Development and considerations for application of a longevity-based prior for the natural mortality rate

The natural mortality rate M is a key parameter in estimating the productivity of a fish stock. However, it is quite difficult to estimate given its correlation with other parameters, variation in processes governing population dynamics and data acquisition, as well as other factors such as ageing error. Meta-analytical approaches relating M to other life history parameters are often used to provide an estimate and/or Bayesian prior for M. Longevity is the one of these life history parameters most directly related to M. The maximum observed or estimated age serves as a proxy for longevity in predicting, or developing a prior for, M to be used within a stock assessment or other population model. Consideration of the age data and factors that might influence its quality for determining a maximum age is important in using a meta-analytically derived relationship between maximum age and M, though these factors are not completely absent from the data used to develop that relationship. Here, we discuss these issues, propose a relationship between maximum age and M, and develop an updated prior for M based upon maximum age.