Conclusion
We demonstrated that shallow shotgun sequencing yields characterization of the Sable Island horse microbiome which is comparable to both deeper shotgun sequencing (when using a read-based profiler) and 16S rRNA gene amplicon datasets. Shallow shotgun sequencing also replicated biological patterns reported in a previously published amplicon dataset, while also providing estimates of microbiome functional potential. Amplicon data may be better suited for some research questions; for example, those pertaining to biological patterns which are hypothesized to be the result of microbiota dispersal. However, the assembly of system-specific MAGs may help to narrow this gap. Researchers should still seek to validate shallow shotgun sequencing for their own study systems, especially if microbiota in their systems are likely to be underrepresented in genome databases. But, based on data from Sable Island feral horse fecal samples, we conclude that shallow shotgun sequencing can provide greater biological insight than amplicon data, at a comparable cost, and with greater long-term value.