Using a Bayesian Framework to explain changes in insecticide use in Canada over 20 years

This file contains R code and data to run the analysis for the article:

Malaj, E., and Morrissey CA. (2021). Increased reliance on insecticide applications in Canada linked to simplified agricultural landscapes Ecological Applications

Depends

To run this analysis the files listed below are required.

1. Main raw data: findatCD.RData in folder data. CleanCensusDat repo explains the data clean process used;
2. Spatial Data #1: camapF2.RData is a fortified Canada polygon file produced by running the code Canada_map_plot.R in folder supportCode. Run that file, save it as camapF2.RData and then continue with the points below;
3. Spatial Data #2: download the shapefile 2016CD_ag.shp from repo explore_agrochemicals;
4. Spatial Data #3: pud_insecticides.RData is a raster file of spatial distribution of insecticide use density as produced in Malaj et al. (2020): https://doi.org/10.1016/j.scitotenv.2019.134765. Data is produced in a similar way as the repo SpatialAnalysis_Mapping;
5. Land Use Data: The csv files aafc_crop_classifications.csv, landuseCDCensus2010.csv, landuseCDCensus2015.csv and SemiNatural.csv are all used for the analysis in correlationsIndex.R;
6. Code #1: INLA_models.R the main R code to run the analysis and produce figures and tables for the Bayesian models;
7. Code #2: correlationsIndex.R R code to run the formatting, extraction and plotting of the correlation between proportion of cropland and proportion of semi-natural habitats, and linear relationship between insecticide area (ha) and insecticide mass (kg);
8. Code #3: Java code to produce SemiNatural.csv data which was used in code correlationsIndex.R
9. Support functions: three R code files in folder supportCode to help run the main analysis in code INLA_models.R.

Results

Folder figures contains all figures generated from this analyis.

R version 4.0.3

Packages: tidyverse, doBy, INLA, rgdal, raster, spdep, egg, cowplot, plyr, glmmTMB, DHARMa, emmeans, MCMCglmm, coda, ape