Connections among soil, ground, and surface water chemistries characterize nitrogen loss from an agricultural landscape in the upper Missouri River basin

Abstract

Elevated nitrate in shallow aquifers is common in agricultural areas and remediation requires an understanding of nitrogen (N) leaching at a variety of spatial scales. Characterization of the drivers of nitrate leaching at the mesoscale level is needed to bridge from field-scale observations to the landscape-scale context, allowing informed water resource management decisions. Here we explore patterns in nitrate leaching rates across a depositional landform in the northern Great Plains within the Upper Missouri Basin, where the predominant land use is non-irrigated small grain production, and nitrate-N concentrations above 10 mg L1 are common. The shallow Moccasin terrace (260 km2) aquifer is bounded in vertical extent by underlying shale and is isolated from mountain front stream recharge, such that aquifer recharge is dominated by infiltration of precipitation through agricultural soils. This configuration presents a simple landform-scale water balance that we leveraged to estimate leaching rates using groundwater nitrate concentrations and surface water discharge, and quantify uncertainty using a Monte Carlo approach based on spatial variation in observations of groundwater nitrate concentrations. A participatory research approach allowed local farmer knowledge of the landscape to be incorporated into the study design, improved selection of and access to sample sites, and enhanced prospects for addressing nitrate leaching through collaborative understanding of system hydrology. Mean landform-scale nitrate-N leaching rates were 11 and 18 kg during the 2012-2014 study for the two largest catchments draining the terrace. Over a standard three-year crop rotation, these leaching rates represent 19 to 31% of typical fertilizer N application rates; however, leaching losses are likely derived not only from fertilizer but also from soil organic N mineralization, and are apparently higher during the post-fallow phase of the crop rotation. Groundwater apparent age is relatively young (0-5 yr) based on tritium-helium analysis, but whole-aquifer turnover time calculations are an order of magnitude longer (20-23 yr), suggesting changes in groundwater may lag behind changes in land management by years to decades.