Differences in Native and Invasive Fish Larval Occupancy and Detection Probabilities in the Upper Mississippi River

ABSTRACT The Upper Mississippi River (UMR) represents one invasion front for bigheaded carps Hypopthalmichthys spp. in North America. Agencies often sample larvae to determine the conditions under which bigheaded carp reproduction occurs, but the ability of conventional ichthyoplankton sampling to detect bigheaded carp reproduction compared to native fishes is unknown. We used occupancy models to estimate larval bigheaded carp and native fish larvae detection probabilities and assessed how habitat type, Julian date, river discharge, and water temperature affect occupancy. We sampled larvae in Pools 18–20 of the UMR every 10 days at the Iowa, Skunk, and Des Moines River confluences in backwater, side channel, and thalweg habitats. Detection probabilities increased with water volume filtered and were lower for larval bigheaded carp than freshwater drum, clupeids, and percids. Freshwater drum and bigheaded carp larvae had higher detection in thalweg and channel margin habitats compared to backwaters. Occupancy of bigheaded carp peaked on June 19 at 20°C, increased with discharge, and declined with the coefficient of variation (CV) of water temperature and discharge. Clupeids and percids occupancy peaked on May 24 and increased with CV of water temperature, while occupancy of freshwater drum peaked on July 3, decreased with CV of water temperature, and increased with water temperature. Our results indicate bigheaded carp are more difficult to detect than native larvae and identified conditions associated with larval occupancy that can be used to maximize detection and better understand when, where, and under what conditions larvae are present.