Use of Alternative Classifications in Studying Broad-Scale Distributional Patterns of Lotic Invertebrates

Multivariate analyses were used to examine relationships between each of three biological classifications of mayfly (Ephemeroptera) larvae and environmental variables recorded at 95 river sites encompassing 18 degrees of latitude in northwestern North America. Biological samples, collected by kick net, were classified by family and superfamily, by size, and by morphological/behavioural group. Environmental variables included measures obtained from topographical maps and hydrological features recorded at each site. Results of canonical correlation analyses showed that all three biotic indicators were significantly associated with environmental variables. Morphological/behavioural classification yielded stronger correlations with environmental data than did taxonomic or size classifications. Multivariate curvilinear regression analyses gave equivalent results and showed that, although several relationships were curvilinear, all were monotonic. Environmental variables independently and significantly associated with mayfly relative abundance included river size, velocity, slope, latitude, and conductivity. Use of alternative classification variables can potentially reveal patterns not evident when standard taxonomic units are applied.