Much can be learned of complex ecological communities by identifying the ecologically-important species and studying the contingencies of their interactions. Recent evidence suggests that a major mediator of species interactions is trait variation within such species. Western mosquitofish (Gambusia affinis) is a globally invasive predator with potent effects on ecosystems and substantial trait variation among populations. In this dissertation, I sought to determine the traits and environmental conditions that mediate variation in the community and ecosystem effects of mosquitofish populations. In Chapter 2 I tested for the effects of mosquitofish sex ratio variation on experimental pond ecosystems. Mosquitofish exhibit sexual dimorphism and their sex ratios vary widely across their global range from highly female-biased to highly male-biased. In experimental ponds, female-biased populations had higher rates of prey depletion and nutrient excretion which induced stronger pelagic trophic cascades, causing significant increases to important ecosystem variables such as water temperature and pH. In Chapter 3 I tested how temperature mediates the importance of mosquitofish trait variation for experimental pond ecosystems. A growing literature demonstrates that top-down effects increase under warming. Thus, I predicted that the differences in ecological role among mosquitofish populations would become larger under warming. In support of my prediction, warming increased top-down effects and increased the ecological divergence between mosquitofish populations. In Chapter 4 I tested how evolutionary adaptation to temperature affects traits and ecological effects of mosquitofish. Mosquitofish exhibited increased mortality, and evolved slower growth and earlier maturity at higher temperatures. Warm-adapted mosquitofish also consumed smaller pelagic zooplankton prey and did not show a strong ontogenetic niche shift to consuming large benthic macroinvertebrates. Because warm-adapted mosquitofish grew slower, reproduced earlier, and continued to consume zooplankton throughout life, they exacerbated warming-induced shifts towards smaller zooplankton communities. This dissertation reveals considerable effects of mosquitofish demographic variation, interpopulation variation, and contemporary evolution on community and ecosystem ecology. These results suggest that the structure and functioning of an ecosystem can depend strongly upon the traits of the ecologically important species present.

Many prey species display behavioral responses to environmental cues from predators. Organisms can inherit fixed behavioral responses to certain predators as a result of evolutionary history with a predator. Developmental plasticity allows individuals to associate predation risk with environmental cues experienced in their lifetimes and respond behaviorally. However, the interactive effects of evolution and plasticity on shaping immediate antipredator behavior is not well known. Here we show that evolution and developmental plasticity affect antipredator behavior of western mosquitofish (Gambusia affinis) through altered food intake in response to predation risk imposed by largemouth bass (Micropterus salmoides). Results show that the effects of evolutionary history and developmental plasticity are substitutive – mosquitofish show a similar level of feeding if they originated from ponds with largemouth bass (evolutionary history) or if they originated from ponds lacking largemouth bass but have been exposed to bass cues during development (developmental plasticity). Mosquitofish lacking both evolutionary history and developmental cues from the predator did not show a behavioral response. For populations lacking an evolutionary history with bass, developmental exposure to the predator was required to induce a behavioral response. In contrast, populations with an evolutionary history with bass displayed antipredator behavior regardless of whether they had been exposed to bass cues during development. Our results suggest a progression of anti-predator feeding responses of prey from no behavioral response, to an inducible response, to a fixed response. This sequence may shift the cascading ecosystem effects of predators introduction from density-mediated to behaviorally-mediated via plasticity to behaviorally mediated via evolution over time.

Variation in life-history traits within and across species is known to reflect adaptations to different environmental drivers through a diversity of mechanisms. Trait variation can also help buffer species and populations against extinction in fluctuating environments and against anthropogenic disturbances. This recognition has led to a growing interest in applying evolutionary principles to inform eco- logical restoration actions. In this dissertation, I explored how widespread variation in migratory life-histories in Pacific salmonids (Oncorhynchus spp.) is driven by a suite of environmental factors with the goals of shedding light on what drives the frequency of these migratory strategies and creating a restoration framework to aid in the restoration of these imperiled species. In Chapter 1, I examined how anthropogenic disturbance in the form of stream modification influences the frequency of anadromy in steelhead (O. mykiss) across California. I found that with an increasing number of instream barriers, the frequency of anadromy was reduced. I then developed an evolutionary restoration framework to guide managers and encourage broader consideration of in situ evolution during the development of habitat restoration projects. In Chapter 2, I quantified distinct juvenile migratory strategies of Chinook salmon (O. tshawytscha) and modeled the environmental drivers that influence their frequency. I found two distinct migratory strategies were widespread across these populations and their frequencies were driven a juvenile density, rearing habitat, and flow. In Chapter 3, I examined what environmental and demographic factors influenced spring juvenile migratory strategies in coastal coho salmon (O. kisutch). I found that juvenile density, flow, and latitude are strong drivers of these migratory behaviors. As in Chinook salmon, a density-dependent mechanism also appears to operate in coho salmon populations, suggesting a form of bet-hedging might be common across out-migrating Pacific salmon. My dissertation links life-history evolution, environmental variation, and anthropogenic modifications to the landscape, demonstrating that alternative life-history strategies are common in Pacific salmonids and suggesting restoration approaches that can be used to conserve this variation.

Species interactions are core to our understanding of communities and ecosystems. Much of the theoretical work that forms the basis of our understanding of species interactions is based on the assumption of constant environmental conditions and the achievement of equilibrium. However, real ecosystems are often characterized by changing conditions and nonequilibrium dynamics. Fluctuations in environmental conditions may affect species interactions in a number of ways: by excluding one partner, by changing the dominance relations of interactors, the magnitude, or even the direction of interaction strength between pairs of species. In this dissertation I studied the effects of fluctuating environmental conditions on predation and competition of threespine stickleback (Gasterosteus aculeatus) in bar-built estuaries along the central coast of California. Due to seasonal variation in rainfall, bar-built estuaries are only intermittently open to the ocean and vary seasonally in important abiotic conditions such as salinity and dissolved oxygen. In Chapter 1 I surveyed threespine stickleback in estuaries of various sizes. I found that larger estuaries supported perennial streamflow, breeding habitat for predatory salmonids and sculpins, and therefore stickleback had higher frequencies of alleles coding for a complete set of armor plates. In Chapter 2 I tested directly for selection on Ectodysplasin-A (Eda), the gene controlling armor plates, by comparing the genotypes of breeding adult stickleback to juveniles recovered from the stomachs of predatory salmonids. I found no evidence of selection over a single season, indicating that the genetic differences in Chapter 1 built up slowly. In Chapter 3 I used empirical dynamic modeling to show that fluctuations in the abundance of endangered tidewater goby (Eucyclogobius newberryi) and threespine stickleback represented independent responses to changes in environmental conditions rather than context-dependent competition. Specifically, stickleback were negatively impacted by breaching, and the peak abundance of both species was offset by several months, potentially indicating that reproductive allochrony alleviates competition. In both the case of predation and competition, long-term effects were evident but short-term effects were not. This pattern suggests that species interactions can and do play an important role in fluctuating environments but that those interactions may be difficult to detect in the short-term or under all conditions.

Dams have been built on most of the world’s large rivers, making the management of environmental flows crucial for meeting the needs of humans and wildlife. Current water management in the mainstem Sacramento River below Keswick Dam focuses on maintaining optimal conditions for early life stages of endangered winter-run Chinook salmon (Oncorhynchus tshawytscha) and providing discharge to support downstream water use. We found that the low temperature and high discharge flows may negatively impact condition of threatened green sturgeon (Acipenser medirostris) larvae rearing in this region. We developed a multi-object optimization model which identified optimal releases of 11.5°C and 150 CMS to balance the three objectives. Reservoir levels required to meet this optimal management strategy occurred in 80% of the last 20 years. This proposed management strategy shifts to multi-object management while meeting downstream water requirements and could be used as a model to provide environmental flows in other impounded rivers.

Migratory animals are ecologically and economically important and are especially vulnerable to predation. Predators affect migratory prey through both consumption and changes to prey behavior scaling up to affect populations to communities. The balance between consumptive and behavioral predator effects are shaped by risk-reward tradeoffs and ecological constraints. The unique biology of migratory animals results in tradeoffs and constraints involving energetic-demands and optimal timing of migration and traversing heterogeneous and unfamiliar habitats. The goal of my dissertation is to examine how predators affect migrating prey behavior, ecology, and evolution. I used diverse approaches to understand the intersection of predation and migration from fine-scale behavioral decisions up to large-scale patterns among populations and species. In chapter 1, I developed a conceptual framework to predict how migrating prey perceive and respond to predation risk—a direct extension of classic economic escape theory. In chapters 1 and 2, I tested this theory empirically using behavioral assays where I timed juvenile salmon swimming downstream with and without predator cues. In two experiments, juvenile salmon changed behavior (speed) in response to predator cues, but the pattern of response was context-dependent on previous predator experience and habitat. In chapter 1, wild salmon with more previous predator experience reacted more strongly to predation risk than hatchery salmon. In chapter 2, salmon responded more strongly to predation risk in the shade compared to direct sun and varied their escape strategy—slowing down in the shade and speeding up in the sun. In chapter 3, I conducted a global synthesis across taxa that examines how predators shape animal migrations from behavior to population dynamics to life history evolution. It further explores how humans are disrupting these predator-prey interactions and provides suggestions on how to conserve healthy predation landscapes. Using diverse approaches, from mechanisms to population consequences to ecosystem effects, this work increases our general ecological understanding of predator effects on migratory prey and informs efforts to conserve threatened migratory prey.

Recent research highlights the prevalence of intraspecific trait variation, even in relatively open ocean habitats. The ecological importance of intraspecific trait variation, however, while shown in freshwater and terrestrial ecosystems, remains unexplored in marine systems. While climates change rapidly and differentially across marine environments at scales within species ranges, population-level trait variation in response to abiotic drivers is inevitable. It is therefore timely and important to explore the climate drivers of intraspecific trait changes and their ecological consequences in marine systems. In this dissertation, I explore these dynamics in a model predator-prey system. The predators, Nucella ostrina-emarginata dogwhelks, exhibit low population connectivity and gene flow due to their life history. The prey, Mytilus californianus, the California mussel, is a foundational mussel that supports high intertidal diversity. These species exist throughout a mosaic of climate conditions in the California Current System, setting the stage for local scale climate effects on Nucella predation that have community consequences. In Chapter 1, I examine climate drivers of population-level variation in size selectivity of Nucella on Mytilus. I find that abiotic variables such as temperature and pH are the strongest drivers of Nucella prey size selectivity rather than neutral genetic relationships among populations, which have no effect. In Chapter 2, I test for population-level differences in the responses to acute exposure to acidified seawater on Nucella size selectivity and consumption time. I find that populations are affected differently by acidification, showing that climate change can affect Nucella predation on local scales. In Chapter 3, I test for community effects of population-level differences in Nucella predation on mussel beds in the field. I find that Nucella predation affects mussel bed size structure and in turn, size structure affects community composition, showing differential predation on a foundation species can alter communities. My dissertation links climate change, trait variation, and community ecology, demonstrating how climate can indirectly alter communities by shaping predator traits on local scales, and expanding the study of population-level trait variation into marine ecosystems.