Modeled Effects of Double-Crested Cormorant Predation on Simulated Reservoir Sport and Forage Fish Populations in Oklahoma

Abstract

Populations of the double-crested cormorant Phalacrocorax auritus have increased rapidly in recent years, eliciting concerns among anglers regarding their predation on reservoir fishes in the southern Great Plains. We adapted a Ricker equilibrium-yield model to evaluate effects of cormorant predation on standing crops and yields of reservoir sport and forage fishes. Measured cormorant abundances were highly variable among reservoirs and between years, but abundances at most reservoirs (14 of 16 reservoir observations; 8 reservoirs × 2 years) were low: mean monthly rate of 2.5 ± 4.2 (SD) cormorant use-days/ha during migration periods (Sep–December and Mar–May). At this level of predation, modeled sport fishery yields in a typical Oklahoma reservoir, based on density-dependent and density-independent fish growth, were reduced by 3.1% and 18.8%, respectively, for channel catfish Ictalurus punctatus, 2.9% and 22.0% for largemouth bass Micropterus salmoides, and 1.8% and 5.3% for white crappie Pomoxis annularis. Standing crop of gizzard shad Dorosoma cepedianum was reduced by 1.4% and 5.6%. These simulations appeared to reasonably approximate conditions found at many reservoirs in Oklahoma. Measured cormorant abundances during the remaining two reservoir observations were high: mean monthly level of 23.4 ± 14.2 cormorant use-days/ha during the migration and wintering (Sep through May) periods. Modeled channel catfish and largemouth bass fisheries collapsed at this level of predation, and yields of white crappie were reduced by 33.3% (density-dependant growth) and 61.4% (density-independent). Gizzard shad standing crop was reduced by 17.4% and 66.8%. However, these high predation rates did not persist at either reservoir because the high cormorant abundances in the first study year were substantially lower the following year. This invalidated the model assumption that all variables were in equilibrium and caused the models to underestimate fishery yields in these special cases.