Climate, embryonic development, and potential for adaptation to warming water temperatures by Bristol Bay sockeye salmon

Abstract

Rapidly warming water temperatures associated with climate change represent a substantial disturbance to the habitat of aquatic ectothermic organisms. For salmonid fishes (family Salmonidae), early life history survival and timing of reproduction and development are closely tied to temperature, such that altered thermal regimes could alter patterns of survival or shift phenology into a mismatch with the environment. Because temperature is the dominant driver of developmental rates, empirical statistical models have been developed to predict the timing of hatching and fry emergence based on incubation temperature. In this thesis I explored how the timing of hatching and emergence may shift in response to warming temperatures and how spawning timing across an Alaskan landscape is shaped by incubation temperatures experienced by sockeye salmon (Oncorhynchus nerka) embryos and alevin. Additionally, I quantified the relative roles of genetics and environmentally induced plasticity on the timing of hatching in two populations of sockeye salmon from the Iliamna Lake system, Alaska by rearing them in common garden conditions in the laboratory. To meet these goals I reformulated a widely cited developmental model to incorporate variability in natural regimes and use it to predict hatching timing over the course of the spawning duration for 25 populations of Bristol Bay sockeye salmon. Additionally, I hind- and forecasted lake temperature based off historical and predicted air temperatures to estimate and predict hatching for a single population. I found that predicted hatching timing for wild populations varied between 58 and 260 days, and was largely variable as a result of habitat thermal heterogeneity and parental spawn time. I also predicted a three-week decrease in hatching timing over the course of the next century for a single beach spawning population, which was just beyond historic variability. Counter to expectations, for a subset of populations hatching and emergence timing variability exceeded that of spawning timing, indicating the relationship between spawning timing and incubation temperature may be weaker than expected. The results of the common garden experiment revealed indistinguishable differences between populations in hatching timing across five temperature scenarios, but strong plasticity as timing differed between 74 and 189 days in the warmest to coolest treatment. Furthermore, I detected family-specific differences in hatching timing both within and among treatments consistent with heritable developmental rates and gene by environment interactions in days to hatch, where the interaction between treatment and family was as high as 10 days difference in hatching. Population or family-specific survival in this experiment did not differ in response to temperature suggesting a lack of thermal adaptation in this regard during this life stage in these populations. Alevin mass and length upon hatching varied little among treatments (<10%), but did significantly decrease with cooling temperatures. Taken as a whole this study indicates that the effects of climate change during the early life history stages may be buffered by phenotypic plasticity and variability in populations and habitats will be important for maintaining diversity in the face of climate change.