Extended Data Figure 5 : Test of phylogenetic dispersion of grassland communities.

From: Phylogenetic structure and host abundance drive disease pressure in communities

Extended Data Figure 5

a, Plant species in the grassland were consistently phylogenetically clustered compared to the pattern expected if species were drawn randomly from the regional pool (529 species on the 810-hectare UCSC campus). Values less than zero indicate phylogenetic clustering, while values greater than zero indicate species are more distantly related than random (overdispersed) (Extended Data Fig. 4). One-tailed t-tests (df = 9, P values given) showed species were more closely related than expected through the first quartile (25%). This is consistent with habitat filtering. b, In contrast, species were less closely related than expected if drawn randomly from the 43 species found in the grassland community, especially at shorter phylogenetic distances (through the 7% quantile). Box-and-whisker plots indicate median (dark line), 25th and 75th quartiles (box) and minimum and maximum (whiskers), with the ten sites as replicates.