Section 4 of 8
Alpha diversity indices
Arthur Joanello Cemin, Vagner Luiz Graeff-Filho, José Pedro Spies Nolibos, Ezequiel Cesar Carvalho Miola, Paulo Mello-Farias, and Vanessa Sacramento Cerqueira · about 1 minutes
Alpha diversity was evaluated for each sampling unit (n = 12) using the vegan package (Oksanen et al. 2026). Two diversity indices were calculated: the Shannon–Wiener index (H') and the Simpson index (D). Differences among olive cultivars were evaluated using generalized linear mixed models (GLMMs) fitted with the glmmTMB package (Brooks et al. 2017), with cultivar identity included as a fixed effect. The Shannon index was modeled assuming a Gaussian distribution, whereas the Simpson index was modeled assuming a beta distribution with logit link after transformation of the original values to avoid boundary values of 0 and 1. Pairwise comparisons among cultivars were performed using Tukey-adjusted estimated marginal means obtained with the emmeans package (Lenth 2026).
Community composition and beta diversity
Beta diversity was assessed using the Bray–Curtis dissimilarity index calculated from spore abundance data. Differences in AMF community composition among olive cultivars were tested by permutational multivariate analysis of variance PERMANOVA with 999 permutations (Anderson 2001). Homogeneity of multivariate dispersion among cultivars was verified prior to PERMANOVA. Community structure was visualized by non-metric multidimensional scaling (Kruskal 1964). All multivariate analyses were performed using the vegan package (Oksanen et al. 2026).
The contribution of individual AMF species to the dissimilarity among olive cultivars was evaluated using similarity percentage analysis (SIMPER; Clarke 1993). Results are expressed as the percentage contribution of each species to the Bray–Curtis dissimilarity between cultivars. Analyses were performed using the vegan package (Oksanen et al. 2026).