Abstract
Principal component analysis is used to develop a technique for predicting the limnological characteristics of a lake from knowledge of its midsummer limnetic crustacean zooplankton community. Patterns of variation in zooplankton community structure are summarized in the principal components extracted, via the species covariance matrix, from the matrix of transformed proportionate numerical abundances of species in a sample of lakes. The relation between these patterns and the limnological characteristics of the lakes is determined from the first-order rank correlations of the limnological variables with the components. In relatively undisturbed lakes of northwestern Ontario, results indicate that lakes characterized by Cyclops bicuspidatus thomasi, Diaptomus minutus, and others are large and clear, whereas those characterized by Tropocyclops prasinus mexicanus, D. oregonensis, and others are smaller and of lower water clarity. These patterns are contrasted with those extracted from the same data using more subjective techniques. In the acid-stressed Killarney region, Ontario, lakes dominated by D. minutus are predicted to be acidic, small, and clear, whereas those dominated by Diaphanosoma leuchtenbergianum, Bosmina longirostris, Mesocyclops edax, and others will be less clear, larger, and with higher pH. In Haliburton lakes, Ontario, those with Diaptomus oregonensis, M. edax, and Ceriodaphnia lacustris are predicted to be small and productive, whereas those with D. minutus, D. sicilis, B. longirostris, and Daphnia dubia will be larger and less productive. The usefulness of the technique and some modifications are discussed. Key words: zooplankton communities, principal component analysis, ELA lakes, Haliburton lakes, Killarney lakes

This publication has 2 references indexed in Scilit: