Raw data analysis and visualisation
Analyses of raw data were performed to graphically visualize the patterns of refuge use by predators and herbivores. For this, we built linear mixed effects models (LMEs) and evaluated the effects of treatments (control vs rolled leaves) in raw data using nlmepackage in R (Pinheiro et al ., 2020). Treatment was the fixed effect, and block was the random effect. We then used these LME models to generate the predictive regression trends graphically usingvisreg (Breheny and Burchett, 2017) and ggplot2 (Wickham, 2016) packages in R environment.