Seasonal variability in the Macroinvertebrate Community Index: Are seasonal correction factors required?

Abstract

Freshwater macroinvertebrate data collected from streams in Taranaki, New Zealand (1981–2006) were used to examine seasonality of biological indices including taxon richness, %EPT (Ephemeroptera, Plecoptera, and Trichoptera) richness, and the hard‐ and soft‐bottomed (‐sb) stream versions of the Macroinvertebrate Community Index (MCI). All indices tested showed modest but statistically significant seasonal variation in hard‐bottomed streams. Seasonal means for the MCI were within ±3.0% of the annual mean. The Semi‐Quantitative MCI (SQMCI) (±4.3%), %EPT richness (±7.4%) and taxon richness (±7.7%) showed greater seasonal variability than the MCI. MCI, SQMCI, and %EPT richness were significantly higher in spring and winter compared with autumn and summer. Taxon richness was lower in winter than in other seasons. In soft‐bottomed streams, SQMCI‐sb showed least seasonality (within ±3.6% of the annual mean, nonsignificant), followed by MCI‐sb and taxon richness (±4.7%), and %EPT richness (±11.2%). Spring and winter MCI‐sb and SQMCI‐sb values tended to be significantly higher than those calculated from summer or autumn samples, although most values were within 5% of the annual site means. Examination of seasonal variation in species traits and their relationships to observed seasonal patterns of biotic indices did not support the suggestion that seasonal variations in life histories of aquatic macroinvertebrates affect index values. Rather, seasons with higher biotic index values were associated with a greater frequency of flow disturbance, which is consistent with the view that the character of New Zealand's stream macroinvertebrate communities with their poorly‐synchronised life histories, has been shaped by the unpredictable physical environment. We recommend the MCI and MCI‐sb for State of Environment reporting in New Zealand for cost‐effectiveness and because data requirements and seasonal variation were less than for quantitative alternative indices. We suggest that seasonal variability is unlikely to confound interpretation, and does not need to be considered.