Data and code from: A natural polymer material as a pesticide adjuvant for mitigating off-target drift and protecting pollinator health

This dataset contains all data and code required to clean the data, fit the models, and create the figures and tables for the laboratory experiment portion of the manuscript:

Kannan, N., Q. D. Read, and W. Zhang. 2024. A natural polymer material as a pesticide adjuvant for mitigating off-target drift and protecting pollinator health. Heliyon, in press. https://doi.org/10.1016/j.heliyon.2024.e35510.

In this dataset, we archive results from several laboratory and field trials testing different adjuvants (spray additives) that are intended to reduce particle drift, increase particle size, and slow down the particles from pesticide spray nozzles. We fit statistical models to the droplet size and speed distribution data and statistically compare different metrics between the adjuvants (sodium alginate, polyacrylamide [PAM], and control without any adjuvants).

The following files are included:

• RawDataPAMsodAlgOxfLsr.xlsx: Raw data for primary analyses
• OrganizedDataPaperRevision20240614.xlsx: Raw data to produce density plots presented in Figs. 8 and 9
• raw_data_readme.md: Markdown file with description of the raw data files
• R_code_supplement.R: All R code required to reproduce primary analyses
• R_code_supplement2.R: R code required to produce density plots presented in Figs. 8 and 9

Intermediate R output files are also included so that tables and figures can be recreated without having to rerun the data preprocessing, model fitting, and posterior estimation steps:

• pam_cleaned.RData: Data combined into clean R data frames for analysis
• velocityscaledlogdiamfit.rds: Fitted **brms** model object for velocity
• lnormfitreduced.rds: Fitted **brms** model object for diameter distribution
• emm_con_velo_diam_draws.RData: Posterior distributions of estimated marginal means for velocity
• emm_con_draws.RData: Posterior distributions of estimated marginal means for diameter distribution

The following software and package versions were used:

• R version 4.3.1
• CmdStan version 2.33.1
• R packages:
• • brms version 2.20.5
  • cmdstanr version 0.5.3
  • fitdistrplus version 1.1-11
  • tidybayes version 3.0.4
  • emmeans version 1.8.9