This document contains supplementary tables and figures for the manuscript. This is the same information as the SI Appendix PDF except this document is written in HTML, can be viewed in most browsers, and contains all the tabular data within the document (rather than separate data files). Tables are horizontally scrollable. You can sort tables and download (or copy) tabular data. Some tables contain hyperlinks.

Contents

Data Availability
Sample Naming
Table S1: Summary of field-based feeding observations.
Table S2: Number of bites observed for each herbivore species at each site.
Table S3: Metadata and microbiome diversity estimates for each sample.
Table S4: Total taxonomic diversity (by Class) of herbivore microbiomes.
Figure S1: Class-level relative abundance of microbial communities from each sample.
Table S5: Results of LEfSe analysis.
Table S6: Results of BLAST analysis for DA ASVs.
Table S7: Accession numbers and unique codes for sequence data in Figure 3.

Data Availability

We made additional data products and processing scripts available through online repositories.

Sample Naming

Raw fastq data files were named using the root format RunQ_GnSpe000_G, where Q was the run number (1, 2, or 3), GnSpe was the host genus and species, 00 was a unique host ID number, and G was the gut segment (F = foregut; M = midgut; H = hind). For example, the file name Run1_SpVir11_M_S147_L001_R2_001.fastq corresponded to: the reverse read; midgut sample; Sparisoma viride; individual 11; Run01. All raw data are publicly available. The 53 individual fish encompassed seven species and three genera. Two species—Sparisoma chrysopterum and Scarus vetula—were only represented by 1 and 2 individuals, respectively. Though we omitted these samples from the analysis due to low sample size, they were sequenced and processed along with the rest of the samples. The data are publicly available.

Table S1

Summary data from observations of individual bites by five species of herbivorous fishes in the Florida Keys. Data include the mean sediment depth and turf height of algal assemblages fed on by each species at each of three sites, as well as the proportion of bites that resulted in a grazing scar on the substrate where reef calcium carbonate had been removed (Prop. grazing scar). The table summarizes the proportion of bites on vertical (> 45 degrees), concave, and convex substrates as well as the relative proportion of bites targeting each of the 10 food types most commonly bitten during the observations. Most macroalgae were aggregated to genus while other food types are summarized by functional group.

Note Table scrolls horizontally.

Table S2

This table summarizes the number of individual bites observed by each of the five species of herbivorous fishes at each of three sites in the Florida Keys.

Table S3

This table summarizes metadata for each host species and sample (sample_ID) used in this study. Includes various details about the host (e.g., weight, length, gut length, etc.) and the associated microbiome (e.g., number of reads, diversity stats, etc). Please note, the table is horizontally scrollable, sortable, and only the first 5 rows are visible. Rows correspond to sample ID and columns correspond to various traits for each sample.

Column descriptions

Host details

  • SampleID
  • Host genus
  • Host species
  • Common name
  • NCBI tAxID NCBI Taxonomic ID of host species.
  • Collection date

Host physiological characteristics

  • Life phase
  • Weight (g)
  • Total length (cm)
  • Foregut length (cm)
  • Midgut length (cm)
  • Hindgut length (cm)
  • Total gut length (cm)

Microbial diversity

  • Total reads Total reads for each sample.
  • Total ASVs Number of ASVs detected in each sample.
  • Chao1 The Chao1 richness estimator.
  • Chao1 (se) Standard error of Chao1 index
  • ACE The ACE richness estimator.
  • ACE (se) Standard error of ACE index
  • Shannon The Shannon diversity index
  • Simpson The Simpson diversity index.
  • InvSimpson Inverse Simpson’s Index
  • Fisher Fisher diversity Index.

Note Table scrolls horizontally.

Table S4

This table shows the Class-level taxonomic diversity of the total microbiome dataset. Data shows the total number of reads and total number of ASVs for each Class plus the relative percent of each in the dataset. Please note the table is sortable and only the first 10 Classes are shown.

Figure S1

In Figure 2A of the main paper, Class-level abundance was displayed by host species. Here the same data is presented for each individual samples.

Supplementary Figure 1: *Class-level relative abundance of microbial communities from each sample. Taxa representing less than 2.5% total relative abundance were conglomerated into **Other**.*

Supplementary Figure 1: Class-level relative abundance of microbial communities from each sample. Taxa representing less than 2.5% total relative abundance were conglomerated into Other.

Table S5

Table showing the results from the LEfSe analysis including Linear discriminant analysis (LDA) scores, P-values adjusted for multiple testing, and False Discovery Rate (FDR) values. Normalized read abundance values (the input for the analysis) for each host species are also given.

Host abbreviations
  • AcCoe: Acanthurus coeruleus
  • AcTra: Acanthurus tractus
  • ScTae: Scarus taeniopterus
  • SpAur: Sparisoma aurofrenatum
  • SpVir: Sparisoma viride

Table S6

Full summary table of BLAST analysis for each of the 59 differentially abundant (DA) ASV. The table includes some details about each ASV (e.g., total reads), details of top BLAST hit(s), and alignment results. The table also includes results from queries to the IMNGS database.

Abbreviations

ND Indicates No Data was provided for given attribute of top BLAST hit.
NR In Top hit acc column indicates Not Recorded. Four ASVs (ASV6, ASV12, ASV224, ASV398) had numerous hits at 100% identity. We did not include top hit data for these ASVs.
NLC No Lifestyle Category from Sullam classification.

Column descriptions

ASV details

  • ASV Query ASV.
  • Putative habitat Proposed habitat preference based on data synthesis.
  • Enriched Host species where ASV was differentially abundant.
  • Total reads Total reads for query ASV in full dataset.
  • Taxon Class-level taxonomic affiliation of query ASV.
  • Num IMNGS hits. Number of hits to the IMNGS database. Value indicates the number of samples that scored a hit to an ASV based on 97% cutoff identity and representing greater than or equal to 0.1% of total reads in a sample from the database.

Details for top BLAST hits

  • Num perfect hits Number of identical matches of query ASV to nr database (out of 50).
  • Top hit acc Accession number of top BLAST hit. In cases of identical percent identity, all top hits are provided.
  • % identity Percent identity of query ASV to top BLAST hit.
  • Isolation source Tissue type of host (where applicable) where top BLAST hit was isolated from.
  • Nat host Scientific name of host (where applicable) where top BLAST hit was isolated from.
  • Common name Common name of host (where applicable) where top BLAST hit was isolated from.
  • Collection year Year sample was collected (where provided) containing top BLAST hit.
  • Country Country and location origin of sample (where provided) containing top BLAST hit.
  • PubMed ID PubMed publication ID (where provided) containing top BLAST hit.
  • Sullam Lifestyle Lifestyle classification of BLAST hit (where applicable) from the Sullam et. al. paper, specifically Table S1.

Alignment details between query ASV and top BLAST hit

  • Alignment length Alignment length between query and subject.
  • Mismatches Number of base pair mismatches in alignment between query and subject.
  • Gap opens Number of open gaps in alignment between query and subject.
  • Q. start Starting base pair position of query sequence in alignment.
  • Q. end Ending base pair position of query sequence in alignment.
  • S. start Starting base pair position of subject sequence in alignment.
  • S. end Ending base pair position of subject sequence in alignment.
  • Evalue The Expect (E) value of alignment.
  • Bit score The bit score of alignment.

Note Table scrolls horizontally.

Table S7

This table provides lookup details for all neighbor sequences from BLAST comparisons used in the tree from Figure 3 (main paper). In the tree, sequences are named after their isolation source plus a unique abbreviated tree code. This table provides (where applicable) the:

  • Accession_number Full accession numbers for each leaf.
  • Tree code Parenthetical code used in the tree.
  • Host, Isolation source, Location Details about source of sequence.
  • Link Hyperlink to sequence page from NCBI Nucleotide database.
  • Sullam lifestyle In our analysis, we found the following categories: 1 Vertebrate gut generalists, 2 Animal and vertebrate gut generalists, 3 Fish gut generalists, 5 Fish gut specialists, 6 Animal generalists, 9 Marine environmental microbes, 12 Microbes from artificial habitats, 13 Undefined, basal clade members.