Data and code for "Push-pull farming system helps insure smallholder maize production under climate change"

Introduction This repository contains the data and code to reproduce the analyses in the paper "Push-pull farming system helps insure smallholder maize production under climate change" published in Agriculture, Ecosystems and Environment. As the detailed description of the data and methods can be found in that paper, the following is a shorter overview of the field data and the climate data used. Field data: Maize yield, stemborer infestation, and striga abundance were measured at local farmer fields across eight regions in Southwestern Kenya between 2005 and 2016. Each site contained a pair of maize fields managed by the same farmer, one under push-pull management and one not utilizing push-pull. Non-push-pull fields were sometimes maize monocultures while others were intercropped with groundnuts, greengrams, or other local crops, but none of the non-push-pull fields included Napier grass, signalgrass, or Desmodium companion crops. All non-push-pull fields were within 100m of their respective push-pull fields. Measurements were taken during both the long rains (March-August) and short rains (October-January) growing seasons. While some fields were sampled consistently between 2005 and 2010, all fields were sampled haphazardly from 2011 to 2016. Climate data: For the long and short rainy seasons, we calculated climate variables using CHIRPS at 0.05° spatial resolution for precipitation and ERA5-Land at approx. 0.07° spatial resolution for daily mean temperature using the grid cells whose center was closest to each location of the field data collection sites for each meteorological data set. We calculated the accumulated rainfall during the rainy seasons, the 90th percentile and 10th percentile of the daily mean temperature. We also assessed the length of dry spells for periods corresponding to different drought-sensitive crop stages.  Below is a description of the files contained in the repository. Results: 3.1 Rainy seasons and long-term trends in temperature and rainfall over time  variables_anon.csv:  Description of the climate variables used in the analysis of long-term trends in temperature and rainfall over time site_gridcells_*.csv: Three files that connect sites with the gridcells for the ERA5 Land, CHIRPS and ESACCI datasets, and the coordinates of the borders of these cells. datpy_clim.RData: Field and climate data; anonymized.  West_Kenya_clim_long.RData: Climate data for use in the analysis of long-term trends in temperature and rainfall over time (contains complete time series, not just in years with field data observations) climate_trend_models.R: R code to run the models Results: 3.2 Interactive effects of push pull and climate on maize yields, striga and stemborer damage  tspp.R: R code to plots some time series for climate data yield_models.R: R code to run the GLMMs for maize yields stem_models.R: R code to run the GLMMs for stemborer damage incidence striga_models.R: R code to run the GLMMs for Striga piecewise.R: R code to run the piecewise structural equation models. Results: 3.3 Maize yield, striga and stemborer means and variability analyses yield_striga_stem_variability.R: R code to run the yield, striga and stemborer variability models