Efficacy of Fecal Sampling as a Gut Proxy in the Study of Chicken Gut Microbiota

Abstract

Despite the convenience and non-invasiveness of fecal sampling, the fecal microbiota does not fully represent that of the gastrointestinal (GI) tract, and the efficacy of fecal sampling to accurately represent the gut microbiota in birds is poorly understood. In this study, we aim to identify the efficacy of feces as a gut proxy in birds using chickens as a model. We collected 1,026 samples from 206 chickens, including duodenum, jejunum, ileum, cecum, and feces samples, for 16S rRNA amplicon sequencing analyses. In this study, the efficacy of feces as a gut proxy was partitioned to microbial community membership and community structure. Most taxa in the small intestine (84.11-87.28%) and ceca (99.39%) could be identified in feces. Microbial community membership was reflected with a gut anatomic feature, but community structure was not. Excluding shared microbes, the small intestine and ceca contributed 34.12 and 5.83% of the total fecal members, respectively. The composition of Firmicutes members in the small intestine and that of Actinobacteria, Bacteroidetes, Firmicutes, and Proteobacteria members in the ceca could be well mirrored by the observations in fecal samples (ρ = 0.54-0.71 and 0.71-0.78, respectively, P < 0.001). However, there were few significant correlations for each genus between feces and each of the four gut segments, and these correlations were not high (ρ = -0.2-0.4, P < 0.05) for most genera. Our results suggest that fecal microbial community has a good potential to identify most taxa in the chicken gut and could moderately mirror the microbial structure in the intestine at the microbial population level with phylum specificity. However, it should be interpreted with caution by using feces as a proxy to study associations for microbial structure at individual microorganism level.

Keywords: chicken; feces; gut microbiota; proxy; spatial relationships.

FIGURE 1

Site origin and inter-individual effects on the shape of microbial community membership and structure. (A) Principal coordinates analysis (PCoA) with unweighted UniFrac distance. Each dot represents a sample from duodenum (D), jejunum (J), ileum (I), cecum (C), or feces (F). PC1 and PC2 represent the top two principal coordinates that captured the most variation, with the fraction of variation captured by that coordinate shown as a percent. (B) PCoA plot with weighted UniFrac distance, similar to (A). (C) Unweighted UniFrac distance (mean ± SEM) between two sampling sites. DJ represents the UniFrac distance between the duodenal and jejunal microbial community, and it was the same as DI, JI, DI, CJ, CI, FD, FJ, FI, and FC. Asterisks indicate the significance of the paired t-test: ^∗∗∗P < 0.001, ^⋅P < 0.1. (D) Weighted UniFrac distance between two sampling sites, similar to (C).

FIGURE 2

OTUs shared across different sampling sites. (A) Venn diagram demonstrating that the taxa overlap among different sampling sites. (B) The percentage of core OTUs and sequences represented by these OTUs in the duodenal (D), jejunal (J), ileal (I), cecal (C), and fecal (F) samples. (C) The percentage of OTUs in feces exclusively contributed by small intestine or cecum, and the percentage of OTUs in feces was below the limit of detection in the gastrointestinal tract.

FIGURE 3

Microbial compositions in feces mirror those in the gastrointestinal tract. Each dot represents a genus. The average relative abundance of each genus in feces is transferred by negative logarithm and shown at x-axis. The average relative abundance of each genus in small intestine (SI) or intestine including small intestine and ceca (SI + C) is transferred by negative logarithm and shown at y-axis. Spearman’s rho was calculated with the negative logarithm-transferred relative abundances between feces and SI (or SI + C).

FIGURE 4

Distribution of Spearman correlations for each genus between two sites. D, J, I, C, and F denote the microbial communities of the duodenum, jejunum, ileum, cecum, and feces, respectively. Only genera with an abundance >0.1% at either site of comparison and significant correlations (P < 0.05) are shown.