applies in full to the metrics developed in the UK and RoI. Portugal is considering using
the metrics only in transitional waters. Germany has only extent data (based on 6 year median of
maximum
for year) referenced across the
whole
(as opposed to the
available)
intertidal habitat. (see
later sections).
Metric - Subtidal Algae (NEA8/9/10)
Reference Conditions
Conditions should fulfil the general criteria for high status as defined by the Water Framework
Directive (WFD): There should be no or only very minor anthropogenic alterations to the values
of the physico-chemical and hydromorphological quality element relative to those associated with
undisturbed conditions. Sites should also fulfil the criteria for high status of the quality element
‘macroalgae and angiosperms’ as defined by the WFD: All disturbance-sensitive macroalgal
and angiosperm taxa associated with undisturbed conditions should be present and the levels of
macroalgal cover and angiosperm abundance should be consistent with undisturbed conditions.
The best-explored benthic vegetation metric is ‘depth limits of selected macroalgal species’ and
the present protocol analyses this metric for Scandinavia. Further metrics ‘macroalgal cover’ and
‘multimetric macroalgal index’ are not yet developed to a state where we can use them in assessment
according to the WFD as e.g. reference levels are not known, but the protocol provides an initial
description of these metrics.
Hydromorphological conditions
Hydromorphological conditions corresponding to no or very minor anthropogenic alteration in
coastal waters involve absence of digging and construction activity.
Physico-chemical conditions
Nutrient concentrations and water transparency are main physico-chemical variables determining
the status of the benthic vegetation. We have attempted to identify reference levels for these
200
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
variables within the NEA8, NEA9 and NEA10 areas, where vegetation data are available. At
present, no Danish coastal areas are considered to represent reference conditions, so we attempt
to define reference levels of total-nitrogen concentrations (TN) and water transparency based on
historical data and modeling. Suggestions for reference conditions have recently been published for
various Danish coastal areas. Reference TN-concentrations of 43 µM have been suggested for the
inner parts and 11-19 µM for outer parts of Randers Fjord based on dynamic modeling (Nielsen et
al. 2003), concentrations of 50-58 µM have been suggested for inner parts of Roskilde Fjord based
on paleoecological reconstruction (Clarke et al. 2003, Andersen et al. 2004), and concentrations
of 23-30 µM have been suggested for Limfjorden (Markager et al. in prep. Christensen et al. xx).
For Øresund, concentrations in the range 5-19 µM have been suggested based on various sources
(Øresundssamarbejdet 2004: Table 1 and references therein).
Reference values for TN and levels of water transparency have been defined for Swedish coastal
waters (Hansson och Håkansson 2006). However, too few data exist from the same area as we have
no depth distribution data for the vegetation. At present no reference conditions for TN and water
transparency exist for Norway within the different areas.
Reference levels can also be defined through modeling. In order to define reference levels of
benthic vegetation metrics through modeling there is a need for dose-response relationships
between physico-chemical variables (nutrient concentration, transparency) and vegetation metrics.
Then reference levels of nutrient concentration and/or transparency are entered in the model and
corresponding levels of the vegetation metrics are calculated.
It is also possible to run hind-cast models based on historical as well as today’s data for nutrient
loads and transparency and adjusting the nutrient input into the model according to the time when
one expected reference conditions. The model can then estimate the new transparency or light
penetration depths. These can then be related to e.g. depth extension of macroalgae: Today’s lower
depth extensions of selected macroalgal species as well as today’s percentage light penetration
at these depths can be obtained from either measurements or the literature. Based on the species
minimum requirement for light, new maximum depth extensions can then be calculated for a
reference situation. However, at present the reference values for benthic vegetation metrics have
been set by use of historical values as well as expert judgement.
Metrics
Metric 1: Depth limit of macroalgal species
This metric describes the depth extension of selected perennial macroalgal species in Scandinavian
coastal waters. The species selected for each type are all 1) perennial, 2) commonly occurring
in the type and 3) easy to determine. The metric is affected by nutrient concentration and water
transparency (Kautsky et al. 2004). Depth limits of macroalgae as a group (Nielsen et al. 2002b)
and depth limits of selected macroalgal species (e.g. Kautsky et al. 1986, Kautsky et al. 2004, 2007)
have been shown to respond to changes in nutrient concentration and water clarity.
Salinity also affects competition among species and thereby the occurrence and depth distribution
of the individual species (e.g. Nielsen et al. 1995. Pedersen and Snoeijs 2001, Torn et al. 2006).
In areas containing large salinity gradients, relationships between depth limits and nutrient
concentration/water clarity should therefore be developed for specific salinity regimes.
201
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
The algae also demand hard substratum for attachment, so the metric is only applicable in areas
where hard substratum occurs to the maximum water depth where light allows growth. Moreover,
the metric should only be applied in relatively deep areas where water depth in itself does not limit
depth penetration of the algae.
We have selected the following macroalgal species for Norwegian and Swedish coastal waters
of Skagerrak– Kattegat: The red algae
Chondrus crispus, Furcellaria lumbricalis, Coccotylus
truncatus/ Phyllophora pseudoceranoides,
Delesseria
sanguinea, Phycodrys rubens
and
Rhodomela
confervoides
and the brown algae:
Halidrys siliquosa
and
Saccharina latissima.
In Danish coastal
waters we have focused on the same species, except for
Delesseria sanguinea
and
Phycodrys
rubens.
Metric 2: Cover of the macroalgal community
This metric describes the cover of the macroalgal community along depth gradients. Algal cover
typically peaks at intermediate water depths while cover is lower in shallow and deep water.
Physical exposure is likely to limit algal cover in shallow water, whereas light limitation is likely to
be responsible for the reduction in algal cover from intermediate depths towards deeper water. With
the purpose of increasing the sensitivity of the metric to changes in water quality we only included
algal cover data from the ‘light regulated part’ of the depth gradient, and excluded data from shallow
water in the analyses of dose-response relationships. We therefore expect the metric to be primarily
regulated by nutrient concentration and water transparency though substratum and salinity also may
play regulating roles (Carstensen et al.2005). The cover of the algal community can be expressed in
two ways:
1) ‘Total cover’: overall cover of the community representing values up to 100 %, or
2) ‘Cumulated algal cover’: summed cover of all individual species, representing values up to
several 100 % as the algae may grow in several stories.
Note: This metric is still under development and will be intercalibrated in Phase II.
Metric3: Composite macroalgal index
This metric describes the macroalgal composition in rocky littoral and sublittoral zones. The metric
includes species richness (or total number of species recorded), proportions of red, green and
opportunist, ratio of ephemeral against perennial algae, ratio between Ecological Status Groups I
and II (ESG ratio), lower depth extension of a few selected species and their abundance as well
as the physical properties of the location. It is based on many elements of the English metric but
includes also sublittoral communities, which are more sensitive to impacts than littoral communities
and abundance of a few key species. The littoral zone is usually inhabited by species with a rather
high resilience toward all kinds of impact. It has been shown that along the Norwegian coast there
has been a change in sublittoral communities within the last 5 years, which has yet to be detected
in the littoral zone (Moy & Christie, in press). This emphasizes the importance of including a
sublittoral element in the metrics.
The metric is based on a combination of the United Kingdom metric for macroalgae (in NEA1 and
NEA26) and the Scandinavian depth limit metric (in NEA8, 9, and 10) and some aspects of the
Danish percentage algal cover. This index is still under development.
Note: As still under development intercalibration will be finalized in Phase II.
202
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
Boundary Setting
Metrics 1: ‘Depth limit of macroalgal species’
Regarding the metric ‘depth limit of macroalgal species’, the dose-response relationships developed
so far seem to show no discontinuities/thresholds, which can be used to define e.g. the good/
moderate boundary. The only distinct theoretical threshold is at a very deteriorated stage (e.g. poor/
bad boundary) when the benthic vegetation has disappeared and no depth limit exists. No distinct
boundaries have been identified based on paired metric assessment. So, for this indicator we should
follow the above example approach for defining boundaries. However, in general the dose-response
relationships on present-time data do not include a range of nutrient concentrations representing
reference levels or high/good boundaries. Reference levels for the benthic vegetation indicators
therefore have to be set using historical data, hind-cast modeling and/or expert judgement as
discussed in Step 1. As reference levels are likely to vary between areas, we suggest that the levels
are defined for each area and that class boundary limits are defined as specified deviations from
reference levels.
Depth limits of macroalgal species
Swedish historical data on depth distribution of individual macroalgal species are available for only
one waterbody i.e. NEA 9. This data set is from the inner Gullmar Fjorden, in 1941, at the Swedish
Skagerrak coast (Eriksson et al. 2002 and references therein). However, Norwegian historical data
for lower depth limits were available for the two other water types e.g. NEA 8 and 10. The Swedish
dataset is limited both in aerial cover and time but can be used to estimate the maximum historical
depth distribution of some commonly occurring, easily recognizable perennial macroalgal species
and thereby to define approximate reference depth limits for the type area represented by the data.
The historical Norwegian datasets are, however, mainly based on dredging and caution must be
exercised. Hence, the historical data were used in combination with modeling (based on Secchi
depths), present depth occurrence of the selected species in areas with low nutrient enrichment,
and expert judgement to estimate reference depth limits for other water body types. The result is
presented in the table 2.2.12, 2.2.13 and 2.2.14. The High/Good boundary for depth limits of nine
selected macroalgal species in each salinity regime is reported within the Swedish coastal water
types. Information can be found at www.naturvardsverket.se/sv/Lagar-och-andra-styrmedel/Lag-
och-ratt/Foreskrifter-och-allmanna-rad/Foreskrifter-utgivningsordning/.
In Table 12 depth limits in NEA9 areas are represented by the only area for which historical data is
available, i.e. Gullmar Fjorden, one profile in 1941 and two recent Norwegian surveys. The high-
good boundary (= reference condition) for depth limits of nine selected macroalgal species has
been defined as a 17-33 % deviation from the estimated reference levels. All depth limits are set
by expert judgement for each of the nine selected macroalgal species since the few available data
do not allow any good statistical treatment. Similarly, good-moderate boundaries have been set by
expert judgement and represent a 42 to 50 % deviation from reference levels.
In table 2.2.13 the lower depth limits have been estimated based on Sundenes historical data
sets (Sundene 1953) and several recent data sets. As for NEA9 both the reference values and the
boundaries are based on expert judgements. The historical values from Sundene (1953) are based
on dredging and are therefore not as reliable as recent surveys carried out by use of diving. Hence,
the new estimated reference values are estimated somewhere between Sundenes findings and recent
observations. For some species the recent findings of lower depth limits exceeds those found by
Sundene. In these cases the recent finding has been used in estimating a new reference value. The
High/Good boundaries vary between 18 and 37 % of the estimated reference value and the Good/
Moderate boundaries vary between 44 and 58 % of the estimated reference values (Table 2.2.13).
203
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0206.png
Table 2.2.12:
Estimated depth limits for NEA 9 – Skagerrak fjords – in Norwegian and Swedish
waters. Historical data by M. Waern 1941 in Eriksson et al. 2002, recent data from Walday et al.
2001ref missing is from Grenlandsfjord 1998-1999, Magnusson et al. 1992 is from Oslofjord
1990.
.2
00
1
20
02
al
.1
99
2
et
al
nu
ss
on
NEA 9
'Skagerrak fjords
Species
al
da
N
EW
ss
on
/G
Survey nr:
1
2
CHOCR
FURLU
HALSI
LAMSA
PHYP+COCTR
RHOCO
DELSA
PHYRU
5
3
7
17
12
17
14
12
13
10
11
9
3
Er
ik
W
M
12.5
15
12.5
10
11
15
16.6
16
12
15
12
12
14
15
17
16
>10
>12
>10
>8
>10
>12
>13
>13
Table 2.2.13. Estimated depth limits for NEA8- Skagerrak inner archipelago - in Norwegian and Swedish waters.
(Walday et al. 2001 is from Grenlandsfjord 1998-1999, Magunsson et al. 1992 is from Oslofjord 1990, Pedersen et
al., is from Chysochromulina surveys in 1988 and 1989, Moy et al. 2005 is from the Norwegian Coastal Monitoring
Program from 1990-2005, Karlsson 1994 -1998 is from the Swedish coastal monitoring program from 1994-1998 and
Sundene 1953 is from surveys in outer Oslofjord in 1947 to 1952. New reference is based on expert judgement and all
these reports mentioned. * uncertainty of viability as dredging was used in collection the seaweeds.)
Within NEA10 there are several extensive monitoring programs both in Norway and in Sweden
that have made it possible to determine a lower depth limits for the same species as in NEA8 and
9. Fredriksen and Rueness (1990) compared the data sets collected at the same stations examined
by Sundene in 1947-1952, and used both dredging and diving. The conclusions were that the lower
depth limits have been reduced since Sundenes explorations. Both these reports have been compared
with recent registrations and new reference levels have been estimated for all species. Within the
Coastal Monitoring Program (1990-) some scattered individuals are found even deeper than the new
reference values, however, these lower depth limits represent individuals that may not survive the
over wintering, hence not a reference level for healthy and reproductive species. The High/Good
boundaries vary between 24 and 29 % of estimated reference level and the boundary level of Good/
Moderate between 40 and 50 % of reference levels (Table 2.2.14)
204
G
oo
d
H
ig
h
ag
/M
od
e
al
et
re
f.
y
oo
et
d
7
8
7
6
8
8
9
8
ra
te
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0207.png
Table 2.2.14:
Estimated depth limits for NEA10- Skagerrak exposed - in Norwegian and Swedish waters.
(Moy et al. 2005 is from the Norwegian Coastal Monitoring Program from 1990-2005, Fredriksen &
Rueness 1990 is from surveys on Sundenes (1953) location in 1989. Karlsson 1994 is from the Swedish
coastal monitoring program from 1994-1998 and Sundene 1953 is from surveys in outer Oslofjord in 1947
to 1952. New reference is based on expert judgement and all these reports mentioned. * uncertainty of
viability as dredging was used in collection the seaweeds).
.2
00
Fr
5
ed
rik
se
n
&
Su
Ru
nd
en
en
es
e1
s,
K
95
19
ar
3
90
lss
*
on
19
94
N
EW
-9
8
re
f.
/G
oo
d
NEA10
exposed
Species
Survey nr:
M
4
8
CHOCR
FURLU
HALSI
LAMSA
PHYP+COCTR
RHOCO
DELSA
PHYRU
9
18
13
14
20
30
24
30
29
14
13
18
15
27
18
20
8
16
35
12
32
27
7
16
14
10
14
16
16
18
16
14
16
30
16
30
29
>13
>12
>10
>12
>22
>12
>22
>22
2.2.4 Results of the comparison
As option 2 has been used the boundaries have been agreed by experts representing all countries in
the GIG macroalgae sub-group. Therefore the final agreed results presented below are the results of
comparing expert views on what the boundaries should be.
Separate MS data were used (option 3) from Portuguese and Spanish rocky shores representing
pressure gradients. Portuguese shores were surveyed on the north coast of Portugal, sites under
pressure from the Douro river, and surveyed after Mieres meeting. Data from Spain were collected
also from rocky shores under different gradient pressures at Liñera, Usgo and Ontón (for details see
earlier reports). To set boundaries, Portugal and Spain used national data from their own pressure
sites, and afterwards both assessed data from the other MS to evaluate the output correspondence of
different assessment methods. Although not very extent, data from both MSs were used to compare
assessment methods and to best fit boundaries to improve EQS classifications.
The Portuguese Marine Macroalgae Assessment Tool (P-MarMAT) operates over a scale range from
0 (impacted) to 1 (non-impacted). Initially, class boundaries were set as equidistant points along the
scale (0.2, 0.4, 0.6 and 0.8).
The CFR index operates over a scale range from 0 (impacted) to 100 (non-impacted), but can
be easily converted to a 0-1 scale by calculating the CFR/100 value. Class boundaries are set as
suggested in REFCOND (2003) 0.2-0.41-0.62-0.83.
These boundaries were then modified to ensure a commitment between the status characteristics
as defined in the Normative Definitions and agreeing as much as possible to other MSs thresholds.
The process of setting and testing the ecological status boundaries was already described in earlier
reports. Portugal and Spain assessed their own data, data from the other MS, and all together from
the common dataset, in order to produce a classification result comparable between each other
205
G
oo
d
H
ig
h
oy
Skagerrak
et
/M
al
9
9
8
9
18
9
18
17
od
e
ra
t
e
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0208.png
and compliant with the expected to study sites. Figure 2.2.1 illustrates the correlations existent
between the compared methods, derived from the process of setting and testing the ecological status
boundaries.
Portugal and Spain assessed data from both MSs in order to produce a classification result
comparable between each other. From this exercise, a weighted Kappa value of 0.84 was reached,
meaning a “very good” correlation between results (Monserud & Leemans, 1992). Table 2.2.15
shows the assessment results from this exercise.
After boundaries optimization, methods showed “excellent” agreement between each other, with a
weighted kappa value of 0.89. Table 2.2.16 shows the assessment results from this second part of
the IC exercise.
Spain
y = 1,5453x - 0,3378
R2 = 0,8852
1
0,9
0,8
0,7
0,6
0,5
0,4
0,3
0,2
0,1
0
0,00
CFR/100
0,10
0,20
0,30
0,40
0,50
0,60
0,70
0,80
0,90
P-MarMAT
Portugal
y = 1,254x - 0,32
R2 = 0,9615
1
0,9
0,8
0,7
CFR/100
0,6
0,5
0,4
0,3
0,2
0,1
0
0,00
0,10
0,20
0,30
0,40
0,50
P-MarMAT
Total
0,60
0,70
0,80
0,90
1,00
1
0,9
0,8
0,7
CFR/100
0,6
0,5
0,4
0,3
0,2
0,1
0
0,00 0,10
0,20
0,30
0,40
y = 1,3472x - 0,2914
2
R = 0,8547
0,50
0,60
0,70
0,80
0,90
1,00
P-MarMAT
Figure 2.2.1:
Correlations between
EQR resulted from SP and PT
methods.
206
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0209.png
Table 2.2.15:
P-MarMAT and CFR assessment results
from data collected in intertidal rocky shores in Portugal
and Spain.
Table 2.2.16:
P-MarMAT and CFR assessment results
from data collected in intertidal rocky shores in Portugal
and Spain, after boundaries optimization.
Although the small number of data available and the possibility of further developments in near
future, after running the IC exercise and the procedure of setting boundaries the best threshold
values for Portugal and Spain are shown in the following table:
From this exercise, it was possible to improve agreement between SP and PT classification methods,
resulting a value of 0.89 (“excellent”) from Kappa analysis.
Although future improvements and updates surely arise, the class boundaries achieved at the
moment to Portugal and Spain are shown in the following table, 2.2.17:
Table 2.2.17:
Class boundaries.
P-MarMAT
H/G
G/M
M/P
P/B
0,816
0,638
0,460
0,282
CFR
0,808
0,568
0,329
0,089
207
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0210.png
2.2.5 Results of the harmonisation – Boundary thresholds and EQR values
Metric - Rocky Shore Macroalgae
CW - NEA1/26a,b,e
UK, Republic of Ireland and Norway
At this time, June 2007, the scoring system for the common indices below have been agreed by
Ireland, Norway, and the UK. These common indices for this intercalibration metric describe species
richness and composition based on a reduced species list devised specifically for the boundary
setting protocol and classification tool development process. There are slightly different between
regions due to the varying levels of diversity and composition as shown in table 2.2.18 below.
Table 2.2.18:
EQRs.
EQR
Quality Class
RSL
0.8 - 1.0
High
35-70
35-69
34-68
33-68
0-12
0-15
0-20
0-20
0.6 - 0.8
Good
25-35
25-35
20-34
20-33
12-20
15-20
20-30
20-30
45-55
45-55
35-45
30-40
0.4 - 0.6
Moderate
17-25
15-25
10-20
10-20
20-30
20-25
30-45
30-45
35-45
40-45
25-35
22-30
0.2 - 0.4
Poor
5-17
5-15
3-10
4-10
0 - 0.2
Bad
0-5
0-5
0-3
0-4
Greens
Scotland
England/Wales/RoI
NI
Norway
Scotland
England/Wales/RoI
NI
Norway
Scotland
England/Wales/RoI
NI
Norway
Scotland
England/Wales/RoI
NI
Norway
Reds
ESG
55-100
55-100
45-100
40-100
1.0-1.2
1.0-1.2
0.8-1.2
0.8-1.2
0-10
0-10
0-15
0-15
30-80
25-80
45-80
45-80
15-35
15-40
10-25
10-22
80-100
80-100
80-100
80-100
0-15
0-15
0-10
0-10
Opportunist
Scotland
England/Wales/RoI
NI
Norway
0.8-1.0
0.8-1.0
0.6-0.8
0.6-0.8
10-15
10-15
15-25
15-25
0.7-0.8
0.55-0.8
0.4-0.6
0.4-0.6
15-25
15-25
25-35
25-35
0.2-0.7
0.2-0.55
0.2-0.4
0.2-0.4
25-50
25-50
35-50
35-50
50-100
50-100
50-100
50-100
0-0.2
0-0.2
0-0.2
0-0.2
A ‘de-shoring factor’ has been incorporated to adjust the level of species richness according to the
overall description of the shore. This uses an exponential-type model of the form:
RICHNESS = a + b
exp (cSHORE)
where a, b and c are parameters to be estimated from the data. Using least squares, these parameters
were estimated to be:
a
= 14.210
b
= 4.925
c
= 0.108
The final metric system works on a sliding scale to enable an accurate EQR value to be calculated
for each of the different parameters, an average of these values is then used to establish the final
classification status.
208
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0211.png
Where a shore description is not available the uncorrected level of species richness is to be put into
the final metric, although the level of confidence in the overall result may be reduced slightly.
EQRs
The above scoring system for the indices for this metric gives EQR boundaries of:
H/G
G/M
M/P
P/B
0.80
0.60
0.40
0.20
The average EQR is used to classify this metric.
The overall scoring and EQR’s show above are currently under review.
Spain have developed preliminary boundary classifications for their CFR metric, which are still to
be fully tested before they are accepted. Details of which can be found in the detailed Milestone 6
Report.
Spain and Portugal:
As explained before, the class boundaries achieved at the moment to Portugal and Spain are shown
in the following table, 2.2.19:
Table 2.2.19:
ES/PT EQRs.
P-MarMAT
H/G
G/M
M/P
P/B
0,816
0,638
0,460
0,282
CFR
0,808
0,568
0,329
0,089
Metric - Opportunistic Macroalgae (NEA1/26a,b,c,e/11)
The Opportunistic Macroalgal bloom tool shown below has been agreed if only by the UK and
Ireland, table 2.2.20 below. Germany is also considering using this tool, full information on German
macroalgae intercalibration can be found in: German Macroalgae & Angiosperm Annex-IC-Report
Macroalgae and Angiosperms NEA GIG - Germany_05-2007.doc (Kolbe, 2007). Portugal is only
considering opportunistic macroalgae in transitional waters.
Other countries are considering if this tool could be used in their waters. Denmark will intercalibrate
at a later date, post June 2007.
The tool is a combination of percent cover by opportunistic green algae, the density of growth with
a defined area and the presence of entrained algae.
The final metric system works on a sliding scale to enable an accurate EQR value to be calculated
for each of the different parameters, an average of these values is then used to establish the final
classification status.
209
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0212.png
Table 2.2.20:
EQRs.
Quality Status
EQR
% Cover of AIH
Area (ha)
Biomass (g m
2
) of AIH
Biomass (g m
2
) of affected area
Presence of entrained algae
(% of Quadrats)
High
0.8 - 1.0
0-5
0 - 100
0 - 100
0 - 100
0-5
Good
0.6 - 0.8
5 - 15
100 - 500
100 - 500
100 - 500
5 - 20
Moderate
0.4 - 0.6
15 - 25
500 - 1000
500 - 1000
500 - 1000
20 - 50
Poor
0.2 - 0.4
25 - 75
1000 - 2500
1000 - 3000
1000 - 3000
50 - 75
Bad
0.0 - 0.2
75 - 100
>2500 (- 6000)
>3000 (- 6000)
>3000 (- 6000)
75 - 100
The above scoring system for the indices for this metric gives EQR boundaries of:
H/G
G/M
M/P
P/B
0.80
0.60
0.40
0.20
Portugal has a long time series of data for one particular estuary and tested the metrics of total area,
percentage and biomass in different combinations. The results are very comparable and comply with
the expert judgement of Portuguese scientists
Currently Germany only has data for the areal cover of opportunistic macroalgal growth on soft
sediments in the intertidal. A recent report has tried to compare this metric to part of the UK/RoI
tool (see below). (Classification tools for biomass and for taxonomic composition might be designed
after further investigations in the field).
A recent German report considered that “UK intercalibration tool on the whole cannot be applied
to the German dataset. Moreover model calculations already show that the class boundaries of the
UK proposal do not reflect the heavy disturbances in algal growth that have been reported from the
German water bodies in the 1990s (ADOLPH 2007).” Therefore, Germany are still in the process of
intercalibrating their data, the progress of which is detailed below.
Currently Germany have described their references conditions as:
Opportunistic macroalgae blooms of anthropogenic origin should be absent or if present should
cover less than 1 % of the total intertidal of the water body.
Coverages with densities ≥1 % rarely ever exceed 15 % of the intertidal; normalised to 100 %
density they stay below 5 % in most of the cases. It is very likely, that differences in morphological
and hydrodynamic conditions between the British and the German water bodies are the underlying
reason for these mismatches.
The class boundaries proposed by Germany are based on the 6 year median of the annual
maximum, defined as the acreage of algal cover (monitored in km²; area with green algal density
≥1 %) as a percent of the total intertidal in a water body at the time of maximal extent. This was
monitored along the entire coast in the respective year. Currently the class boundaries are calculated
separately for each waterbody (based on historical reference values and recent information from
problem areas). The EQRs that match the percentage cover data are calculated by means of linear
interpolation within the respective equidistant EQR class ranges (KOLBE 2007). Germany’s EQRs
were not downgraded if total acreage exceeded a certain total patch size. See below table 2.2.21.
210
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0213.png
Table 2.2.21:
DE EQRs.
Metric
Percentage intertidal
area overgrown with
opportunistic macroalgae
(green algae) (density ≥1 %)
High
0-1%
EQR
1 – 0.8
Good
1.1 - 1.5 %
EQR
0.799 – 0.6
Moderate
1.6 - 5 %
EQR
0.599 – 0.4
Poor
5.1 -20 %
EQR
0.399 – 0.2
Bad
> 20 %
EQR
0.199 – 0
In an attempt to compare with parts of the UK/RoI method, a “normalization” process to 100 %
density, and a downgrading option have been applied to German data. After normalization these data
were classified using the German class boundaries. The resulting EQRs were downgraded according
to the UK-method if the total acreage exceeded a certain total patch size.
The results of this analysis were compared with the results of the German method. It was shown that
the use of the UK density basis but DE-class boundaries can, if downgraded, may lead to the same
outcome as the DE-tool. This only holds true for 6 (NEA 26 and 4) of the 13 German water bodies
tested (46 %).
For 7 (NEA 1, 3, 11) of the 13 water bodies (54 %) the classification lead to a status rank which is
still one or two classes higher than the De-tool and does not reflect the eutrophication status in those
water bodies (KOLBE 2007).
Future information and work will assist in completing the intercalibration. The intercalibration
therefore has been completed only for the sub-types 1/26a, b, e.
Metric - Subtidal Algae (NEA8/9/10)
Metric 1: ‘Depth limit of macroalgal species’
After the intercalibration procedure where new values for references conditions, and the depth limit
between High/Good and Good/Moderate ecological status for the selected number of macroalgal
species was agreed upon the following EQR –calculations for area NEA 8, 9 and 10 and the
following scoring system was used.
For the depth limit of the selected set of easily determined species the following expert EQR-scale
was decided on the areas intercalibrated as well as nationally:
1.0 - 0.81 = High status
0.80 - 0.61 = Good status
0.60 - 0.41 = Moderate status
To calculate and intercalibrate the ecological status, i.e. the EQR values in NEA 8, 9 and 10 we have
agreed to use the intercalibrated depth limits given in tables 12, 13 and 14 respectively.
An EQR-value is calculated by transforming the depth value (in meter) to a classification value
(score) for each of the included species, ( i.e. 5, 4 or 3. where 5 corresponds to reference/high status
, 4 to a depth distribution above Good status and 3 to a depth distribution in the studied transect
above Moderate status. After giving each of the occurring species the appropriate score the mean
value is calculated and divided by 5, i.e. the reference condition. For example, if 3 species score
5, 2 species score 4 and 2 species score 3; the total is 32 giving an average score of 4; which when
divided by 5 gives an EQR of 0.8; thus equating to Good Status.
211
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0214.png
Table 2.2.2:
Depth limits (m) for a set of selected macroalgal species used in the intercalibration and Calculation of
EQR for NEA 8, 9 and 10.
NEA 8
Species
Furcellaria lumbricalis
Phyllophora pseudoceranoides
Rhodomela confervoides
Chondrus crispus
Delesseria sanguinea
Halidrys siliquosa
Saccharina latissima
Phycodrys rubens
NEA 9
Furcellaria lumbricalis
Phyllophora pseudoceranoides
Rhodomela confervoides
Chondrus crispus
Delesseria sanguinea
Halidrys siliquosa
Saccharina latissima
Phycodrys rubens
NEA 10
Furcellaria lumbricalis
Phyllophora pseudoceranoides
Rhodomela confervoides
Chondrus crispus
Delesseria sanguinea
Halidrys siliquosa
Saccharina latissima
Phycodrys rubens
16
30
16
18
30
14
16
29
12
22
12
13
22
10
12
22
9
18
9
9
18
8
9
17
5
9
5
5
9
4
5
9
15
14
15
12
17
12
12
16
12
10
12
10
13
10
8
13
8
8
8
7
9
7
6
8
4
4
4
4
5
4
3
4
Reference
value
16
22
16
12
25
10
16
22
Score 5
if >than
10
18
12
8
18
8
10
15
Score 4
if >than
7
12
7
5
12
5
7
10
Score 3
if >than
4
6
4
3
6
3
4
5
An EQR-value is calculated by transforming the depth value (in meter) to a classification value
(score) for each of the included species, ( i.e. 5, 4 or 3. where 5 corresponds to reference/high status,
4 to a depth distribution above Good status and 3 to a depth distribution in the studied transect
above Moderate status. After giving each of the occurring species the appropriate score the mean
value is calculated and divided by 5, i.e. the reference condition. For example, if 3 species score
5, 2 species score 4 and 2 species score 3; the total is 32 giving an average score of 4; which when
divided by 5 gives an EQR of 0.8; thus equating to Good Status.
212
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
2.2.6 Open issues and need for further work
Gaps for the future:
Subtidal macroalgae tools are to be developed
Spain will attempt to separate macroalgae and seagrass from their macrophyte tool, for
intercalibration purposes
Questions of how to combine different sub-metrics into a higher level overall EQR.
Development of macroalgae tool for transitional waters (or use of the existing Spanish tool)
and intercalibration
Completion of the intercalibration process for subtype 1/26c.
3 Discussion
3.1 Comparability between GIGs
Metrics
Intercalibration for the macroalgae quality element has been successfully performed by all of
the four Coastal Waters GIGs, using indices (MEDGIG) or a selection of unaggregated metrics
(NEAGIG).
Types
Almost all types for each GIG have been covered.
Option
Option 2 (MS agreed on the common metrics, created data sets relating MSs assessment
methods to the common metrics, agreed on High/Good and Good/Moderate class boundaries and
established relationships between common and national metrics) approach has been adopted for the
intercalibration.
Boundary Setting Procedure
All Coastal Waters GIGs are providing results for the H/G and G/M boundaries (at least), applying
similar approaches based on the expert supervised selection of appropriate percentiles as boundaries
best fitting with evident changes in ecological status (90
th
percentile is the favorite choice).
Reference Condition
There’s no shared view among GIGs (but also among MSs) on how to fix Reference Conditions: in
some cases real reference sites have been identified and selected as a reference, in some others the
maximum possible values for the selected indices/metrics have been used.
3.2 Open issues and need for further work
A shared task for the next intercalibration round is to refine typology (or include types non
considered at this stage) in order to maximize the representativeness towards different waterbodies
and ecological status of the Quality Element.
213
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
4 References
Adolph, W., 2007. Classification of opportunistic Macroalgae blooming in Germany compared to the UK-
Intercalibration proposal. Report for Intercalibration, Mieres, 2007.
Andersen JH, Conley DJ, Hedal S. 2004. Paleoecology, reference conditions and classification of ecological
status: the EU Water Framework Directive in practice. Marine Pollution Bulletin 49: 283-290.
Arévalo R., Pinedo S., Ballesteros E. 2007. Changes in the composition and structure of Mediterranean rocky-
shore communities following a gradient of nutrient enrichment: descriptive study and test of proposed methods
to assess water quality regarding macroalgae. Marine Pollution Bulletin 55: 104-113.
Ballesteros E. 1988. Estructura y dinámica de la comunidad de
Cystoseira mediterranea
Sauvageau en el
Mediterráneo noroccidental.Investigación Pesquera 52: 313-334.
Ballesteros E. 1992. Els vegetals i la zonació litoral: espècies, comunitats i factors que influeixen en la seva
distribució. Arxius Secció Ciències I.E.C. 101, 1-616.
Ballesteros E., Pérez M., Zabala M. 1984. Aproximación al conocimiento de las comunidades algales de la
zona infralitoral superior en la costa catalana. Collect. Bot. 15: 69-100.
Bellan G., Bellan-Santini D. 1972. Influence de la pollution sur les peuplements marins de la région de
Marseille. In: Ruivo M. (Ed.) Marine Pollution and Sea Life, Fishing News Ltd. Survey, 396-401.
Bellan-Santini D. 1965. Étude quantitative du peuplement à
Mytilus galloprovincialis
Lamarck en eau
moyennement polluée. Rapports Commission International de la Mer Méditerranée 18: 85-89.
Bellan-Santini D. 1966. Influence des eaux polluées sur la faune et la flora marines benthiques dans la région
Marsellaise. Techniques et Sciences Municipales 61: 285-292
Bellan-Santini D. 1968. Influence de la pollution sur les peuplements benthiques. Rev. Intern. Oceanogr. Med.
10: 27-53.
Belsher T. 1977. Analyse des répercussions de pollutions urbaines sur le macrophytobenthos de Méditerranée
(Marseille, Port-Vendres, Port-Cros). Thèse Doctorat. Université d’Aix-Marseille II, Marseille, pp. 287.
Boudouresque C.F. 1969. Étude qualitative et quantitative d’un peuplement algal à
Cystoseira mediterranea
dans la région de Banyuls sur Mer. Vie Milieu 20, 437-452.
Boudouresque C.F. 1971. Contribution à l’étude phytosociologique des peuplements algaux des côtes varoises.
Vegetatio 22: 83-184.
Carstensen J, Krause-Jensen D, Dahl, K, Middelboe AL. 2005. Development of macroalgal indicators of
water quality
In
Dahl, K., Andersen, J., Krause-Jensen, D., Josefson, A., Kjerulf-Petersen, J., Hansen, O.S.,
Middelboe A.L. (Eds) Scientific and technical background for intercalibration of Danish coastal waters.
National Environmental Research Institute, Denmark xx pp. – NERI Technical report no. 563. http://technical-
reports.dmu.dk
Clarke A, Juggins S, Conley DJ. 2003. A 150-year reconstruction of the history of coastal eutrophication in
Roskilde Fjord, Denmark. Marine Pollution Bulletin 46: 1615-1629.
Coppejans, E., 1980. Phytosociological studies on mediterranean algal vegetation: Rocky surfaces of the
photophilic infralittoral zone. In:Price, J.H., Irvine, D.E.G., Franham, W.E. (Eds.), The shore environment,
Ecosystems, Vol. 2. Academic, London, pp. 371-393.
Cottalorda JM., Meinesz A., Thibaut T., Chiaverini D. 2004. Représentation cartographique de l’abondance de
quelques algues et invertébrés sur le littoral des îlots du Rascas et de la Gabinière (Parc National de Port Cros,
Var, France). Sci. Rep. Port-Cros Natl. Park, Fr. 20: 195-209.
Edwards, P., 1975.
An assessment of possible pollution effects over a century on the benthic marine algae of
Co. Durham, England.
Bot. J. Linn. Soc. 70, 269-305.
Eriksson KB, Johansson G, Snoeijs P. 2002. Long–term changes in the macroalgal vegetation of the inner
Gullmar fjord, Swedish Skagerrak coast. Journal of Phycol. 38: 284-296.
Feldmann J. 1937. Recherches sur la ve´ge´tation marine de la Me´diterrane´e. La coˆ te des Albères. Revue
Algologique 10: 73-254
214
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
Giaccone G. 1993. The vertical zonation along the phytal system in the Mediterranean sea and the effects of
municipal and industrial wastewater disposal on phytobenthos communities. Fifth Optima Meeting, Istanbul
5: 47-56.
Gibert A. 1918. Flora algològica marina de les aigües i costes occidentals de Catalunya. Publicacions de
l’Agrupacio´ Excursionista, pp. 65.
Golubic S. 1970. Effect of organic pollution on benthic communities. Marine Pollution Bulletin 1: 56-57.
Gros C. 1978. Le genre
Cystoseira
sur la côte des Albères. Répartition – Écologie–Morphogenèse. Thése
Doctorat. Université P. et M. Curie, Paris VI, pp. 115.
Guinda, X., Juanes, J.A., Puente, A., Revilla, J.A., 2008. Comparison of two methods for quality assessment of
macroalgae assemblages, along different pollution gradients. Ecological Indicators 8 (5), 743-753.
Hoffmann L., Clarisse S., Detienne X., Goffart A., Renard R., Demoulin V. 1988. Evolution of population of
Cystoseira balearica
(Phaeophyceae) and epiphytic Bangiophyceae in the Bay of Calvi (Corsica) in the last
eight years. Bulletin de la Socie´te´ Royale de Liège 4-5: 263-273.
J
aklin
, S., P
eterSen
, B., a
dolPh
, W., P
etri
, G. & W. h
eiBer
(2006). Aufbau einer Matrix für die Gewässertypen
nach EG-WRRL im Küstengebiet der Nordsee, Schwerpunkt Flussgebietseinheiten Weser und Elbe.
Abschlussbericht Teil 1: Nährstoffe, Fische, Phytoplankton, Makroalgen und Seegras. Berichte des NLWKN
2006.
Juanes, J.A., Guinda, X., Puente, A., Revilla, J.A., 2008. Macroalgae, a suitable indicator of the ecological
status of coastal rocky communities in the NE Atlantic. Ecological Indicators 8 (4), 351-359.).
Kautsky L, Andersson C, Dahlgren S. 2004. Framtagning av nya bedömningsgrunder för kust och hav
enligt ramdirektivets krav. Förslag till Ekologisk Kvalitetskvot (EQR) för den biologiska parametern –
makrovegteation. Report to the Swedish EPA. In Swedish. 23pp.
Kautsky N, Kautsky H, Kautsky U, Waern M. 1986. Decreased depth penetration of
Fucus vesiculosus
L. since
the 1940s indicate eutrophication of the Baltic Sea. Mar. Ecol. Prog. Ser. 28: 1-8.
Kautsky, L, Wibjörn, C, Kautsky, H. 2007. Bedömningsgrunder för kust och hav enligt krav i ramdirektivet
vatten – makroalger och några gömfröiga växter. Report to the Swedish EPA, in Swedish, 50 pp.
kolbe, k.
(2006). Bewertungssystem nach WRRL für Makroalgen und Seegräser der Küsten- und
Übergangsgewässer der FGE Weser und Küstengewässer der FGE Elbe. Studie im Auftrag des Niedersächsischen
Landesbetriebs für Wasser-, Küsten- und Naturschutz, Oldenburg/Brake, 99 S.
Kolbe, K., 2007. Assessment of German Coastal Waters (NEA 1/26, NEA 3/4) and Transitional Waters (NEA
11) by Macroalgae and Angiosperms. Intercalibration Report (NEA GIG).- (postal to D. Coates, D. Jowett).
Lipej L., Mozetič P., Orlando-Bonaca M., Mavrič B., Šiško M., Bettoso N. 2006. Evaluation of the Ecological
Status of Coastal Waters in accordance with the European Water Framework Directive (Water Framework
Directive, 2000/60/EC). Final national report in Slovenian, Marine Biology Station Piran, National Institute
of Biology, October 2006, 180 pp.
Littler M.M., Littler D.S. 1980. The evolution of thallus form and survival strategies in benthic marine
macroalgae: field and laboratory tests of a functional form model. American Naturalist 116: 25-44.
Littler M.M., Murray S.N. 1975. Impact of sewage on the distribution, abundance and community structure of
rocky intertidal macroorganisms. Marine Biology 30: 277-291
MacArthur R.H, Wilson E.O. 1967. The theory of island biogeography. Princeton University Press,
Princeton.
Magnusson J, Johnsen T.M, Beyer F, Gjøsæter J, Lømsland E.R, Sollie, Aa, 1999. or 1992 som i tabellen
vilket år skall det vara?? Pollution monitoring in the Inner Oslofjord in 1998. NIVA-report. OR-4058. 63pp. In
Norwegian (http://www.niva.no/)
Mangialajo, L., Ruggieri, N., Asnaghi, V., Chiantore, M., Povero, P., Cattaneo-Vietti, R., 2007. Ecological
status in the Ligurian Sea: The effect of coastline urbanisation and the importance of proper reference sites.
Marine Pollution Bulletin 55: 30-41.
Meinesz A., Javel F., Cottalorda J.M., Susini M.L., Capiomont A., Levi F., Robert A. 2004. Cartographie des
espèces médiolittorales et infralittorales supérieures des falaises de Bonifacio (Corse du Sud) – Mission 2003
215
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
– Rapport intermédiaire. Contrat Office de l’Environnement de la Corse & Laboratoire Environnement Marin
Littoral. Laboratoire Environnement Marin Littoral publ., Nice: 1-14.
Meinsez A., Javel F., Chiavérini D., Cottalorda J.M., Giletta L. 2004. Représentation cartographique de
l’abondance de quelques algues et invertébrés du littoral de la facez ouest du Parc National de Port-Cros.
Laboratoire Environnement Marin Littoral publ., Nice: 1-26.
Monserud, R., Leemans, R., 1992. Comparing global vegetation maps with the Kappa statistic. Ecological
modelling 62, 275-293.
Murray S.N., Littler M.M. 1978. Patterns of algal succession in a perturbated marine intertidal community.
Journal of Phycology 14: 506-512.
NFS: Information can be found at www.naturvardsverket.se/sv/Lagar-och-andra-styrmedel/Lag-och-ratt/
Foreskrifter-och-allmanna-rad/Foreskrifter-utgivningsordning/.
Nielsen R, Kristiansen A, Mathiesen L, Mathiesen H. 1995. Distributional index of the benthic macroalgae of
the Baltic Sea area. Acta Bot Fenn 155: 1-51.
Nielsen SL, Sand-Jensen K, Borum J, Geertz-Hansen O. 2002b. Depth colonization of eelgrass (Zostera
marina)
and macroalgae as determined by water transparency in Danish coastal waters. Estuaries 25: 1025-
1032.
Nielsen SL, Sand-Jensen K, Borum J, Geertz-Hansen,O. 2002a. Phytoplankton, nutrients and transparency in
Danish coastal waters. Estuaries 25: 930-937.
NIENBURG, W. (1927): Zur Ökologie der Flora des Wattenmeeres I. Teil. Der Königshafen bei List auf Sylt.
- Wiss. Meeresunters. (Abteilung Kiel) Neue Folge, 20(2): 145-196.
Opportunistic Green Macro algae References (Types NEA 1/26a,b,c, NEA11, NEA3/4).
Orfanidis S. 2007. Comments on the development of new macroalgal indices to assess water quality within the
Mediterranean Sea. Marine Pollution Bulletin (in press).
Orfanidis S., Panayotidis P. 2005. Implementation of Water Framework Directive (WFD) for coastal waters
by using the Ecological Evaluation Index-EEI: the case of Kavala’s and Maliakos Gulfs, Greece. Proceedings
12th Panhellenic Congress Ichthyologists, Drama, Greece, pp. 237-240.
Orfanidis S., Panayotidis P., Stamatis N. 2001. Ecological evaluation of transitional and coastal waters: a
marine benthic macrophytes-based model. Mediterranean Marine Science 2: 45-65.
Orfanidis S., Panayotidis P., Stamatis N. 2003. An insight to the ecological evaluation index (EEI). Ecological
Indicators 3: 27-33.
Orfanidis S., Stamatis N., Tsiagga E. 2005. Ecological status assessment of Delta Nestos Lagoons by using
biological and chemical indicators in agreement to Water Framework Directive (WFD 2000/60). Proceedings
12th Panhellenic Congress of Ichtyologists: 245-248 (In Greek with English summary).
Orfanidis, S., Panayotidis, P. and Stamatis, N., 2001.
Ecological evaluation of transitional and coastal waters:
A marine benthic macrophytes-based model.
Mediterranean Marine Science. 2/2, 45-65.
Orlando-Bonaca M., Lipej L., Orfanidis S. 2008. Benthic macrophytes as a tool for delineating, monitoring
and assessing ecological status: the case of Slovenian coastal waters. Marine Pollution Bulletin 56 (4): 666-
676.
Panayotidis P., Montesanto B., Orfanidis S. 2004. Use of low-budget monitoring of macroalgae to implement
the European Water Framework Directive. Journal of Applied Phycology 16: 49-59.
Panayotidis P., Orfanidis S., Tsiamis K. 2007.
Cystoseira crinita
community in the Aegean Sea. Rapp. Comm.
int. Mer Médit., 38 (in press).
Panayotidis P., Siakavara A., Orfanidis S., Haritonidis S. 2001. Identification and description of habitat types at
sites of interest for conservation. Study 5: Marine habitats. Final Technical Report, Athens October 2001.
Pedersen M, Snoeijs P. 2001. Patterns of macroalgal diversity, community composition and long-term changes
along the Swedish west coast. Hydrobiol. 459: 83-102.
Polo L. 1978. Estudio sobre las algas bentónicas de la costa catalana. Tesis Doctoral, Universidad de Barcelona,
pp. 222
216
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
REFCOND, 2003. Guidance on establishing reference conditions and ecological status class boundaries for
inland surface waters. Working group 2.31 Reference conditions for inland surface (REFCOND), Common
Implementation Strategy of the Water Framework Directive, European Commission.
Rodríguez-Prieto C., Polo L. 1996. Effects of sewage pollution in the structure and dynamics of the community
of
Cystoseira mediterranea
(Fucales, Phaeophyceae). Scientia Marina 60: 253-263.
Sauvageau C. 1912. Á propos des
Cystoseira
de Banyuls et de Guéthary. Bulletin de la Station Biologique
d’Arcachon 14: 133-556.
Scanlan, C.M., Foden, J., Wells, E. and Best, M.A., 2007. The monitoring of opportunistic macroalgal blooms
for the water framework directive.
Marine Pollution Bulletin
55: (1-6): 162-171.
Seoane-Camba J.A. 1975. Algas bentónicas españolas en los herbarios Thuret-Bornet y Sauvageau del Museum
National d’Histoire Naturelle de Paris. II. Anales Instituto Botánico Cavanilles 32: 33-51.
Soltan D., Verlaque M., Boudouresque C.F., Francour P., 2001. Changes in macroalgal communities in the
vicinity of the Mediterranean sewage outfall after the setting up of a treatment plant. Marine Pollution Bulletin
42: 59-70.
Sundene O, 1953. The algal vegetation of Oslofjord. Mat. – Naturv. Klasse. No. 2 244 pp.
Thibaut T., Hereu B., Susini M. L., Cottalorda J.M. 2005. Inventaire et cartographie des Fucales du Parc
National de Port-Cros. Contrat Parc National de Port-Cros et Laboratoire Environnement Marin Littoral.
Laboratoire Environnement Marin Littoral publ., Nice: 1-30.
Thibaut T., Pinedo S., Torras X., Ballesteros E. 2005. Long term decline of the populations of Fucales
(Cystoseira,
Sargassum)
in the Albères coast (northwestern Mediterranean). Marine Pollution Bulletin 50:
1472-1489.
Torn K, Krause-Jensen D, Martin G. 2006. Present and past depth distribution of bladderwrack (Fucus
vesiculosus)
in the Baltic Sea. Aquatic Botany 84: 53-62.
Wells, E. and Wilkinson, M., 2002b.
Intertidal seaweed biodiversity in relation to environmental factors –
a case study from Northern Ireland.
Marine Biodiversity in Ireland and Adjacent Waters, Ulster Museum,
Belfast.
Wells, E. and Wilkinson, M., 2003.
Intertidal seaweed biodiversity of Orkney.
Coastal Zone Topics. 5, 25-30
Wells, E., 2002a.
Seaweed Species Biodiversity on Rocky Intertidal Seashores in the British Isles.
Ph.D. Thesis,
Heriot-Watt University, Edinburgh.
Verlaque, M., Tine, J., 1979. Végétation marine de la rade de Toulon (Var, France): la Rade-abri. Revue de
Biologie et Ecologie Méditerrannée 5, 67-86.
Wilkinson, M. and Tittley, I., 1979.
The marine Algae of Elie: A reassessment.
Botanica Marina. 22, 249-256.
Wilkinson, M., Fuller, I.A., Telfer, T.C., Moore, C.G. and Kingston, P.F., 1988.
A Conservation Oriented Survey
of the Intertidal Seashore of Northern Ireland.
Institute of Offshore Engineering, Heriot-Watt University,
Edinburgh.
Øresundsvandsamarbejdet - Öresundsvattensamarbetet. 2004. Jämförelse mellan ‘urtida’ och ‘nutida’
näringsnivåer i Öresund – beräkningar utförda med MIKE3-modell. www.oresundsvand.dk
217
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0221.png
Section 5 – Angiosperms
1 Introduction
National classification tools for the angiosperms quality element have been intercalibrated in two
coastal GIGs. Intercalibration of angiosperm-specific assessment tools was possible to carry out
between Denmark and Germany in the Baltic Sea GIG, and between Germany, Ireland, Netherlands,
and UK in the North East Atlantic in the North East Atlantic GIG. For these GIGs a sub-set of
common types was selected where angiosperms were present. Information about the types and
countries with each type in each GIG is described in the sections below.
The intercalibration of the angiosperm quality element for the Mediterranean GIG was not finalised
during the first phase and will be continued in the future.
The Option 2 was chosen in this phase of the intercalibration process. Member States participating
in the intercalibration have developed metrics as part of their own classification schemes. Selected
metrics have been chosen for assessment and agreement of boundaries in this phase of the
intercalibration process. At this stage not all the metrics that make up Member States’ schemes can
be intercalibrated, so it is not possible to produce EQRs for the whole quality element. Therefore
boundaries have been agreed for the selected metrics.
2 Methodology and Results
2.1 Baltic GIG
2.1.1 Intercalibration approach
The Baltic Sea Geographical Intercalibration Group (GIG) includes the whole or parts of the
coastline of the following countries: Germany, Denmark, Estonia, Finland, Latvia, Lithuania, Poland
and Sweden (Table 2.1.1).
The common coastal water types are characterised by the descriptors of the System B typology.
The typology factors are based on the common typology framework presented in the guidance
on the typology for the coastal and transitional waters
27
. In the Baltic Sea GIG, the common
intercalibration types were characterized using basic salinity and exposure with further delineation
based on depth and number of ice cover days (Table 2.1.1). One transitional water type (TW B 13)
was identified. All countries agreed to focus the intercalibration on the quality elements that are
sensitive to eutrophication pressures.
The common intercalibration types were characterized by the following descriptors:
Salinity (using practical salinity scale): low (0,5-3) and high (3-6) oligohaline, mesohaline (6-22)
27
Guidance document No. 5
‘Transitional and Coastal Waters - Typology, Reference conditions, and Classification sys-
tems’.
Common Implementation Strategy of the Water Framework Directive, Available at: http://forum.europa.eu.int/Pub-
lic/irc/env/wfd/library
219
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0222.png
Depth: all shallow (<30 m)
Exposure (using agreed Pan-European scale
28
):
exposed, sheltered and very sheltered
Duration of ice cover: >150 days/ year, 90-150 days/ year, no or very short ice cover
Table 2.1.1:
Description of Baltic Sea Common intercalibration types that are included in the intercalibration exercise.
Type
CW B0
CW B2
CW B3 a
CW B3 b
CW B12 a
Eastern Baltic
Sea
CW B12 b
Western Baltic
Sea
CW B13
CW B 14
TW B 13
Salinity psu
0.5- 3
3-6
3-6
3-6
5-8
Exposure
Sheltered
Sheltered
Sheltered
Exposed
Sheltered
Depth
Shallow
Shallow
Shallow
Shallow
Shallow
Ice days
> 150
> 150
90-150
90 -150
-
Other Characteristics
Sites in Botnian Bay (Northern Quark)
Sites in Bothnian Sea
Sites in the area extending from the
southern Bothinian Sea to the Archipelago
Sea and the western Gulf of Finland
Sites in the Gulf of Riga
Sites at the Southern Swedish coast and
the South western Baltic Sea open coast
along Denmark and Germany
Sites along the coast of the Estonia, Latvia
and Lithuania, the Polish coast and the
Danish island “Bornholm”
Lagoons
Transitional water. Sites along the coast of
Lithuania and Poland
8 - 22
Sheltered
Shallow
-
6-22
6-22
6-22
Exposed
Sheltered
Exposed
Shallow
Shallow
Shallow
-
-
Countries sharing types that have been intercalibrated:
Types CWB0, CWB2, CWB3a, CWB3b:
Finland, Sweden.
Type CWB12a:
Estonia
Type CWB12b:
Germany, Denmark, Sweden.
Type CWB13:
Denmark, Estonia, Lithuania, Latvia, Poland.
Type CWB14:
Denmark, Poland
Type TWB13:
Lithuania, Poland.
Angiosperms were only intercalibrated between Denmark and Germany for the type B12b.
Angiosperms are not intercalibrated in the other types, because the vegetation is scarce and
distribution scattered. For the angiosperms quality element a hybrid approach between option 2 and
3 was agreed bilaterally between Denmark and Germany. Reference levels are based on historical
data, expert judgment and modeling. Two approaches have been used for classification: 1) percent
deviation from reference conditions (3 scenarios) and 2) modeling. Denmark used the maximum
depth of 5 % eelgrass (Zostera
marina)
cover to define the depth limit. Germany used historical
records of
Zostera marina
depth limit to define reference and light modeling to define depth limits
28
According to the definitions of the common European exposure categories; Guidance document No. 5
220
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
2.1.2 National methods that were intercalibrated
Increased nutrient concentrations stimulate the growth of phytoplankton and opportunistic
macroalgae which cause light attenuation in the water column to increase. Angiosperms like
seagrasses but also perennial macroalgae and eventually the entire benthic vegetation become
shaded and vegetation cover, biomass and depth distribution decline (Duarte 1991, Nielsen et al
2002, Valiela et al. 1997). Moreover, increased nutrient load is expected to increase the abundance
of opportunistic algae, which may further shade seagrasses. A reduction in the vertical and
horisontal distribution of eelgrass beds vegetation is likely to cause increased resuspension of
bottom sediments, which may further shade the remaining benthic vegetation. Increased nutrient
load also tends to increase the risk of anoxic events, which may further harm the benthic vegetation
(e.g. Duarte 1995).
The selected metric is ‘depth limit of eelgrass’. It describes the depth extension of eelgrass (Zostera
marina),
which is the dominant seagrass in Scandinavian coastal waters. The metric is affected by
nutrient concentration and water transparency (Nielsen et al. 2002, Krause-Jensen et al. 2005).
A large compilation of environmental data on eelgrass distribution from Danish coastal waters
demonstrated that eelgrass depth limits increase significantly as nitrogen concentrations decline and
water clarity increases (Nielsen et al 2002, Boström et al. 2003). Similar trends have been shown for
other sea grass species in a worldwide compilation (Duarte 1991). Examples of these relationships
are illustrated by Nielsen et al. 2002 (see Figures 3 and 4, in Nielsen et al. 2002). It should be noted
that in order to illustrate the relationship between depth limit and nitrogen concentration a double
logarithmic scale was used denoting that depth limits vary considerably for given levels of nutrients
or Secchi depths (Nielsen et al. 2002). This means that the models may be useful to describe general
relationships but cannot be used to predict depth limits precisely in specific areas.
2.1.3 Reference conditions and class boundary setting
Reference conditions
Modelling
At present, no Danish and German coastal water bodies are considered to represent reference
conditions, so we attempt to define reference levels of total-nitrogen concentrations (TN) and water
transparency based on historical data and modeling.
In order to define reference levels of benthic vegetation metrics by modeling there is a need for dose-
response relationships between physico-chemical variables (nutrient concentration, transparency)
and vegetation metrics. The reference levels of nutrient concentration and/or transparency are
entered in the model and corresponding levels of the vegetation metrics are calculated. In this
approach the reference levels of nutrient concentration and transparency has been identified within
Danish waters, where vegetation data are available i.e. Øresund.
The model by Nielsen et al. (2002) was used to hind-cast reference depth limits based on reference
TN levels. If reference TN levels are defined as e.g. 14 µM along open Danish coasts, then the
corresponding reference eelgrass (Zostera
marina)
depth limits are 7.7 m. If the high-good boundary
is defined as 25 % of reference levels, then the high/good boundary for eelgrass is 5.8 m.
However, historical data suggest differences in reference depth limits between various coastal
waters/estuaries so it may be an advantage to use historical information on reference depth limits or
to use area-specific values of reference total-N to model reference depth limits.
221
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0224.png
Historical data
The reference depth limit is defined as the values >90 % of the historical values. Furthermore, it is
assumed that the depth limit represents the eelgrass (Zostera
marina)
main distribution rather than
the maximum depth limit.
The most extensive historical data set of eelgrass depth distribution was reported by Ostenfeld
(1908) around year 1900. This dataset contains raw data on eelgrass depth distribution in Danish
coastal waters representing samples taken from a ship with information on location, water depth
and whether or not eelgrass was present. Based on these raw data Ostenfeld assessed the depth limit
of eelgrass in various Danish coastal waters (e.g. Kattegat, Øresund, Little Belt) as the maximum
depth where eelgrass was observed. We defined the reference depth limit for eelgrass as equaling the
high-good boundary and defined this to represent 90 % of the historical maximum. This definition
thus assumes that values above 90 % of the historical maximum represent a high ecological status/
reference situation.
It is observed that for the reference conditions there is a reasonable compliance between the
historically-based and the modeled values, but there is a tendency that the modeled values are a bit
lower than the data based on history (Table 2.1.2).
German data show reference values of a “closed Zostera vegetation” at depth of 8.0-10.0 m
Table 2.1.2:
Historical level and the high/good boundary of the eelgrass (Zostera marina) depth limits defined as
reference condition in Danish and German Waters in the Baltic Sea. The historical depth limits are based on Ostenfeld
(1908) - for the western part of the Baltic Sea it is assumed that the depth limit is the same as for Kattegat, and we have
excluded the Ostenfelds data from the Bøgestrømmen as the depth limit as it seems to be determined by the maximal
limit of the area. The reference depth limits are defined as the values >90 % of the historical values.
German values are after Schories et al. (2006)
Danish and German water areas
In the Baltic Sea GIG
Storebælt and Langelandsbælt
Lillebælt
Smålandshavet – open part
Østersøen, Fakse Bay
Hjelm Bay
Zostera marina
depth limit (m)
Historical level
10,4
8,5
8,5
Class boundary High/good
(Reference)
9,4
7,7
German coast (German type B3)
10,4
10,0
7,5
7,7
6,7
9,4
8,0
Boundary setting
Denmark:
The good-moderate boundary for eelgrass depth limits in Danish coastal waters has been calculated
in three scenarios representing a 15 %, a 20 % and a 25 % deviation from reference levels (Krause-
Jensen 2005). Based on an expert judgment the 15 % and the 20 % deviations are the best scenarios
to comply with the normative definition of good ecological state (Krause-Jensen 2006). Populations
of eelgrass growing at depths of 4.4 – 4.0 m in Limfjorden and a population of eelgrass in Kattegat
growing at 8.1-7.6 m depth are examples of good growing conditions. Accordingly, it was
concluded that 25 % deviation from the reference condition is not congruent with “slight changes of
disturbance”.
222
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0225.png
To underpin which of the 3 scenarios describing the boundary between good and moderate is most
appropriate, a comparison with a model scenario was carried out. The model scenario enters the
total nitrogen concentration, believed to represent the boundary between good and moderate into the
empirical relationship between nutrient concentration and eelgrass depth limit (Nielsen et al. 2002).
Jacob Carstensen (NERI) estimated the total nitrogen concentrations (TN) describing the reference
condition and the good/moderate boundary for a number of Danish coastal waters. Similar values have
been calculated for the German coast. The empirical relationship (Nielsen et al. 2002) is based on
average TN-concentrations for the period March-October, while the Carstensen’s TN-concentrations
are from the period January-June. Therefore, to correct the period a reduction is made (reducing the
January-June data with 22.5 % compared to the March-October concentrations, Table 2.1.3).
Table 2.1.4 compares the reference depth limits and the depth limits for good/moderate state
calculated partly by the historical data and the modeled TN-concentrations. The empirical model
used is not area-specific, but a general relationship for all Danish coastal waters, i.e. it does not take
into account differences in the relationship between TN and depth limit between areas.
Table 2.1.3:
Total nitrogen concentrations, which define respectively, reference condition and the boundary between
high/good and the good/moderate state. Danish data: Average values for the period January-June from Carstensen et al.
(2006). German data from sub-working group “Physico-chemical quality elements” of Bund-Länder-Messprogramm
(BLMP) Working group “WFD”. The good status was estimated as 50 % increase of reference.
Baltic Sea GIG –
Type B 12
Water
Fakse Bay
Total Nitrogen concentrations (μM)
Reference
16.6
16.6
9
Class boundary (high/good)
17.6
17.7
12
Class boundary (good/moderate)
19.6
19.8
16
Hjelm Bay
German coast
Table 2.1.4:
Eelgrass (Zostera
marina) reference
depth limit and depth limit for good/moderate state. Calculations based
on historical eelgrass-data and modeled values derived from TN-concentrations. The empirical relation origins from
Nielsen et al. (2002): lnZ
c
=6.039-0.755*lnTN, where Z
c
is the eelgrass depth limit and TN represent the period March-
October. TN-values are calculated by Carstensen et al. (2006) – corrected by 22.5 % to adjust to March-October level.
Baltic Sea GIG – Type B 12
Danish Coastal waters
Fakse Bay
Østersøen Falster/Hjelm Bay
Zostera marina
depth limit (m)
based on historical data
Ref.
>6.7
>9.4
15 %
5.7
8.0
20 %
5.4
7.5
Zostera marina depth limit (m)
modeled based on TN-concentration
Ref.
8.3
8.3
High/good Good/moderate
8.0
7.9
7.3
7.3
Good/moderate (% of ref)
25 %
5.0
7.1
German coast (German type B3)
10.0
8.0
7.0
In Danish and German coastal waters, the dose-response relationships on present-time data do not
include a range of nutrient concentrations representing reference levels or high/good boundaries.
Reference levels for the benthic vegetation indicators therefore have to be set using historical data,
hind-cast modeling and/or expert judgment. Reference levels have been shown to vary between
areas in Denmark (Table 2.1.5), so we suggest that reference levels are defined for each area and
that class boundary limits are defined as specified deviations from reference levels.
The good/moderate boundary representing a 25-30 % deviation from reference levels and the
resulting depth limits are shown in Table 2.1.6.
223
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0226.png
Table 2.1.5:
Eelgrass depth limit in the Danish coastal waters in the Baltic Sea: The historical depth limits are based on
Ostenfeld (1908) - for the western part of the Baltic Sea it is assumed that the depth limit is the same as for Kattegat,
and we have excluded the Ostenfelds data from the Bøgestrømmen as the depth limit as it seems to be determined by
the maximal limit of the area. The reference depth limits are defined as the values >90 % of the historical values.
For the Danish data the class boundary between good and moderate is described in 3 scenarios: 15 %, 20 % and 25 %
of the reference value (reference Krause-Jensen (2005)) corresponding to EQR values = 0.85, 0.80 and 0.75.
In addition a fourth scenario representing 30 % deviation was included.
Zostera marina
depth limit (m)
Danish water areas
In the Baltic Sea GIG
Storebælt and Langelandsbælt
Lillebælt
Smålandshavet – open part
Østersøen, Fakse Bay
Hjelm Bay
Historical
level
10,4
8,5
8,5
7,5
10,4
Class boundary
High/good
(Reference)
9,4
7,7
7,7
6,7
9,4
Class boundary good/moderate
15 % of
ref.
8,0
6,5
6,5
5,7
8,0
20 %
Of ref.
7,5
6,2
6,2
5,4
7,5
25 %
of ref.
7,1
5,8
5,8
5
7,1
30 %
of ref.
6,6
5,4
5,4
4,7
6,6
Table 2.1.6:
Eelgrass (Zostera marina) depth limits (m) for the reference conditions and for the good-moderate
boundary in the Danish Baltic Sea intercalibration sites. The good-moderate boundary is defined as a 25-30 % deviation
from reference conditions.
Baltic Sea GIG- Type B 12b
Danish intercalibration sites
Coastal area
Reference
Depth limit (m)
9,4
6,7
Good/moderate boundary
Depth limit (m)
7,1 – 6,6
5 – 4,7
Baltic Sea – Hjelm Bay
Baltic Sea – Fakse Bay
Germany
In Germany from historical records eelgrass “stands” were defined as >50 shoots/m², which is
the minimum end of a range of 50->2500 shoots/m² for the Baltic Sea (Schories et al. 2006). The
historical depth limit of
Zostera marina
was assessed as 10 m for stands, while for single plants
few records of deeper occurrence exist (down to 17 m). However, it has to be noted that seeds of
Zostera
may germinate at depths where germination is possible due to stored energy, but persisting
occurrence of adult plants is not. Also, as in Denmark, the genuine historical depth limit of single
plants remains unclear, since data were obtained with historical techniques.10 m is thus assumed as
the historical depth limit.
Through light-modelling (compare Blümel et al. 2002) and under the assumption that 10 % of the
incident light is necessary to maintain
Zostera
stands, Schories et al. (2006) calculated depth limit
for the WFD classes. Here they arbitrarily set certain percentages of light reduction (compared to
pristine conditions) by enhanced attenuation and calculated border depth limits for
Zostera
stands
and single plants (see Table 2.1.7).
Schories et al. 2006 have for the German coast defined the eelgrass depth limit for good status to 7.0
– 8.0 m and moderate status to 4.5 – 7.0 m for the intercalibrated type B12 (German type B3) with
an application of the Danish 90 % rule to historical data, these values show a good fit to the Danish
boundaries.
224
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0227.png
For the German coastal waters, intercalibration type B12b, TN concentrations representing High
status were estimated to vary between 9 and 12 µM, and for Good status between 11 and16 µM
(sub-working group “Physico-chemical quality elements” of Bund-Länder-Messprogramm (BLMP)
Working group “WFD”). The good status was estimated as 50 % increase of reference.
Table 2.1.7:
German depth limits of Zostera marina from Schories et al. (2006) under the assumption of 10 % incident
light necessary to maintain Zostera. As basis attenuation coefficient, K = 0.23025 was used.
Classes
High
Good
Moderate
Poor
Bad
Reduction of light,
single plants
0–
1–
1%
5%
Depth limits of
single
Zostera
plants
9.63 – 10.00
8.39 – 9.62
4.89 – 8.38
1.11 – 4.88
0.00 – 1.10
Tolerance reduction
of
Zostera
stands
0–
6%
6 – 10 %
10 – 30 %
30 – 80 %
80 – 100 %
Depth limits of
stands
8.12 – 10.00
7.21 – 8.11
4.32 – 7,20
0.63 – 4.31
0.00 – 0.63
5 – 25 %
25 – 75 %
75 – 100 %
4.1.4 Results of the comparison
Comparison of Danish and German values
based on experts’ opinions:
Eelgrass depth limits for reference status and good-moderate boundary in the Danish and German
Waters in the Baltic Sea.
Coastal area
Danish Baltic 12b- GIG areas
Baltic Sea – Fakse Bay
Baltic Sea – facing Falster
German Baltic 12b- GIG areas
Geltinger Birk
Darß-Zingst-Außenküste
8-10* m
8-10* m
7* m
7* m
0.88
0.88
6,7 m
9,4 m
5 – 4,7 m
7,1 – 6,6 m
0.75 – 0.70
0.75 – 0.70
Reference
Good/moderate
boundary
Preliminar
EQR values
* These values represent the recommended thresholds for field monitoring, where the estimation of depth to the first or
second decimal is futile. They are rounded and thus slightly different from those in Table 7.
225
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0228.png
4.1.5 Results of the harmonisation – Boundary EQR values
INTERCALIBRATION RESULTS
Biological Quality Element
Results
BALTIC SEA GIG: Angiosperms intercalibration results.
Depth limit eelgrass
Zostera marina
Reference
conditions m
Type B12 Denmark
9.4 (8 – 10.4)
Germany Open coast
Boundary
H/G m
8,5 (8 – 9.4)
Boundary
G/M M
7 (6.6 – 7.1)
EQR H/G
0.90
EQR G/M
0.74
Angiosperms
Final Results
Biological Quality Element
Angiosperms
Angiosperms: parameter indicative of abundance (Depth limit of eelgrass
Zostera marina)
Results: Ecological quality ratios and parameter values
Type and country
Ecological Quality Ratios for the national
classification systems
High-Good boundary Good-Moderate
boundary
CW B 12 b Denmark
and Germany Open
coast
0.90
0.74
Parameter values/ranges Depth limit (m)
eelgrass
Zostera marina
High/-Good
boundary
8.5 (8.0 – 9.4)
Good-Moderate
boundary
7 (6.6 – 7.1)
4.1.6 Open issues and need for further work
Results have been achieved at parameter level ( depth limit of eelgrass,
Zostera marina)
for
coverage and density by two countries (DK and DE) covering one type.
The metric used for angiosperms is not relevant for other countries in the GIG
More widely usable methods for coverage and density should be explored in most types.
The need for assessment methods related to taxonomic composition should be analysed.
226
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0229.png
2.2 NE Atlantic GIG
2.2.1 Intercalibration approach
In the NE Atlantic seven basic intercalibration types have been agreed. These are shown in table 1
below:
Table 2.2.1:
NEA GIG typology.
New Type ID
CW –NEA
1/26a,b,c,d,e
CW – NEA3/4
CW – NEA7
CW – NEA8
CW – NEA9
Name
Exposed or sheltered,
euhaline, shallow
Polyhaline, exposed or
moderately exposed
(Wadden Sea type)
Salinity
(PSU)
Tidal range
(m)
Depth
(m)
Current
velocity
Exposure
Exposed or
sheltered
Exposed or
moderately
exposed
Mixing
Fully mixed
Fully mixed
Fully mixed
Partially
Stratified
Permanently
Stratified
Residence
time
Fully saline
(> 30)
Polyhaline
(18 - 30)
Fully saline
(> 30)
Mesotidal
(1 - 5)
Mesotidal
(1 - 5)
Mesotidal
(1 - 5)
Shallow
(< 30)
Shallow
(< 30)
Deep
(> 30)
CW – NEA10
TW – NEA11
(will be split
into sub-
types)
Transitional waters
Polyhaline, microtidal
exposed, deep (Skaggerak
outer arc type)
Fjord with a shallow sill
at the mouth with a very
deep maximum depth
in the central basin with
poor deepwater exchange.
Polyhaline, microtidal,
sheltered, shallow
(Skagerrak inner arc type)
Deep, low current,
sheltered
Polyhaline
(18 - 30)
Polyhaline
(18 - 30)
Microtidal
(< 1)
Microtidal
(< 1)
Shallow
(< 30)
Deep
(> 30)
low
(< 1 knot)
low
(< 1 knot)
low
(< 1 knot)
Medium
(1 - 3
knots)
Medium
(1 - 3
knots)
Days
Days
Days
Sheltered
Sheltered
Sheltered
Days-
Weeks
Weeks
Polyhaline
(18 - 30)
Oligo-
Euhaline
(0 - 35)
Microtidal
(< 1)
Mesotidal
(1 – 5 )
Deep
(> 30)
Shallow
(< 30)
low
(< 1 knot)
Medium
Exposed
Sheltered
or
moderately
Exposed
Permanently
Stratified
Partially- or
Permanently
Stratified
Days
Days-
Weeks
The above types occur in Member State’s waters as detailed below in table 2.2.2:
Table 2.2.2:
Member States sharing types.
CW - NEA1/26a
CW- NEA1/26c
CW-NEA1/26e
CW – NEA3/4
CW – NEA7
CW – NEA8
Type
BE
X
DK
FR
X
X
DE
IE
X
NL
X
NO
X
PT
ES
X
SE
UK
X
X
CW - NEA1/26b
CW- NEA1/26d
X
X
X
x
x
x
X
x
x
x
X
x
TW – NEA11
CW – NEA10
CW – NEA9
X
x
x
x
X
X
X
X
x
X
X
X
Option 2 has been used in this phase of the intercalibration process. Two metrics have been selected, Intertidal Seagrass:
Abundance (Areal extent and density) and species composition (in types NEA1/26a,b,c/3/4/11) and Subtidal Seagrass:
Abundance (Areal extent and density) and species composition (in type NEA8). Boundaries have been agreed for some
countries for each metric where applicable. This is described below.
227
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0230.png
2.2.2 National methods that were intercalibrated
Table 2.2.3 below indicates the components of the national methods that were considered for
intercalibration for intertidal seagrass.
It is important to note that:
Member states have different seagrass species and different numbers of species. Member states
are considering standardising their percentage loss descriptors.
The Netherlands considers waterbody types 1 and 26 separately. Type 26 represents saline
lakes.
Spain might only able to intercalibrate on taxonomic composition (from the three metrics
proposed) because the Spanish area and density (% cover) metrics incorporate actual data (for
seagrass, saltmarsh and macroalgae) not potential or historical data.
German seagrass data are derived from aerial monitoring and ground truth investigations.
At present no differentiated data are available concerning species composition and coverage/
density (per species). Accordingly only the bed extent of intertidal seagrass is ready to be
classified. Future monitoring programs will help to close this gap.
Table 2.2.3:
National Methods.
Member
states
Angiosperm metric
Change in
taxonomic
composition
Yes
Yes
Yes
Change in
density
Yes
Yes
Yes
Change in
bed extent
Yes
Yes
Yes
Yes
Under
consideration
Under
consideration
Under
consideration
Notes
UK
Republic of
Ireland
Netherlands
Germany
Spain
Portugal
France
Intertidal seagrass
Zostera noltii
and
Z. angustifolia
&
Ruppia
sp.
Intertidal seagrass
Zostera noltii
and
Z. angustifolia
&
Ruppia
sp.
Intertidal seagrass beds; Zostera
noltii
& intertidal
Z. marina
Intertidal seagrass beds; Zostera noltii
& intertidal Z. marina
Habitat code 1110-A, 1110-B & 1140
Below is a summary, table 4, of member states and waterbody types for intertidal seagrass
intercalibration.
Denmark and Sweden have assessed subtidal seagrass bed depth limits in the NEA8 typology (see
below).
228
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0231.png
Table 2.2.4:
Member States sharing types relevant to intertidal seagrass
UK
CW-NEA1/26a, b
Waterbody
types
CW-NEA7
Republic of
Ireland
Netherlands
CW-NEA4
Germany
CW NEA 4
CW-NEA1/26c
Spain
Portugal (for future
intercalibration)
CW-NEA1/26e
CW-NEA1/26a, b CW-NEA1/26b
TW-NEA11
TW-NEA11
TW-NEA11
TW-NEA11
TW-NEA11
TW-NEA11
2.2.3 Reference conditions and class boundary setting
The following intercalibration work between Netherlands, Ireland and the UK has been published
by Foden and de Jong (2007).
Metric - Intertidal Seagrass: Abundance (Areal extent and density) and species composition
(NEA1/26a,b,c/3/4/11)
This metric is only being fully intercalibrated between the Netherlands, Ireland, and UK. Germany
currently only has bed extent data and is currently only intercalibrating on Metric 2 (Seagrass
Abundance: acreage/bed extent), however when species data is available Germany should be able
to participate in all the metrics. Spain have developed a separate metric which has not yet been
intercalibrated, this will be completed Phase II of Intercalibration (see end of section for more
details). This metric has been declared as not applicable by the other GIG Member States.
Reference Conditions
The reference conditions for seagrass for each of the chosen sub-metrics in the co-operating
countries are defined as below. The assumption is made that these occur in unimpacted areas with
unpolluted water quality and no hydromorphological alterations to the shore or seabed. Dutch
waterbodies are embanked and may be classed as heavily modified. Although the waterbodies
are managed and protected by engineering works, habitats such as seagrass beds have established
naturally within them. Potential Reference Conditions (P-REF) and Potential Good Ecological
Status (P-GES) are the highest two classes heavily modified waterbodies can attain, and scientists
in the Netherlands have set values for these by focusing on the current situation in the waterbodies
concerned (de Jong, 2004).
German Reference conditions can be found in German Macroalgae & Angiosperm Annex: IC-
Report Macroalgae and Angiosperms NEA GIG - Germany_05-2007.doc (Kolbe, 2007)
Metric 1. Species Composition (NL, UK, IE)
Table 2.2.5:
NL species composition.
Number of seagrass species
REF
2 species
GES
1 species
Moderate
-
Poor
-
Bad
-
229
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0232.png
Table 2.2.6:
UK and IE species composition.
Number of seagrass
species
High/Ref
No loss of
species
Good
¼ to
1
/
3
loss
of species
Moderate
Loss of �½ of
species.
Poor
Loss of
2
/
3
to
¾ of species
Bad
Loss of all
species
Metric 2. Seagrass Abundance: acreage/bed extent Composition (NL, UK, IE, DE)
Table 2.2.7:
NL seagrass acreage (hectares).
Water type
CW-NEA4
CW-NEA4
TW-NEA11
TW-NEA11
Waterbody
Wadden Sea total
Oosterschelde
Ems-Dollard
Westerschelde
REF
250
1000
100
3
GES
150
750
50
2
Moderate
<25 % below GES
>112.5
>563
>37.5
>1.5
Poor
25-50 % below GES
112.5- 75
563-375
37.5-25.0
1.5-1.0
Bad
>50 % below GES
< 75
< 375
< 25.0
< 1.0
Table 2.2.8:
UK AND IE seagrass bed extent: historical data and expert judgement establish the reference conditions
for a seagrass bed to be compared with its current extent.
High/Ref
No loss in seagrass bed extent
– at maximum potential and
in equilibrium (within natural
variability)
Good
Extent < 30 %
loss from highest
recorded
Moderate
30 – 50 % loss of
bed extent
Poor
50 – 70 % loss of
bed extent
Bad
> 70 % loss of
bed extent
Metric 3. Seagrass Abundance: coverage/density Composition (NL, UK, IE)
Table 2.2.9:
NL seagrass coverage as mean % cover per species. The present positions of the waterbodies are indicated
(WZ = Wadden Sea, OS = Oosterschelde, ED = Ems-Dollard, WS = Westerschelde)
REF
Common Eelgrass (Zostera marina)
Dwarf grass-wrack (Zostera noltii)
Water
Table 2.2.10:
UK and IE seagrass density (% cover).
High/Ref
Bed density at or above ~highest
previously recorded
Good
Density <30 % loss
Moderate
Density <50 %
Poor
Density <70 %
Bad
Density >70 %
>/= 30
>/= 60
GES
>/= 20
>/= 40
WS
Moderate
>/= 10
>/= 30
OS, ED
Poor
>/= 5
>/= 20
WZ
Bad
<5
< 20
230
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0233.png
Sub-metric to support Metrics 2 & 3
Trends in seagrass abundance
Table 2.2.11:
NL trends in seagrass acreage (bed extent) and coverage ( % density). The present trends of the
waterbodies are indicated (WZ = Wadden Sea, OS = Oosterschelde, ED = Ems-Dollard, WS = Westerschelde).
REF/GES
Trend
Water
Positive/neutral
ED WZ
Moderate
Negative
WS
UK
(Foden & Brazier, 2007)
Classification status for density is determined by the underlying trend over a period of 5-6 years,
where data exist, to coincide with the WFD’s reporting cycle. The trend for an individual bed and
the loss or gain, as compared with a maximum recorded density, can be used to identify whether the
seagrass bed is in a state of degradation or recovery.
Boundary Criteria
Metric 1. Species Composition
Most seagrass beds in the UK will comprise 1 or 2 species. Consequently, the NL and UK metrics
are similar. The main difference is that no distinction can be made between High/Ref and Good
for the UK metric because, for example, there are sublittoral beds of
Z. marina
that are naturally
mono-specific and are at High status. The NL’s metric is not able to define conditions less than GES,
whereas the UK metric has boundaries between Good and Moderate, and between Moderate and
Poor/Bad. As the tool testing examples show (below) in most cases the outcomes of the NL metric
and UK metric are generally the same.
Metric 2. Seagrass Abundance: acreage/bed extent
There are significant similarities between the NL and UK metric boundary conditions between each
ecological status class for seagrass acreage/bed extent. With only four waterbodies the NL have been
able to use modeling and expert judgement to set precise bed areas for REF and GES for each of
those waterbodies. The average difference between REF and GES is ~30 % which is broadly in line
with the UK’s more generalised boundary of a 30 % decrease in bed extent between High/Ref and
Good. The mean difference between the NL’s Moderate and REF for all four waterbodies is ~50 %,
between Poor and REF is ~70-75 % and between Bad and REF is >70 %. All of these boundaries are
broadly in common with the UK/IE and Germany metric’s boundary conditions (see section 5.3.5).
Metric 3. Seagrass Abundance: coverage/density
As with bed extent, there are significant similarities between the NL and UK metric boundary
conditions between each ecological status class for seagrass coverage/density. With only four
waterbodies the NL have been able to use modeling and expert judgement to set precise density
ranges for
Z. noltii
and
Z. angustifolia,
for REF and GES. The difference between NL’s REF and
GES for both species is ~30 % which is broadly in line with the UK boundary of a 30 % difference
between High/Ref and Good. For
Z. noltii
the difference between The NL’s Moderate and REF is 50
%, which corresponds with the 50 % difference between Moderate and High for the UK metric. For
Z. angustifolia
there is a greater difference between The NL’s Moderate and REF (⅔) than between
Moderate and High for the UK metric. However, only the Ems-Dollard waterbody will be assessed
against this criterion because the other 3 waterbodies either comprise solely of
Z. noltii
or
Z. noltii
is the dominant species present.
231
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0234.png
There is a difference of ⅔ between NL’s Poor and REF and >⅔ between Bad and REF, for
Z. noltii.
These boundaries are broadly in common with the UK metric’s boundary conditions of <70 % loss
of seagrass for Poor and >70 % loss for Bad.
Sub-metric to support Metrics 2 & 3 Trends in seagrass abundance
Both NL and UK agree that the underlying trend in seagrass abundance should show a stable
seagrass bed (at the maximum potential identified for that site/waterbody). If abundance is less
than would be expected for High/Reference conditions then abundance should show a positive
underlying trend, indicative of recovery. Conversely, a negative trend in seagrass abundance is
undesirable, indicative of degradation, and would signal a potential deterioration in ecological class.
Furthermore, the Member States agree that the ideal period over which to consider the trend in
abundance is ~6 year, designed to coincide with the WFD reporting cycles.
Testing Dutch waterbody data against NL and UK metrics; data for 2004
NB. The analysis of German metric against NL and UK metrics information can be found in
German Macroalgae & Angiosperm Annex: IC-Report Macroalgae and Angiosperms NEA GIG -
Germany_05-2007.doc
Wadden Sea, NL
Table 2.2.12:
Wadden Sea CW-NEA4.
1 – species composition
NL REF
NL GES
Current situation
NL outcome
UK outcome
2 species
1 species
2 species
REF
No species lost = High
2 – acreage/bed extent
250 ha.
150 ha.
47 ha.
Bad
19 % of reference
conditions = Bad
3 – coverage/bed density ( %)
Z. marina
var.
angustifolia
= ≥30
Z. noltii
= ≥60
Z. marina
var.
angustifolia
= ≥20
Z. noltii
= ≥40
Z. noltii
≈40 %
GES
Previous highest recorded density was
58 % in 1997. = Good
The outcomes for the NL and UK metrics are comparable (Table 12). The only difference is for
species composition where the loss of
Z. angustifolia
(littoral eelgrass) indicates a Moderate
outcome for the UK metric rather than a GES outcome. The underlying trends in both seagrass bed
extent (acreage) and density could be described as positive or neutral (Figure 1); so although the
size of the bed is very small compared with the REF of 250 ha. It is possibly in a recovery phase.
232
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0235.png
70
Seagrass area (ha. @ >5% cover) and
coverage (% density)
60
50
40
30
20
10
y = 1.1408x - 2250.5
R
2
= 0.0838
Wadden Sea seagrass
y = 0.6435x - 1243.3
R
2
= 0.0216
A rea (ha.)
0
1990
1992
1994
1996
1998
Year
2000
Coverage (% density)
2002
2004
2006
Figure2.2.1:
Temporal trends in
Wadden Sea seagrass acreage
(bed extent) and coverage (% density).
Oosterschelde, NL
Table 2.2.13:
Oosterschelde CW-NEA4.
Species composition
NL REF
NL GES
Current
situation
NL outcome
UK outcome
2 species
1 species
2 species
REF
No species lost = High
Acreage/bed extent
1000 ha.
750 ha.
94 ha.
Bad
Coverage/bed density (%)
Z. marina
var.
angustifolia
= ≥30
Z. noltii =
≥60
Z. marina
var.
angustifolia
= ≥20
Z. noltii
= ≥40
Z. marina
var.
angustifolia
= ≥30 in
Roggenplaat, <10 % elsewhere
Z. noltii
≈ 62 %
GES/Moderate boundary
Previous highest density was 50 % in 1979-
1983; decline of ~35 % = Moderate
~10 % of reference
conditions = Bad
900
Area (>5% cover) and coverage (%
density)
800
700
600
500
400
300
200
100
0
7
3
Oosterschelde seagrass
Area (ha.)
y = -60.187x + 653.32
R
2
= 0.7284
Coverage (% density)
y = -0.9649x + 46.448
R
2
= 0.2138
8
2
7
1
-0
97
19
-8
6
-9
5
75
73
79
-9
-9
-9
9
-8
-9
0
82
86
84
91
19
19
19
19
19
19
88
93
95
19
Year ranges
Figure 2.2.2:
Temporal trends in Oosterschelde seagrass acreage (bed extent) and coverage
(% density).
19
19
19
19
99
-0
3
-7
9
-7
-8
233
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0236.png
The outcomes for the NL and UK metrics are broadly the same for the three metrics (Table 13).
Both
Z. noltii
and
Z. angustifolia
are present. The underlying trend in seagrass bed extent (acreage)
is strongly negative between 1973 and the beginning of the 1990s (Figure 2). Since then it has
become neutral, with possibly a very slight increase in recent year.
Z. noltii
is the species considered
for the density metric as it dominates the species composition;
Z. angustifolia
(littoral eelgrass)
constitutes a very small proportion of the overall seagrass area at very low density therein. The most
recent mean density calculation (62 %) lies on the boundary between the NL’s GES and Moderate.
The trend in density could be described as neutral as there has been consistently low percentage
cover (Figure 2).
Ems-Dollard, NL
Table 2.2.14:
Ems-Dollard TW-NEA11.
Species composition
NL REF
NL GES
Current situation
NL outcome
UK outcome
2 species
1 species
1 species
GES
(No record of
Z. noltii)
= High
Acreage/bed extent
100 ha.
50 ha.
14 ha.
Bad
~15 % of reference
conditions = Bad
Coverage/bed density (%)
Z. marina
var.
angustifolia
= ≥30
Z. marina
var.
angustifolia
= ≥20
Z.
marina
var.
angustifolia
≈ 13 %
Moderate
Previous highest density was 30 % in
1988; decline of ~55 % = Moderate
The outcomes for the NL and UK metrics are the same in all three instances (Table 14).
Z. noltii
has
not been recorded in this waterbody and the overall seagrass area is comprised of
Z. angustifolia
(littoral eelgrass) at low densities. The underlying trend in seagrass bed extent (acreage) is positive
from 1986 to 2003, but drops sharply in 2004. The trend in density could be described as negative
with the 2004 figure being one of the lowest since 1988.
Ems_Dollard seagrass
100
90
Area (ha. @>5% cover) and coverage
(% density)
80
70
60
50
40
30
20
10
0
1986
1988
1990
1992
Area (ha.)
Coverage (% density)
y = 2.8547x - 5655.3
R
2
= 0.215
y = -0.7837x + 1583.7
R
2
= 0.4825
1994
1996
Year
1998
2000
2002
2004
2006
Figure 2.2.3:
Temporal trends in Ems-Dollard seagrass acreage (bed extent) and coverage
(% density).
234
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0237.png
Westerschelde, NL
The
Z. noltii
bed is a small bed, situated in the harbour of Flushing-east, with an uncertain history.
It might be that
Z. noltii
was present in the past (before the construction of the harbour) in the inlet
(Sloe) between the two islands Walcheren (western side) and South Beveland (eastern side), but no
records are available. Equally it might that it appeared there after the construction of the harbour.
It is only mapped when a vegetation map is made of the small salt marsh that is present there. The
cover is low, as it is a patchy bed in between the
Spartina
tussocks on the seaward side of the marsh.
Estimated mean cover is between 5 and 20 %.
Table 2.2.15:
Westerschelde TW-NEA11.
Species composition
NL REF
NL GES
Current situation
NL outcome
2 species
1 species
1 species
GES
Acreage/bed extent
3 ha.
2 ha.
2 ha.
GES
Coverage/bed density ( %)
Z. marina
var.
angustifolia
= ≥30
Z. noltii
= ≥60
Z. marina
var.
angustifolia
= ≥20
Z. noltii
= ≥40
Z. noltii
≈ 5-20 %
Poor (possibly Bad)
Previous highest density data
unavailable, but cover is patchy.
Therefore less than ‘Good’ –
precautionary Moderate or Poor
Moderate or
Poor
UK outcome
(No record of
Z. marina
var.
angustifolia)
= High
~66 % of reference
conditions = Good
The outcomes for the NL and UK metrics are the same for species composition and acreage (bed
extent), but differ for bed density (Table 15). In fact this latter difference is of minor significance
because; (a) there is a lack of raw data to allow a more confident setting of class, and (b) both The
NL and UK outcomes are less than Good, meaning a programme of investigative measures would
be undertaken. The absence of raw data prevents a temporal trend plot over recent years for the
Westerschelde.
Testing UK waterbody data against NL and UK metrics
Strangford Lough, UK (Portig, 2004)
Historical data are scarce, but there has clearly been a marked decline in the distribution of
seagrasses in Northern Ireland since 1930s. This has been coupled with a change in the dominant
Zostera
spp. present in the intertidal areas with
Z. marina
in its perennial form dominant in the
1930s being replaced by
Z. noltii
and
Z. angustifolia
by 1970.
There has been a general improvement in the status of
Zostera
spp. in the northern end of Strangford
Lough during the last 10 years. The necessary data are lacking, however, to determine whether these
changes are part of ongoing cyclical processes or longer term changes. The mean seagrass density
for all N.I. (Northern Ireland) loughs is 52 %
235
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0238.png
Table 2.2.16:
Strangford Lough CW-NEA26 (CW8).
Species composition
Z. marina
var.
angustifolia,
Z. noltii,
Z. marina,
Ruppia
spp.
REF
High
Acreage/bed extent
924 ha. Literature and qualitative
sources suggest the beds are in a
‘recovery’ phase; i.e. expanding,
but at less than maximum
potential; assume >50 % and <70
% of maximum.
Moderate
Moderate
Coverage/bed density ( %)
53 % mean density for all species.
No baseline data are available, but
this figure is comparable with N.I.
records of seagrass bed densities in
other loughs.
REF
High
Current situation
NL outcome
UK outcome
Strangford Lough seagrass
60
50
Cover (% density)
40
30
20
10
0
2001
2002
Year
2003
2004
Figure 2.2.4:
Temporal trends in
Strangford Lough seagrass % density.
The outcomes for the NL and UK metrics are the same in all three instances (Table 16). The lack of
raw data makes precise statements regarding seagrass acreage (bed extent) and coverage (density)
difficult. The underlying trend in seagrass density is positive over the 4 survey years, confirming the
judgement of the seagrass being in a ‘recovery’ phase.
Fleet lagoon, UK (Bunker et al., 2004)
Distribution changes in seagrass species of Fleet Lagoon, as surveyed in 2002 (Bunker
et al.,
2004):
Z. marina
; Lost from Swannery Basin since 1983, but a north-westward extension of range in
West Fleet since 1999
Z. noltii
; Lost from Swannery Basin since 1983, but qualitative reports of unchanged
distribution since 1999.
Ruppia
spp.; Lost from Swannery Basin and west of Berry Coppice since in 1983, but
distributions in the rest of the Fleet remain broadly unchanged since 1999.
Variability in the data is very high because of the transect survey method and absence of mapping.
There are three issues that affect the final decision of Fleet’s ecological status:
1. Standard Deviation at some stations was almost as large as the mean of the 12 quadrats
(Figure5)
236
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0239.png
2. Power analysis shows a high % change in density (>50 % at the majority of stations) would
have to occur to detect a statistically significant change; analysed for Type I and Type II errors
(percentage change in mean density required in order to be 95 % confident that a statistically
significant change has occurred).
3. Previous surveys do not provide raw quadrat data so statistical comparison between data sets
is not possible.
100
90
80
Z. noltii density (% cover)
70
60
50
40
30
20
10
0
0
5
10
15
20
25
Station number
30
35
40
45
50
Figure 2.2.5:
Standard deviation examples.
2.2.4 Results of the comparison
As option 2 has been used the boundaries have been agreed by experts representing all countries in
the GIG angiosperms sub-group. Therefore the final agreed results presented below are the results
of comparing expert views on what the boundaries should be.
2.2.5 Results of the harmonisation – Boundary thresholds and EQR values
Metric - Intertidal Seagrass: Abundance (Areal extent and density) and species composition
(NEA1/26a,b/3/4/11)
The Netherlands, Ireland and the UK have agreed a common matrix for allocating status to intertidal
seagrass assessments on the basis of the table below. This matrix combines both loss of species and
degradation in the % cover (measured as the number of seagrass shoots in a quadrat or % cover
of seagrass within a quadrat). The matrix covers both situations where naturally either two or three
species of seagrass are found within either a type or where there are differences within types in
specified geographic areas. The appropriate selection from the matrix is made at the waterbody level.
237
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0240.png
Table 2.2.17:
Intertidal Seagrass: Abundance and Species composition classification boundary matrix.
Density (% cover)
No spp. lost
1 spp. lost, 2 remain
Species
1 spp. lost, 1 remains
2 spp. lost, 1 remains
All spp. lost, therefore
0 % cover
0 - 10 % lost
High
Good
Good
Moderate
10 - 30 % lost
Good
Good
Moderate
Moderate
30 - 50 % lost
Moderate
Moderate
Poor
Poor
Bad
50 - 70 % lost
Poor
Poor
Bad
Bad
>70 % lost
Bad
Bad
Bad
Bad
The above matrix does not allow the true mathematical derivation of EQRs so the generic approach
of allocating equal EQR ranges for status classes (as allowed by the boundary setting protocol) will
apply; the EQR boundaries and ranges are shown below.
Status
High
Good
Moderate
Poor
Bad
EQR
0.9 (0.8 – 1.0)
0.7 (0.6 – 0.8)
0.5 (0.4 – 0.6)
0.3 (0.2 – 0.4)
0.1 (0.0 – 0.2)
France and Portugal are to consider using the above matrix and hope to decide by June 2007.
Germany’s seagrass data are derived from aerial monitoring and ground truth investigations. At
present no differentiated data are available concerning species composition and coverage/density
(per species). Consequently Germany cannot yet participate in the species composition component.
Future monitoring programs will help to close this gap.
Metric - Intertidal seagrass area (acreage/bed extent) (NEA1/26a,b,c/3/4/11)
This scheme has been agreed by UK, IE, NL and DE and is tabulated below:
Table 2.2.18:
Intertidal seagrass area (acreage/bed extent) classification boundary values.
Member
states
UK
Republic of
Ireland
Netherlands
Description of seagrass
Intertidal
seagrass beds
Intertidal
seagrass beds
Z. noltii
and
mixed beds
Z. noltii
and
mixed beds
Change
in area
Change
in area
Change
in area
Change
in area
High
(REF in NL)
0 - 10 % loss
0 - 10 % loss
0-10 % loss
(allowing for
natural variation)
0 - 10 % loss
Good
(GES in NL)
11 - 30 % loss
11 - 30 % loss
Moderate
31 - 50 % loss
31 - 50 % loss
31 - 50 % loss
from REF
31 - 50 % loss
Poor
51 - 70 % loss
51 - 70 % loss
51 - 70 % loss
from REF
51 - 70 % loss
Bad
>70 % loss
>70 % loss
>70 % loss
from REF
>70 % loss
11-30 % loss
Germany
11 - 30 % loss
238
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
The values in table 17 translate into the EQR boundary values and ranges shown below.
Status
High
Good
Moderate
Poor
Bad
EQR
0.9 (0.8 – 1.0)
0.7 (0.6 – 0.8)
0.5 (0.4 – 0.6)
0.3 (0.2 – 0.4)
0.1 (0.0 – 0.2)
Spain:
Have developed an Angiosperm Intertidal/Subtidal Quality Index for their NEA Type 11 waters.
Other countries have not yet tested the Spanish Angiosperm quality index and no comparison
between the developed metrics has been made. Therefore the intercalibration for Spain will not be
complete until testing has been carried and boundaries agreed for June 2007. Details of the metric
proposed can be found in the detailed Milestone 6 Report.
2.2.6 Open issues and need for further work
Gaps for the future:
Spain will attempt to separate macroalgae and seagrass from their macrophyte tool, for
intercalibration purposes
Development of saltmarsh tools is needed for most member states.
Questions of how to combine different sub-metrics into a higher level overall EQR.
Spain has to decide the metric to use for seagrass and improve it
Germany has to test and to decide using the combined metric intertidal seagrass abundance
(density)/species composition
Intercalibration for Saltmarsh will need to be undertaken at a future date.
239
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
3 References
Blümel C, Domin A, Krause J, Schubert M, Schiewer U, Schubert H (2002). Der historische Makrophytenbewuchs
der inneren Gewässer der deutschen Ostseeküste. Rostock. Meeresbiol. Beitr. 10:5-111.
Boström C, Baden SP, Krause-Jensen D. 2003 The seagrasses of Scandinavia and the Baltic Sea. In Green EP,
Short FT (Eds) World atlas of seagrasses. California University Press. 310 pp.
Bunker, F., T. Mercer & C. Howson, 2004. Fleet lagoon and tidal rapids survey 15th–22nd July 2002. Report
to English Nature, 80 pp.
Carstensen, J., Conley, D. J., Andersen, J., Ærtebjerg, G. (2006) Coastal eutrophication and trend reversal: a
Danish case study. Limnology and Oceanography 51, pp. 398-408
de Jong, D.J., 2004. Water Framework Directive: determination of the reference condition and Potential-
REF/Potential-GES and formulation of indices for plants in the coastal waters CW-NEA3 (K1), CWNEA4
(K2), CW-NEA1 (K3), transitional water, TW-NEA11 (O2), and large saline lakes, NEA26 (M32), in The
Netherlands. (Working document RIKZ/OS/2004.832.x; final draft).
Duarte CM. 1991. Seagrass depth limits. Aquat. Bot. 40: 363-377.
Duarte CM. 1995. Submerged Aquatic Vegetation in Relation to Different Nutrient Regimes. Ophelia 41: 87-112.
Foden, J. and Brazier, D.P., 2007. Angiosperms (seagrass) within the EU water framework directive: A UK
perspective. Marine Pollution Bulletin 55: 181–195.
Foden, J. and de Jong, D.J., 2007. Assessment metrics for littoral seagrass under the European Water Framework
Directive; outcomes of UK intercalibration with the Netherlands. Hydrobiologia 583(1): 383-383
Hansson M, Håkansson B. 2006 Förslag till Vattendirektivets bedömningsgrunder för pelagiala vintertida
näringsämnen och sommartida effektrelaterade i kust- och övergångsvatten
Krause-Jensen D. 2005. Scenarier for klassifikation af kystvande vha. ålegræssets dybdegrænse
Chapter 3.4 in Dahl K, Andersen JH, Riemann B (Eds) Carstensen J, Christiansen T, Krause-Jensen D,
Josefson AB, Larsen MM, Petersen JK, Rasmussen MB, Stand J 2005: Redskaber til vurdering af miljø-
og naturkvalitet i de danske farvande. Typeinddeling, udvalgte indikatorer og eksempler på klassifikation.
Danmarks Miljøundersøgelser. 158 p. – Technical report from NERI no. 535. (In Danish) http://technical-
reports.dmu.dk
Kolbe, K., 2007. Assessment of German Coastal Waters (NEA 1/26, NEA 3/4) and Transitional Waters (NEA
11) by Macroalgae and Angiosperms. Intercalibration Report (NEA GIG).- (postal to D. Coates, D. Jowett).
Krause-Jensen D, Greve TM, Nielsen K. 2005. Eelgrass as a bioindicator under the Water Framework Directive.
Water Resources Management 19: 63-75.
Magnusson J, Johnsen T.M, Beyer F, Gjøsæter J, Lømsland E.R, Sollie, Aa, 1999. Pollution monitoring in the
Inner Oslofjord in 1998. NIVA-report. OR-4058. 63pp. In Norwegian (http://www.niva.no/).
Nielsen SL, Sand-Jensen K, Borum J, Geertz-Hansen O. 2002. Depth colonization of eelgrass (Zostera
marina)
and macroalgae as determined by water transparency in Danish coastal waters. Estuaries 25: 1025-1032.
Ostenfeld DH 1908.’Ålegræssets (Zostera
marina)
udbredelse I vore farvande’, in CGJ Petersen (ed), Beretning
til landbrugsminiteriet fra den danske biologiske station XVI, Copenhagen, Centraltrykkeriet.
Portig, A.A., 2004. The distribution of intertidal Zostera spp. in Northern Ireland, 2003. Report for Environment
and Heritage Service p. 84.
Schories D, Selig U, Schubert H. 2006. Testung des Klassifizierungsansatzes Mecklenburg-Vorpommern
(innere Küstengewässer) unter den Bedingungen Schleswig-Holsteins und Ausdehnung des Ansatzes
auf die Außenküste. Küstengewässer-Klassifizierung deutsche Ostsee nach EU-WRRL. Teil A: Äußere
Küstengewässer.
Valiela I, McClelland J, Hauxwell J, Behr PJ, Hersh D, Foreman K. 1997. Macroalgal blooms in shallow
estuaries: Controls and ecophysiological and ecosystem consequences. Limnol. Oceanogr. 42:1105-1118.
Øresundsvandsamarbejdet - Öresundsvattensamarbetet. 2004. Jämförelse mellan ‘urtida’ och ‘nutida’
näringsnivåer i Öresund – beräkningar utförda med MIKE3-modell. www.oresundsvand.dk
240
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
European Commission
EUR 23838 EN/3 – Joint Research Centre – Institute for Environment and Sustainability
Water Framework Directive intercalibration technical report - Part 3: Coastal and Transitional waters
Alessandro Carletti, Anna-Stiina Heiskanen
Luxembourg: Office for Official Publications of the European Communities
2009 – 240 pp. – 21 x 29,7 cm
EUR – Scientific and Technical Research series – ISSN 1018-5593
ISBN 978-92-79-12568-3
DOI 10.2788/19561
Abstract
This Technical Report gives an overview of the technical and scientific work that has been carried out in
the intercalibration of coastal and transitional waters ecological classification systems across the European
Union as required by the Water Framework Directive (WFD). The results of this exercise were published in the
Official Journal of the European Union as Commission Decision 2008/915/EC of 30 October 2008.
L 68 - 2015-16 - Endeligt svar på spørgsmål 62: Spm. om oversendelse af al korrespondance mellem kvælstofudvalget og forskere og forskningsinstitutioner m.fl., til miljø- og fødevareministeren
1601764_0244.png
14
The mission of the JRC is to provide customer-driven scientific and technical support
for the conception, development, implementation and monitoring of EU policies. As
a service of the European Commission, the JRC functions as a reference centre of
science and technology for the Union. Close to the policy-making process, it serves
the common interest of the Member States, while being independent of special
interests, whether private or national.
LB-NB-23838 EN/3-C
LB-NC-23838-EN-C