Evaluating alternative harvest policies for yellow perch in southern Lake Michigan

In the southern basin of Lake Michigan, yellow perch (Perca flavescens) are ecologically and economically important. However, there is no explicit harvest policy for the management of this resource, the authority for which is shared among four U.S. states. We used decision analysis and projections from a stochastic simulation model to aid managers in formulating a harvest policy. In workshops that included management agency personnel and other experts, critical uncertainties relevant to the population (e.g., alternatives for future stock–recruitment relationships and mixing of recruits among management areas) were identified as well as potential harvest policies (using constant fishing mortality or state-dependent control rules) and associated performance statistics. Our simulation model acknowledged uncertainty in the stock–recruitment relationship, parameter uncertainty given such a relationship, stochastic process variation, and uncertainty associated with assessment and implementation errors. We used the model to project age-, sex-, size-, and spatial-dynamics of the yellow perch population, and thus predicted likely distributions of performance statistics for different harvest policies. Performance statistics included time averages of recreational harvest, remaining spawning stock biomass (SSB), and length of harvested fish as well as the frequency of how often such measures were below desirable thresholds. Results indicate that state-dependent policies produce higher average harvests and lower frequency of years with low SSB, but sometimes more frequent years with low harvest, than constant-F policies that lead to similar depletion of average SSB.