Efficiency of a Quadrat Sampling Technique for Estimating Riffle-Dwelling Crayfish Density ERIC R. LARSON 1 Arkansas Cooperative Fish and Wildlife Research Unit, Department of Biological Sciences, University of Arkansas, Fayetteville, Arkansas 72701, USA ROBERT J. DISTEFANO Missouri Department of Conservation, Resource Science Center, 1110 South College Avenue, Columbia, Missouri 65201, USA DANIEL D. MAGOULICK* U.S. Geological Survey, Arkansas Cooperative Fish and Wildlife Research Unit, Department of Biological Sciences, University of Arkansas, Fayetteville, Arkansas 72701, USA JACOB T. WESTHOFF 2 Department of Biological Sciences, Tennessee Technological University, Cookeville, Tennessee 38505, USA Abstract.—Quantitative sampling of crayfish can be challenging as a result of clustered distributions produced by a variety of factors. We tested the efficiency of a 1-m 2 quadrat sampler for estimating the density of riffle-dwelling crayfish. Sampling efficiency was evaluated in an Ozark Plateau stream through a mark– recapture study. We stocked three marked crayfish per square meter into 12 enclosed riffle sections (range, 46–53 m 2 ) and then randomly sampled five 1-m 2 quadrats in each of the 12 sections. Sampling efficiency was determined by comparing the estimates of crayfish density to the known density of marked crayfish. We found crayfish sampling efficiency (mean, 69%) to be comparable to that in studies evaluating the quadrat sampler for stream fish, although efficiency across our study riffles ranged widely, from large underestimates (27%) to positively biased overestimates (140%). Coefficients of variation revealed that replicating habitat units within a stream or stream reach may provide more precise crayfish density estimates than replicating quadrat samples within a single habitat unit. None of the physical habitat or water chemistry variables that we measured were correlated with sampling efficiency. A power analysis of confidence interval (CI) precision demonstrated that moderate precision (i.e., CI half-width ¼ 20% of mean) was achievable at a reasonable sampling effort (40 quadrats). The 1-m 2 quadrat sampler can be effective for quantitatively sampling lotic crayfish. This information should benefit researchers and managers interested in estimates of crayfish density in streams. Crayfish are keystone species or ecological domi- nants (Simberloff 1998) in many lotic ecosystems, where they are prey for aquatic and terrestrial predators, process organic matter, modify the physical environment, and strongly influence the benthic community as consumers of algae and predators of other invertebrates (Creed 1994; Momot 1995; Whit- ledge and Rabeni 1997; Statzner et al. 2003; DiStefano 2005). Additionally, crayfish are among the most imperiled North American taxa (Taylor et al. 2007) and have also been used as bioindicators for metal contamination and in water quality bioassessments (Sylvestre et al. 2002; Besser et al. 2007; Allert et al., in press). Despite their recognized importance, crayfish remain difficult to sample quantitatively owing to clustered distributions produced by habitat preferences, behavioral responses to predation, and environment- mediated activity levels (Stein and Magnuson 1976; Rabeni 1985; Somers and Stechey 1986). Sampling methods used to obtain quantitative estimates of stream crayfish populations or communities by themselves or in combination with other methods include electrofish- ing, quadrat sampling, 1-m 2 seines, and swimming or underwater (scuba) direct observation along transects. However, none of these methods produces reliable estimates in all situations (Westman et al. 1978; Rabeni et al. 1997; Williams 2006), and there are no published studies that evaluate the crayfish sampling efficiency of these methods. Sampling efficiency is the ratio of individuals * Corresponding author: danmag@uark.edu 1 Present address: School of Aquatic and Fishery Sciences, University of Washington, Seattle, Washington 98195, USA. 2 Present address: Missouri Cooperative Fish and Wildlife Research Unit, University of Missouri, Columbia, Missouri 65211, USA. Received February 14, 2007; accepted January 11, 2008 Published online August 18, 2008 1036 North American Journal of Fisheries Management 28:1036–1043, 2008 Ó Copyright by the American Fisheries Society 2008 DOI: 10.1577/M07-029.1 [Article]