Occurrence records of cashew pests and its environmental drivers in cashew plantations in Guinea-Bissau

This dataset comprises records of 21 abiotic factors, 2 biotic factors, and 3 variables related to pest occurrence across 151 cashew plantations (Anacardium occidentale L.) in Guinea-Bissau (West Africa). Abiotic variables were obtained from CHELSA v2.1 (Karger et al., 2021) as monthly data for 1981–2010 at 30 arc-second (~1 km) resolution. These included temperature variables (Bio1–Bio8), precipitation variables (Bio12–Bio19), vapor pressure deficit (kPa; monthly maximum—VaporH, monthly minimum—VaporL, annual mean—VaporA), wind speed (m s⁻¹; monthly maximum—WindH, monthly minimum—WindL, annual mean—WindA), and relative humidity (%; monthly maximum—RH). Distance to the coastline (m; Coast) was derived from the 2020 coastline map (Digital Earth Africa, 2023). The records are in Excel (xlsx) format. This document contains four worksheets: “Data” contains all the complete data and was used for the Kruskal-Wallis test, principal component analysis (PCA), and boxplot graphs. The remaining 3 sheets correspond to the data used by agroecological zone for the a Generalized Additive Models (GAMs) model framework (“DataChelsa_variables_East,” “DataChelsa_variables_North,” “DataChelsa_variables_South”). Biotic variables, including tree density (trees ha⁻¹; Tree_density) and age of the oldest trees (years; Older_trees), were collected during fieldwork (2021–2023). Pest occurrence was quantified as total morphospecies richness (Tpest). Insects were further classified into two guilds: (i) reproductive pests (RPest), damaging reproductive structures (inflorescences) and/or harvestable products (cashew nut and apple); and (ii) non-reproductive pests (NRPest), affecting vegetative organs (leaves, shoots, branches, stems) or acting as disease vectors (Dent and Binks, 2020). The dataset also includes four R scripts (R Core Team, 2024) used to analyze pest diversity and its environmental drivers in cashew plantations. The script “1_Kruskal-Wallis_test” performs non-parametric comparisons of abiotic and biotic factors among the three agroecological zones. The script “2_Boxplot” generates boxplots for these variables across zones. The script “3_PCA” conducts principal component analysis to evaluate relationships between abiotic and biotic factors and pest occurrence. Finally, “4_GAM_models” fits generalized additive models (GAMs) for the East, North, and South zones, and extracts model performance metrics, coefficient estimates with 95% confidence intervals, and Moran’s I statistics for residual spatial autocorrelation.