1. Introduction
Whole milk (M), waste milk (WM), and milk replacer (MR) are the most common liquid feeds for calves on dairy farms. WM is comprised of colostrum, milk obtained from mastitic cows, and milk from cows treated with antibiotics. The growth performance has been shown to be similar for calves fed WM and M [
1,
2], whereas the growth rate of calves fed MR is influenced by the ingredient composition and nutrient intake of the MR [
3,
4]. As most liquid feed flows directly into the abomasum upon suckling action, which closes the esophageal groove, the intestine, rather than the rumen, is the major digestion site for pre-weaned calves. Therefore, previous research has mainly focused on the effects of different liquid feeds on intestinal microbiota and development. Colostrum was shown to have a positive effect on gastrointestinal tract development and function in calves, not only through the provision of nutrients, but also due to the high concentration of growth factors and biologically active peptides [
5,
6]. Calves fed pasteurized WM had a more diverse bacterial community in feces [
7], whereas calves fed MR containing soy flour had a less acidic abomasal environment [
8,
9] and slower intestinal development [
10].
As the rumen develops and is colonized by microorganisms, a calf physiologically transits from a non-ruminant to ruminant state. The postnatal period may be the most critical window for rumen manipulation, and the early feeding regime may lead to long-lasting changes in the rumen microbial composition [
11,
12,
13]. Diets could modify the establishment of the bacterial community of lambs around the time of weaning, and this modification persists over four months [
14]. De Barbieri (2015) [
15] also found that ruminal bacterial communities of lambs can be altered by the diet of the maternal ewes and lambs or by inoculums from donor ewes fed with different diets, and the difference lasts until five months of age. However, not all nutritional interventions in the early life of calves can promote the establishment of different microbial populations in the rumen of the young animals. Dietary supplementation with sanguinarine and resveratrol did not yield long-term effects on rumen fermentation patterns or the bacterial community in calves [
16]. Therefore, the aim of the current study was to evaluate the effects of the most common liquid feeds on calf ruminal development, fermentation, and the bacterial community during the weaning period, and to then follow them up to six months of age to determine the persistence of any effects.
4. Discussion
In calves that have not been weaned, most of their liquid feed intake bypasses the rumen and enters the abomasum directly since the esophageal groove is closed. Abe (1979) [
34] reported that the reflex closure of the esophageal groove occurs efficiently and independently of the feeding method (nipple-feeding or bucket-feeding) when calves are familiar with either method. Wise (1984) [
35], however, indicated that the reticular groove reflex is more efficient when calves suck nipples rather than drink from buckets. Changes in the feeding method from a floating nipple to bucket have been shown to successfully improve the plasma metabolic and endocrine profiles of ruminal drinking calves [
36]. More recent research found that a considerable amount of liquid feed may leak into the rumen, an amount that has been estimated to be 17%–35% of the total intake in bucket-fed calves [
37,
38,
39], as opposed to 0%–20% in nipple-fed calves [
40,
41]. The bucket feeding method has been widely used because it is labor-saving and easy to perform. This method, however, increase the risk of rumen drinking, which may lead to ruminal fermentation disorders and metabolic acidosis [
42]. Subacute ruminal acidosis (SARA) is characterized by a sustained depression in ruminal pH below a value of 5.6 [
43] or 5.8 [
44] in dairy cows. However, the rumen pH of dairy calves is lower than that of mature cows, averaging between 5.09 and 5.31 [
37] and 5.19 and 6.16 [
45]. This is primarily because maximum dry feed intake is generally encouraged in preweaning calves in order to promote rumen development. In addition, the leakage of liquid feed into the rumen may result in the bacterial fermentation of milk and lead to ruminal acidosis. In this trial, the ruminal pH was within the normal pH range for calves, possibly suggesting that no abnormal rumen function occurred in the experimental calves.
Waste milk comprises colostrum, milk obtained from mastitic cows, and milk from cows treated with antibiotics. The content of milk protein that entered the rumen might have been greater in the WM group than in the M group. Branched-chain VFA (BCVFA) are synthesized by microorganisms in the rumen via oxidative deamination and the decarboxylation of branched-chain amino acids [
46]. Isobutyrate, isovalerate, and 2-methylbutrate are generated from valine, leucine and isoleucine, respectively. The relatively greater concentration of isovalerate in the rumen of the WM group may be associated with the greater protein content of WM. Supplementation of isovalerate increases the population of
Butyrivibrio fibrisolvens in calf and steer feed [
47,
48]. This was confirmed by our finding that
Butyrivibrio fibrisolvens was positively correlated with the concentration of isovalerate in two-month-old calves. Additionally, the populations of
Ruminococcus albus,
Ruminococcus flavefaciens and
Fibrobacter succinogenes have been found to linearly increase with increasing isobutyrate or isovalerate supplementation [
47,
48,
49]. No differences in the abundance of these bacteria were detected in this trial. The increases in these bacterial populations occur mainly because these bacteria firmly adhere to plant tissues when initiating cellulose degradation [
50]. In this study, however, we sampled ruminal liquids, which did not include the bacterial populations that attach to feed particles. Deng (2017) [
51] reported that the number of OTUs in the rumen digesta of calves fed pasteurized waste milk was higher than that in calves fed untreated whole milk. We observed that, compared to untreated whole milk, feeding pasteurized waste milk increased the ruminal bacterial richness. This may be associated with the concentration of isovalerate in the rumen, as a greater population of total bacteria has been detected in the rumen of steers fed isovalerate [
47].
The extraruminal effects of BCVFA have been detected by feeding BCVFA to dairy cows: the level of GH increases and that of insulin and NEFA decreased [
52,
53,
54]. Serum GH and IGF-1 for both pre- and post-weaning calves increased linearly with increasing isovalerate supplements [
55]. A similar result was observed in our trial: a greater concentration of ruminal isovalerate may induce increases in serum GH and IGF-1 but decreases in insulin and NEFA in the WM2 group. BCVFA receptors are believed to be present in ruminal and hepatic membranes and might perturb the function of hormone-regulated systems, such as those involving insulin and IGF-1 [
56,
57]. Additionally, bovine colostrum is characterized by high levels of IGF-1 and EGF, which have been found to be resistant to pasteurization [
58]. Therefore, the bioavailability of these growth factors in colostrum may contribute to greater concentrations of serum IGF-1 and h-EGF in calves fed WM. Rauprich et al. (2000) [
59] reported that calves fed colostrum had greater plasma IGF-1 levels than those fed MR. Insulin, IGF-1, and EGF have been implicated as possible mediators of rumen epithelial cell proliferation and thus play an important role in accelerating rumen development in calves [
60,
61].
Antibiotic susceptibilities of ruminal bacteria have been previously determined [
62]. Chlortetracycline, oxytetracycline, tylosin, and monensin exhibit a strong inhibitory effect on in vitro cellulose digestibility and VFA production in mixed-rumen cultures [
63]. Monensin and virginiamycin can alter in vivo rumen microbial populations [
64,
65]. Feeding calves milk with very low concentrations of ampicillin, ceftiofur, penicillin and oxytetracycline affects the composition of the microbial population in feces [
66]. The concentration of antibiotic residues in waste milk cannot be reduced by pasteurization [
67]. Therefore, it is likely that a substantial number of bacteria may be sensitive to the presence of antibiotic residuals in WM, which may lead to a distinctive rumen bacterial community of calves in the WM2 group. Gentamicin was the only antibiotic residue detected in waste milk: its concentration was 0.067 ± 0.042 mg/L (mean ± SD). The decreased relative abundance of
Prevotella 7 in the WM2 group might be explained by its susceptibility to gentamicin, a hypothesis supported by a recent study reporting that gentamicin inhibited 90% of
Prevotella intermedia [
68]. A decreased abundance of the genus
Prevotella was also reported in calves fed waste milk containing 0.024 mg/L penicillin, 0.025 mg/L streptomycin, 0.10 mg/L tetracycline, and 0.33 mg/L ceftiofur [
69].
Feeding calves a limited amount of liquid nutrition to encourage rumen development is an effective strategy that promotes the transition from a non-ruminant to ruminant state. The nutrient value of MR was lower than that of WM and M. The digestibility of milk protein ranged from 90% to 97% [
70], whereas that of MR containing soy protein ranged from 70% to 78% [
4]. Therefore, the amount of digestible protein in calves fed MR may be lower than those fed M or WM, which would account for the slower growth rate. Starter feed intake is a good indicator of rumen development. Morrill (2012) [
71] reported that starter feed intake of 700 to 1000 g/d for 3 consecutive days was an adequate weaning criterion. In this trial, we observed that calves in WM consumed less starter feed than the above weaning requirement, which was in accordance with the data of stomach weight obtained from male calves. The starter feed intake was similar for calves in the M and MR groups, although a much greater TVFA level was observed in the MR group. This may suggest poor-efficient VFA absorption in the MR group. The soy protein in the MR would impair intestinal epithelial development, primarily due to the existence of antinutritional factors [
3] and increased intestinal abnormalities [
72]. As mentioned above, some of the MR may directly leak into the rumen, thus slowing rumen papillae growth. Moreover, MR containing soy protein negatively affects calf growth, metabolic status, and small intestine development, which are effects that may inhibit rumen development indirectly [
10]. Starter feed intake is generally considered to be the main stimulator of rumen development. Our results, however, suggest that rumen development is also closely associated with the type and composition of liquid feed.
The postnatal period may be the most critical window for rumen manipulation, and the early feeding regime may lead to permanent changes in the rumen microbial composition [
11,
12,
13]. It has been reported that the structure of the bacterial community established in lambs was affected by the diet fed around the time of weaning; an effect that persists for over four months [
14]. Ruminal bacterial communities of lambs can be modified by the diets of the maternal ewes and lambs or by inoculation treatment: this modification lasts until five months of age [
15]. We noted a long-lasting effect on rumen fermentation and bacterial community at the age of six months for the calves fed MR. The greater pH and ratio of acetate to propionate in the MR6 group might have a close relationship with the rumen bacterial composition. The species
Prevotella sp. DJF CP65 is a member of the genera
Prevotella 7, and its presence positively correlated with the molar proportion of propionate and negatively correlated with the pH value, the molar proportion of acetate, and the acetate/propionate ratio. As the abundance of
Prevotella 7 was significantly lower in the rumen of calves fed MR, it might explain the differences found for the rumen fermentation parameters. Conversely, a significant increase in the presence of the species
rumen bacterium NK4A214 might contribute to the increased pH, acetate/propionate, and decreased propionate. Reddy (2017) [
73] indicated that the early diet affects rumen papillae development, which may result in the differences observed in later carcass traits of beef. Therefore, we speculated that the ruminal imprinting effect in calves fed an MR diet might be associated with the development of the rumen epithelium. Moreover, although the feeding, management, and field conditions for all calves were the same from the ages of two to six months, various barn effects may influence the rumen environment of calves, thus impairing the imprinting evidence from the early feeding regime in this trial. Not all nutritional interventions in the early life of a calf can promote the establishment of different microbial populations in the rumen of the young animal. Providing supplemental plant extracts to calves did not yield long-term effects on rumen fermentation patterns or the bacterial community [
16]. Similarly, in our results, we observed that differences in ruminal fermentation and the bacterial community diminished at six months of age.