Evaluating the Influence of Spawning Dynamics and Sampling Design on Estimates of Fecundity, through Simulation Modeling

In many federally managed species, estimating population fecundity is challenging, because of complex reproductive biology. Despite growing interest in this topic, it is often unclear how realistic fecundity estimates are and how to best apply research effort.

Using a simulation model, we formalized the current understanding of reproductive dynamics of indeterminate batch spawning species. We applied the model to evaluate current methods of estimating annual fecundity and examine how these estimates affect biological reference points important to management.

We developed two base model scenarios representing very different levels of knowledge. The Perfect scenario was set to behave in the most predictable way to produce nearly deterministic results assuming limitless resources, while the Realistic scenario was set to include as many important real elements as possible, based on empirical research. We then sequentially added realistic elements to the Perfect scenario, and found that biases readily occurred in annual fecundity and the size of the spawning stock at maximum sustainable yield. To address these biases we sequentially added elements to improve the Realistic model, and found that changes in the temporal pattern of sampling had the greatest effect for some outputs, but biases in other outputs could not be avoided.