Fig. 4: Jaccard dissimilarity across elevational zones.

Jaccard dissimilarity for each taxonomic group across the three elevational zones. Jaccard dissimilarity was calculated between each plot pair of a different developmental stage within each elevational zone, modelled and predicted using individual Bayesian multilevel models with a beta probability distribution, separately for each taxonomic group. Posterior distributions were summarized by means of the MAP and 95% and 50% (thick bars) HDI (explanation in methods). Different point shapes indicate substantial differences in Jaccard dissimilarity between elevational zones. Supplementary Fig. S13 includes data points.