In this dissertation, I examine how environmental and anthropogenic factors and autocorrelated disturbances affect the dynamics of an exploited renewable resource. The analysis focuses on Long Island Sound (US) for which environmental factors play an important role. This objective is addressed first by investigating the relative importance of high temperatures, low dissolved oxygen, and overexploitation in the die-off experienced by the fishery after the fall of 1999. Second, I analyze the effects of temporally correlated environmental disturbances on the stationary distribution of the dynamics of a harvested population with an application to the LIS lobster fishery.

In the fall of 1999, the American lobster population in Long Island Sound experienced a significant die-off. Biologists attributed the cause to a period of prolonged high temperatures that in concert with hypoxia suppressed lobsters' immune system making them more susceptible to diseases and infections. The relative importance of environmental factors like high temperatures and anthropogenic factors like overfishing and hypoxia is investigated. Data from annual trawl surveys and the fishing industry are combined to construct a model of lobster population dynamics and fishing effort. The model is used to investigate the relative importance of high temperatures, low dissolved oxygen, and incentives for more intensive fishing effort experienced in 1999. Simulations show that low dissolved oxygen and economically induced fishing pressure were the major factor underlying the die-off. These results indicate that stricter regulations of nutrient emissions and resolving institutional failure may be the most effective way to protect fisheries like the Long Island Sound lobster fishery. Furthermore, to mitigate the effects of global climate change, it may be advisable to impose stricter regulations of nutrient emissions in anticipation of more acute episodes of hypoxia.

Standard models of renewable resource allocation under uncertainty typically assume that environmental disturbances are identically and independently distributed. When weather patterns impact environmental conditions, shocks may be serially correlated. This serial correlation has implications for the long run conservation of harvested renewable resources. This issue is analyzed by investigating the dynamics of a harvested, open-access renewable resource whose productivity is influenced by serially correlated random environmental disturbances. The main question that is addressed is: how does the expected value of stock escapement depends on the parameters that determine the distribution of environmental shocks? In answering this question I also characterize how the maximum and minimum escapement policy functions depend on these parameters.

An application of the conceptual framework to the American lobster fishery of Long Island Sound is used to analyze these issues qualitatively and quantitatively. In the application, the model is parameterized using an econometric model of population dynamics for the Long Island Sound lobster fishery. Results show that shocks are negative correlated and transient so a high current productivity shock decreases the probability of high future shocks. Further, the results suggest that population variability is increasing in the degree of autocorrelation of the random shocks as well as the underline uncertainty of the environmental disturbances. As a result, population may experience ample fluctuations which could increase the likelihood of extinction. This suggests the adoption of management strategies that focus on maintaining a minimum population level and increasing the stability of the natural resource in the face of environmental variability.