• Ecological interactions among important groundfishes in the Gulf of Alaska

      Barnes, Cheryl L.; Beaudrea, Anne H.; Dorn, Martin W.; Holsman, Kirstin K.; Hunsicker, Mary E.; Mueter, Franz J. (2019-12)
      Complex ecological interactions such as predation and competition play an important role in shaping the structure and function of marine communities. In fact, these processes can have greater impacts than those related to fishing. We assessed ecological interactions among economically important fishes in the Gulf of Alaska - a large marine ecosystem that has recently undergone considerable shifts in community composition. Specifically, we developed an index of predation for Walleye Pollock (Gadus chalcogrammus) to examine spatiotemporal changes in consumption, quantify portfolio effects, and better understand diversity-stability relationships within the demersal food web. We also evaluated the potential for competition between two important pollock predators, Arrowtooth Flounder (Atheresthes stomias) and Pacific Halibut (Hippoglossus stenolepis). We found highly variable predation intensity on Gulf of Alaska pollock. The combination of a single dominant predator and synchronous consumption dynamics indicated strong top-down control in the region. Spatial heterogeneity, however, may offset trophic instability at the basin scale. Assessments of resource partitioning provided little indication for competition between Arrowtooth Flounder and Pacific Halibut of similar lengths. Morphological differences between the two flatfish predators prompted an exploration into whether our conclusions about resource partitioning were dependent upon the size metric used. From this study, we found a relatively early onset of piscivory for Arrowtooth Flounder. Relationships between predator size and prey size also suggested gape limitation among Pacific Halibut sampled. Trophic niche separation was more pronounced for fishes with larger gapes, indicating greater potential for competition among smaller Arrowtooth Flounder and Pacific Halibut in Southeast Alaska. Reexamining basin-scale relationships between spatial and dietary overlap according to gape size would further elucidate the effects an increasing Arrowtooth Flounder population has had on changes in Pacific Halibut size-at-age. Results from this dissertation improve our understanding about the impacts of complex ecological interactions on population and community dynamics, and how those interactions may change in time, space, and under different environmental conditions.
    • Exploring the potential role of late stage predation and Chinook salmon age structure

      Manishin, Kaitlyn A.; Seitz, Andrew C.; Westley, Peter A. H.; Cunningham, Curry J.; Goldman, Kenneth J. (2018-12)
      Chinook salmon (Oncorhynchus tshawytscha) populations across the North Pacific have displayed a decrease in body size-at-return resulting from declines both in age- and body size-at-maturity. These changes have precipitated the loss of the oldest age classes in some populations and have occurred throughout the range of this species, suggesting a shared - yet currently unknown - driver in the common marine environment. A hypothesis for the cause of these changes is intense and/or selective predation marine mortality after the first winter in the ocean, potentially from predators selectively removing relatively large sub-adult Chinook salmon. Here I consider the question: under what circumstances could predation on large sub-adult individuals by salmon sharks (Lamna ditropis) change the age structure of a Chinook salmon population? To address this question, I first estimated total per capita prey consumption by salmon sharks - an increasingly acknowledged predator of salmon on the high seas - using three methods: 1) daily ration requirement, 2) bioenergetic mass balance, 3) and a von Bertalanffy growth model. Second, I examined the effects of additional predation on an indicator Chinook salmon population from the Yukon River by simulating alternative predation scenarios with a stage-structured life cycle model. Scenarios described the strength and selectivity of predation, and the resulting simulated age structure was then compared to observed demography. The selectivity and intensity of removals required to produce this change in age structure were considered in the context of top predators, focusing on salmon sharks. The daily ration method yielded individual salmon shark consumption estimates of 1461 and 2202 kg·yr-1, the mass-balance method produced estimates of 1870 kg·yr-1, 2070 kg·yr-1, 1610 kg·yr-1, and 1762 kg·yr-1, depending on assumed diet, and the growth model output estimates of 16,900 kg·yr-1 or 20,800 kg·yr-1, depending on assumed assimilation efficiency. The per capita prey consumption estimates from the mass-balance method may be the most realistic because they incorporated life history data specific to salmon sharks and did not produce extreme values. Taken as a whole, these estimates suggest salmon sharks have energetic requirements similar to those of piscivourous marine mammals and corroborates conclusions of previous research suggesting that endothermic fishes exhibit metabolic rates similar to marine mammals. The simulated mortality scenarios that most closely mimicked observed shifts in age structure of the indicator Chinook salmon population focused intense and selective predation on the third year of Chinook salmon residence in the ocean. This simulated predation is corroborated by emerging results from an independent electronic tagging study in which tagged Chinook salmon experienced high predation rates, and research suggesting that killer whales (Orcinus orca) selectively prey upon Chinook salmon in their third year at sea. In summary, salmon sharks likely have high energetic requirements that could result in a large biomass of prey consumed, Chinook salmon populations are sensitive to predation during the third ocean year, and salmon sharks and other predators appear to frequently consume fish at that ocean stage. Taken together, these lines of evidence point to a potentially important mechanism for top down pressure on Chinook salmon populations that may explain observed changes in age-at-return, which in turn can affect population productivity. Future work and more robust data on predator distributions and abundances are needed to explore this finding further.
    • Impacts of a top predator (Esox lucius) on salmonids in Southcentral Alaska: genetics, connectivity, and vulnerability

      Jalbert, Chase S.; Falke, Jeffrey; Westley, Peter; López, J. Andrés; Dunker, Kristine (2018-12)
      Worldwide invasion and range expansion of northern pike (pike; Esox lucius) have been linked to the decline of native fishes and new techniques are needed to assess the effects of invasion over broad geographic scales. In Alaska, pike are native north and west of the Alaska Mountain Range but were introduced into Southcentral Alaska in the 1950s and again in the 1970s. To investigate the history of the invasion into Southcentral Alaska, I identified 7,889 single nucleotide polymorphisms (SNPs) from three native and seven introduced populations in Alaska and examined genetic diversity, structure, and affinities of native and invasive pike. Pike exhibited low genetic variability in native populations (mean heterozygosity = 0.0360 and mean π = 0.000241) and further reductions in introduced populations (mean heterozygosity = 0.0227 and mean π = 0.000131), which suggests a bottleneck following introduction. Population differentiation was high among some populations (global FST = 0.424; max FST = 0.668) when compared to other freshwater fishes. I identified five genetically distinct clusters of populations, consisting of three native groups, a single Susitna River basin invasive group, and a Kenai Peninsula group, with little evidence of admixture among groups. The extremely reduced genetic diversity observed in invasive northern pike populations does not appear to affect their invasion success as the species range Southcentral Alaska continues to expand. To assess the vulnerability of five species of Pacific salmon (Oncorhynchus spp.) to the invasion, I combined intrinsic potential habitat modeling, connectivity estimates, and Bayesian networks across 22,875km of stream reaches in the Matanuska-Susitna basin, Alaska, USA. Pink salmon were the most vulnerable species, with 15.2% (2,458 km) of their range identified as "highly" vulnerable. They were followed closely by chum salmon (14.8%) and coho salmon (14.7%). Finally, analysis of the intersection of vulnerable salmon habitats revealed 1,001 km of streams that were highly vulnerable for all five Pacific salmon. This framework is easy to implement, adaptable to any species or region, and cost effective. With increasing threats of species introductions, fishery managers need new tools like those described here to efficiently identify critical areas shared by multiple species, where management actions can have the greatest impact.
    • Navigating the predator gauntlet: consumption of hatchery- and wild-born juvenile chum salmon (Oncorhynchus keta) by common nearshore marine fishes in Southeast Alaska

      Duncan, Douglas H.; Beaudreau, Anne H.; McPhee, Megan V.; Westley, Peter A. H. (2018-12)
      Juvenile chum salmon (Oncorhynchus keta) undergo extensive mortality at marine entry, a period which is believed to be a potential population bottleneck. Although this early mortality has been consistently observed, our understanding of the mechanisms responsible is limited. Furthermore, the implications of large-scale salmon hatchery releases for the ecology of juvenile chum salmon and their consumers is another important knowledge gap. To better understand the predation responses of abundant consumers to hatchery- and wild-born juvenile chum salmon, we examined the diets of Pacific staghorn sculpin (Leptocottus armatus) and Dolly Varden (Salvelinus malma) near Juneau, Alaska, in 2016 and 2017. Chum salmon composed 4.5% and 19.6% of the diets of staghorn sculpin and Dolly Varden by weight, respectively, and 88% of chum salmon individuals consumed were of hatchery origin. Chum salmon prey were shorter than average when compared to chum salmon concurrently collected by beach seine and hatchery releases of chum salmon. Regression analyses indicated that occurrence of juvenile chum salmon in diets varied primarily by date and site. Predation generally occurred more frequently at sites closer to hatchery release areas. The quantity of chum salmon in staghorn sculpin stomachs was related to predator length, chum salmon catch-per-unit-effort (CPUE), and the proportion of hatchery fish present; however, date was the only important predictor explaining quantity of chum salmon in Dolly Varden stomachs. To translate diet data into consumption rate, we experimentally determined gastric evacuation rate for staghorn sculpin and implemented a field-based consumption model. Average daily consumption of chum salmon was low relative to all other prey groups. Estimates of average seasonal consumption of juvenile chum salmon by staghorn sculpins suggest that predator populations would have to be implausibly large to consume even 1% of local hatchery chum salmon production. Together, these results yield new insights into the interactions between the predators of wild-born and hatchery-born salmon during the critical stage of marine entry.