Assessment of Razorback Sucker Monitoring in the Lower Colorado River Basin

Authors: Brian Kesner, Aaron Burgad, Paul Marsh

Year: 2018

Abstract

Stocking of hatchery-raised fish or repatriation has been the primary means of maintaining populations of the endangered Razorback Sucker Xyrauchen texanus throughout its historic range. In the lower Colorado River basin, populations in the thousands have been established and maintained in Lake Havasu and Lake Mohave by stocking over a half million subadult to adult Razorback Sucker over the last three decades. Since 2006 all Razorback Sucker released in these reservoirs are injected with a 134.2 kHz passive integrated transponder (PIT) tag. These fish are monitored using traditional capture methods (e.g. annual netting activities) and remote sensing via deployment of remote PIT tag scanning units. Remote PIT tag scanning has resulted in orders of magnitude more contact data than traditional methods. To date, 19,713 Razorback Sucker have been contacted via remote PIT scanning in the lower Colorado River basin. These data have been used to refine estimates of population size and post-stocking and adult survival using mark-recapture models. However, there has been little formal assessment of how these estimates correlate to results from traditional capture methods. A discrete time population model was developed to compare estimates of post-stocking and adult survival and frequency distributions of annual Razorback Sucker stockings in Lake Havasu and Lake Mohave with single-census estimates of abundance. We also extend the population model to include tag loss and compared ratios of tagged to untagged fish calculated from netting events to determine if natural recruitment at low levels was likely occurring without detection. The results were then compared to model results for Lake Mead, the third reservoir in the lower Colorado River basin which has no formal stocking program and limited natural recruitment.