How well do New Zealand's stream‐monitoring indicators, the macroinvertebrate community index and its quantitative variant, correspond?

Abstract

The Macroinvertebrate Community Index (MCI) and its quantitative version (QMCI) were developed to provide measures of organic pollution in streams on the Taranaki Ring Plain, New Zealand. Subsequently, they have been used widely to evaluate “river health” in many running water environments throughout New Zealand, although their suitability as general‐purpose monitoring tools needs evaluation. In the present study, the MCI and QMCI were calculated from fixed‐count macroinvertebrate data for 230 stream and river sites in Canterbury, New Zealand. Values ranged from 60 to 135 (MCI) and 2.0 to 8.8 (QMCI). The two indices ranked sites similarly (rs = 0.86), but the distribution of sites among published degradation bands (Stark 1998) differed. The MCI placed most sites in the “doubtful quality, or possible mild pollution” and “probable moderate pollution” categories, whereas the QMCI assigned most sites to the “clean water” and “probable severe pollution” groups. This implies either that category assessment was more conservative with MCI, or that the boundaries between groups are not equivalent. Since the two indices are alternative procedures with the same purpose, and results need to be transmitted to managers and others, descriptively, we consider this needs rectifying. Proposals for the modification of category boundaries are provided. Further, because the MCI and QMCI are influenced by factors other than water and habitat quality, and these are not well understood, index scores may provide an unrealistically precise measure of stream health in some circumstances. Low‐level identification (order, class, phylum) of fixed‐count data ranked sites in a similar way to conventional taxonomic resolution, but this relationship was stronger for QMCI (rs = 0.93) than for MCI (rs = 0.81).