Geology and climate influence rhizobiome composition of the phenotypically diverse tropical tree Tabebuia heterophylla

while sites with very high temperature had significantly higher amounts of Actinobacteria (as did dry sites of all geologies to some extent), Crenarchaeota (mostly in dry karst and volcanic sites) and CCM11b (mostly in the dry karst) (Fig 4B, S4 Fig). Elevation (and humidity) also affected phylum-level abundances, with low elevation sites exhibiting a greater abundance of Actinobacteria, Acidobacteria, and Gemmatimonadetes (S4C Fig). Effects of geology and elevation were more apparent at lower taxonomic levels (class and genus). At the parent material scale, volcanic sites are enriched in Verrucomicrobial genus MC18 and Rhodoplanes (Fig 5). Karst sites harbor significantly more Rhodocyclales, CCM11b, Sinobacteraceae while at serpentine sites Cyanobacteria (Microcoleus), Acidobac- teria (Chloracidobacteria) and Bradyrhizobium (Fig 5, S5 Fig). Dry karst sites had significantly higher read abundance of OTUs within order Rhodocy- clales, Rubrobacter or Solirubrobacter while humid karst sites had significantly higher amounts of Rhizobium, CCM11bPH and OTUs within family Rhodospirillaceae. Serpentine high eleva- tion sites had significantly higher amounts of OTUs in class Betaproteobacteria, Chloracido- bacteria, and Phyllobacterium. Drier serpentine sites had significantly higher quantities of certain Cyanobacteria, such as Microcoleus and OTUs within family Phormidiaceae (Fig 5,S5 Fig). Volcanic dry sites had higher dominance of MC18 (Verrucomicrobia)—both greater abundance among all genus, as well as higher abundances in volcanic sites compared other

Fig 4. Composition of the bacterial communities across soil and rhizosphere samples per geology and altitude. Panel shows a phyla-level barplot, with samples sorted by geology and elevation considering only OTUs present in at least 40% of the samples, and with relative abundances greater than 3%. Panels B and C show significant different phyla across geology; panel B shows discriminant phyla-level analyses (FDR p-values <0.05) according to geology and altitude, calculated with LEfSe (Linear discriminant analysis Effect Size) algorithm. The algorithm employs non-parametric Kruskal-Wallis rank sum test to detect phyla with significant differential abundance for geology, and panel B shows discriminating taxa per geology and altitude. https://doi.org/10.1371/journal.pone.0231083.g004 PLOS ONE Rhizobiota of the tropical tree Tabebuia heterophylla PLOS ONE | https://doi.org/10.1371/journal.pone.0231083 April 7, 2020 9 / 16