ABSTRACT Monitoring physical and biological conditions in the open ocean is an inherently difficult task, particularly when monitoring toxic and harmful compounds. Efforts to use biomonitor species to measure and assess ocean and organismal health require considerable information on habitat use and other life history characteristics to contextualize and interpret contaminant data. A large proportion of biomonitoring research focuses on a single species, a single site, and/or a small range (i.e. 1-3) of contaminant classes. While informative, the scope of such studies can limit their applicability, which is concerning as data suggests the abundance of organic and heavy metal contaminants in the ocean is increasing. Contaminants can have sub-lethal effects that affect population viability, and new, unknown contaminants enter the environment with little knowledge of their possible effects and limited ability to monitor these emerging contaminants. Seabirds serve as effective biomonitors for contaminants for multiple reasons. First, seabirds feed at high trophic levels, where toxicants are biomagnified. Because most seabird species are long-lived, birds may bioaccumulate toxicant types into tissues (e.g., fat) over a lifetime, much like humans. Contaminants also accumulate in seabird tissues that can be sampled non-lethally and at low cost, including blood, feathers, and non-viable eggs. Thus, it is also easy to assess the sub-lethal effects of low, but chronic levels of contaminants on a variety of ecological parameters, especially when paired with long-term datasets. Additionally, many seabird tissues have enough volume with which to test for multiple contaminant classes in a single sample. Lastly, seabirds exhibit high breeding site fidelity, so individuals can be sampled and re-sampled with regularity. My research explores how seabirds may be used as biomonitors for a rapidly-changing ocean environment. In Chapter 1, I show that seabird tissues can be used to indicate the magnitude and extent of a wide range of contaminants at the regional scale in the Southern California Bight. In contrast to single species, single site monitoring, regional assessments maximize the ability to use biomonitoring efforts to meet mandated monitoring objectives, prioritize site remediation, and trace the dispersal and uptake of toxicants in marine food webs. The results suggest at least one species, the California least tern (Sterna antillarum browni), may be a robust indicator of contaminant patterns in this region. Chapter 2 investigates blood mercury concentrations as a function of foraging distribution in western gulls (Larus occidentalis) nesting at three colonies off the California and Oregon coast. We found that ocean-foraging gulls had elevated mercury concentrations and also lower fidelity to geographic foraging areas, confirming work that suggests aquatic foragers have greater exposure to methylmercury. As mercury exposure and likely exposure to other contaminants differs across the land-sea gradient, species that forage in marine and terrestrial environments may be used to better understand the ecological consequences of contaminant-associated diet. Chapter 3 uses an established nontargeted approach to examine the presence and patterns of legacy and frequently unmonitored halogenated organic compounds in three albatross species that range the North Pacific. The majority of contaminants we found are currently unmonitored, which demonstrates the extent of chemical contamination beyond urbanized coastal areas to remote coastal regions. We also found support for previous research that suggests differences in broad-scale foraging areas impact contaminant abundance among these three species.

Climate shifts and increasing variability are anticipated to alter ocean systems, significantly impacting marine ecosystems as well as the communities, businesses, and fisheries reliant on them. Apart from the gradual effects of climate change (i.e. long-term ocean warming), marine organisms and fisheries are also experiencing episodic extreme oceanic warming events, known as marine heatwaves (MHWs), that can push ecosystems into new environmental conditions, further challenging already stressed social-ecological systems. The impact of MHWs may be particularly pronounced for highly migratory pelagic species who shift their distributions during anomalous events in response to unfavorable conditions, which can lead to ecological disruptions, such as predators being separated from their prey, and economic disruptions if target fish populations decline or move out of the range of the fishers who catch them. Understanding and predicting how such disruptions will affect both ecological and human communities will enhance the ability of conservation and management efforts to keep pace with future MHWs as they unfold.Species distribution models (SDMs) have emerged as a fundamental tool to predict how species will respond to changing environmental conditions and have served a critical role in marine conservation and management. Despite their known utility, there is considerable uncertainty as to how useful SDM predictions can be given the current and continued changes in ocean conditions. Digital and technological advancements have greatly expanded the volume and diversity of data available for developing SDMs, which, when combined, can offer complementary information on species distributions and improve model inference and prediction. Development of data integration for SDMs has advanced over the last decade, showing great potential in enhancing model performance as they are able to account for data-specific biases while retaining the strengths of each dataset. However, despite these advantages, such applications to integrate disparate data sources have been scarce in marine systems, particularly for highly migratory pelagic species, limiting our knowledge on how best to leverage diverse data sources for modeling species distribution under a changing climate. My research focuses on modeling species and fishery spatiotemporal distributions and exploring the capacity of data integration in improving model performance under MHWs. In Chapter 1, we take a bi-coastal approach to explore how MHW properties influence the redistribution of U.S. pelagic fisheries. Our study provides the first evidence that MHW size, rather than MHW intensity or duration, exerts the largest impact on fishing vessels, with southern regions on both coasts losing fishing grounds as size increased. The extent of fleet displacement in response to MHWs, however, varies between coasts, as the Atlantic longline fleet displaced farther in southern regions whereas peripheral regions of the Pacific troll fleet shifted farther. Our results suggest that MHWs drive differential impacts across fishing fleets and regions, and characterizing how fleets will respond can be used to identify regions of resilience or vulnerability in the face of future MHW events. Chapter 2 evaluates approaches to synthesizing multiple data types for SDMs, comparing traditional data pooling and ensemble techniques to a model-based integrative framework using diverse data sources for blue sharks across the north Atlantic Ocean. Results revealed that integrated SDMs, which account for data-specific biases and seasonal variability, generate more accurate and ecologically realistic predictions, though they are computationally intensive. Our findings highlight trade-offs in data integration techniques and the importance of leveraging diverse data sources for marine conservation and management. Building on these findings, chapter 3 seeks to explore how model-based integration improves spatiotemporal forecasts for an important fishery species, albacore tuna, during a period of unprecedented MHWs in the northeast Pacific Ocean. Results indicate that traditional SDMs generally struggle with predictive accuracy and realism as environmental novelty increases, while integrated SDMs (iSDMs) maintain greater predictive skill and ecological realism. The findings suggest that leveraging diverse data sources through appropriate integration methods is crucial for generating accurate ecological forecasts and supporting climate-ready management and conservation strategies as episodic climate events become more prevalent.

Background

Results

Conclusions