Microbiome Sequencing and Bioinformatic Processing
Individual swab DNA extracts were barcoded, pooled and sequenced on the Illumina MiSeq platform (250 bp paired-end reads) in two assays: (1) barcoded 16S primers (515F and 806R) (38) were used to examine bacterial diversity; and (2) barcoded 18S v4 primers (TAReuk454FWD1 and TAReukREV3) (39) were used to examine microeukaryote diversity. 16S libraries were constructed at the Universidade Estadual Paulista (BR) and sequenced at the Tufts University Core facility (USA) while 18S library preparation and sequencing was performed at the University of Michigan (USA). Negative (template-free) controls were run simultaneously with each sequencing library to ensure there was no contamination from PCR or sequencing reagents.
Sequences were quality-filtered and processed using the Quantitative Insights into Microbial Ecology (QIIME) MiSeq pipeline using default settings (40). As no mock community was included as a positive sequencing control, low abundance OTUs were filtered from the dataset using a conservative abundance threshold (<0.005% of all reads) (41). Sequences were clustered into operational taxonomic units (OTUs) using a 97% similarity threshold and compared against reference databases (rdp GreenGenes for 16S, Silva 97 for 18S) to assign taxonomy using the BLAST search algorithm. Chimeras were identified and filtered using UCHIME2 (42). For 18S data, sequences assigned to frog or other non-target non-microbial species (e.g. , Streptophyta) were filtered from the dataset. Sequences were rarefied to 1000 per sample for 18S data and 2000 per sample for 16S data. These values were selected based on visual examination of histograms and read accumulation curves constructed for all samples.
To infer ecological effects of bacteria on fungi and protists, bacterial OTU representative sequences from the T. taophora samples were compared against a reference database containing bacteria that were previously isolated from amphibian skin and evaluated for effects onBd growth in co-culture experiments (43). The BLAST algorithm was implemented and an E-value threshold of E < 1e-20 was used to identify OTU matches with the reference database. Matching T. taophora skin bacteria were categorized as Bd enhancing,Bd inhibiting, or having no effect on Bd growth.