ABSTRACT
Fecal microbiota variability and individuality are well studied in humans and also in farm animals (related to diet- or disease-specific influences), but very little is known for exotic zoo-housed animals. This includes a wide range of species that differ greatly in microbiota composition and variation. For example, herbivorous species show a very similar and constant fecal microbiota over time, whereas carnivorous species appear to be highly variable in fecal microbial diversity and composition. Our objective was to determine whether species-specific and individual-specific clustering patterns were observed in the fecal microbiota of wildebeest (Connochaetes taurinus) and tigers (Panthera tigris). We collected 95 fecal samples of 11 animal individuals that were each sampled over eight consecutive days and analyzed those with Illumina MiSeq sequencing of the V3-V4 region of the 16SrRNA gene. In order to identify species or individual clusters, we applied two different agglomerative hierarchical clustering algorithms - a community detection algorithm and Ward's linkage. Our results showed that both, species-specific and individual-specific clustering is possible, but more reliable results were achieved when applying dynamic time warping which finds the optimal alignment between different time series. Furthermore, the bacterial families that distinguish individuals from each other in both species included daily occurring core bacteria (e.g., Acidaminococcaceae in wildebeests or Clostridiaceae in tigers) as well as individual dependent and more fluctuating bacterial families. Our results suggest that while it is necessary to consider multiple consecutive samples per individual, it is then possible to characterize individual abundance patterns in fecal microbiota in both herbivorous and carnivorous species. This would allow establishing individual microbiota profiles of animals housed in zoos, which is a basic prerequisite to quickly detect deviations and use microbiome analysis as a non-invasive and cost-effective tool in animal welfare.
ABSTRACT
BACKGROUND: Through the rapid development in DNA sequencing methods and tools, microbiome studies on a various number of species were performed during the last decade. This advance makes it possible to analyze hundreds of samples from different species at the same time in order to obtain a general overview of the microbiota. However, there is still uncertainty on the variability of the microbiota of different animal orders and on whether certain bacteria within a species are subject to greater fluctuations than others. This is largely due to the fact that the analysis in most extensive comparative studies is based on only a few samples per species or per study site. In our study, we aim to close this knowledge gap by analyzing multiple individual samples per species including two carnivore suborders Canoidea and Feloidea as well as the orders of herbivore Perissodactyla and Artiodactyla held in different zoos. To assess microbial diversity, 621 fecal samples from 31 species were characterized by sequencing the V3-V4 region of the 16S rRNA gene using Illumina MiSeq. RESULTS: We found significant differences in the consistency of microbiota composition and in fecal microbial diversity between carnivore and herbivore species. Whereas the microbiota of Carnivora is highly variable and inconsistent within and between species, Perissodactyla and Ruminantia show fewer differences across species boundaries. Furthermore, low-abundance bacterial families show higher fluctuations in the fecal microbiota than high-abundance ones. CONCLUSIONS: Our data suggest that microbial diversity is significantly higher in herbivores than in carnivores, whereas the microbiota in carnivores, unlike in herbivores, varies widely even within species. This high variability has methodological implications and underlines the need to analyze a minimum amount of about 10 samples per species. In our study, we found considerable differences in the occurrence of different bacterial families when looking at just three and six samples. However, from a sample number of 10 onwards, these within-species fluctuations balanced out in most cases and led to constant and more reliable results.
