Omprising 2 of total richness), and, Proteanae, Santalanae, conifers (superorder Pinidae), Dillenianae
Omprising two of total richness), and, Proteanae, Santalanae, conifers (superorder Pinidae), Dillenianae, Chloranthanae and Ranunculanae, each with of total variety of species. The 0 a lot more frequent species in the dataset were, in decreasing order, Casearia sylvestris (Salicaceae), Myrsine umbellata (Myrsinaceae), Cupania vernalis (Sapindaceae), Allophylus edulis (Sapindaceae), Matayba elaeagnoides (Sapindaceae), Casearia decandra (Salicaceae), Zanthoxylum rhoifolium (Rutaceae), Campomanesia xanthocarpa (Myrtaceae), Guapira opposita (Nyctaginaceae) and Prunus myrtifolia (Rosaceae). We discovered 946 species in Mixed forests, ,36 in Dense forests and ,87 in Seasonal forests. ANOVA outcomes showed that different forest kinds did not show considerable variation in relation the number of species (Fig. a). This discovering provides support towards the significant variation identified in relation to the 3 phylogenetic structure metrics analyzed. Mixed forests showed larger standardized phylogenetic diversity (Fig. b) and reduce NRI values, indicating phylogenetic overdispersion, than the other forest kinds (Fig. c). By its turn, Seasonal forests showed reduced standardized phylogenetic diversity and greater NRI values, indicating phylogenetic clustering. Dense forests presented intermediary values in between Mixed and Seasonal forests. In relation to NTI, SeasonalPLOS A single plosone.orgforests showed larger values than the other two forest kinds, indicating phylogenetic clustering (Fig. d), although Mixed and Dense forests did not vary in relation to one another. Mantel tests showed that dissimilarities PubMed ID:https://www.ncbi.nlm.nih.gov/pubmed/23467991 computed determined by matrix P had significant Mantel correlations with all other phylobetadiversity approaches. The highest correlation was involving phylogenetic fuzzy weighting and COMDIST (r 0.59; P 0.00), followed by Rao’s H (r 0.48; P 0.00), COMDISTNT (r 0.48; P 0.00) and UniFrac (r 0.39; P 0.00). MANOVA indicated that species composition of floristic plots varied substantially (P,0.00) between all forest kinds (Table two). Nonetheless, the model fit for species composition was worse than for virtually all phylobetadiversity techniques (exception for COMDIST, see Table two), indicating that phylobetadiversity patterns observed within this study have been robust, and not merely an artifact on the variation in species composition involving forest forms. Amongst the phylobetadiversity methods, phylogenetic fuzzy weighting showed the most effective model fit (R2 0.42; F 73.four). While PERMANOVA showed significant final results for the other four procedures, their model fit varied in line with the properties with the method. COMDIST, a phylobetadiversity technique that MedChemExpress GSK583 captures patterns related to additional basal nodes, showed an extremely poor (despite the fact that statistically considerable) match, although the other three metrics, which capture phylobetadiversity patterns associated to terminal nodes showed better match, particularly Rao’ H. Taking into account only the two approaches with ideal model match (phylogenetic fuzzy weighting and Rao’s H), we identified that most phylobetadiversity variation (larger Fvalue) was observed among Mixed and Seasonal forests. On the other hand, though phylogenetic fuzzy weighting showed a larger phylogenetic similarity among Dense and Seasonal forests (decrease Fvalue), Rao’s H showed a larger similarity among Mixed and Dense (Table two). The ordination of matrix P enabled us to explore the phylogenetic clades underlying phylobetadiversity patterns (Fig. two). The 4 very first PCPS axes contained much more than five of total information.