Fisheries Research 99 (2009) 226–233 Contents lists available at ScienceDirect Fisheries Research journal homepage: www.elsevier.com/locate/fishres Assemblage level monitoring of stream fishes: The relative efficiency of single-pass vs. double-pass electrofishing Péter Sály a , Tibor Er ˝ os b,∗ , Péter Takács b , András Specziár b , István Kiss a , Péter Bíró b a Szent István University, Department of Zoology and Animal Ecology, Páter K. u. 1., H-2103 Gödöll˝ o, Hungary b Balaton Limnological Research Institute of the Hungarian Academy of Sciences, Klebelsberg K. u. 3., H-8237 Tihany, Hungary article info Article history: Received 15 April 2009 Received in revised form 10 June 2009 Accepted 11 June 2009 Keywords: Fish assemblages Lowland streams Autosimilarity Sampling effort Sample representativeness abstract The relative efficiency of single-pass electrofishing of longer stream sections vs. double-pass electrofishing of shorter reaches was evaluated in small (<5 m wide) wadable lowland streams in the Lake Balaton basin, Hungary. Two hundred meters long stream sections at 8 sites were divided into ten 20m long sampling units each. These units were used to estimate the representativeness of species richness, species compo- sition and relative abundance data at each level of sampling effort (single vs. double-pass, and number of sampling units pooled) using rarefaction and similarity-based approaches. Assemblage variables showed strong response to the length of the stream sampled (number of sampling units pooled). However, no differences were found between the single- and double-pass methods at any level of sample size for any assemblage variable. Estimates of species richness and species occurrence distributions required more sampling effort than estimates of species relative abundances, using any evaluation method. If a proxy estimate of sample representativeness cannot be obtained in the field, a minimum sampling of single- pass electrofishing of 100 m long sections may be necessary even in small wadable streams with low level of habitat and assemblage diversity to get a relatively unbiased picture on assemblage characteristics. © 2009 Elsevier B.V. All rights reserved. 1. Introduction The accurate estimation of assemblage level variables (e.g. species richness and composition, relative abundance) is a funda- mental requirement in environmental monitoring and assessment (e.g. Cao et al., 2003; Kennard et al., 2006; Er ˝ os et al., 2009). How- ever, resources for field surveys are limited, and therefore sampling programmes should usually be optimized in a way that maximizes sample representativeness and minimizes the cost of the sampling (i.e. human effort, time and financial resources). In the monitoring of stream fish assemblages single-pass surveys with electrofishing of variable reach lengths are the most preferred methods, but of course, the method used depends largely on the study objectives (see e.g. Meador et al., 1993, 2003; CEN, 2003; Er ˝ os, 2007; Hughes and Peck, 2008; Reid et al., 2009). Several studies have dealt with the optimal length of the stream reach to be sampled for accurately estimating assemblage variables, and species richness especially (e.g. Lyons, 1992; Angermeier and Smogor, 1995). The general conclusion of these studies was that true ∗ Corresponding author. Tel.: +36 87 448 244; fax: +36 87 448 006. E-mail address: ertib@tres.blki.hu (T. Er ˝ os). values usually cannot be obtained even at extreme efforts, but the sampling of a diverse array of habitat types and ca. 30–40 times of the mean wetted stream width may yield an efficiency that can be used for monitoring purposes, especially if extreme rare species are omitted from the evaluations (Lyons, 1992; Paller, 1995). In Europe, the single-pass electrofishing of 10–20 times the wetted width, but a minimum of 100m long sections was suggested for wadable streams by the FAME Consortium (2004). Much less is known about the relative efficacy of double or multiple-pass removals compared with the reach length optimization based single-pass method. In fact, most studies suggest that single-pass surveys of longer stream reaches are more effective than multiple-pass depletion of shorter sections for assessing reach level assemblage attributes (Paller, 1995; Reynolds et al., 2003; Bertrand et al., 2006; Reid et al., 2009), while others argue for the opposite sampling strategy (Kennard et al., 2006). A drawback of multiple-pass removals is the more intensive mesoscale level perturbation of both the habitat and the biota. Time constraints are also an important factor and it can be easily envisaged that for the same sampling distance multiple-pass based surveys are more time consuming. Therefore, multiple-pass sur- veys may be necessary for estimating population densities if that is the objective of the study (see e.g. Cowx, 1983; Meyer et al., 2006; Penczak and Głowacki, 2008 and references herein), but are not really practical for assemblage level regional scale monitoring pur- 0165-7836/$ – see front matter © 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.fishres.2009.06.010