We performed a one-way ANOVA with blocking to contrast the environmental variables(e.g., microclimate and soil characteristic parameters) between land use types. Differentsurvey visits affect responses by creating variation and therefore are part of the error ornoise aspect of the analysis. We treated them as nuisance factors where the effect of surveyvisits was included as the experimental block in this analysis. This provides a reliablerepresentation of the microclimate condition of each land use and additionally minimizesthe variation effects caused by the localized climate conditions. Tukey’s post hoc tests weredone once significant differences were established to compare each land use group meanwith every other group mean in a pairwise manner. The assumptions of the ANOVAconcerning the residual or error terms (e.g., the normality, variance homogeneity, andindependence) were examined before proceeding with the analysis. The data distributionAnamulai et al. (2019), PeerJ, DOI 10.7717/peerj.7656 5/16was examined using Shapiro–Wilk test and Bartlett’s test for normality and homogeneityof variances, respectively. GenStat version 15 software was used for all analyses (VSNI;Hemel, Hempstead, UK). The raw data is available in the Supplemental File.