Abstract

Agriculture currently applies annually more than 100 terra grams of Nitrate fertilizer to crops, with major impacts on water quality, freshwater ecosystems, and human health. Meeting future food demands will require an increase in agricultural productivity, and thus closing the yield gap through increased fertilizer application on currently nutrient limited croplands. Yet, the researchers currently lack tools and data that allow for examining how these changes will impact nitrate concentrations in fresh and surface water, human exposure to dangerous levels of nitrate at a global scale To close this knowledge gap, we built a global dataset of 20,322 observations of nitrate concentration in ground and surface water. After some spatial averaging, we used these observations to train a random forest model that uses results of a global nutrient balance model with 300 m resolution (InVEST Nutrient Delivery Ratio model) and other global datasets of relevant environmental covariates to predict nitrate concentrations in surface and ground water. This framework enables us to (1) characterize current spatial patterns of nitrate concentration in ground and surface water, (2) quantify human exposure to drinking water nitrate levels commonly considered unhealthy (> 50 mg/l) using gridded estimates of population and (3) forecast future trends under different scenarios of agricultural expansion and intensification.
After training and spatial cross-validation, the modeling framework reaches an R2 of 0.766 for surface and 0.784 for groundwater. Our preliminary results reveal that up to 35 % of the world’s population might be exposed to unhealthy nitrate levels, with global hotspots in South Asia and North America. Expanding and intensifying agriculture to all suitable croplands would increase this percentage to more than 40%, with significant increases in the Global South, notably sub-Saharan Africa. Scenario analysis showed that implementing better agricultural practices, notably buffer strips along streams, in intensified food systems would allow for increasing food production with only a marginal increase in exposed population. Our framework highlights opportunities to study global water quality, with implications for aquatic conservation and agricultural policies, and highlights the role of nature-based solutions and best management practices to reduce the impacts of intensified agro-food systems.

Coauthors: Rafael Schmitt, Ian Madden, Ginger Kowal, James Douglas, Peter Hawthorne, James Gerber, Kate Brauman, Gretchen Daily, Justin Johnson, Lisa Mandle, Issoufou Ouedraogo, Atharva Rajesh Ekatpure, Mary Ruckelshaus, Jeffrey Smith, Stephen Polasky, Adrian Vogl, Becky Chaplin Kramer