How does seasonal variability in growth, recruitment, and mortality affect the performance of length-based mortality and asymptotic length estimates in aquatic resources?

Abstract

Abstract Hufnagl, M., Huebert, K. B., and Temming, A. 2013. How does seasonal variability in growth, recruitment, and mortality affect the performance of length-based mortality and asymptotic length estimates in aquatic resources? – ICES Journal of Marine Science, 70: 329–341. We tested the sensitivity of eight methods for estimating total mortality from size frequencies (modified Wetherall; Powell; Beverton and Holt; Jones and van Zalinge; Hoenig; Ssentongo and Larkin; seasonal and non-seasonal Length Converted Catch Curve) to violations of basic assumptions, such as seasonal growth, mortality, recruitment and variable asymptotic length L∞ or growth parameter K. For each method, bias was estimated by simulating length frequency distributions with different combinations of known L∞, Z and K values, calculating θ (Z/K) and L∞ estimates, and comparing the true input with the estimated output values. Input mortality was generally underestimated by all methods and in 27% of all simulations no method provided estimates within θ ± 1. Spring recruitment especially negatively influenced the mortality estimate. A decision tree was developed that provides general guidance in selecting appropriate methods despite violated assumptions, but species-specific case studies are recommended. An example of a species-specific study is provided for the brown shrimp, Crangon crangon. Despite inherent limitations for all methods, the results illustrate that estimates of θ and Z for brown shrimp can be improved substantially by selecting suitable methods and correcting for observed bias.