Climate change influence on brook trout populations in the Central Appalachians

Examining variation in fish populations over time is strengthened when climatic variations are understood and incorporated into analyses. We used a 18-year brook trout (Saivelinus fontinalis) data set with samples across a similar to 4800 km(2) spatial area in the Central Appalachian Mountains, combined with PRISM climate data at the HUC-12 subwatershed level to investigate temporal trends of each. We found significant increases in air temperature (p <.01) with no significant trends in precipitation across time, including no trends in changes in variation of temperature and precipitation also. Extreme rainfall events (# of days above 95% percentile for total daily precipitation in a year) and consecutive dry days (cumulative # of consecutive days with <3 mm precipitation in a year) were also examined arid produced no significantly changing trends over time. Using random forests, seasonal climate variables identified as changing over time explained 35.1% of the variability in young-of-the-year brook trout abundance over the 18-year monitoring period. The same procedure explained 25.1% of the variability in juvenile abundance and only 5.1% in adult abundance. Variable selection following random forest generation illustrated significant effects for winter and spring precipitation (both negative on young-of-the-year abundance. Spring degree days and spring precipitation both had significant negative effects on juvenile abundance. Random effects for stream and year explained a maximum of 27.7% and 10.8% of the variation in abundance across all models. A significant, positive interaction between spring degree days and precipitation was also identified, lending support to the important balance between temperature and flow in these systems that is critical to brook trout persistence. This work provides long-term evidence to help understand the dynamics of these sentinel headwater fish populations as they experience a changing climate.