Numbers of selected bacterial species in pig faeces do not accurately represent their numbers in the ileum

Although it is known that the microbial population of faeces differs from that of the gastrointestinal tract (GIT) (Looft et al. 2014), monitoring the microflora of the pig GIT requires euthanasia. Identifying correlations between selected bacterial species in faeces and ileal mucosa would allow extrapolation of GIT bacterial numbers from faeces. This study tested the hypothesis that numbers of selected bacterial groups would differ between the ileum and faeces, and that they would be significantly correlated. Clostridium perfringens and Escherichia coli (part of the Enterobacteriaceae family) were selected as prominent pathogens, whilst Lactobacilli are important commensal bacteria.

Paired faecal and ileal mucosal scrapings (without intestinal contents) were collected from three nursery and nine weaner pigs. Bacterial numbers were determined by quantitative polymerase chain reaction (qPCR) on extracted DNA (MagMAX Pathogen Kit), as previous studies demonstrated a good correlation with culture techniques (Castillo et al. 2006). The qPCRs targeted the 16S or 16S-23S rRNA intergenic region of selected and total bacteria (Collins and Bowring 2014). The percentage of selected bacterial groups relative to total bacteria was calculated to overcome variation in water content of samples, and then log₁₀ transformed for normality. Bacterial numbers and percentages were analysed using the paired t-test and Pearson correlations were performed on percentages (Genstat, 17th Edition; UK).

Differences in Cl. perfringens, Lactobacilli and total bacterial numbers were observed in ileal mucosa and faeces (P < 0.025) (Table 1). However, when bacterial numbers were expressed as percentages of total bacteria only Cl. perfringens remained significantly different (P = 0.015), along with a trend towards a reduced percentage of Lactobacilli in mucosa (P = 0.076). Adhesion to the intestinal mucosa is a characteristic feature of pathogenic Cl. perfringens, which may partly explain the increased percentage of Cl. perfringens in mucosa and its underrepresentation in faeces. Looft et al. (2014) also observed an increased relative abundance of Cl. perfringens in ileal mucosa compared to faeces using microbial sequencing. Linear correlations between bacterial numbers in faeces and ileal mucosa were not demonstrated, suggesting that other factors may affect the relative abundance of bacteria.

Expressing selected bacterial numbers as a percentage of total bacteria was critical for comparing the two different sample types, which varied in their water content. The absence of significant correlations between percentages of selected bacteria in faeces and mucosa may be explained by the small sample size and the dramatic changes occurring in the GIT associated with weaning and disease. Larger sample sizes are needed to identify correlations between bacterial numbers in faeces and mucosa. Good correlations between bacterial numbers in faeces and mucosa would enable approximation of bacterial numbers in the ileum, avoiding animal sacrifice and allowing repeated sampling over time. Regardless of correlations, faeces remain a valuable, non-invasive sample for quantifying pathogen excretion and potential disease transmission.

Table 1.  Log₁₀ numbers of Cl. perfringens, E. coli, Enterobacteriaceae, Lactobacilli and total bacteria (mean ± SE), and percentages of selected bacteria in ileal mucosa and faeces