Evaluation of low impact pressure waves as a quantitative sampling method for small fish in shallow water Martin Snickars a, ⁎ , Alfred Sandström b , Antti Lappalainen c , Johanna Mattila a a Husö Biological Station, Department of Biology, Åbo Akademi University, Akademigatan 1, FI-20500 Turku, Finland b Swedish Board of Fisheries, Institutes of Coastal and Freshwater Research, Stångholmsvägen 2, SE-17893 Drottningholm, Sweden c Finnish Game and Fisheries Research Institute Viikinkaari 4, FI-00791 Helsinki, Finland Received 8 June 2006; received in revised form 2 November 2006; accepted 13 December 2006 Abstract Low impact pressure waves (LIMP) generated by small (b 1 g) underwater ignition capsules have been used to qualitatively and semi-quantitatively, as catch per unit effort, sample young-of-the-year (YOY) and juvenile fish in shallow (0–3 m) waters. In order to evaluate this method on small fish, four categories, 0+ and 1+ year Eurasian perch, 1+ year roach and 0+ to 1+ year pike, were subjected to the LIMP at distances of 0.5–5 m from the detonation point. At a distance b 1 m all specimens were stunned and either floated to the surface or sank to the bottom, whereas almost all were unaffected at distances N 3 m. The efficiency radius of the LIMP varied between 1.5 and 3.0 m, and the effective area was 7–28 m 2 depending on habitat and species. The average effective area for the two extensively studied categories (0+ perch and 1+ roach) was 17.9 m 2 . Both depth and vegetation affected the effective area. The smallest area was obtained at b 1 m depth with vegetation present, and the largest at 1–2 m depths with no vegetation present. Within each fish category, the method sampled fish independently of length suggesting that a size/year-class is sampled equally. The proportion of affected fish that sank was large and species-dependent. In addition to the netting of surfaced fish, collecting fish from the bottom is recommended for achieving an accurate quantitative estimate. The tested parameters of the LIMP, P max and Impulse, declined as a function of distance following a negative exponential growth model with a baseline. The similar-shaped responses among the fish categories suggest that the method works at depths b 3 m also in heterogeneous environments, and that the method can be used to quantitatively estimate abundance of small fish with swim bladders. A model for quantitative abundance estimates based on the evaluation is presented. © 2007 Elsevier B.V. All rights reserved. Keywords: Esox lucius L.; Fish sampling; Method evaluation; Perca fluviatilis L.; Rutilus rutilus (L.); Swim bladder 1. Introduction Precise information on abundance and presence/ absence is necessary in many type of studies on fish assemblages. For example, it is important to get accurate and unbiased information on the early life-stages of fish, since recruitment often is a key mechanism in fish production. Abundance studies of small fish in shallow brackish or marine waters have traditionally been conducted with a variety of active sampling methods such as beach seines, small trawls and drop-samplers, as well as visual surveys (Mankki and Vauras, 1974; Thorman and Wiederholm, 1983; Aneer et al., 1992; Brosse et al., 2001). These methods are, however, less suitable in several types of habitats. Particularly shallow Journal of Experimental Marine Biology and Ecology 343 (2007) 138 – 147 www.elsevier.com/locate/jembe ⁎ Corresponding author. Tel.: +358 2 215 3426; fax: +358 2 215 3428. E-mail address: masnicka@abo.fi (M. Snickars). 0022-0981/$ - see front matter © 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.jembe.2006.12.008