Over the past 150 years, much of the Midwestern US has been converted from prairies and wetlands to intensive row crop productions of corn and soybeans. The effects of agricultural land use, including riparian vegetation removal, substrate homogenization, and altered hydrologic regimes, manifest as a complex array of physical and chemical changes to stream ecosystems that influence biogeochemical cycling. As a result, agricultural watersheds, also known as "agroecosystems", export excess nitrogen (N) and phosphorus (P), causing numerous water quality problems including drinking water contamination, loss of biodiversity, and eutrophication of downstream freshwaters and coastal zones. My dissertation research examined linkages between terrestrial land cover and stream ecosystem function by determining the role of abiotic controls on ecosystem processes, including nutrient uptake and ecosystem metabolism. I also contrasted denitrification and ecosystem respiration in restored and naturalized floodplains in an agricultural ditch to examine the role of hydrologic connectivity in an agricultural stream. Finally, I examined the effect of shifting land cover at the watershed-scale on N and P loss in a small, agricultural watershed.I found that size and heterogeneity of benthic substrate influenced nutrient processing in streams by controlling spatial and temporal variability of nutrient demand by biological assemblages. Stable, inorganic substrate promoted algal biofilm development under stable flow conditions, enhancing removal of N and P from the water column. Subsequently, algal biofilm development contributed to benthic organic matter accrual over time and demonstrated the critical link between gross primary production and ecosystem respiration in open canopy streams where light is abundant. These findings emphasized the role that managing benthic substrate can play in restoring ecosystem function to agricultural streams, particularly in the face of increasing disturbance events. I also found that restored, in-set floodplains supported higher rates of denitrification than naturalized floodplain components, i.e., formed via hydrologic variation. Floodplain stability and frequent inundation promoted organic matter accumulation on restored floodplains compared to naturalized, effectively jumpstarting the establishment of characteristics needed to increase N removal. Finally, I found that planting cover crops increased N and P retention in agricultural fields thereby preventing nutrient loss to adjacent streams and downstream systems. Hydrologic variability masked the signature of cover crops at coarse temporal scales (i.e., months, years), hower, cover crops effectively managed tile inputs and storm export of both NO3- and SRP in an agricultural watershed.