Rusa deer microbiota: the importance of preliminary data analysis for meaningful diversity comparisons

The microbiome is an important consideration for the conservation of endangered species. Studies provided evidence of the efect of behavior and habitat change on the microbiota of wild animals and reported various inferences. It indicates the complexity of factors infuencing microbiota diversity, including incomplete sampling procedures. Data abnormality may arise due to the procedures warranting preliminary analysis, such as rarefaction, before downstream analysis. This present study demonstrated the efect of data rarefaction and aggregation on the comparison of wild rusa deer’s gut microbial diversity. Eighty-fve feces samples were collected from 11 deer populations inhabiting three national parks in Java and Bali islands. Using the Illumina Nova-Seq platform, fragments of 16s rRNA gene were sequenced, and raw data of 51,389 reads corresponding to 2 domains, 22 phyla, 45 classes, 83 orders, 182 families, and 460 genera of bacteria were obtained. Data rarefaction was applied at two diferent library sizes (minimum and fxed) and aggregation (11 populations into 3 research sites) to investigate its efect on the microbial diversity comparison. There are signifcant diferences in alpha diversity
between populations, but not research sites, at all library sizes of rarefaction. A similar fnding is also found in beta diversity. Moreover, data rarefaction and aggregation result in diferent values of the diversity metrics. This present study shows that statistical analysis remains a substantial concern in microbiome studies applied to conservation biology. It suggests reporting a more detailed data normalization in microbiome studies as an inherent control of suboptimal sampling, particularly when involving feces.