Correctly characterizing growth of fish within a population is a crucial component of fish biology and fishery management because, among other things, it informs population dynamics that affect management decisions. Size-at-age is a common metric of fish growth and is often measured at the population level with the assumption that, on average, all fish of a given age are a given size. Over time, several studies have shown that ignoring individual variability in growth can influence population parameter estimates and these inaccuracies can be propagated in population models that are used to calculate reference points for management. In the first chapter we develop a hierarchical, mixed-effects statistical growth model that measures individual variability in growth model parameters and partitions it into two sources. We fit this model to length-at-age data of lake trout (Salvelinus namaycush) from six populations in Lake Superior and show that individual-level variability exceeds population-level variability for this system, and persistent error contributes more to variability in length-at-age. In our second chapter, we simulate a population of fish and predict biological reference points, yield-per-recruit, and spawning stock biomass-per-recruit curves from the population using a 'standard' method that ignores individual variability and a 'true' method that accounts for size-selective mortality and its interaction with individual fish. We show that ignoring individual variability in these models results in overestimation of yield-per-recruit and the biological reference points F0.1 and FMAX. Further, spawning stock biomass-per-recruit is underestimated at low levels of fishing intensity and overestimated at high levels of fishing intensity when individual variability is ignored.
Read
