A toolbox for calculating net anthropogenic nitrogen inputs (NANI)

Abstract

The “Net Anthropogenic Nitrogen Input” (NANI) to a region represents an estimate of anthropogenic net nitrogen (N) fluxes across its boundaries, and is thus a measure of the effect of human activity on the regional nitrogen cycle. NANI accounts for the fluxes of atmospheric N deposition, fertilizer N application, agricultural N fixation, and net food and feed imports, each of which represents a potentially important source of nitrogen in watersheds. In watersheds of sufficient size for the accounting to make sense, NANI has been shown to have a strong relationship with riverine nitrogen export. Since its first introduction, it has been applied to various regions of the US and the world by numerous researchers. Comparison and synthesis of these studies have been somewhat problematic due to discrepancies in the data sources and assumptions applied in their calculations. To address this problem, it would be useful to have a set of tools for calculating NANI from a standardized dataset allowing the user to clearly see the assumptions behind the calculations and alter them to test their sensitivity. Here we present a prototype of such a toolbox, the NANI Calculator Toolbox. The toolbox comprises three sets of tools: (1) a GIS tool overlaying a watershed map with various data maps to calculate overlaying proportions, (2) a set of Excel-based tools extracting necessary data from various standardized datasets (Agricultural Census, Census, USGS fertilizer input estimates, and CMAQ N deposition estimates), and (3) an Excel-based tool calculating the NANI budget and other related components (e.g., animal excretion) from the extracted data. All the datasets are included in this package, allowing the user to calculate NANI anywhere in the US from the watershed map of interest. As a demonstration, we present county-scale NANI estimates for the entire continental US and also for all the US watersheds where NANI have been reported in the earlier literature. Possible improvements of the toolbox and extension to European datasets are discussed.