Lack of Evidence for Pervasive DOC Concentration Increases in the Mississippi River Basin

Friday, 19 December 2014
Sarah M Stackpoole1, Douglas A Burns2, Edward Stets3, Robert G Striegl3, David W Clow4, Irena F Creed5, Robert M Hirsch6, Hjalmar Laudon7 and Brian A Pellerin8, (1)USGS-Branch of Reg Research, Denver, CO, United States, (2)USGS, Troy, NY, United States, (3)National Research Program Boulder, Boulder, CO, United States, (4)USGS Colorado Water Science Center Denver, Denver, CO, United States, (5)University of Western Ontario, London, ON, Canada, (6)USGS Headquarters, Reston, VA, United States, (7)SLU Swedish University of Agricultural Sciences Umeå, Umeå, Sweden, (8)USGS California Water Science Center Sacramento, Sacramento, CA, United States
Several recent studies have reported widespread increases in dissolved organic carbon (DOC) concentrations and loads in streams and rivers worldwide, but few studies have focused on temporal patterns in large rivers that drain landscapes influenced by agriculture. We analyzed annual changes in flow-adjusted concentrations (FACs) of DOC between 1997 and 2013 at eight sites in the Mississippi River Basin (MRB) using the Mann-Kendall Trend Test. The FACs represent the residuals of simple linear regression models evaluating the relationship between stream flow and DOC concentration. Our results indicate that statistically significant increases in DOC FACs occurred for only two of the eight sites, at Clinton, IA and Hermann, MO. Trends in DOC concentration were also determined using the Weighted Regressions on Time, Discharge, and Season (WRTDS) model, which is designed to examine the nature and extent of temporal trends in flow normalized (FN) concentrations but does not provide p-values for departure from the null hypothesis of stationary behavior. The results of the WRTDS model corroborate the results of the Mann-Kendall test, and the only considerable increases in FN DOC concentrations occurred at the sites near Clinton, IA and Hermann, MO, which had 26% and 15% increases in DOC FN concentrations during median to high flows, respectively. Changes for the six other sites were much lower, ranging from -4.9% to +6.9%, with the two most downstream sites at St. Francisville and Melville, Louisiana showing the decreasing changes. Since the WRTDS model eliminates the impact of variable hydrology, it allows us to focus on other factors that may drive changes in concentration over time. We couple the output from the WRTDS model with variables related to declining acid deposition, land management, and climate change in an attempt to explain the pattern in DOC concentration trends in the MRB.