Microbial biogeography is a growing subdiscipline in the field of ecology. This is due in no small part to rapidly declining technical and analytical barriers and costs associated with the characterization of environmental samples using parallel sequencing technologies. Ideas persist in the literature that “everything is everywhere;” that is, that microbes are not principally limited by dispersal but that environmental filtering is the primary driver of species distributions. While climatic and other biotic and abiotic factors are clearly important, history and contemporary movement likewise strongly determine patterns of occurrence within and among regions. Here we report on a comprehensive analysis of the drivers of diversity, distribution and composition of the ecologically and economically important genus, largely plant-pathogenic Phytophthora (Latin for “plant-killer”). Environmental DNA from 414 soil samples collected across Australia was sequenced (on a 454 pyrosequencing platform) with Phytophthora-specific ITS primers and blasted against a local, highly curated database for the genus. We assessed 22 environmental and seven categorical variables as potential predictors of Phytophthora species richness, α and β diversity, and spatial autocorrelation, including both phylogenetically and non-phylogenetically explicit methods. We also classified each species as putatively native or introduced and examined aspects of distribution with respect to origin.
Results/Conclusions
Phytophthora taxa were detected in 65% of 640 sites across Australia and represented 68 distinct phylotypes. Of these, 21 were identified as potentially undescribed taxa and 25 were new detections in natural areas and/or new introductions to the country. Introduced forms comprised over 54% of Australia’s Phytophthora diversity and were highly overrepresented in soils from crop habitats as well as urban landscapes and forestry. Nonnatives were strongly asymmetric among Phytophthora clades but not among ecoregions. The two most widespread species (P. multivora and P. cinamomi) were introduced, though five of the top ten most widely distributed species were native. Species composition was most strongly predicted by bioregion (R2=0.29) and ecoregion (R2=0.26) suggesting that aggregate measures of habitat and climate show greater predictive power of soil microbial diversity than single or even combined climate, soil and vegetation parameters at a landscape scale. Precipitation of warmest quarter, mean temperature of the wettest quarter and latitude were also highly significant and described approximately 21%, 14% and 13% of variation in Phytophthora composition. We found weak but significant evidence for phylogenetic over-dispersion with respect to key climate variables. This study provides a strong baseline for understanding biogeographical patterns in this important plant pathogenic genus.