The spatial organization of CAFOs and its relationship to water quality in the United States

Abstract

• Applied Nearest Neighbor Index and Moran’s I to quantify the spatial pattern of AFOs. • Watersheds with higher AFO NNI were associated with elevated nutrient levels. • Clustering of smaller farms (less regulated) also corresponded to high nutrients. • Results help to inform management of AFOs and other pollutant sources. Large animal operations in the United States (US) have expanded in recent decades as a solution for meeting increasing food demand. These operations can release large amounts of organic matter and nutrients into the environment, potentially impacting water quality conditions of nearby waterbodies. Often referred to as animal feeding operations (AFOs) or CAFOs (concentrated AFOs) based on nomenclature in federal regulations, these operations are usually clustered in space, which may intensify the impacts observed. However, the influence of the known spatial aggregation of CAFOs on water quality conditions has not been evaluated. This study investigates the spatial organization of CAFOs and its relationship to water quality conditions. We hypothesized that watersheds with clustered CAFOs are likely to be associated with higher concentrations of Total Phosphorus (TP) and Total Nitrogen (TN) in the US. We gathered and digitized CAFO locations in 16 US states and used TP and TN concentration data to evaluate our hypothesis. We calculated the flow-weighted mean concentration of TN and TP from 2005 to 2014 and averaged those values per season. We used the nearest neighbor index (NNI) to determine the spatial pattern (i.e., clustering or dispersed) of CAFOs per watershed and its significance level. Additionally, for states with animal number information per farm, we performed a Local Moran’s I analysis to understand if the clustering of CAFOs with a low or high number of animals relates to the TP and TN concentrations of each nearest water quality station. Overall, we found that watersheds with significant clustering patterns were associated with higher TP and TN levels. Seasonally, this finding also held principally during spring and summer seasons, when manure tends to be land applied. Our Local Moran’s I analysis also revealed that the clustering of smaller and medium-sized farms, usually less regulated, is associated with elevated TP and TN concentrations. This reveals that the small, yet potentially unregulated farms could be highly impactful to the environment, which warrants further study. Given the lack of information on the influence of CAFO clustering (and other pollutant source clustering) on water quality and that most states do not regulate emitters based on this aspect, our outcomes advance the knowledge of how these entities may be driving changes to the environment. This study also brings insights into new water quality modeling approaches and supports future policy decisions.