DNA extraction and 16S rRNA gene sequencing
DNA was extracted from cloacal swabs using DNeasy PowerPlant Pro kits (Qiagen). Both the swab and the RNAlater (Invitrogen, Thermo Fisher Scientific) that the swab was stored in were added to the PowerBead tubes, then the extraction proceeded following the instructions in the manufacturer’s protocol. We quantified the DNA concentration for each sample using a Qubit fluorometer and froze samples at -20℃ until samples could be shipped for sequencing. The V4 region of the 16S rRNA gene was sequenced at the University of Texas’s Genome Sequencing and Analysis Facility (GSAF) using an Illumina MiSeq platform. Samples with less than 1,000 total sequences were excluded from subsequent analyses. After quality filtering, we had 63 samples for further analyses in the diet manipulation experiment (experiment 1). Of these samples, 22 were from birds fed a high fat diet (day 0: n=11, day 8: n=11), 21 were from birds fed diets with equal ratios of protein and lipid (day 0: n=9, day 8: n=12), and 20 were from birds fed a high protein diet (day 0: n=11, day 8: n=9). For the second experiment, some samples failed during sequencing or after quality filtering, yielding 69 total cloacal microbiota samples for further analyses. Of these samples, 38 were from injected birds (LPS day 0: n=9, LPS day 5: n=8; saline day 0: n=10, saline day 5: n=11) and 31 were from focal birds (LPS-focal day 0: n=6, LPS-focal day 5: n=8; saline-focal day 0: n=7, saline-focal day 5: n=10). We used DADA2 (v.1.16) in R (v.4.2.0) to process our 16S sequence data. ASVs were assigned to taxonomy using the Silva 132 bacterial reference database (Quast et al. 2013). Sequences identified as a chloroplast or mitochondria were removed from the dataset. TheDECIPHER package (v. 2.24.0; Wright 2015) was used to create a multiple sequence alignment and a generalized time-reversible maximum likelihood tree of the remaining ASVs was constructed with thephangorn package version 2.9.0 (Schliep 2011). The ASV table, taxonomic information, phylogeny, and sample metadata were joined for bacterial community analyses using the package phyloseq(McMurdie and Holmes 2013).