Microhabitat–Preference Curves of Blackfly Larvae (Diptera: Simuliidae): A Comparison of Three Estimation Methods

Abstract

Three analytical techniques are available to describe the observed response of stream-dwelling organisms to variations in their physicochemical environment: the incremental method, polynomial regression on a single factor, and multiple regression. The efficacy of these tools as descriptors of the responses has not been compared. We used the three methods to describe density and microhabitat–preference curves for Prosimulium mixtum/fuscum, Stegopterna mutata, and Simulium aureum in response to distance from the lake, current velocity, and water depth in a stream draining a Laurentian lake. The incremental method yielded the least precise estimates of density and biased estimates of optimal current velocity for two of the three species; multiple regression yielded the most precise estimates of density and unbiased estimates of optimal conditions, whereas polynomial regression on a single factor was intermediate. From this comparison, we suggest that the multiple-regression approach to estimate microhabitat–preference curves be used in developing optimal management stategies. Further, we suggest that rare species should be excluded from such analysis because of the low precision of density estimates for rare organisms.