Climate change is predicted to increase water temperatures in many lotic systems, but little is known about how changes in air temperature affect lotic systems heavily influenced by groundwater. Our objectives were to document spatial variation in temperature for spring-fed Ozark streams in Southern Missouri USA, create a spatially explicit model of mean daily water temperature, and use downscaled climate models to predict the number of days meeting suitable stream temperature for three aquatic species of concern to conservation and management. Longitudinal temperature transects and stationary temperature loggers were used in the Current and Jacks Fork Rivers during 2012 to determine spatial and temporal variability of water temperature. Groundwater spring influence affected river water temperatures in both winter and summer, but springs that contributed less than 5% of the main stem discharge did not affect river temperatures beyond a few hundred meters downstream. A multiple regression model using variables related to season, mean daily air temperature, and a spatial influence factor (metric to account for groundwater influence) was a strong predictor of mean daily water temperature (r2 = 0.98; RMSE = 0.82). Data from two downscaled climate simulations under the A2 emissions scenario were used to predict daily water temperatures for time steps of 1995, 2040, 2060, and 2080. By 2080, peak numbers of optimal growth temperature days for smallmouth bass are expected to shift to areas with more spring influence, largemouth bass are expected to experience more optimal growth days (21 – 317% increase) regardless of spring influence, and Ozark hellbenders may experience a reduction in the number of optimal growth days in areas with the highest spring influence. Our results provide a framework for assessing fine-scale (10 s m) thermal heterogeneity and predict shifts in thermal conditions at the watershed and reach scale.
The imperiled crayfish Orconectes williamsi is found only in the upper White River drainage in Missouri and Arkansas, and is known from only nine localities in four Missouri counties. Its distribution has never been thoroughly assessed. Knowledge of distribution and habitat associations is essential to conservation and management of rare species. Our major objectives were to estimate the distribution of O. williamsi in Missouri streams and identify associations between this crayfish's presence and selected multi-scale environmental variables. We used a probabilistic method and stratified (by stream order) random design to survey 50 of 223 total stream segments in the drainage during 2002 and 2003. We sampled 21 additional segments postsurvey (2004 and 2005). This crayfish was detected at an overall rate of 0.34 (±0.12, 95% confidence interval), and at a total of 27 sites, but we failed to collect it at two historical sites. Many stream segments harboring O. williamsi were determined to be effectively isolated from similarly populated segments by reservoirs. At the microhabitat scale, O. williamsi was associated with high current velocities, shallow depths, and cobble and small boulder substrates. At the reach scale, O. williamsi was positively associated with stream channels that were narrower and deeper relative to channels at segments where we did not detect the crayfish. This species was also negatively associated with emergent macrophyte (Justicia sp.) densities. No watershed-scale variables showed statistically significant associations with this crayfish, but the variable “drainage area” appeared to be the most likely of these to influence its distribution. Study results and our survey method will focus future searches or surveys for this crayfish. Whereas threats to O. williamsi remain, we suggest that its state rank be downgraded from “critically imperiled” to “imperiled” and that conservationists consider a reduction in this species' assigned global conservation status.