Multimetric assessment of freshwater macroinvertebrate communities in Flanders, Belgium

One of the objectives of the European Water Framework Directive (WFD; EU, 2000) is to attain a good status for all surface waters in the European Union by the end of 2015. To this end, the ecological status of natural surface waters should be assessed, based on a number of biological quality elements which depend on the category of surface water (rivers, lakes, transitional waters or coastal waters). For each of these elements, member states must choose or develop a classification method, taking into account a set of parameters depending on the quality element and on the surface water category (EU, 2000). The assessment system must be differentiated among the types of water bodies within a category. The method must be in agreement with an Ecological Quality Ratio (EQR) showing relative proportion of the index compared to the reference conditions. This EQR ranges from zero to one, respectively corresponding to a bad and a very good ecological status. This interval is divided into five classes reflecting bad, poor, moderate, good and high ecological status (EU, 2000). For the categories rivers and lakes, one of the relevant biological quality elements is the “benthic invertebrate fauna” (EU, 2000), commonly referred to as macroinvertebrates. For this quality element, the parameters “taxonomic composition and abundance”, “ratio of disturbance sensitive to insensitive taxa” and “diversity” should be taken into account. In Flanders, Belgium, the Belgian Biotic Index (BBI; De Pauw and Vanhooren, 1983), based on macroinvertebrates, has been applied in routine monitoring schemes by the Flemish Environment Agency (VMM) since 1989, confirming the reliability and robustness of this biological quality index. However, with respect to the application of this index for rivers and lakes within a WFD context, not all technical requirements are met by the BBI. The abundance, which is one of the relevant parameters imposed by the WFD, is not taken into account in the BBI calculation, and this index is not explicitly based on a reference condition approach. Furthermore, it is not a type-specific method, in other words, all types of rivers are evaluated by means of the same criteria, and the BBI was intended as an assessment system for watercourses only, not for lakes (De Pauw and Vanhooren, 1983). Two general problems are identified that are associated with taxonomic resolution in water quality assessment based on macroinvertebrates. These are explored by means of analysis of the BBI index calculation method. A first difficulty is caused by possible changes in taxonomy over time, giving rise to inconsistencies in index calculation. A second problem is due to the introduction of exotic species. Both problems can introduce a bias in calculation of the index. To avoid this problem in future assessment methods, it is proposed to use a taxa list to which no changes are made except for the addition of newly introduced exotic species. Since the BBI does not meet all the requirements of the WFD, a new index, the Multimetric Macroinvertebrate Index Flanders (MMIF) for assessing rivers and lakes is proposed. This index is developed using the database of macroinvertebrate samples provided by the VMM. The MMIF is calculated based on macroinvertebrate community data obtained using the same sampling and identification procedure as the BBI. The index calculation is a type-specific multimetric system based on five equally weighted metrics, which are taxa richness, number of Ephemeroptera, Plecoptera and/or Trichoptera taxa, number of other sensitive taxa, the Shannon-Wiener diversity index, and the mean tolerance score. The final index value is expressed as an EQR ranging from zero for bad status to one for high status. The MMIF combines the robustness of the BBI and the long-term experience in Flanders with the flexibility of multimetric indices, while at the same time taking into account the technical requirements of the WFD. To ensure that boundary values for the national biological assessment methods are comparable along all member states in Europe, regional intercalibration exercises were envisaged by the WFD for each quality element and for each category of water body. For river macroinvertebrates, the relevant regional intercalibration exercise for Flanders was coordinated by the so-called Central-Baltic Geographical Intercalibration Group (CB-GIG, 2006). Flanders contributed to this intercalibration exercise, using VMM sampling data, in order to compare the Flemish boundary values to those of the other participating countries and regions and, if necessary, to adjust them to ensure inclusion in the intercalibration decision of the European Commission (EU, 2007). This exercise for river macroinvertebrates essentially consisted in a regression of each participating member states’ national method against a common, generally applicable index. This regression is carried out separately by each member state using a national dataset. Based on the obtained regression equation, each country converts its national boundary values into values on the scale of the common index for subsequent comparison. The mean value of the converted national boundaries is calculated and the rescaled national boundaries should be within a certain range of this mean value in order to be considered as comparable (CB-GIG, 2006). Two major problems emerged with regard to the Flemish contribution. First, the derivation of reference values for the metrics of the common index was not possible using field data because no reference sites are present in Flanders. To overcome this problem, an alternative and transparent way of deriving reference values was used in order to have reference values comparable to those of the other member states (Gabriels, 2007). The second problem concerned the boundary values, which clearly fell below the comparability interval. Therefore, alternative boundary values were proposed. This adjustment of the MMIF class boundaries for rivers was formally accepted by the steering group of the intercalibration exercise (CB-GIG, 2007). The adjusted values were adopted in the draft version of the intercalibration decision of the European Commission (EU, 2007), of which the official publication is foreseen for the end of 2007