Comparison of Gears for Sampling Littoral-Zone Fishes in Floodplain Lakes of the Lower White River, Arkansas

Abstract

Gear comparison studies are useful to fisheries managers because many aquatic systems require multiple gears to assess fish assemblages. Many previous studies have emphasized comparisons of catch per unit effort or basic community measures such as richness and diversity. Our objectives were to (1) compare fish assemblage richness, diversity, and evenness across sampling gears, (2) assess the similarity of fish assemblage structure as depicted by the different gears, and (3) compare fish assemblage–environment relationships depicted by the different gears. Assessments were conducted in littoral-zone areas of 16 floodplain lakes in the lower White River, Arkansas. Four sampling gears were used: 500-V, 60-Hz (high-pulse) and 1,000-V, 15-Hz (low-pulse) boat-mounted nighttime electrofishing, mini-fyke nets, and experimental gill nets. Species richness and diversity were significantly greater using boat-mounted electrofishing (both pulses) than the other gears. Experimental gill nets exhibited the lowest richness of all gears (<50% of the other gears), and mini-fyke nets exhibited significantly lower evenness owing to the tendency for 1–2 species to dominate catches. Procrustean analyses indicated that the lake-specific fish assemblages depicted by the different gears were significantly different during all paired-gear comparisons except between the two electrofishing configurations. Regardless of gear, multivariate direct gradient analyses indicated that lake depth, water clarity, and morphoedaphic index were consistently important variables associated with the structuring of littoral-zone fish assemblages. The results support the idea that multiple-gear approaches are useful in assessing floodplain lake fish assemblages and that, in the case of lower White River floodplain lakes, the assemblage information collected from the different gears was not redundant. Additionally, if a large-river sampling program is using only one or two gears, major fish–environment relationships that are depicted may not be gear dependent.