1. Results of the Intercalibration in the ALPINE RIVER GIG
MACRO
ZOOBENTHOS

2. Introduction: Common Intercalibration Type: Alpine Rivers
From: ECOSTAT WG 2.A, 2004: Overview of common Intercalibration types

3. HYBRID OPTION Option 1: Common WFD Assessment method
Introduction: The 3 Options for the Intercalibration Process
Option 1: Common WFD Assessment method
Option 2: Use of a common metric(s) method identified specifically for the purposes of the intercalibration exercise
Option 3: Direct comparison of national methods at intercalibration sites
HYBRID OPTION
From: ECOSTAT WG 2.A, 2004: Guidance on the Intercalibration Process

4. Introduction: The 3 Options for the Intercalibration Process
HYBRID OPTION
establish boundary values with national assessment methods (as in Option 3)
subsequent comparison of boundary values with common metrics method (as in Option 2).
Comparison: ICMialpine (qualitative, quantitative)
Harmonisation: Median boundary value

5. ICMi - qualitativeAlpine GIG
Introduction: METHODS
ICMi - qualitativeAlpine GIG
KEY METRICS
Indivative for:
total number of taxa
number of EPT taxa
number of selected (sensitive) taxa
taxa richness
general degradation
habitat degradation
ASPTIberian
Organic pollution
CALCULATION:
simple average of EQRs

6. ICMi - quantitativeAlpine GIG
Introduction: METHODS
ICMi - quantitativeAlpine GIG
Type
Metric
Weight
Tolerance
ASPTIberian - 2
0,333
Abundance/Habitat
Log 10
(sel_sens_taxa)
Log (sum abundance of selected sensitive families)
0,266
RETI
Rithron feeding type index
0,067
Richness & Diversity
Total number of taxa
0,167
Number of EPT-taxa
0,083
Shannon-Wiener Diversity Index
Calculation of ICM: weighted sum

7. QE Macroinvertebrates
National classification methods
QE Macroinvertebrates
Assessment Method
Status (choose A-F)
AUSTRIA
Multimetric Indices for General Degradation (Structural Diversity, nutrients,…), Saprobic Index (Moog, O. & Ofenböck, T.(2006), in print) B
FRANCE
IBGN Indice Biologique Global Normalisé), AFNOR NF T , Includes Number of taxa and sensitive taxa. A
Agreed National method and classification (but revision planned to meet WFD requirements)
GERMANY
Handbuch zur Untersuchung und Bewertung von Fließgewässern auf der Basis des Makrozoobenthos vor dem Hintergrund der EG-WRRL, April 2005, C
Finalized: National method in verification
ITALY
IBE – Indice Biotico Esteso (Extended Biotic Index). A minimum number of individuals is necessary for each SU to be counted. The method foresees a combination of two different metrics: total number of SU and presence/absent of the most sensitive taxa (in order Plecoptera, Ephemeroptera, Tricoptera…). The Quality Class is related to the IBE are 5 from high quality to bad quality.
Agreed national method. Compulsory by law and applied by all environment offices
SLOVENIA
Multimetric index (Hydromorphology), Saprobic Index D
In development
SPAIN
IBMWP-Iberian BMWP
Frequently used method

8. Number of reference sites
Reference conditions
The selection of reference sites was based on common criteria (see Annex B).
The reference value was calculated by using the median of reference sites
Number of sites/coutry:
Country
Number of reference sites
R-A1
R-A2
Austria 7
France alpine 4 21
France pyrenean - 16
Germany 2
Italy 14
Slovenia 5
Spain
12*

9. ICMi: Minimum Quality Criteria
Introduction: Quality checks
ICMi: Minimum Quality Criteria
minimum : 20 sites covering widest range of quality classes
reference state compliant to the REFCOND guidance
Pearson R²: national index vs. ICMi: >=0.64 (at a=0,05)

10. ICMi: Quality Criteria
Introduction: METHODS
ICMi: Quality Criteria

11. ICMi: Standardization of Calculation
Introduction: METHODS
ICMi: Standardization of Calculation
4 metrics (qual. ICM) / 6 metrics (quant. ICM)
reference value: median of reference sites
EQR for every value = value / reference
ICM: average (qual.) / weighed sum (quant.)
Regression of EQRnational vs EQRICM
Regression formula, R²
Transformation of national EQR boundary values into EQR ICM - values

12. Key metrics and ICMi: Example SPAIN
Introduction: METHODS
Key metrics and ICMi: Example SPAIN

13. ICM – causes of variation:
Results
ICM – causes of variation:
ICM is a simplifying approach
linear relationship is a simplification
less accuracy / confidence than nat. methods
typology
ICM typology coarse – more simple than nat. types
“problems” with min. qual. criteria
new methods – lack of data and experience
quality of our streams is “to good”
conclusion: “accepted variation” of boundary values
instead of fixed value is necessary

14. Acceptable range of variation
As a consequence the Alpine GIG suggests to use an „acceptable range of variation“ rather than a fixed value alone.
As value for this „acceptable range of variation“ the GIG proposes ¼ of the median status class width of the participating member states.

15. Intercalibration procedure
Results
Of the ICM
Intercalibration procedure
QUALITATIVE
ICM

16. Results R-A1: calcareous type

17. Results

18. Discriminatory power of the ICMi class boundaries
Results
Discriminatory power of the ICMi class boundaries
National ranges of status classes expressed with ICMi values. Boxplots: 25th percentile – median - 75th percentile. Whiskers above and below the box indicate the 90th and 10th percentiles.

19. Results

20. Discriminatory power of the ICMi class boundaries
Results
Discriminatory power of the ICMi class boundaries
National ranges of status classes expressed with ICMi values. Boxplots: 25th percentile – median - 75th percentile. Whiskers above and below the box indicate the 90th and 10th percentiles.

21. Results

22. Intercalibration procedure
Results
Of the ICM
Intercalibration procedure
QUANTITATIVE
ICM

23. Correlation ICMiqualitative - ICMiquantitative
for member states with available quantitative data.

24. Conclusion CONCLUSION ICM is a proper instrument for comparison
Quantitative and qualitative ICM show comparable results
“accepted range of variation” is necessary due to several sources of variation
Boundaries seem to be in an acceptable range, especially for the type R-A2
In some cases slight changes of national boundary values are probably necessary (IT, ES)
Data quality and quantity will enhance in future with monitoring programmes – increasing the scientific accuracy of the intercalibration

25. Expected time of finalisation for QE
Expected finalisation date
Main open issues
Macroinvertebrates
July 2006 
Further improvement of accuracy and precision with future monitoring data
Phytobenthos
2007
Data available but organisation and methodology not clear yet (follow approach of CBGIG)
Macrophytes -
Macrophytes are not considered relevant for alpine streams

26. Thanks to all of you!!! Alpine GIG people
(Homo alpinus var. Intercalibrensis)
Thanks to all of you!!!

27. Background of this proposal:
The results of the assessment methods are subject to several sources of variation. Thus the status assessment is somehow more significant in the middle of a status class than compared to the transitional zone to the neighbouring status classes. This “insecure” zone of assessment is assumed to be ¼ of the status class width (more detailed estimates of accuracy and precision is lacking in most countries at the moment).

28. As an alternative approach the bands of “accepted variation” were also calculated as median confidence limits of all member states
From the statistical point of view this approach does not seem to be appropriate, as the confidence limits are decreasing with increasing data quality, but the sources of variation if the results are still present