Growth and slaughter performance
As shown in Table 1, DM intake, average daily gain (ADG), body weight (BW), and the ratio between DM intake and ADG were similar between the two diets. The slaughter BW, carcass weight, and dressing percentage were similar between the two groups. The weight of head and spleen in alfalfa silage diet (AS) group lambs were greater than those in mulberry leaf silage diet (MS) group (P < 0.05). The weight of kidney in AS group lambs had a tendency to be lower than that in MS group (P = 0.053).
Meat quality characteristics
Meat quality characteristics were similar between the two diets, as shown in Table 1. Compared with the AS diet group, the lightness (L*) value at 45 min had a tendency to be lower in the MS diet group compare to AS group (P = 0.10).
Nutrients metabolism, antioxidant activity, and immune response
Overall, serum lipid metabolite levels including the concentrations of cholesterol, triacylglycerol, and the high density lipoprotein cholesterol (HDL) and low density lipoprotein cholesterol were un-affected by the MS diet (Table 2). The serum blood urea nitrogen (BUN) concentration was greater in the MS group than in the AS group (P < 0.01) but not for the serum albumin and globulin. As shown in Table 2, the activity of serum catalase (CAT), glutathione peroxidase (GSH-PX), superoxide dismutase (SOD), and total antioxidant capacity (T-AOC) were significantly greater in the MS group than in the AS group (P < 0.05). The content of malondialdehyde (MDA)was significantly lower in the MS group than in the AS group (P < 0.05). Compared to the AS group, the serum interferon-γ (IFN-γ) concentration was significantly increased in the MS group (P < 0.05). No significant difference was found for the tumor necrosis factor-α (TNF-α) concentration between the two groups.
Fermentation characteristics
As shown in Table 3, rumen pH value and the concentrations of ammonia-N and total volatile fatty acids were similar between the AS and MS groups. The molar proportion of the isobutyrate had a tendency to be lower in the MS group than in the AS group (P = 0.09).
Change in ruminal bacterial communities
The Good’s coverage of all samples was above 0.99. The Chao, Sobs, Shannon, and Ace indexes of bacterial richness and diversity were different between AS and MS (Table S2). The NMDS plots (Fig. 1A) showed that the clouds derived from the AS and MS data were clearly separated from each other with a significant stress value (0.026). There were 1,046 OTUs that were identified in both of the AS and MS groups, while 973 and 478 specific OTUs were observed in AS and MS groups, respectively (Fig. 1B). The anosim (analysis of similarities) based on bray-curtis distances showed significant different between the two groups (P = 0.007).
Five bacterial phyla were identified in the rumen samples that had relatively higher abundances (>1%), including Bacteroidetes, Firmicutes, Proteobacteria, Actinobacteria, and Kiritimatiellaeota, among them, Bacteroidetes was significantly lower in MS group compare with AS group (Fig. 2A). There were 106 bacterial taxa identified at the genus level, and top 10 genera were present with relatively high abundances (Fig. 2B), including Prevotella_1, Selenomonas_1, Rikenellaceae_RC9_gut_group, Prevotella_7, Succiniclasticum, Bifidobacterium, Prevotellaceae_UCG-001, Succinivibrionaceae_UCG-001, Ruminococcus_2, and Veillonellaceae_UCG-001. We identified 23 genera as different bacteria based on the Wilcoxon rank-sum test (Fig. 3A), in which, 12 genera were more abundant in the MS samples, including Advenella, Tannerella, Succinivibrionaceae_UCG-001, Schwartzia, Bifidobacterium, Succinivibrio, Mitsuokella, Shuttleworthia, Olsenella, Howardella, Syntrophococcus, and Ruminococcus_gauvreauii_group. and 11 genera were with lower abundance in MS group, such as Prevotella_1, Erysipelotrichaceae_UCG-004, Ruminococcaceae_UCG-010, Fibrobacter, Ruminiclostridium_6, Oscillospira, Prevotellaceae_UCG-003, rumen_bacterium_YS3, Ruminococcaceae_UCG-001, Lachnospiraceae_XPB1014_group, and Lachnospiraceae_NK4A136_group. As shown in the non-strict version of LEfSe analysis (Fig. 3B, C), fourteen clades were more abundant in the MS samples, in which six genera were shown out as Howardella, Pelagibacterium, Protochlamydia, Shuttleworthia, Advenella, Planctomicrobium. Two clades were more abundant in the AS samples, in which the genus Lachnospiraceae_NK4A136_group was enriched.
It was found that the ruminal bacteria communities were related to the blood antioxidant activity ad immune function indices (Fig. 4). In detail, Lachnospiraceae_XPB1014_group and Ruminococcaceae_UCG-010 were both positively correlated with MDA (r = 0.73, P < 0.01; r = 0.70, P < 0.05), but negatively correlated with CAT (r = -0.63, P < 0.05; r = -0.66, P < 0.05), GSH-PX (r = -0.76, P < 0.01; r = -0.64, P < 0.05), SOD (r = -0.73, P < 0.01; r = -0.71, P < 0.05), and T-AOC (r = -0.77, P < 0.01; r = -0.70, P < 0.05). Fibrobacter was positively correlated with MDA (r = 0.60, P < 0.05), but negatively correlated with SOD (r = -0.61, P < 0.05) and T-AOC (r = -0.60, P < 0.05). The rumen_bacterium_YS3 was positively correlated with MDA (r = 0.68, P < 0.05), but negatively correlated with GSH-PX (r = -0.71, P < 0.01), SOD (r = -0.66, P < 0.05), and T-AOC (r = -0.69, P < 0.05). The Schwartzia was positively correlated with GSH-PX (r = 0.61, P < 0.05), CAT (r = 0.61, P < 0.05), and T-AOC (r = 0.58, P < 0.05). Ruminococcus_gauvreauii_group was positively correlated with SOD (r = 0.59, P < 0.05). Prevotellaceae_UCG-003 was positively correlated with MDA (r = 0.62, P < 0.05), but negatively correlated with SOD (r = -0.63, P < 0.05), T-AOC (r = -0.62, P < 0.05), and IFN-γ (r = -0.64, P < 0.05). Bifidobacterium (r = 0.76, P < 0.01), Howardella (r = 0.72, P < 0.01), and Olsenella (r = 0.62, P < 0.05) were positively and Prevotella_1 (r = -0.75, P < 0.01), was negatively correlated with IFN-γ.