Stream temperature response to changing glacier coverage in coastal watersheds of southeast Alaska
Jason Fellman1, email@example.com
Sonia Nagorski1, Andy Vermilyea1, Sanjay Pyare1, Durelle Scott2, Eran Hood1
1University of Alaska Southeast; 2 Virgina Polytechnic Institute
Glacial environments are sensitive to changes in climate because of linkages between atmospheric conditions, glacier mass-balance and runoff, and water quality. Changes in glacier mass-balance that result in declining summer flows could therefore have substantial hydroecological effects, especially for stream temperature. We measured stream temperature continuously during the 2011 summer runoff season (May through October) in nine watersheds of coastal southeast Alaska that provide spawning habitat for Pacific salmon. Five of the watersheds have glacier coverage ranging from 9 to 55%. Our goal was to determine how air temperature and watershed land cover, especially glacier coverage, influence stream temperature across the seasonal hydrograph. Our results provide insight into how changing glacier coverage may influence stream thermal suitability for Pacific salmon.
Mean stream temperature for the runoff season ranged from a low of 3.4°C in the highly glacial Mendenhall River to a high of 14.7°C in the non-glacial Auke Creek, which has a large lake upstream of the sampling site. Stream temperature strongly reflected contributions of glacial meltwater during the summer months of July through September with the highest temperatures observed in the non glacial-fed streams and lowest temperatures observed in the highly glaciated watersheds. Watershed glacier coverage was negatively correlated with mean monthly stream temperature (r2 values ranged from 0.52 to 0.89, all p<0.04) for all months except May (r2=0.13, p=0.33). Mean monthly stream temperature was also unrelated to mean monthly air temperature for all sites except two non-glacial streams. Furthermore, the maximum weekly average stream temperature (MWAT, an index of thermal suitability for fish species) was less than 8.0°C for the five glacial-fed streams, which is well below the optimum of 12.8–14.8°C for coho salmon growth. A linear regression between MWAT and watershed glacier coverage (r2=0.97, p<0.01) predicted an increase in MWAT of 0.8°C for every 10% reduction in glacier coverage. Under this scenario, some future reduction in glacial meltwater input may enhance salmon survival in heavily glaciated streams in northern southeast Alaska through greater food availability and improved thermal suitability for salmon physiology.