Fisheries Research 96 (2009) 102–108 Contents lists available at ScienceDirect Fisheries Research journal homepage: www.elsevier.com/locate/fishres Small – but not easy: Evaluation of sampling methods in floodplain lakes including whole-lake sampling P. Jurajda a,∗ , M. Janᡠc a,b , S.M. White c , M. Ondraˇ cková a a Institute of Vertebrate Biology Academy of Sciences CR v.v.i., Kvetna 8, 603 65 Brno, Czech Republic b Department of Botany and Zoology, Masaryk University, Kotlarska 2, 611 37 Brno, Czech Republic c Oregon State University, Department of Fisheries and Wildlife, Oregon Cooperative Fishery Research Unit, 104 Nash Hall, Corvallis, OR 97331, USA article info Keywords: Sampling strategy Floodplain lake Electrofishing Beach seining Fish behaviour abstract We tested the efficiency of electrofishing and beach seining with several sampling strategies in small (<1ha) floodplain borrow pits of the Dyje River (Danube basin). The study sites were characterized as rectangular-shaped with maximum depths of 2m, having uniform habitat with minimal shelters and gravel bottom with organic sediment. We conducted point abundance sampling and continual electrofish- ing along the shoreline and also sampled in a zigzag pattern in the mid-zone from a boat. Sub-sampling and whole-lake sampling was conducted using a beach seine (40 m length, 10 mm mesh size). Each of the four sampling strategies (point sample electrofishing, continuous electrofishing, seining sub-samples, and whole-lake seining) was conducted on separate days in two sampling events in November 2005 and May 2007. A total of 22 species was registered in both sites. Electrofishing was more efficient for determining species richness in both sites compared to seining in the November sampling, but this finding was reversed in May 2007. Some species (chub, weatherfish) were recorded only by electrofishing, while other species (perch, bitterling, tench) were recorded mainly by electrofishing. Asp and blue bream were recorded only by seining. Pelagic species (bleak, bream, white bream) were highly underestimated by electrofishing in both sites and sample periods. The accurate representation of the fish community using just one sampling method and strategy is not feasible even in a small floodplain lake. Whole-lake seining through the entire borrow-pit site was more representative for pelagic species but less efficient for shelter-seeking species. Regarding the ability to capture representative samples, the behaviour of particular fish species seems to be a more significant factor than fish size. © 2008 Elsevier B.V. All rights reserved. 1. Introduction The number and abundance of fish species occurring at a partic- ular site and time are of considerable interest to fish biologists and fisheries managers. Results are often dependent on how sufficiently species richness and composition of fish assemblages are character- ized by sampling (Cao et al., 2001). Many sampling methods have been designed to achieve reliable estimates; however, each sam- pling method has its own shortcomings (Brosse et al., 2001). For example, many studies attempted to find the most representative sampling method for lake habitats and to evaluate the accuracy of electrofishing (e.g. Perrow et al., 1996; Fago, 1998; Cao et al., 2001; Meador, 2005; Lapointe and Corkum, 2006b), beach seining (Lyons, 1986; Fago, 1998; Pierce et al., 1990; Bayley and Herendeen, ∗ Corresponding author. Tel.: +420 543422523; fax: +420 543211346. E-mail address: jurajda@brno.cas.cz (P. Jurajda). 2000; Lapointe and Corkum, 2006a), visual observation (Brosse et al., 2001; MacRae and Jackson, 2006), various traps (Jackson and Harvey, 1997; Lapointe and Corkum, 2006a; MacRae and Jackson, 2006) fyke nets (Weaver, 1993) or gill nets (Weaver, 1993; Jackson and Harvey, 1997). Although there is a wide range of approaches to sampling and the type of data collected, there is little information to guide researchers as to whether the degree of sampling is sufficient to provide good quantitative or even qualitative estimates of fish abundance. Many studies underestimate the species composition within lakes due to insufficient effort and restricted use of sam- pling gears (Jackson and Harvey, 1997). Most studies conclude that no single method describes the full species richness, nor do single methods adequately describe the fish assemblage compo- sition. These problems with estimating fish community structure are not restricted to large lakes; unbiased estimates of fish com- munity structure are problematic even in small standing water bodies (Basler and Schramm, 2006) and in small streams and rivers (Hankin and Reeves, 1988). 0165-7836/$ – see front matter © 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.fishres.2008.09.005