Controls on nitrogen inputs, loads, and in-stream concentrations in the Altamaha River, Georgia, and beyond

Abstract

Human activity has increased the availability of reactive nitrogen (N), an element of critical importance to life. The broad goal of this dissertation was to understand which sources of N to a watershed reach rivers. Nitrogen input budgets were calculated for 18 watersheds on the United States west coast. Fertilizer was the most important source of new N, with atmospheric deposition second. N export was best correlated with streamflow variations, which explained 66% of the variance. Including inputs explained an additional 16% of the variance. Riverine N export averaged 12% of inputs. Percentage export was also best related to streamflow. These results were likely due to the large range of streamflows across the study region, which may have overwhelmed other factors contributing to N export. An in-depth examination of the Altamaha River watershed (Georgia, USA) found that riverine N concentrations and cumulative loads were best related to population density, rather than input factors. Concentrations were highest in the upper watershed and lowest in blackwater streams and a sampling station downstream of a dam. Isotopic analysis of nitrate suggested that most in-stream nitrate is derived from sewage or manure. Where the concentration was particularly low, an atmospheric signal was observed, suggesting that background nitrate is of atmospheric origin. N in this system appeared to be lost primarily on the watershed surface, as only a small proportion of inputs reached the stream, N was transported downstream conservatively, and an isotopic mixing model fit the data well. The Altamaha is comparable to other watersheds worldwide in its nitrate concentrations and population density. A metadata analysis of medium-sized (2,000-50,000 km2) watersheds, which are under-studied relative to large watersheds, revealed that nitrate concentrations are generally low (<25 µM). We found a relationship between population density and riverine NO3- concentration, which varied among continents and latitudinal zones. A combination of population density, population density without access to improved sanitation, temperature, precipitation, slope, and fertilizer use, explained 45% of the variation in nitrate concentrations. The relationship had less explanatory power than in large watersheds, potentially due to increased variation in NO3- concentrations with decreased watershed size.