Simple Modeling to Inform Harvest Strategy Policy for a Data-Moderate Crab Fishery

Attempts to model the giant mud crab Scylla serrata fishery in the Northern Territory (Australia), have often been complex and the results difficult to interpret, leading to divergent estimates of fishing mortality. This has hindered the development of meaningful management policy. Additionally, analyses based on the entire Northern Territory fishery have masked the extreme variation in catches observed along the Gulf of Carpentaria coast. We applied a structurally simple model to visualize the historical patterns in stock size, recruitment, fishing mortality, and fishing mortality at maximum sustainable yield in the western Gulf of Carpentaria mud crab fishery (WGOCMCF) from 1983 to 2017. We also projected future catch and female spawning stock biomass (FSSB) under positive, neutral, and negative recruitment scenarios for three closure periods contained in the fishery harvest strategy (which start in October, if triggered) and compared these with the results of equivalent closures beginning in September. This exercise was undertaken because of known and significant changes in the proportion of fishing effort across different months as well as large variations in the proportion of females harvested each month (with both factors being particularly low in December). These differences were annualized and incorporated into the yearly time step of the model. Predicted catch and FSSB were similar for shorter closure periods (3 or 6 weeks), irrespective of the starting month. However, initiating a 3-month closure in September rather than October could lead to a 16-17% increase in FSSB under negative and average recruitment anomalies, imposing a 9-10% reduction in predicted catch. Based on our experience applying a simple model to the WGOCMCF, we also describe processes and practices that could improve the quality of assessment data for this and other data-moderate crab fisheries.