The objectives of this study were to characterise the early maturation of oral and faecal microbiota in beef and dairy calves, and to highlight if microbes may have been shared between anatomical niches on the dam and their calves. By four-weeks of age, the oral microbiota of calves was composed of microbes which were more similar to those found in the oral microbiota of adult cows, whereas the faecal microbiota was composed of microbes which bore little resemblance to those in the faeces of adult cows. This was similar in both beef and dairy calves despite dairy calves having no contact with their dam after the first few hours of life. This suggests the development of the gastrointestinal microbiota in calves may not be dependent on continued maternal contact.
Oral microbiota changes in calves and cows
The oral microbiota of newborn calves changed in the first four weeks of life to contain microbes which were similar to those found in the oral microbiota of adult cows. This was also described by Alipour et al. (2018) in calves which were separated from the dam 24 hours after parturition . In our study, this trend occurred in both beef and dairy calves and therefore did not appear to be affected by whether or not the calf was separated from the dam. PCoA indicated there were separate beef and dairy clusters in oral microbiota at four-weeks of age (Figure 4), but both appeared to contain microbes which were similar to those in the adult cow oral microbiota. Additionally, set analysis indicated that in beef and dairy calves, the oral microbiota at four-weeks of age had a large number of ASVs exclusively in common with the cow oral microbiota at calving (Figure 5), which included families which had the greatest relative abundance in the calf oral microbiota such as Neisseriaceae, Streptococcaceae and Pasteurellaceae. These differences also support the assertion that the oral microbiota of calves progresses towards the adult microbiota over the first four weeks of life, in a similar manner in both beef and dairy calves. The number of ASVs that were present exclusively in both the four-week old calf oral microbiota and the oral microbiota of the cow at calving was greater in beef animals than dairy animals. This may suggest that more microbial transfer occurred between the cow and calf oral microbiota in beef animals than dairy animals immediately postpartum, possibly reflecting the reported differences in mothering behaviours between beef and dairy breeds .
The similarity between cow and calf oral microbiota appeared to be loosely dependent on the sampling time point, although there was a moderate degree of inter-animal variation in these samples. The oral microbiota of the newborn calf contained microbes which were most similar to those in oral microbiota of the cow at calving, whereas at four-weeks of age the calf oral microbiota constituted microbes were more similar to those in the cow oral microbiota at four-weeks postpartum, and indeed 4-8 weeks prepartum. There were a large number of ASVs in the oral microbiota of adult cows at calving which were not present in the oral microbiota in pre- or postpartum samples (Supplementary Figure 3). This trend could suggest that the newborn calf alters the oral microbiota of adult cows in the periparturient period, although the oral or faecal microbiota of the calf do not appear to influence this, or the oral microbiota of the cow changes for another reason.
Oral microbiota of calves and teat-skin microbiota of cows
The changes in the relative abundances of Moraxellaceae in the calf oral microbiota over the first four weeks of life were different in beef and dairy calves (4.4% to 19.2% in beef calves; 11.9% to 4.2% in dairy calves). One explanation for the relative abundance of Moraxellaceae increasing in the oral microbiota of beef calves, but decreasing significantly in dairy calves, is the continued sharing of microbes between the cow teat skin and calf oral cavity in beef cows through suckling. This is supported by the set analysis which indicated that Moraxellaceae was the most prevalent family represented by the ASVs common to both the cow teat skin and calf oral microbiota. The teat-skin of beef cows had a high relative abundance of Moraxellaceae which was also high prepartum and therefore more likely to a be source of Moraxellaceae in the oral microbiota of the calves, rather than a consequence. The relative abundance of Moraxellaceae on the teat-skin of dairy cows was low, consistent with previous reports [38, 39].
Faecal microbiota changes in calves
The faecal microbiota of calves changed significantly in the first four weeks of life, consistent with previous studies [5, 6, 8, 9]. However, we observed limited phylogenetic similarity between microbes in the faecal microbiota of calves and adult cows (Figures 3 and 4). Furthermore, the general changes in the faecal microbiota of calves were similar in both beef and dairy animals, suggesting they may not be influenced by continued cow-calf contact.
There was a decrease in beta-diversity of faecal microbiota in beef and dairy calves by four-weeks of age, as has been previously reported [5, 17, 33]. In general, microbes in faecal microbiota of all calves at four-weeks of age were most similar to those in the oral microbiota of four-week old calves and adult cows. This is consistent with a previous study , in which authors suggest the oral microbiota of the cow seeds the faecal microbiota in calves. However, the set analysis indicated that few, if any, ASVs present in the cow oral microbiota at calving persisted in the calf faeces four weeks later. Nevertheless, the phylogenetic similarity between components of the calf faecal and adult oral microbiota appeared to be comparable in both beef and dairy calves. This suggests that if seeding from cow to calf does occur, it occurs in the first few hours after parturition before dairy calves are separated.
More dairy calves had a faecal microbiota that consisted of microbes similar to those in their oral microbiota at four-weeks of age than beef calves, the same trend was also evident in the set analysis (Figure 6). This could be a consequence of the more intensive housing conditions of dairy calves compared to beef, the use of artificial feeding equipment, or due to behaviours such as navel sucking [40, 41].
Immediately after calving, there were a large number of ASVs that were common to the microbiota of calf faeces and adult cow vagina, faeces, and teat-skin (Figure 5), but there was little, if any, indication of these ASVs still being present in calf faeces at four-weeks of age. These results suggest that although microbes may be transferred between multiple niches on the dam and the calf faeces at calving, they do not persist, and it has been reported that the faecal microbiota of calves does not establish until after weaning . Our results are in contrast to the results reported by Yeoman et al. (2018) who found operational taxonomic units (OTUs) were common to multiple sites in the calf GIT and the dam vagina, udder skin, and colostrum over the first three weeks of life, however, the degree of sharing appeared to peak within the first week and then decline for faecal samples . Lima et al. (2019) also found OTUs were present in calf faeces at 3, 14 and 35 days old which were present in the faeces of the dam despite prompt separation of the calf and dam . It should be noted that caution is required when comparing conclusions drawn from OTU analysis with ASV analysis, as OTU methods are less exact and more prone to errors than ASV methods [42, 43].
Overall, the most abundant family in adult faecal samples was Ruminococcaceae, as reported by others [9, 44], but it was present in a low relative abundance in newborn calves. After four weeks, Ruminococcaceae had increased in beef calves to a similar relative abundance to adult cows but only to approximately half the adult level in dairy calves. At four weeks the Ruminococcaceae genus with the greatest relative abundance in calves was Faecalibacterium. The relative abundance of Faecalibacterium was low (<0.01%) at calving in both beef and dairy calves but by four weeks, it was 11.9% in beef calves but only 6.3% in dairy calves. A high prevalence of Faecalibacterium in the first week of life has previously been associated with higher weight gains and a reduced risk of diarrhoea in dairy calves .
Set analysis indicated that only ASVs which were common to both the vaginal microbiota and the faecal microbiota of the adult cow were also present in the oral or faecal microbiota of calves. It is therefore difficult to unravel which of these anatomical niches is most responsible for the potential transfer of maternal microbes to the calf. Yeoman et al. (2018) observed common OTUs in both the vaginal microbiota of the dam and the gastrointestinal tract of the calf , although the faecal microbiota of the adult cow was not sampled. Within adult cows, the two anatomical niches which shared the greatest number of different ASVs were the vaginal and faecal microbiota. It is possible that this is the result of contamination during sampling but may also reflect genuinely similar microbial environments in cattle and other studies have also reported the microbiota of the faeces and vaginal mucosa to be compositionally similar [8, 33].
There were a large number of ASVs common to the calf oral and faecal microbiota at calving that were present in dam samples including colostrum. In beef animals, these included 20 ASVs which were exclusively present in cow colostrum and the calf oral and faecal microbiota. By four-weeks, however, there were no ASVs present in calf faecal and oral microbiota that were present in the colostrum unless they were also present in other adult cow niches at calving (Figure 5). In dairy animals, the only ASVs that were common to the faecal and oral microbiota of newborn dairy calves and dairy cow colostrum were also present in the teat skin and oral microbiota of dairy cows and by four weeks of age very few, if any, of these ASVs were still present. There is, therefore, nothing in these results to suggest the calf gastrointestinal tract is seeded exclusively from colostrum. This is consistent with the findings reported by Klein-Jöbstl et al. (2019)  but in contrast to a previous study which suggested microbes in colostrum contribute to the gastrointestinal microbiota for the first weeks of life . It is possible that microbes which do not persist in the gastrointestinal microbiota are still relevant to the development of the immune system; the gastrointestinal microbiota has been shown to be highly correlated with mucosal gene expression in calves, specifically genes that are related to immune function . Therefore, transient changes in the gastrointestinal microbiota may have longer-lasting effects.
A limitation of this study was the small number of animals included, unfortunately this is a common problem in many microbiome studies in cattle [1, 8, 9, 46–48]. The limited power reduces the scope of our study to perform robust statistical analysis which could provide more conclusive results. Additionally, the study design limited the interpretation of results specific to the influence of the dam on the microbiota of calves. The study was designed to describe the changes in cows and calves in the periparturient period on different farming systems. The beef and dairy farms used in this study employed different, but typical, approaches to calf management and therefore the opportunity existed to explore the influence of the dam, but important factors such environment, breed and diet were not able to be controlled.
Faecal swabs, and to a lesser extent oral swabs, are frequently used as the basis of exploring the microbiome of the calf gastrointestinal tract. Studies which used more invasive techniques indicate similarities between faecal samples and other gastrointestinal niches such as the colon and caecum, but little similarity to proximal intestines [1, 5, 46]. Equally, a high degree of correlation between the oral and rumen microbiota was described by Tapio et al. (2016) ; although oral samples in that study were collected immediately following regurgitation which was not the case in our study. Using oral and faecal samples to assess gastrointestinal microbiota is a compromise between invasive sampling, which provide the most precise results, and non-invasive sampling which allows animals to be sampled repeatedly over time in farm production settings . Changes in relative abundances should be interpreted cautiously and not conflated with absolute abundances; furthermore, comparing sample composition based on family can create erroneous impressions of similarities compared to comparisons made with lower taxonomic levels. Finally, if ASVs were identified in different microbial niches in cows and calves it was inferred that this represented common, and possibly shared, microbes. However, this assertion is based on 16S rRNA sequencing which is less accurate than strain-level metagenomic techniques which have been used to more robustly demonstrate the vertical transmission of specific bacterial strains in humans .