<p> The impacts of a 20,993 L diesel and gasoline spill on the benthic macroinvertebrate (BMI) and diatom communities in Cold Creek and the Santa Ana River (below its confluence with Cold Creek) were assessed by comparing species richness, diversity and compositions between fuel-impacted and reference sites. BMIs and diatoms were sampled four times (13&ndash;26 months after the spill) in sites located upstream from the spill (reference) and in sites located 0.5, 1.3, 2.7 and 3.0 km downstream from the spill (impact). BMI communities up to 2.7 km below the spill and in Cold Creek had (1) lower species richness and diversity for at least 25 months after the spill and (2) dissimilar species compositions for up to 26 months after the spill when compared to the reference site. Diatom communities up to 2.7 km below the spill and in Cold Creek had (1) similar or higher species richness and diversity when compared to the reference site and (2) dissimilar species compositions when compared to the reference site for the entirety of the study. BMI and diatom communities located 3.0 km below the spill and in the Santa Ana River had similar species richness, diversity and compositions when compared to the reference site. These results provide evidence that the fuel spill had an impact on the BMI and diatom communities in Cold Creek and that these communities are still changing.</p>

<p> Information on the movement of migratory demersal fishes such as Pacific halibut, Pacific cod, and sablefish is needed for management of these valuable fisheries in Alaska, yet available methods such as conventional tagging are too coarse to provide detailed information on migration characteristics. In this dissertation, I present methods for characterizing seasonal and annual demersal fish movement at multiple scales in space and time using electronic archival and acoustic tags. In Chapter 1, acoustic telemetry and the Net Squared Displacement statistic were used to identify and characterize small-scale movement of adult female Pacific halibut during summer foraging in a Marine Protected Area (MPA). The dominant movement pattern was home range behavior at spatial scales of less than 1 km, but a more dispersive behavioral state was also observed. In Chapter 2, Pop-up Satellite Archival Tags (PSATs) and acoustic tags were deployed on adult female Pacific halibut to determine annual movement patterns relative to MPA boundaries. Based on observations of summer home range behavior, high rates of year-round MPA residency, migration timing that largely coincided with winter commercial fisheries closures, and the demonstrated ability of migratory fish to return to previously occupied summer foraging areas, the MPA is likely to be effective for protecting both resident and migrant Pacific halibut brood stock year-round. In Chapter 3, I adapted a Hidden Markov Model (HMM) originally developed for geolocation of Atlantic cod in the North Sea for use on demersal fishes in Alaska, where maximum daily depth is the most informative and reliable geolocation variable. Because depth is considerably more heterogeneous in many regions of Alaska compared to the North Sea, I used simulated trajectories to determine that the degree of bathymetry heterogeneity affected model performance for different combinations of likelihood specification methods and model grid sizes. In Chapter 4, I added a new geolocation variable, geomagnetic data, to the HMM in a small-scale case study. The results suggest that the addition of geomagnetic data could increase model performance over depth alone, but more research is needed to continue validation of the method over larger areas in Alaska. In general, the HMM is a flexible tool for characterizing movement at multiple spatial scales and its use is likely to enrich our knowledge about migratory demersal fish movement in Alaska. The methods developed in this dissertation can provide valuable insights into demersal fish spatial dynamics that will benefit fisheries management activities such as stock delineation, stock assessment, and design of space-time closures. </p><p>

<p> Wetlands exhibit high primary productivity and play a significant role in the global carbon cycle. Brackish and salt marshes co-occur in Southern California; yet, restoration designs often eliminate remnant brackish marshes, along with their ecosystem functions, without evaluation. Vegetation, soil organic matter, and carbon flux were compared between brackish and salt marsh habitats in the Huntington Beach Wetlands. Newland Brackish Marsh had more carbon aboveground in denser and taller vegetation than the other two marshes. Brookhurst Salt Marsh sediments had more organic matter than the other two marshes. CH₄ emissions were negligible at all sites, and there were no differences in CO₂ flux or aerobic and anaerobic microbial respiration among sites. Although these components of the carbon cycle were similar among sites in this project, such quantitative functional evaluations should be part of the restoration planning process.</p>

<p> In large, seasonally dynamic and spatially complex watersheds, the availability and relative importance of various food resources for stream fishes can be expected to vary substantially. While numerous studies have attempted to uncover the trophic linkages that support stream salmonids, much of these efforts have occurred at small scales that disregard variability of food resources inherent in lotic systems. This study aimed to determine large-scale patterns in the contributions of freshwater, terrestrial, and marine-derived food resources to juvenile Chinook and Coho salmon (<i>Oncorhynchus tshawytscha</i> and <i>O. kisutch</i>) in the large, glacially influenced Susitna River, Alaska. I quantified diet patterns both spatially, across different macrohabitat types positioned along a 169-km segment of the river corridor, and temporally, from June to October, using stable isotope and stomach content analyses. To further resolve energy pathways from basal carbon sources to juvenile salmon, I determined the relative roles of terrestrial organic matter and freshwater periphyton food sources to aquatic benthic invertebrate diets. The latter analysis showed that invertebrate consumers were more reliant on freshwater periphyton than on terrestrial organic matter. Bayesian stable isotope mixing models indicated that juvenile salmon in the middle Susitna River were, in turn, largely supported by freshwater invertebrate prey regardless of spatial and temporal context. The relative contribution of marine-derived prey (salmon eggs) to juvenile salmon diets was greatest in the fall within tributary mouth and off-channel macrohabitats during both years of the study. Terrestrial invertebrate prey contributions were generally greatest during mid-summer within all macrohabitat types sampled, however this pattern varied across years. No upstream to downstream diet pattern was apparent from the data. These results underscore the importance of freshwater energy pathways for sustaining juvenile Chinook and Coho salmon in the Susitna River and provide further spatial and temporal context for the importance of pulsed marine and terrestrial prey subsidies. As Pacific salmon stocks continue to decline, management and mitigation efforts should operate on knowledge gained from studies that encompass the large-scale spatial and temporal variability inherent in riverine landscapes.</p>

<p> Gamefish populations in North America are experiencing increasingly elevated recreational fishing pressures; however, the impacts of current fishing regulations have not been evaluated for most marine species. Mandatory catch and release often results from size and bag limits with the assumption of fish survival and recovery with minimal sublethal effects. Depending on the intensity and duration of the stressor, acute stress from angling and handling can have deleterious physiological and behavioral impacts with consequences for fish health and fitness. We evaluated the short-term sublethal effects of angling-related stressors on kelp bass, <i>Paralabrax clathratus</i>, one of the most popular gamefish in southern California. Collaborating with recreational anglers to capture fish using standard hook and line practices, we evaluated biochemical responses and recovery by collecting blood samples at different time points after capture (10-120 min), after fish were released and some were recaptured (3h to 186 d). Blood cortisol, glucose, and lactate were significantly elevated and steadily increased in the hour following capture and holding (p &lt; 0.001). Fish caught, released, and recaptured after varying time at liberty were found to have returned to basal levels in &lt; 24 h (p > 0.05). Physiological stress responses were positively correlated with angling and handling duration, but negatively correlated with increasing fish size (p &lt; 0.05), suggesting that larger individuals may be more resilient to capture stress. Likewise, kelp bass exhibited no signs of cumulative or chronic stress from repetitive angling/recapture events and recovered to basal levels in &lt;24 h (p > 0.05). Kelp bass were also tagged with acoustic accelerometer transmitters and tracked to evaluate fine-scale behavioral effects of angling stress on their normal diel movement and activity. Reduced frequency of high acceleration movements (i.e. rapid feeding strikes) was observed for 33 h post-release, followed by recovery and evidence of elevated feeding behavioral activity. Rate of movement and area use size showed high individual and temporal variation; however, reduced movement immediately after release steady increased to normal levels over the first 10 h post-release. There was no discernable effect of angling on area use size, but areas of high relief substratum may be important refuge habitat for angled kelp bass during recovery. By integrating physiological and behavioral endpoints for kelp bass to standard angling induced stress we conclude that this species is able to recover in &lt;24 h; however, smaller individuals may be more susceptible to predation during this recovery period. Based on these findings, it is likely that kelp bass may still fair well under current size and bag limits.</p><p>

Understanding the relationship between heterogeneity, biodiversity and ecosystem function is an active focus of ecological research that has direct applications to the formulation of sustainable, science-based, watershed conservation plans. Here, I applied ecological theory on heterogeneity to the expansion of North American beaver to test hypotheses about physical habitat and fish biodiversity at a riverscape scale. To test these hypotheses (Chapter 4), I first addressed two methodological issues (Chapter 2, 3). By evaluating three types of gear at three levels of effort in a randomized block design over 4 replicate days, I show that 10 minnow traps, 2 hoop nets and 20 m of electrofishing captured most fish species within a 30-m sampling area (Chapter 2). Multiple statistical measures provided similar information, therefore I used general indices (richness, diversity), ecological guilds (flow based), and select multivariate analyses (DCA) to summarize fish communities (Chapter 3). I used these methodological insights to test ecological hypotheses by collecting habitat and fish data at all beaver dams (n = 15) and select control sites (n = 9) in Fish Brook, a coastal watershed in northeastern Massachusetts. From these data, I gained six basic and applied insights. First, beaver dams were distributed throughout the stream network. Second, at a local scale, beaver dams created more habitat heterogeneity than control sites. Specifically, beaver dams created four types of habitat alterations based on upstream-downstream differences in stream width, depth, velocity, and substrate. Third, richness and diversity of fish species around beaver dams were linked to habitat heterogeneity. Fourth, the mechanisms by which beaver dams altered fish biodiversity were mediated through habitat changes at the beaver dam patch boundary. Upstream of the dam macrohabitat guilds occupied the lentic areas, while below dams, fluvial fish guilds used shallow, faster water. Fifth, fluvial species responded the most dramatically to these habitat changes. Finally, in a system depauperate of lotic habitat, fluvial habitats created below beaver dams provided an important refuge for native stream fish. These source areas can increase resiliency and maintaining them may be useful for sustainable watershed conservation plans in these types of systems.

<p> Eulachon smelt <i>Thaleichthys pacificus</i> are anadromous forage fish of the North Pacific Ocean that annually spawn in coastal rivers of North America in late winter and early spring. These spawning runs range from northern California to southwestern Alaska and provide important resources to nearby communities, indigenous cultures, and wildlife predators. However, eulachon life history is not well understood or documented throughout their range. In recent years, concerns for eulachon population abundances in the southern portions of their range have led to federal protection. Though there are no federal listings in Alaska, there have been local concerns documented for eulachon runs of the Chilkat and Chilkoot rivers since approximately 1990. However, eulachon run timing and abundance trends are difficult to detect due to limited available data and variability in eulachon runs. To document baseline information and explore patterns of eulachon runs of the Chilkat and Chilkoot rivers, we sought local and traditional knowledge from residents of nearby communities to document information about local uses, run timing, abundance, and wildlife observations related to eulachon runs. Observations of eulachon runs were integrated with historical records from newspaper articles and scientific reports to construct temporal trends in eulachon run timing and abundance. </p><p> Based on the findings of this study, annual eulachon runs of the Chilkat and Chilkoot rivers generally occur for about a week or two between mid-April and mid-May. The arrival dates of eulachon runs often vary from year to year, but the timing appears to have shifted earlier, from mid-May to mid-April, over the past couple of decades. Abundance records were not sufficient to quantify trends. However, qualitative information regarding abundance did not suggest any clear trends in eulachon abundances of the Chilkat and Chilkoot rivers over the years, nor did there appear to be prominent local concerns about abundance declines. Many respondents suggested that eulachon populations were naturally too variable to be able to describe trends in abundance. Interviews also provided insight into local perspectives on eulachon life history and ecology. These results suggest that variability in eulachon run timing and abundance could be related to environmental conditions, including tidal height, river habitat, and water temperature. For a data-limited species like eulachon, integrating local observations and historical records offers a promising approach to documenting baseline information and improving the scientific understanding of eulachon runs and other environmental phenomena.</p>