1. Reference Conditions Cross-GIG Group Report on consistency in reference criteria application IC steering group

2. Why are we reviewing reference conditions in intercalibration?  The RC review work is an essential part of intercalibration  The need for consistency checking and recommendations for improving reference concepts was mandated by ECOSTAT  The new draft IC guidance requires consistent reference setting across all GIG QE groups.

3. Summary of the RC WG work 1.Checking consistency of reference condition setting in Phase 1 Intercalibration; evaluation of threshold values used for selecting reference sites in IC Phase 1 [2009-10, completed] 2.Future development of concepts/approaches [2010-11]

4. Summary of the RC WG work: Checking consistency of Reference Condition setting in Phase 1  Report “Revision of the consistency in Reference Criteria application in the phase 1 of the European Intercalibration exercise”  Prepared by RC group  Covering all water categories  Report was distributed to ECOSTAT as final draft  To be published as JRC technical report spring 2010

5. Checking consistency of reference condition setting in Phase 1  Do the GIGs have agreed reference criteria?  How did MS assess compliance with these criteria?  Based on existing GIG information (all categories)  Do the pressure data associated with current reference sites demonstrate consistency with agreed criteria?  Based on additional information requested from Member States (Lakes and Rivers)

6. Outline of this presentation  Consistency analysis Rivers (WvdB)  Consistency analysis Lakes (SP)  Consistency analysis Coastal/Transitional (WB)

7. Rivers

8. RIVERS - approaches

9. River fish– RC approach  Common database with pressure and biological data  Defined criteria applied to common databases - ‘undisturbed sites’  MS defined ‘reference sites’ using national criteria  IC reference sites should be both ‘undisturbed sites’ and national ‘reference sites’

10. River macrophytes IC  High variability in ‘reference sites’  Common notion of natural reference communities (expert judgment)  needs to be validated  RC derived from benchmark data set covering degradation gradient

11. River phytobenthos– recommendations on RC  Reference sites selection criteria need to be validated and published  Testing IC typology priority  Common format for collecting key environmental data needed

12. River invertebrates IC – application of RC criteria  Common approach across GIGs (except EC): MS check sites against 42 REFCOND criteria  Differences in application:  Choice of criteria (not all were used)  Detail in answers  Yes/No (AL, NO)  Measured vs. field inspection vs. expert judgment vs. other criteria

13. Rivers invertebrate IC - differences in application of criteria between GIGs

14. Rivers invertebrate IC - comparability of pressure information  Point sources, diffuse sources: mostly ‘measured’, but much missing info

15. Rivers invertebrate IC - comparability of pressure information  Morphological alterations: information mainly at reach scale, missing at basin scale

16. Application of agreed reference/rejection thresholds  Do reference sites meet agreed numerical criteria  Was not checked in ‘round 1’  Data request to MS April 2009  Data collected from 13 MS – analysis based on incomplete information

17. Present situation on data request to Member States

18. UK, Spain: delivered data, but are not on the map Present situation on data request to Member States: Reference sites request

19. REF SITES Drivers/Pressures 50% 30% Present situation on request to Member States for data from reference sites Problem #1: Data completeness

20. Artificial land use: many ‘reference sites’ exceed the rejection threshold!

21. Intensive agriculture land use: most ‘reference sites’ below reference thresholds

22. Problem #2: Water chemistry: More spot values than mean values SPOT > MEAN > MEDIAN > 90% > 70% Present situation on request to Member States for data from reference sites

23. Mean nitrate values: large differences between MS; missing data

24. Spot measure nitrate values: more data – but difficult to use because thresholds are defined as mean values…

25. River invertebrates – checking of suitability of the criteria  Looking for no-impact-thresholds using pressure and biological data  Currently not enough data to come to firm conclusions  RC group intends to continue working on this – further data expected

26. Lakes  Sandra

27. LAKES 1.How Reference conditions were set ? Approaches, criteria,,,, 2.How criteria were applied ? Checking and comparison of pressure data

28. LAKES Common concept of reference lakes Screening based on common datasets Pressure criteria for screening Impact criteria for confirming Broadly following REFCOND But – different approaches / criteria

29. Summary on approaches Northern and Mediterranean GIGs – criteria not harmonized within the GIG In many cases not possible to quantify stressor criteria (e.g., land use in Finland) not all REFCOND criteria used – not considered important for eutrophication – lack of data/information in the central databases Both pressure data and in-lake nutrient data were used for selection of reference lakes GIGs used different criteria/ approaches - no guarantee for comparable reference conditions across Europe;

30. Consistency analysis - LAKES Collection of pressure data Reference lakes from IC phase 1 427 lakes, 19 MS Land use data 385 lakes Population data 272 lakes

31. Natural land use

32. Population density

33. TP values What is more important: landuse or TP values ?,, probably TP !

34. TP values - comparison Nordic – Low/similar TP – 4-8  g TP/l Alpine – Low/similar TP – 4-8  g TP/l

35. TP values - comparison Central-Baltic – variable TP between MS

36. Different strategies – different outcomes Nordic reference lakes – low artificial land cover (95th percentile 0.46%) and intense agriculture (95th percentile 0.13%) – high share of natural and semi natural land cover (median value 94.1%), mostly corresponding to reference criteria; – Low and coherent TP values

37. Different strategies – different outcomes Alpine reference lakes – highly variable land use pattern, frequently exceeding reference thresholds – Alpine median value of natural landcover 80 %, while ref threshold 80-90%; – Low and coherent TP values – TP used for selection of reference lakes based on paleodata, modelling, historical data

38. Different strategies – different outcomes Central Baltic reference lakes – highly variable land use pattern, frequently exceeding reference thresholds – median value of natural land cover only 70%, while ref threshold 90% – Variable TP values within type between MS

39. Reference criteria ? What is the right tool to select reference lakes – pressure data or TP ?

40. Chl-a – % natural land in reference lakes The only sign difference between 50-60 and 60-70% No significant difference between groups The only sign difference between 50-60 and 60-70% No relationship in range 60-100% natural land use

41. 2. Nutrient ‘production‘ in the catchment area Retention Reduction Export 1. Land use 3. Nutrient load into the lake 4. Nutrient concentration in the lake 5. Phytoplankton abundance / taxonomic composition Loading model (Vollenweider …) Good data availability  Use as proxy for land use Natural run-off Diffuse point sources (agriculture) Point sources (industry, inhabitants, …)

42. Summary – Reference Lakes Common concept but slightly different approaches Gaps of information and harmonisation Need to develope common framework Lack of reference lakes in many regions of Europe Not always a clear link between pressure criteria (e.g. land use) and in-lake nutrient levels Nutrient levels can be used as a proxy of eutrophication pressure Still this approach has to be used with a caution, including a well-grounded setting of reference TP levels

43. COASTAL/TRANSITIONAL  Wendy

44. COAST – approaches general overview BQEIC phaseWFD approach to Reference condition PhytoplanktonIExpert judgment, modeling, in some cases reference sites, abiotic and biotic Macro-algae & angiosperms IReference sites / best available sites in MED GIG, biotic (in relation to abiotic in Baltic Sea) Benthic invertebrate fauna IExpert judgment of the best values of a biological index in a long time series or at increasing distance from a pollution source, biotic in relation to abiotic FishIIDirect comparison in reference estuary /lagoon and estuaries/lagoons with specific pressures

45. Phytoplankton Baltic Sea GIG: Eutrophication is the only pressure assessed for all BQEs No current reference sites but historical measurements of secchi depth available (expert judgment and modeling) Chlorophyll a biomass: –2 approaches: reference concentration based on relationship between Secchi depth - TN – chl a (DK, FI, EE) or relationship between hind- casted nitrogen inputs - TN - chl a (DE), SE and PO compared both approaches –Secchi depth DK: 1925 – 1934 FI: 1903 - 1959 PO: summer 6 m SE: 10 m for Baltic propper, adjusted for coastal areas –TN referenceDE: 10 μM SE: 15.3 μM Summary on reference conditions in the GIGs

46. Phytoplankton NEA GIG: Chlorophyll a biomass –Type 1/26a: Reference sites in Ireland, UK, France (1992-2006) and Spain –Type 1/26b North Sea: boundary G/M and H/G has been defined by expert judgment, modeling, literature –UK: expert judgment H/G boundary 10 μg/l based on offshore measurements –NL: measurements 70ies – relation freshwater runoff estimates – UK and DE –BE: average diatom pre-bloom biomass and Phaeocystis biomass –FR: based on offshore measurements 1992 - 2006 –Type 1/26e: reference sites / current measurements for ES and PT Mediterranean Sea GIG Chlorophyll a biomass: expert judgment on measurements at reference (high status) sites of monitoring programmes (for 3 of 4 types) Black Sea GIG –BU: Phytoplankton biomass: historical biovolume data of relatively pristine period (1954-1970) and 1983-2005 (10th and 25th percentile), related to measurements of chl a (1990-2006) –RO: Phytoplankton biomass: historical biovolume data of relatively pristine period (1960-1970) and 1986-1997 Summary on reference conditions in the GIGs

47. Conclusions Phytoplankton Often lacking for compliance/consistency check: Mentioning time periods of measurements/description of dataset used for derivation of reference conditions Description on choice of reference sites where applied (some info from WFD Art.5 pressure analysis !?) Explanation on levels of difference between Ref, H/G and G/M boundary in relation to provisions of OSPAR – HELCOM Illustration of relationships with physico-chemical parameters like nutrients (was well described in Baltic, 1-1 relationship is not possible) Relationship between hind-casted estimates of nitrogen inputs – TN – chlorophyll a should be a wider applied approach (link with reference nutrient concentrations in rivers possible ?)

48. Comparison OSPAR COMPP = the Comprehensive Procedure and the Water Framework Directive Assessment levels are based on a justified area-specific % deviation from background levels not exceeding 50%.

49. Macroalgae NEA-GIG –1st phase: not all the metrics that make up MS’ schemes can be intercalibrated, not possible to produce EQRs for the whole quality element. Boundaries have been agreed for the selected metrics –No current reference sites –Assessed for pressure morphological alteration –Type 1/26a: biological reference = expert judgement on maximum possible values of species richness and percentage of green / red / opportunistic – sensitive algae + characteristic species (ES) –Intertidal rocky shore algae (only ES also subtidal) –Type NEA 8,9,10: reference depth limit for 9 selected macroalgae species based on historical data (not allowing good statistical treatment) with expert judgment and modelling –Subtidal rocky shore algae Summary on reference conditions in the GIGs

50. Macroalgae Mediterranean Sea GIG –EEI (Greece, Cyprus, Slovenia) and CARLIT (Spain, France, Italy) includes relative abundance/cover and disturbance sensitive metrics. No description included, only references. –EEI: ESGI and ESGII percentages = disturbance sensitive metric –CARLIT: spatial evaluation of communities (disturbance sensitive or not) –Macroalgae community described from reference sites (marine reserves) intercalibrating EEI and CARLIT –Littoral and upper sublittoral rocky shore algae –BENTHOS ICCM in Option 2 –Most difficult composition part was done very well –Relationship with pressure illustrated for Greece and Spain (water quality, nutriens N and P, urban gradient)

51. Conclusions Macroalgae Mentioning time periods of measurements/description of dataset used for derivation of reference conditions still to be done Choice of reference sites OK for MED-GIG, but explain how the reference sites were selected (some info from WFD Art.5 pressure analysis), CARLIT still needs some amendment No distinction of types in NEA 1/26, but PT applies higher proportion of red algae species (>70%) as reference than other countries (40-55%), but lower species richness (>25) in relation to others (>33-35) Subdivision of types necessary ? Illustration of relationships with hydro-morphological parame- ters in NEA? In WISER questionnaires some mentioned.

52. Angiosperms Baltic Sea GIG –No current reference sites –Angiosperms for one type (DK – DE): historical data, expert judgement (percentage deviation scenarios from reference tested) compared with results of modelling relationship with TN –Reference depth of 8 to 10 m for seagrass stands –Illustration of relationship with physico-chemical parameters (nutrients like TN) tested for the Baltic Sea NEA GIG –Assessed for pressure morphological alteration –Angiosperms taxonomic composition, bed extent and bed density: reference = expert judgement on potential biological reference, some historical data Summary on reference conditions in the GIGs

53. Angiosperms Mediterranean Sea GIG (reported, not included yet in 1st technical report) –POMI (Spain) includes seagrass abundance/cover (2 metrics: meadow cover, shoot density) and disturbance sensitive metrics (at population, individual and physiological level) –PREI (France) includes seagrass abundance/cover (2 metrics: depth and type of the lower limit, shoot density) and disturbance sensitive metrics (shoot leaf surface area, E/L ratio (Epiphytic biomass/leaf biomass)) –Posware (Italy) includes seagrass abundance/cover (1 metric: shoot density) and disturbance sensitive metrics (width of the intermediate leaf, leaf production, rhizome production, rhizome elongation) –ICCM with 2 metrics: lower limit type of meadows and leaf surface area per shoot: common understanding of reference conditions must be made very clear

54. Conclusions Angiosperms Mentioning time periods of measurements/description of dataset used for derivation of reference conditions still to be done Choice of reference condition OK for MED-GIG – but explain how the best biological situation is defined for the POMI and the other methods (some info from WFD Art.5 pressure analysis) Baltic Sea GIG: reference depth for Zostera marina is 8 – 10.4m, while reference Secchi depth was set at 6 m (coastal) – 10 m (offshore) for phytoplankton reference: cross-validation? 25% deviation from the H/G boundary was considered not congruent with normative definitions, but finally the G/M boundary was defined as a 25-30% deviation from H/G Illustration of relationship with physico-chemical parameters like nutrients (like TN) tested for the Baltic Sea, not for NEA where general degradation of seagrass fields is described, for MED for separate metrics

55. Benthic invertebrate fauna Baltic Sea GIG + NEA GIG + MED GIG + Black Sea GIG -Reference = expert judgment of the best values of a biological index in relation to nutrients (Greece) /organic enrichment (Spain) or at increasing distance from the pollution source (Cyprus) – (so not ALL pressures for MED GIG) Baltic Sea GIG (for 4 most northern types) –FI: The median of the 10% highest BBI-values recorded. When S < 9, individuals < 20, method result cannot be high or good –SE: Best available data from areas without local discharges for each type representing at least good ecological status: the upper third of the span exceeding the good–moderate boun- dary was reserved for the status of high –No description available what high status means concerning faunal composition/diversity/abundance Summary on reference conditions in the GIGs

56. Conclusions benthic invertebrate fauna Mentioning time periods of measurements/description of dataset used for derivation of reference conditions still to be done (Difference in) Classification of the species can be documen- ted? As included for MEDOCC and AMBI. Ecotone of good and high status class can be compared? Was explained in scientific paper but not in technical report. Mediterranean Sea GIG: Greece, Cyprus, Slovenia and Spain use maximal species richness of 100-110 and Italy 30-42 ? Illustration of national relationships with physico-chemical parameters established for the Baltic, NEA and MED, not for Black Sea Black Sea does not have sufficient data yet for proper statistical treatment (RO 21 and BU 95 samples, mostly not replicated)

57. COAST Summary - overview of findings Lack of reference sites in many regions of Europe because of human pressures and impacts and focus of monitoring programmes previously centred on polluted areas Gaps of documentation on used reference sites and other issues → The use of an alternative biological benchmark is rather common instead of the exception for coastal waters (already predicted in 2004) No common concept for (alternative) benchmarking but direct ecological characterization of class boundaries has often been made Need to develope common framework / benchmarking Few or no data for some of the biological quality elements Derivation of reference conditions that encompass the full natural variability found within a water body type is likely to take many years → Every six years from 2013, Member States must review reference conditions Not always a clear link between pressure indicators and methods