Effect of hay supplementation timing on rumen microbiota in suckling calves

Abstract An animal feeding trial was conducted on 18 seven‐day‐old Holstein dairy bull calves weighing 42 ± 3 kg each. Calves were randomly assigned into three groups (n = 6 each). The dietary treatments were as follows: (1) milk and starter for the control group (MS), (2) supplementation of oat hay from week 2 on the basis of milk and starter (MSO2), and (3) supplementation of oat hay from week 6 on the basis of milk and starter (MSO6). All animals were fed starter and oat hay ad libitum. The major phyla in the different groups of rumen fluid included Firmicutes, Actinobacteria, Bacteroidetes, Proteobacteria, and Euryarchaeota. The major genera were identified, and major genera proportions in the three groups were as follows: Methanobrevibacter (Euryarchaeota), 2.1%, 1.7%, and 2.1%; Olsenella (Actinobacteria), 23.9%, 17.7%, and 12.8%; Prevotella (Bacteroidetes), 10.5%, 16.5%, and 19.2%; Dialister (Firmicutes), 3.3%, 4.1%, and 2.8%; Succiniclasticum (Firmicutes), 3.8%, 4.7%, and 9.2%; and Sharpea (Firmicutes), 0.4%, 2.5%, and 0.2%, respectively. There were no significant differences in the various phyla among the three groups (p > .05). The results showed that calves hay supplementation time did not affect the diversity of the rumen microbiota in the suckling calves. However, the hay supplementation altered the proportion of the various microbial populations, supplementation of oat hay from week 2 on the basis of milk and starter could improve calves rumen pH.


| Oat hay
The hay was an imported high-quality oat hay. The nutritional ingredients of the hay are shown in Table 1.

| Rumen fluid sampling and analysis
Rumen fluid samples were collected from the calves at 63 days of age using a rumen fluid collector. Sampling was conducted within 2-3 hr after calves were fed with the starter. Fluid samples were dispensed into two 50 ml centrifuge tubes and immediately frozen in liquid nitrogen for storage. The samples were delivered to the laboratory and slowly thawed at 4°C. Total DNA was extracted from the rumen fluid using the Stool DNA Isolation Kit (Tiangen, Beijing, China). The DNA samples were sent to RiboBio (Guangzhou, China) for V4 region of the 16s rDNA gene sequencing.

| Data analysis
Raw data were sorted using Excel (Microsoft Corp., Redmond, WA, USA). Statistical analysis was performed using the ANOVA module in SAS 8.0 (SAS Institute Inc., Cary, NC, USA). Differences in the means were tested using Duncan's multiple comparison test. A p < .05 indicates a significant difference, and a p < .01 indicates a highly significant difference in sequencing of the V4 region of the 16s rDNA gene. If the two paired end reads overlapped, the consensus sequence was generated by FLASH. Tags were redundant with Mothur (version 1.27.0). Then, we selected unique tag. Unique tag was preclustered by means of SLP, the tags were then clustered into OTU with a 97% threshold by using UCLUST (Edgar, 2010;Huse & Mark Welch, 2010).

| Alpha diversity analysis
The alpha diversity was evaluated using Chao, PD_whole_tree, Shannon, and Simpson indices ( Figure 1). The four groups of rarefaction curves tended to flatten out or reach a plateau, indicating that the sequencing depth generally covered all species.
In Table 2

| Beta diversity
We used a beta diversity analysis to examine the difference in the microbial community structure among the samples. The closer the beta diversity value is to 0, the smaller the difference is in species diversity between the two samples. Table 3 shows beta diversity values between the corresponding samples in the horizontal and vertical directions. The larger the value is, the greater the difference between the two samples, and vice versa (Table 3; Figure 2). The results suggested that there were bigger difference between sample 12 and 15, 5, 10, 2, 1, 11, 13, between sample 9 and 2, 4, 11, 15, 16. Note: treatments MS,c1,1-6; MSO2,c2,7-12; MSO6,c3,13-17.

| Heatmap of OTU richness
The information of species composition at the optimal classification level was clustered according to the distance between the samples and the species. Horizontal (sample clustering) and vertical clustering

| Microbiota composition analysis
The OTU sequences were subject to species annotation using the existing 16s database, followed by taxonomic classification into phylum, class, order, family, genus, and species. The number of tags at the different classification levels was counted for each sample. Based on the results, we identified the optimal classification level. Richness analysis was conducted according to the number of tags of the different species on annotations at each classification level. As shown in    Tables 5 and 6 compare  A few studies have suggested that Bacteroidetes play a role in the normal development of the digestive tract (Thomas, Hehemann, Rebuffet, Czjzek, & Michel, 2011). Additionally, Bacteroidetes are involved in the metabolism of bile acids and toxins or their conversion into mutagenic compounds (Smith, Rocha, & Paster, 2006

| CONCLUSIONS
Hay supplementation time did not affect the diversity of the rumen microbiota in the suckling calves with the condition that supplementation of oat hay from week 2 calves or week 6 on the basis of milk and starter.
However, it altered the proportion of the various microbial populations, and their structural differences indirectly affected the rumen pH.

ACKNOWLEDGMENTS
The study was financially supported by the fund from the Modern Values with different letters indicate significant difference (p < .05).