John Wiley & Sons, Ltd.

A probabilistic approach for estimating the spatial extent of pesticide agricultural use sites and potential co‐occurrence with listed species for use in ecological risk assessments

0
A crop footprint refers to the estimated spatial extent of growing areas for a specific crop, and is commonly used to represent the potential “use site” footprint for a pesticide labeled for use on that crop. A methodology for developing probabilistic crop footprints to estimate the likelihood of pesticide use and the potential co‐occurrence of pesticide use and listed species locations was tested at the national scale and compared to alternative methods. The probabilistic aspect of the approach accounts for annual crop rotations and the uncertainty in remotely sensed crop and land cover datasets. The crop footprints used historically are derived exclusively from the National Land Cover Database (NLCD) Cultivated Crops and/or Pasture/Hay classes. This approach broadly aggregates agriculture into two classes, which grossly overestimates the spatial extent of individual crops that are labeled for pesticide use. The approach also does not use all the available crop data, represents a single point in time, and does not account for the uncertainty in land cover dataset classifications. The probabilistic crop footprint approach described herein incorporates best available information at the time of analysis from the National Agricultural Statistics Service (NASS) Cropland Data Layer (CDL) for 5 years (2008–2012 at the time of analysis), the 2006 NLCD, the 2007 NASS Census of Agriculture, and 5 years of NASS Quick Stats (2008–2012). The approach accounts for misclassification of crop classes in the CDL by incorporating accuracy assessment information by state, year, and crop. The NLCD provides additional information to improve the CDL crop probability through an adjustment based on the NLCD accuracy assessment data using the principles of Bayes' Theorem. Finally, crop probabilities are scaled at the state level by comparing against NASS surveys (Census of Agriculture and Quick Stats) of reported planted acres by crop. In an example application of the new method, the probabilistic crop footprint for soybean resulted in national and statewide soybean acreages that are within the error bounds of the average reported NASS yearly soybean acreage over the same time period, whereas the method using only NLCD resulted in an acreage that is over four times the survey acreage. When the probabilistic crop footprint for soybean was used in a co‐occurrence analysis with listed species locations, the number of potentially proximal species identified was half the number based on the standard NLCD crop footprint method (276 species with the probabilistic crop footprint versus 511for the conventional method). The probabilistic crop footprint methodology allows for a more comprehensive and representative understanding of the potential pesticide use footprint co‐occurrence with endangered species locations for use in effects determinations. This article is protected by copyright. All rights reserved

Customer comments

No comments were found for A probabilistic approach for estimating the spatial extent of pesticide agricultural use sites and potential co‐occurrence with listed species for use in ecological risk assessments. Be the first to comment!