Ecologically and evolutionarily motivated models of fish populations for conservation and management

Models traditionally used in fish management rarely incorporate important ecological or evolutionary factors. In this thesis, I present ecologically and evolutionarily informed models addressing questions relevant to fish conservation and management. In Chapter II, an age-structured, density-dependent, population specific computer simulation model incorporating environmental and demographic stochasticity is used to explore the viability of an Oregon chinook salmon stock under alternative assumptions about future habitat conditions and values of critical demographic parameters. The results indicate that habitat degradation is the most important determinant of the future viability of this stock. In Chapters III and IV, analytical models are used to explore the implications of length selective harvesting for the population dynamics and evolution of length in the exploited population. These models incorporate relevant aspects of ecology, evolution, and life history that are widely applicable to many fish species, and define the harvest strategies by both their length specificity and their demographic yield or escapement goals. In Chapter III, results from models with fixed harvest goals demonstrate that length selective harvesting can produce changes in the stock's equilibria for mean length, abundance, and yield from harvesting. They also demonstrate that, for a constant set of evolutionary and demographic parameters, some harvesting strategies always produce a characteristic equilibrium for mean length and abundance regardless of the initial attributes of the population, while others produce alternate stable equilibria, with the initial mean length and abundance determining which equilibrium is achieved. Chapter IV uses variants of models in Chapter III to explore the consequences of length selective harvesting when harvest goals are set by a simple process of adaptive management and fishing is permitted only if pre-harvest abundance exceeds a chosen escapement cutoff. These conditions generate long-term non-equilibrium dynamics for mean length, abundance, or both under a wide variety of parameterizations. The non-equilibrium dynamics take the form of multi-point cycles in mean length and abundance that, while often of small amplitude, translate into large-scale variation in biomass yield. Taken together, the result presented demonstrate the importance of incorporating both evolutionary and ecological factors into research used to set management strategies and goals.