Dissolved organic carbon (DOC) constitutes a small yet important part of a watershed's carbon budget. While DOC generally comprises less than 1% of the overall carbon budget, it is important because it is the most mobile and biologically reactive form of carbon. The primary vegetation present within a watershed, rainfall intensity, and duration of storm events may impact the concentration of DOC in runoff water and the amount of DOC exported from watersheds. Agricultural practices which promote carbon sequestration may also influence DOC concentrations and load in surface runoff, consequently impacting stream ecosystem processes.

In a long-term experiment at the Neal Smith National Wildlife Refuge in Jasper County, Iowa, USA, selected native vegetation perennial cover treatments were randomly assigned to twelve small agricultural watersheds in a balanced incomplete block design. Treatments applied to the watersheds consist of native perennial vegetation (NPV) strips varying in location and percentage of the total area within each agricultural watershed. One of four treatments was randomly assigned to each watershed. Three watersheds were planted in 100% row-crops, three with 10% NPV only in the footslope position, three with 10% of their area in NPV divided into two strips; one on the hillslope and one in the footslope position, and three watersheds with 20% in NPV with strips on the hillslope and footslope positions. Two additional watersheds planted in 100% NPV located in the Neal Smith National Wildlife Refuge were also monitored but are not part of the balanced incomplete block design.

Samples from 2008-2010 were analyzed for DOC concentrations and correlated with flow data to determine flow weighted DOC concentrations and total flux per watershed. All three years of the study experienced higher than normal precipitation. From analysis over the full three year study, for flow weighted DOC concentrations, treatment was significant (p = 0.09) only between 10% NPV at the footslope and 20% NPV in contours watersheds. During an extreme storm event August 8-11, 2010, flow weighted DOC concentrations from the 100% agricultural watersheds was significantly higher than from all of the NPV treatment watersheds (p = 0.008).

Watersheds planted in 100% agriculture exported greater DOC loads than from the 10% NPV at footslope watersheds over the three year study (p = 0.04) and during the storm event August 8-11, 2010 (p = 0.07). Results from this study show that of the four treatments, the conversion of 10% of an agricultural watershed's area into NPV in the footslope position significantly increased DOC concentrations but decreased export when compared to 100% agricultural watersheds. Results indicate that the incorporation of NPV as buffer strips may be a valuable land management tool to reduce DOC loading to levels exported from tallgrass prairie watersheds.