A Rapid Assessment Procedure for the Enumeration of Salmonine Populations in Streams

Abstract
Population enumeration is a key component of fisheries investigations for riverine salmonines. We examined the reliability of a rapid population assessment technique for stream salmonines that depends on a single episode of electrofishing rather than traditional multiple fishing, removal, or mark–recapture methods. We show that for 12 sites sampled in 1992 on Wilmot Creek, a small, coldwater tributary to Lake Ontario, the catch of salmonines from a single electrofishing episode predicted the population estimate obtained with a more time-consuming multiple-pass removal method. When we collected similar data from eight additional Lake Ontario tributary sites in 1994, the relationship was not significantly different from that obtained in 1992. We also conducted nonparametric analyses of covariance on subsets of these data and found no significant differences in the above relationship for rainbow trout Oncorhynchus mykiss versus brown trout Salmo trutta or for age-0 versus older fish, although the statistical power of the species comparison was low because of the small number of samples containing brown trout. The regression model developed from the pooled 1992 and 1994 data predicts population sizes that only slightly exceed estimates obtained with the removal method at 12 other sites throughout southern Ontario. We use the rapid assessment technique to obtain rainbow and brown trout population estimates with measurable precision for an entire catchment and show that the error in our site-specific estimator of population size is small relative to among-site sampling error, even though our sampling fraction is relatively large compared to that of typical surveys. Finally, we provide suggestions to potential users of the proposed methodology regarding data collection and analytical techniques.

This publication has 0 references indexed in Scilit: