Download presentation
Presentation is loading. Please wait.
1
Angel Borja Coordinator of the Group
NEA-GIG Macroinvertebrates group (ECOSTAT, Brussels, October 2011) Intercalibration of transitional water macroinvertebrates within the NEA-GIG Angel Borja Coordinator of the Group
2
Introduction Steps in IC Types Data Intercalibration Second Phase of the intercalibration ( ) All countries, having transitional waters, are represented in the working group Active members Apologies Non-participant Absent, since 2010
3
Introduction Steps in IC Types Data Intercalibration The following steps for intercalibration were agreed upon by this group: (i) to establish common water body types across Europe; (ii) to compile a common dataset; (iii) to harmonise the taxonomy of the dataset; (iv) to collate human pressures from each estuary; (v) to set reference conditions for each type; (vi) to calculate Ecological Quality Ratios for each of the methods proposed for IC; (vii) to interpret the response of these methods to different anthropogenic pressures; (viii) to determine boundaries for each of the 5 quality classes (from bad to high status), using all the selected methods; and (ix) final agreement in the assessment and intercalibration.
4
Initial National Types collated
Introduction Steps in IC Types Data Intercalibration Initial National Types collated
5
Common NEA Types proposed
Introduction Steps in IC Types Data Intercalibration Common NEA Types proposed Data obtained for types C to F Types C and D merged for IC
6
From 59 estuaries under different pressure conditions
Introduction Steps in IC Types Data Intercalibration Samples available From 59 estuaries under different pressure conditions
7
Harmonisation using UK lists, ERMS and WoRMS
Introduction Steps in IC Types Data Intercalibration Samples available Taxonomy Harmonisation using UK lists, ERMS and WoRMS Oligohaline taxa to be checked against register of freshwater list ( Harmonised taxon list includes Benthic Quality Index (BQI) sensitivity scores and AMBI ecological group for each species A total of 1939 harmonised taxa for NEA estuaries Single metrics calculated by sample and sieve size. Then harmonised to a unique sample size (0.1 m2 in subtidal, 0.01 m2 in intertidal). The harmonisation includes also the sieve size (0.5 mm and 1 mm)
8
Methods proposed & pressures
Introduction Steps in IC Types Data Intercalibration Samples available Taxonomy Methods proposed & pressures
9
Methods proposed & pressures
Introduction Steps in IC Types Data Intercalibration Samples available Taxonomy Methods proposed & pressures Pressures not available for all samples and/or water bodies Same pressure value for different samples and years, within the same water body
10
Excluding pressure data Including pressure data
Introduction Steps in IC Types Data Intercalibration Samples available Taxonomy Methods proposed & pressures Response of benthos to pressures: Multivariate analysis between pressures and benthic abundance (BioEnv) to test the effect of: i) Environmental data alone (salinity, grain size, coordinates) Ii) Environmental data + NEAGIG pressure data Results (Spearman correlation): Benthos dominated by natural conditions Weak correlation to pressure data (<0.1 increase) Complex multi-pressure systems (pressures difficult to quantify) Problematic for setting reliable benchmarks on pressure data Data set Excluding pressure data Including pressure data Subtidal 0.5mm 0.507 0.598 Subtidal 1mm 0.447 0.546
11
Methods proposed & pressures
Introduction Steps in IC Types Data Intercalibration Samples available Taxonomy Methods proposed & pressures
12
Partial reference sites,
Introduction Steps in IC Types Data Intercalibration Samples available Taxonomy Methods proposed & pressures Reference conditions Partial reference sites, Expert judgment (to define ‘virtual’ reference conditions) 95th percentile (referred to habitats: salinity, sediment type, intertidal/subtidal) UK: metric reference condition values adapted in response to environmental gradients (salinity and grain size) Reference conditions were derived for most of the ecotopes and all types for M-AMBI, BAT and TaSBeM.
13
Case 1: option 3 Not common level of benchmarks for all countries
Introduction Steps in IC Types Data Intercalibration Case 1: option 3 Not common level of benchmarks for all countries Inssufficient number of benchmark for all countries High number of samples without EQR for several methods
14
Hence, we decided to use another approach
Introduction Steps in IC Types Data Intercalibration Case 1: option 3 Some conclusions: Tasbem has to be excluded since the slope was (impossible to reduce the bias) AeTV was excluded since the correlation was negative with the two methods for which there is coincidence (M-AMBI, BEQI) The IC results were not satisfactory, because of the absence of good benchmark sites, benchmark sites for all countries and low number of sanples with EQR for all methods Hence, we decided to use another approach
15
Case 2 First phase approach
Introduction Steps in IC Types Data Intercalibration Case 1: option 3 Case 2 First phase approach
16
Case 2 First phase approach
Introduction Steps in IC Types Data Intercalibration Case 1: option 3 Case 2 First phase approach
17
Case 2 First phase approach
Introduction Steps in IC Types Data Intercalibration Case 1: option 3 Case 2 First phase approach TYPE D AETV excluded (negative correlation) M-AMBI and IQI to be adjusted TYPE E QSB to be adjusted TYPE F BQI excluded (low correlations with other methods) M-AMBI to be adjusted
18
Case 2 First phase approach
Introduction Steps in IC Types Data Intercalibration Case 1: option 3 Case 2 First phase approach TYPE D TYPE E TYPE F
19
Case 2 First phase approach
Introduction Steps in IC Types Data Intercalibration Case 1: option 3 Case 2 First phase approach TYPE D: sample level TYPE D: WB level
20
Some conclusions: Case 1: option 3 Case 2 First phase approach
Introduction Steps in IC Types Data Intercalibration Case 1: option 3 Case 2 First phase approach Some conclusions: Some pairwise analyses have problems, probably due to the different approach in deriving these methods, together with the high variability of the transitional waters At sample level we obtained more consistent results than at WB level, probably because of the number of samples (>2500 and 100, respectively) After testing IQI with M-AMBI reference conditions, kappa values increased significantly There is an increased variability resulting from different approaches to setting reference conditions by each MS and the effect on the final correlations (we expected increased correlation if same reference condition methods used). Regarding the pressure-response analysis, the approach used does not rely on the highly variable pressure data. Test if the boundaries are correct in terms of pressures and if they are dependent or not on habitats. IC also at the water body level, improving the integration system for each method and using more data. Solve the differences in the boundaries for some methods when IC at sample or water body level In general, we consider the results satisfactory, especially taking into account the many problems faced
21
Thanks for your attention!
Angel Borja
Similar presentations
© 2024 SlidePlayer.com. Inc.
All rights reserved.