Interactions of anglers and walleyes (Stizostedion vitreum vitreum) we
re analyzed by fitting difference equations to a 33-yr time series fro
m Escanaba Lake, Wisconsin. Fish population dynamics were density-depe
ndent, and harvest by anglers followed a sigmoid functional response.
The deterministic form of the model was stable and suggested that the
fishery could withstand relatively high exploitation rates. The stocha
stic form of the model compared management strategies using the probab
ilities of increased or decreased fish and angler populations and the
probability that fish populations will fall below a threshold level. T
hese calculations exposed high risks associated with management strate
gies based on deterministic, equilibrium equations. The probabilities
of sustaining both the fish stock and angler effort increase if harves
t rates are reduced below the optimum derived from deterministic model
s. Consideration of angler dynamics reduced the prediction error of fi
sh stock forecasts and substantially altered risk assessments and mana
gement recommendations. If angler dynamics were ignored, the optimal e
xploitation rate and the risk of overfishing would be overestimated. L
ake sport fishery assessments can be improved by coupling models of fi
sh dynamics at the scale of individual lakes with angler models at the
larger scale of lake districts.