Riverscape and Geophysical Drivers
Physical and geologic covariates we hypothesized drove variation in flow regimes were collected for each watershed (Table 1). We used the National Hydrography Database Plus Version 2 data layer in ArcGIS Pro (ESRI Redlands, CA) to obtain the slope of each stream segment, and the area and mean slope for the watershed upstream of each site. We used the USDA Web Soil Survey (NRCS 2023) for McCurtain County to calculate geological metrics for each watershed, including mean depth to water table and depth to impermeable layer, because groundwater inputs and geologic permeability exert influences on streamflow (Costigan et al. 2015; Bourke et al. 2021). Additionally, we calculated the percentage of each watershed that is forested, because vegetative type and cover influence water movement, water use and evapotranspiration patterns (Chappell and Tych 2012; Ellison et al. 2012). We imported a 2020 land cover type dataset from the LandFire database (LANDFIRE 2020) into ArcGIS Pro and determined the cumulative percentage of each watershed that was covered by any category of forest.