A SQUARE-METER ELECTROFISHING SAMPLER FOR BENTHIC RIFFLE FISHES

We describe the design and field-testing of a pre-positioned 1-m2 elec trofishing sampler with a novel electrode configuration. The sampler p roduces relatively uniform voltage gradients (0.36-7.00 mean volts/cm: water conductivity = 100 muMHO, 153 VAC, 600 W) that are sufficient t o stun fish and are largely restricted to the sampling frame. Voltage gradients averaged 0.1 mean volts/cm 10 cm outside the frame regardles s of water conductivity and declined exponentially with increasing dis tance from the frame. We compared the effectiveness of our sampler wit h adjacent kick-seine samples in the upper, middle, and lower regions of a riffle in Green River, Kentucky, using stratified random sampling . Species rank concordances were statistically similar regardless of m ethod or habitat; both methods yielded similar information on benthic species composition. Statistically similar estimates of benthic fish a bundance and number of benthic fish species were derived regardless of habitat or method. The frame sampler was more precise than kick-seini ng for estimates of fish density only in the upper riffle, a habitat c haracterized by relatively shallow water and small, more uniform subst rate than found in other sampled habitats. Fewer frame samples (9) tha n kick-seine samples (17) were required to characterize species richne ss of the benthic fish fauna. In each habitat, more benthic species we re collected using electrofishers than with kick-seines; Shannon's spe cies diversity and evenness were larger when estimated from electrofis hing samples than kick-seine samples in each habitat. We conclude that our sampler allows repeatable, equal-effort sampling of benthic speci es from well-defined sampling areas, small enough to allow fine-graine d measurement of microhabitat. Importantly, this is achieved without p roducing voltage gradients strong enough to bias adjacent samples.