Using Macroscopic Traits to Hindcast and Predict Maturity in Southern Flounder

Southern flounder Paralichthys lethostigma is a highly valued commercial and recreational resource in North Carolina that is presently assessed as a “depleted” stock.  In order for managers to improve the stock’s status, new and updated biological information is needed so that appropriate harvest limits can be established.  Because spawning takes place offshore and no direct observation of reproduction is available, maturity data needs to be assessed from inshore individuals, which often lack traditional evidence of maturation (e.g., high gonadosomatic indices). Following an extensive study of southern flounder maturity, we used histologically-validated maturity information to develop cross-validated models based on a suite of macroscopic variables in order to both hindcast variability in maturation as well to provide a tool to assess maturity in future years.  Hindcast results suggest that annual southern flounder L₅₀s have varied >80 mm over the last two decades. Best-performing models using histologically-validated data typically include macroscopic classification, and had prediction successes of 72%.  Modest improvements in prediction success were seen when additional, novel macroscopic traits (e.g., gonad color) were included in the models. Given the likelihood of temporally variable maturity schedules, predictive models are useful in fisheries management, as they can both improve the explanatory power of existing data as well as provide a possible substitute for more expensive and time consuming methods of maturity assessment, such as histology.