RIVPACS - Ecological Quality Indices

To enable the biological conditions of completely different types of river sites across a region and between ecoregions to be compared in a standardised manner, there is a need to summarise all the differences between the observed and expected fauna into one or a few classified states. This is achieved in RIVPACS by classifying sites into a small number of biological quality grades or "Ecological Status Classes" as they are referred to in the WFD (Figure 1).

Figure 1. Ecological quality indices and the EU Water Framework Directive ecological status classes

This is done through the intermediate step of calculating Ecological Quality Indices (EQI), defined as the ratio of observed to expected values (O/E) of each biotic index being used in the grading process. The intention of using such ratios is that it provides a means of standardising the biotic indices, so that a particular value of the EQI ratio implies the same ecological quality for that index, no matter what the type of river or stream.

It is largely because of the success of the RIVPACS type approach, using O/E ratios, and its acceptance as a robust tool for standardisation within the scientific freshwater community, that the WFD prescribes the calculation and use of O/E ratios for reporting monitoring results.

Table 1. Lower limits of EQI values of BMWP-ASPT and number of BMWP taxa for each of the UK biological quality grades

The Ecological Status of river sites in the UK based on macroinvertebrates is currently taken as the lower of two grades based on the EQIs for BMWP-ASPT and number of BMWP taxa. This rule, which requires the use of the lower of a number of scores, has implications for assessing the uncertainty or reliability of the overall classification.

It is important to remember that the predicted index value for any site, including the reference sites themselves, is derived from a weighted average of “environmentally similar” reference sites. Therefore, by definition, the average O/E value for the reference sites is forced to be around one, but roughly half of the reference sites will be better than average and have O/E values >1 and roughly half will have O/E values <1 (Figure 2).

Setting the lower limit for "Very Good" grade of EQI to 1.0, as in currently done in the UK, forces half of the RIVPACS reference sites into the "Good" grade or worse. This does not mean they are not good Reference sites, it is merely an artefact of the chosen grade boundaries.

Figure 2. Variation in EQI-ASPT for UK RIVPACS reference sites