Root and Rhizosphere Bacterial Phosphatase Activity Varies with Tree Species and Soil Phosphorus Availability in Puerto Rico Tropical Forest

Kristine G. Cabugao, Collin M. Timm, Alyssa A. Carrell, Joanne Childs, Tse-Yuan S. Lu, Dale A. Pelletier, David J. Weston, Richard J. Norby

Tropical forests generally occur on highly weathered soils that, in combination with the immobility of phosphorus (P), often result in soils lacking orthophosphate, the form of P most easily metabolized by plants and microbes. In these soils, mineralization of organic P can be the major source for orthophosphate. Both plants and microbes encode for phosphatases capable of mineralizing a range of organic P compounds. However, the activity of these enzymes depends on several edaphic factors including P availability, tree species, and microbial communities. Thus, phosphatase activity in both roots and the root microbial community constitute an important role in P mineralization and P nutrient dynamics that are not well studied in tropical forests. To relate phosphatase activity of roots and bacteria in tropical forests, we measured phosphatase activity in roots and bacterial isolates as well as bacterial community composition from the rhizosphere. Three forests in the Luquillo Mountains of Puerto Rico were selected to represent a range of soil P availability as measured using the resin P method. Within each site, a minimum of three tree species were chosen to sample. Root and bacterial phosphatase activity were both measured using a colorimetric assay with para-nitrophenyl phosphate as a substrate for the phosphomonoesterase enzyme. Both root and bacterial phosphatase were chiefly influenced by tree species. Though tree species was the only significant factor in root phosphatase activity, there was a negative trend between soil P availability and phosphatase activity in linear regressions of average root phosphatase and resin P. Permutational multivariate analysis of variance of bacterial community composition based on 16S amplicon sequencing indicated that bacterial composition was strongly controlled by soil P availability (p-value < 0.05). These results indicate that although root and bacterial phosphatase activity were influenced by tree species; bacterial community composition was chiefly influenced by P availability. Although the sample size is limited given the tremendous diversity of tropical forests, our study indicates the importance of roots and bacterial function to understanding phosphatase activity. Future work will broaden the diversity of tree species and microbial members sampled to provide insight into P mineralization and model representation of tropical forests.