Supplemental for "Differential Effects of Social Learning, Independent Learning, and Dropout in Scaling Conservation Uptake"

Community-based initiatives are recognised as key to achieving global conservation targets; however, the social mechanisms underpinning their uptake and persistence remain poorly described. Previous work modelling uptake dynamics pays little attention to dropout and relies on dissimilar case studies, limiting comparative inference around the importance of distinct mechanisms. We contribute a theoretical demonstration of how small changes in rates of social learning, independent learning, and dropout produce highly differential long-term uptake. We show these theoretical dynamics are consistent with the empirical patterns of uptake in a rangeland conservation initiative across five South African communities. In theory and practice, we show small changes in independent learning rates have disproportionate effects on early adoption, while small changes in dropout have large downstream effects on total uptake. We recommend that practitioners explicitly engage with these distinct mechanisms to inform scaling strategy, working with communities to understand how these dynamics can be improved. This repository contains the project files, R scripts, and associated materials for "Differential Effects of Social Learning, Independent Learning, and Dropout in Scaling Conservation Uptake". The working directory should be set to the "Project" folder for the scripts to run correctly. The adoption datasets analysed during this study are not publicly available, as they were provided under a data-sharing agreement between Conservation South Africa and Imperial College London and contain site-level programme and spatial information. Data may be made available from the corresponding author upon reasonable request, subject to approval by the data custodians (Conservation South Africa). Datasets to reproduce supplemental materials are included as these from open access sources (see Appendix 3a for references). To forecast your own adoption curves using Script 3, note the structure of the original raw dataset: "Year" and "Quarter" columns were mutated into a single "YQ" (year–quarter) column; "Total_People_RSA" recorded the total number of people engaged with an initiative at each YQ; and "No_dropped_out" recorded the number of people who left the initiative at each YQ. These transformations, along with renaming for clarity, are handled within Script 3 itself. Reviewing that script is recommended for anyone wishing to reproduce the workflow with their own dataset, since it documents the expected column structure and naming conventions.