Dynamics of Drift and Microdistribution of Two Mayfly Populations: A Predictive Model

Abstract

A model is derived to predict changes in larval microdistribution of Ephemerella inermis Eaton and Baetis tricaudatus Dodds (Ephemeroptera) that result from drift. Transport by drift was assumed to be a Markov process. Estimates of departure from the substrate, downstream movement, and lateral transport were combined to produce transition matrices from which changes in microdistribution, magnitude of drift, and mean daily downstream displacement of live and dead animals were determined. Predictions from the model were compared with observations made in a river similar in composition to the area modelled. There were marked differences between expected drift of living versus dead animals; estimates for drift of living animals corresponded most closely to values observed in the field. Predicted drift either corresponded to or was less than observed levels. Observed changes in benthic microdistribution were minimal, but the model predicted large-scale relocation of animals toward river margins. Thus, magnitude and spatial pattern of drift could be adequately modelled but benthic distribution could not. Drift is an important redistribution mechanism but animals must have a complementary ability to actively travel towards the river center to maintain constant benthic distribution patterns. Inhibition of movement during floods should produce increased abundance at margins. During stable flow conditions, drift and benthic activity probably act together to produce and maintain microdistributional patterns among vagile organisms. In contrast, drift of invertebrates possessing poor locomotory ability probably reflects larger-scale, permanent displacement events that are a normal aspect of development, or the outcome of aggressive interactions.