Does variation in sampling effort during a long term monitoring project matter (Lees Ferry, Colorado River)?

Authors: David L. Rogowski

Year: 2013 (xlv)

Abstract

With the construction of Glenn Canyon Dam the Colorado River immediately downstream of the dam was severely altered and since 1964 the Glen Canyon reach (aka: Lees Ferry reach) has been managed by the Arizona Game and Fish Department as a nonnative Blue Ribbon rainbow trout (Oncorhynchus mykiss.) fishery. Creel surveys and electrofishing have been the mainstays of evaluating this fishery. Current standard protocols for electrofishing have been in place since 2002, and sites were selected using a stratified random sampling approach taking into account shoreline habitat characteristics (cliff ledge, sand bars, debris fans, cobble bar, and talus) and river reach (upper, middle, lower). For standard sampling this stretch of the river was dived into 360 separate sampling sections (180 per river side) these varied in length from 131-259 m. From June 2002 through October 2012, 1085 sites were sampled. As these electrofishing sites are fixed with starting and end points identified via a combination of GPS and orthophotographs, sampling efforts (time/site length) for a particular site should remain relatively stable accounting for differences in flow. Preliminary results show that there was a difference in sampling effort by habitat, season (spring, summer and fall) and flow. Significantly less time per meter was spent electrofishing cobble bars and cliff ledges, as well as in summer. The number of rainbow trout captured (#fish/time/site length) differed via habitat, and season, with a significant habitat by season interaction. Surprisingly flow was not a significant factor in the number of fish captured. Capture rates for rainbow trout were higher for talus habitats and the least in debris fan habitats.