Data and code from: Selection of the least drifting spray nozzles for pesticide application to protect insect pollinators

This dataset includes all raw data and statistical software code required to reproduce the analyses presented in the manuscript:

Kannan, N. and Q. D. Read. 2025. Selection of the least drifting spray nozzles for pesticide application to protect insect pollinators. Journal of the American Society of Agricultural and Biological Engineers, in review (citation and DOI pending).

Pesticide exposure, including off-target drift, is troubling the thriving beekeeping industry in Mississippi. The selection of the least drifting spray nozzle(s) is one way to protect insect pollinators. This dataset is associated with a manuscript presenting the results of a study whose major goal was to identify nozzles with low drift for pesticide spray applications. Specifically, we a) observed the size, speed and flow rate data from seventeen broadcast nozzles; b) estimated the proportion of spray volume with driftable size droplets; c) estimated the size and speed relationships for all the nozzles. These results enabled us to identify the nozzles that are less likely to drift during pesticide applications. The dataset includes raw observations for each nozzle at pressures ranging from 30-50 psi and at 1” height intervals from 9” to 20”, by droplet diameter class. It also includes summary statistics calculated from the data, specifically volume-based percentiles by diameter and by speed, Sauter mean diameter, and proportion of droplets < 150 µm diameter. In addition, we include R statistical software code to calculate the summary statistics, create data visualizations, fit linear models comparing the summary statistics by nozzle type within and across pressure levels, and fit a nonlinear Bayesian regression comparing the maximum slope of the size by speed relationship for each nozzle. The findings we present here may be beneficial to producers for selecting the nozzle that is optimal for both increasing the efficiency of pesticide application and minimizing collateral damage to honeybees and other pollinators.

Included files:

• all_nozzle_data.csv: Spreadsheet containing raw observations by droplet diameter class for each nozzle type, pressure level, and height level
• nozzle_summary_statistics.csv: Spreadsheet containing summary statistics calculated for each nozzle type, pressure level, and height level, as well as aggregated over height levels
• all_nozzle_distribution_comparison.html: Rendered output of code notebook including code, figures, tables, and text description
• all_nozzle_distribution_comparison.Rmd: RMarkdown source of notebook
• speed_bnlmm_asymp.rds: Pre-fit model object for the Bayesian nonlinear regression so that the results can be reproduced without having to refit the model
• column_metadata.txt: Text file explaining contents of the two CSV files