Spatial heterogeneity and controls of ecosystem metabolism in a Great Plains river network

Gross primary production and ecosystem respiration together define ecosystem metabolism and help indicate the importance of internal and external carbon sources. Spatial variability of these processes is poorly characterized in rivers. We measured metabolism in the Kansas River: (1) at 10 locations over 100 s of km in tributaries within the watershed and (2) over 20 km with detailed sampling in the main stem. Whole-river metabolism at the larger scale was decoupled from light, algal growth, and nutrient limitation, and was positively related to nutrients. Smaller-scale main stem sampling revealed almost as much variance over a few kilometers as the larger scale sampling. Local processes seemed to dominate dissolved oxygen dynamics, since diurnal dissolved oxygen patterns were better correlated with absolute time than data corrected for travel times. A single-station method compared against two-station metabolism methods indicated that local hotspots of metabolism occur at scales less than 1 km and that single-station estimates average out this variance. The main stem data provide support to the idea that functional processing zones control characteristics used to estimate system metabolism, but the nutrient effect at the whole watershed level indicates that transport from upstream can also be important.