Date of this Version
The Prairie Naturalist 52:21-30; 2020
Understanding limitations of larval fish capture gears is critical for developing appropriate sampling protocols and interpreting catch data. We evaluated genera richness, genera diversity, assemblage similarities, abundance indices (i.e., density or catch per unit effort [CPUE]), and sample size requirements between a surface slednet and glow-stick light traps used in 2014 and 2015 and a benthic slednet and light-emitting diode light (LED) traps used in 2015 in the Minnesota River. The surface slednet captured the greatest number of larval fish genera (15) while the LED light trap captured the fewest (1). Similarities of assemblages sampled was highest between surface and benthic slednets (58%) and lowest between the benthic slednet and LED light trap (0%). All evaluated gears had low and variable catch rates; the highest variability was observed for the LED light trap (CV = 800), and the lowest variability was observed for surface slednets (CV = 173). Slednets required less effort to detect a 25% change in total larval fish abundance compared to light traps. Low CPUEs or densities were possibly the result of suspended sediment loads (85.3 ± 8.5 Nephelometric Turbidity Units) that blocked light trap entrance slots and clogged net pores. Further, not targeting habitats critical to adult spawning and larval rearing (e.g., log jams or shallower or inside bends of meanders) may have influenced CPUEs and densities. We recommend modifications to evaluated sampling gears (e.g., nets with larger mesh sizes) or the evaluation of additional larval fish sampling methods (e.g., larval seines or pumps) coupled with a stratified random sampling protocol that incorporates complex habitats for sampling larval fish within the main channel of the Minnesota River or other river systems with similar high turbidity levels.