Vulnerability of Recently Recharged Groundwater in Principle Aquifers of the United States To Nitrate Contamination

Recently recharged water (defined here as <60 years old) is generally the most vulnerable part of a groundwater resource to nonpoint-source nitrate contamination. Understanding at the appropriate scale the interactions of natural and anthropogenic controlling factors that influence nitrate occurrence in recently recharged groundwater is critical to support best management and policy decisions that are often made at the aquifer to subaquifer scale. New logistic regression models were developed using data from the U.S. Geological Survey’s National Water-Quality Assessment (NAWQA) program and National Water Information System for 17 principal aquifers of the U.S. to identify important source, transport, and attenuation factors that control nonpoint source nitrate concentrations greater than relative background levels in recently recharged groundwater and were used to predict the probability of detecting elevated nitrate in areas beyond the sampling network. Results indicate that dissolved oxygen, crops and irrigated cropland, fertilizer application, seasonally high water table, and soil properties that affect infiltration and denitrification are among the most important factors in predicting elevated nitrate concentrations. Important differences in controlling factors and spatial predictions were identified in the principal aquifer and national-scale models and support the conclusion that similar spatial scales are needed between informed groundwater management and model development. ■ INTRODUCTION Nitrate (NO3 −) is the most ubiquitous nonpoint-source (NPS) contaminant of groundwater resources worldwide. The reduction of NO3 − in groundwater can release the greenhouse gas nitrous oxide 2 and pose well-known ecological and humanhealth risks. The U.S. Environmental Protection Agency (USEPA) maximum contaminant level (MCL) for NO3 − in drinking water is 10 mg/L (as Nitrogen (N)). Health concerns, including methemoglobinemia or “blue baby” disorder, spontaneous abortion, and increased risk of nonHodgkins lymphoma, have been linked to drinking water with NO3 − concentrations as low as 2.5 to 4 mg/L. The annual cost for U.S. water treatment to meet federal NO3 − standards has been estimated in the hundreds of millions to billions of dollars. In the U.S., more than 40 million people, including most rural populations, obtain drinking water from domestic selfsupply (private) wells that generally receives no treatment prior to use, and about 40% of the Nation’s public water supply is from groundwater in principle aquifers (PAs). PAs are regionally extensive aquifers and aquifer systems of national significance because of their high productivity or use and are critically important sources of potable water. In 2000, total groundwater withdrawals from PAs were estimated at more than 289 million m per day. A recent (2010) national assessment of PA groundwater quality reported 50% of the 5,101 wells sampled had NO3 − >1.0 mg/L, which has been proposed as a nationwide relative background concentration or threshold indicative of anthropogenic sources. Aquifer-scale studies indicate that the distribution of NO3 − concentrations in groundwater vary by PA, with many PAs having median NO3 − >1.0 mg/L and higher percentiles that exceed the 10 mg/L MCL. Understanding the governing natural and anthropogenic factors of groundwater vulnerability to NPS NO3 − contamination within and among PAs is fundamental to developing effective policy and management practices to reduce N inputs and protect groundwater as a safe drinking-water source. Groundwater management is often implemented at the local or aquifer scale. However, prior vulnerability assessments have been conducted at the aquifer or national scales and along various points within the flow system that represent a range of apparent groundwater ages and residence times, including shallow (young) and deep (old) groundwater. Although NO3 − concentrations are highly variable at the Received: February 20, 2012 Revised: May 2, 2012 Accepted: May 14, 2012 Published: May 15, 2012 Article