When the pandemic forced researchers to rethink their fieldwork plans, we saw an opportunity. Could existing data provide important information about how these relatively pristine coastal wetlands are and will continue to respond to climate change? That question set the foundation for our recent study, published as “Freshwater subarctic wetlands are vulnerable to future thermal stress from climate warming.”
A temperature story written in data
The USDA Forest Service collected water temperature data from 20 freshwater ponds across southcentral Alaska’s Copper River Delta and Yakutat Forelands from 2012 to 2020. These remote sites are critically important since they serve as freshwater rearing habitats for several species of Pacific salmon before they journey to the ocean to mature.
The data revealed an elegant but concerning pattern. Unlike salmon streams buffered by cold groundwater or glacial meltwater, these wetland ponds showed remarkably strong coupling with air temperature. Their water temperature closely tracks air temperature, making them perfect candidates for climate forecasting but also revealing their vulnerability. If air temperature rises, these ponds have few mechanisms to resist warming.
Building a forecast with a purpose
With guidance from an ecological forecasting expert on our team, we used Bayesian state-space modeling to project pond temperatures through the end of the century under two emissions scenarios (RCP 4.5 and RCP 8.5). The challenge was to ensure that our models reflected reality, so we developed bias corrections to align the climate models with actual conditions, ensuring our forecasts reflected the specific conditions of these coastal systems.
A warming mosaic
The results were stark. Under RCP 8.5, we projected that pond water temperatures in the Copper River Delta will warm by 11°C by the end of the century, with the Yakutat Forelands warming by nearly 10°C. Even under the more conservative RCP 4.5 scenario, warming of 7°C and 6°C, respectively, is expected.
The greatest temperature increases occurred in fall and winter, meaning less ice cover and warmer conditions during salmon egg incubation, potentially accelerating development in ways that could create mismatches with food availability when juveniles emerge. But more urgently, we found that pond temperatures are likely to reach the growth cessation threshold of salmon by the end of the century with an 80-90% probability under RCP 8.5 and xx-xx% under RCP 4.5. In the absence of rapid evolution by salmon to tolerate warmer conditions, our findings suggest dire conditions for a species that is ecologically, economically, and culturally foundational to Alaska and the entire Pacific Rim.
Remote science, collaborative spirit
This project exemplifies the power of collaboration across institutions and expertise. Our laboratory has collaborated with the USDA Forest Service for decades, building strong relationships with the Cordova and Yakutat Ranger Districts of the USFS. Therefore, we had crucial context about these systems, including their hydrology, their role in salmon life cycles, and the management challenges facing these landscapes. To this team, we brought climate modeling expertise, fisheries experience, and novel insights on aquatic ecosystems.
Why it matters beyond salmon
These coastal ponds support complex food webs and migratory birds, such as Red-Necked Grebes and Aleutian Terns, several listed species of birds, and a myriad of plants, invertebrates, fish, and mammals. Our findings provide a template for understanding how static water bodies across warming Arctic and subarctic regions will respond to climate change.
Looking ahead
We hope this work encourages others to examine similar systems with the same forecasting lens. We also hope it sparks research into how salmon and other species are already responding to ongoing climate change by adjusting their timing, habitat use, and physiology. This project demonstrated that meaningful science can emerge from asking the right questions of existing data and collaborating across entities and disciplines. The future, while uncertain, is not unknowable; we can see it coming in the data and models. The ponds will keep warming, the question is whether we will give their inhabitants the habitat diversity, connected landscapes, and management foresight they need to adapt.