Effect of age, severity of diarrhoea, number of pathogens present and blooming of E. coli on metagenomic characteristics of stools from Danish dairy calves with diarrhoea

BACKGROUND: Calf diarrhoea causes substantial welfare and economic losses, and it is one of the major drivers of antimicrobial use. This study aimed to characterize the faecal microbiome of diarrhoeic calves, with a specific focus on Escherichia coli, and to assess whether microbial profiles vary with age, diarrhoea severity, and high E. coli abundance in the absence of other detectable enteric pathogens. METHODS: Stool samples from Danish diary calves (n = 32) below 4 weeks of age were collected from 11 herds and were analysed using direct long-read sequencing (mgt) as well as analyses of a subset of samples by swiping microbiota from faecal samples grown on McConkey agar plates (plate-swipe). Metagenomes were analysed to characterise community structure (Shannon α-diversity; Bray-Curtis PCoA with PERMANOVA) and to assess differential abundance at the species level while adjusting for sample type (mgt/plate swipe), herd, age, number of other pathogens detected by qPCR (rotavirus, coronavirus, Cryptosporidium parvum, Salmonella Dublin, Clostridium perfringens A, B, C, Eimeria and Escherichia coli F5) and recorded as presence/absence and summarised into infection classes (None/Mono/Co-2/Co-3 +). Binning was performed to build metagenome assembled genomes (MAGs) of E. coli. RESULTS: Microbiome structure was dominated by methodological and contextual factors: sample type (direct metagenomic vs plate swipe) and herd explained far more variation than clinical severity and age. Metagenomic species profiles from plate swabs were comparatively homogeneous and E. coli-rich, whereas direct metagenomes captured higher diversity. Differential abundance identified species enriched with increasing diarrhoea severity and with infection classes, while pathogen-specific contrasts (e.g., C. perfringens A-positive vs negative) revealed discrete sets of bacterial co-occurrences. Classical pathotype markers (virulence-genes) were uncommon among E. coli MAGs. CONCLUSIONS: Long-read metagenomics revealed insignificant influence of severity of diarrhoea, age below 4 weeks and number of pathogens detected in stool samples on diversity and microbial communities in diarrheic dairy calves. In contrast, large variation was observed between herds. On average, E. coli constituted about half of the microbiota. MAGs generated by binning indicated non-specific blooming of strains without particular virulence genes.