Climate change, natural disturbances, and human activities interact to influence and alter coastal and marine ecosystems in many ways. Knowledge of these interactions and their outcomes may allow managers and decision-makers to anticipate ecosystem change in the context of extreme events and changing climate. Yet, major questions remain regarding how spatial patterns of coastal, benthic landscapes vary across driver regimes and whether driver thresholds exist that result in abrupt changes in spatial configuration. Understanding and predicting when landscapes may experience changes in not only configuration but also extent, directly informs decisions involving resource allocation especially for extractive industries that depend on seagrass ecosystems as nurseries and/or refuge for the target resources. Extractive industries, such as commercial fishing are in turn affected by extreme events and climate and that interaction can result in large amounts of marine debris being contributed from those fisheries to the environment in the form of gear loss. In addition, the relationship between marine debris generated from ocean-based human activities and a changing climate remains relatively understudied. This dissertation addresses knowledge gaps in disturbance dynamics for marine and coastal systems by using a combination of remote sensing, empirical field data, scenarios, and statistical models to evaluate spatial pattern in seagrass ecosystems of coastal North Carolina as well as the spatial and temporal dynamics of marine debris generation in two separate commercial fisheries based in Florida and Hawaii.
Combining visual photointerpretation with a semi-automated image classifier (linear spectral unmixing), I demonstrated that seagrass and bare substrate can be effectively distinguished in remotely sensed images to produce fine-scale seagrass maps. By identifying small, individual seagrass patches and eliminating bare substrate from within the boundaries of manually mapped seagrass polygons, the linear spectral unmixing classifier effectively improved seagrass maps, moving beyond ‘seagrass habitat’ extent and allowing for estimation of actual seagrass area and calculation of landscape pattern metrics for seagrass. Applying this classification technique and including empirical data on hydrodynamic drivers in coastal North Carolina, I identified change points in wave energy above and below which seagrass spatial configuration significantly differed. Furthermore, observed patterns in the frequency distribution of percent cover provided moderate support for the existence of alternate states (bistability) in seagrass landscapes. In addition to developing predictions for how landscape changes and the subsequent changes in resource distribution affect extractive industries, understanding the conditions that influence seagrass landscape pattern and coverage directly informs mitigation efforts in areas marked for restoration to provide compensatory services.
Using empirical data on wind speed to model rates of commercial spiny lobster trap loss, I developed three plausible but contrasting future scenarios of tropical cyclone activity and fishery effort to evaluate the consequences of inputs of lost gear to south Florida and the potential implications for benthic resources. The scenarios suggest that were existing fishing effort to be maintained in the coming decades, tropical cyclone-related trap loss could exceed 11 million over 60 years depending upon the rate of storm intensification. In these simulations, trap loss was greatly exacerbated under scenarios of increasing tropical cyclone intensity, underscoring the importance of including these factors in climate change adaptation planning. Fishery models of catch-per-unit-effort were applied to an empirical data set of marine debris counts made by fishery observers aboard vessels operating in the Hawaii-based pelagic longline fishery to estimate debris abundance in the north Pacific Ocean near the Subtropical Convergence Zone. The majority of debris caught by longline gear was derelict nets from other commercial fisheries. Despite considerably less effort (number of hooks per set), vessels operating in shallow water encountered 4x more debris than vessels fishing in deep water. Greater amounts of debris were encountered within the Subtropical Convergence Zone, highlighting the importance of ocean circulation in accumulating debris. Collectively, this research provides insights into disturbance dynamics in coastal and marine systems and emphasizes the importance of considering interacting drivers to inform coastal and marine management and policy concerns.