UBC Theses and Dissertations
Properties of age structure models for harvested populations with applications to yellowfin tuna Thunnus albacares (Bonaterre) and harp seals Pagophilus groenlandicus (Erxleben) Allen, Robin Leslie
Dynamic population models that incorporate the age structure of the population can be used as a powerful tool for fisheries management. In this thesis some of the properties and potential uses of a class of these models is examined. This class can be described as a life table operating on a population with a stock-recruit relation that is formed by multiplying the egg production by a function of a linear combination of the numbers of fish of different ages. The study of the properties included an analysis of the stability and the non-equilibrium behaviour of the model. This behaviour was related to some types of fluctuations observed in fish populations, special attention was given to the limit cycles that occur in populations with an unstable equilibrium and this is considered as a possible a mechanism causing the regular fluctuations known as cyclic dominance in sockeye salmon Oncorhyncus nerka (Walbaum). The fisheries for yellowfin tuna in the eastern Pacific Ocean, and harp seals in the western North Atlantic were used as examples to show how these models might be used by a management body. In the case of the yellowfin tuna the model agreed with results obtained using surplus production models and in addition suggested it is possible to increase both catch and catch per unit effort by changing the open season from the beginning of the year to later in the year. The yield could also be increased by using a fishing method that increased the age of first capture. However, this would suffer from the disadvantage of producing a much greater variability in the yield because of the conseguent reduction in the number of year classes making a substantial contribution to the yield. The analysis of the harp seal population showed that this population is being over-harvested, and that if the size of the pup harvest is not reduced the entire pup production would be harvested by 1980. If the harvest rates are substantially reduced a sustained harvest could be taken, the best estimate of a sustained kill of pups only was 172,000. Because the estimates of the parameters are not very precise an analysis of the sensitivity of the results to changes in the parameters was made, and this showed if the estimates are not very accurate' considerable errors could be made in the predictions of the model. This analysis was used to compare several possible observations that could be used to check the effect of a reduced harvest on the population, and of these the catch curves of the younger age groups offered the best compromise between sensitivity to changes in population size and ease of measurement.
Item Citations and Data