The most harmful diseases are mastitis and foot rot , which reduce the quality of dairy products, inhibit the development of the dairy industry, and have a negative impact on food security and human health. Bovine mastitis is a major disease affecting the dairy industry worldwide with huge economic losses and decreased animal health. Furthermore, it is a common but complicated disease in high-yielding dairy farms. During the course of dairy cow breeding, it is quickly spread and difficult to cure, resulting in serious economic losses. Farm management, particularly proper treatment of manure, has become an issue of concern in many farms. China has the largest population of livestock animals of any country in the world; however, many livestock farms have poor animal manure management facilities for the treatment and disposal of manure. It is therefore suggested that manure management needs to be improved in China. An increasing number of studies have found that poor farm management can lead to severe environmental problems such as the pollution of air, water, and land. The milk industry is one of the five leading strategic industries in the Ningxia region of China. However, few studies have examined the effect of poor dairy farm management on microbial community compositions and diversity among the dairy farm matrix in this region. By using pyrosequencing of metagenomic 16S rDNA, grow trays the objective of this study was to characterize bacterial diversity in feces, manure, and soils in dairy farms in the Ningxia region of China.Three dairy farms representative of typical dairy farm operations in the Ningxia region were enrolled in this study. Farms 1, 2, and 3 were located in the suburbs of Xingqing , Jinfeng , and Xixia districts of Yinchuan City, respectively. Between April and July of 2016, fresh feces, manure, and soil samples were collected from these farms.
Fresh fecal samples were collected within 30 seconds of excrement from lactating cows. Manure samples were collected from piles of accumulated manure without further treatment. Soil samples were collected from around the farms at different depths . To eliminate error caused by individual differences or unrelated factors, fresh fecal samples were a blend of at least six cow feces and manure and soil samples were a blend of at least six sampling sites or depths. Each sample from each farm was mixed separately, with a total of 27 samples being generated after mixing. All samples were refrigerated immediately after collection and during transportation to the laboratory. Upon arrival at the laboratory, samples were stored at −80° C until processing.Metagenomic DNA was extracted from all types of samples using a QIAampR DNA Stool Mini Kit according to the manufacturer’s instructions. Sequencing by synthesis was performed on an Illumina HiSeq 250 platform .Sequence analysis was performed using the Sparse software . Sequences with ≥97% similarity were assigned to the same operational taxonomic unit . The representative sequence for each OTU was screened for further annotation. Sample reads were assembled using mothur v1.3213. Moreover, high-quality sequences were aligned against the SILVA database. Sequences were further qualitatively trimmed using a 2% cluster error, and chimeras were removed using UCHIME. Assignment of OTUs was performed at 97% identity using the furthest neighbor algorithm. Taxonomic assignments were made against the Ribosomal Database Project database. For comparisons, groups were normalized to include 27 samples, each randomly subsampled to 25,000 sequence reads . For determination of the percentages of sequence reads and OTUs unique to each group, no normalization was performed.
The UniFrac distances were calculated using QIIME software , and these data were used to build the UPGMA sample cluster tree. Jackknifed beta diversity included both unweighted and weighted UniFrac distances calculated with 10 times subsampling, and these distances were visualized by principal coordinate analysis. Principal component analysis , PCoA, and non-metric multidimensional scaling analysis graphs were drawn using the R software . Taxonomy assignment of OTUs was performed by comparing sequences to the Greengenes database . The Mann–Whitney U test was used to test for the significance of alpha diversity. Two-sided Student’s t-test was conducted to determine the significance of beta diversity between sample groups. Linear discriminant analysis coupled with effect size was performed to identify the bacterial taxa represented between groups at the genus or higher taxonomy levels. The functional profiles of microbial communities were predicted using PICRUSt. The bootstrap Mann–Whitney U test with 1000 permutations wasalso used to identify gene pathways or OTUs with significantly different abundances between groups. The R packages “phyloseq” and “heat map” were used for data analysis and plotting .In total, 2.2 million strands of 16S rDNA amplicon data were generated from the 27 samples using pyrosequencing. After trimming and cleaning, this number was reduced to 69,065 high-quality reads with a median length of 253 bp. Only 1.5% of the sequences were identified as chimeras and were excluded from further analysis. The number of sequences in the 27 filtered samples was in the range of 47,924 to 83,624 sequences, and after homogenizing these sequences, the sequences were concentrated to around 45,000. An OTU table was generated by clustering all of the sequences into OTUs with a 97% similarity level. The species observed among the samples showed that the same samples from different farms possessed the same patterns, with the number of microbial notes in soil being higher than that in other samples.
The samples were grouped by category, and the main annotations are shown in Table 1. A bacterial community bar chart of all samples was constructed at the phylum level, from which 47 units were annotated from fecal samples, 44 units were annotated from manure samples, and 50 units were annotated from soil samples . The percentages of each of the top 10 phyla in all of the samples are shown in Table 1. These phyla were abundant and accounted for >94.63% of the entire bacterial communities in all of the samples. Therefore, these 10 phyla of bacteria were chosen for further analysis. In the class-level analysis, 75 classes of bacteria were detected among the 10 phyla . At the order level, 63 bacterial orders were noted from the Proteobacteria.The distribution of the relative abundances of bacteria at the phylum level varied among fresh feces, manure, and soil samples. The dominating phylum of bacteria in fresh feces was Firmicutes , followed by Bacteroidetes , Proteobacteria , and Spirochaetes . In manure, the dominating phylum was Proteobacteria , followed by Firmicutes , Bacteroidetes , and Actinobacteria . In soil, the dominating phylum was Proteobacteria , followed by Bacteroidetes , Actinobacteria , and Firmicutes . Data show that Proteobacteria, Bacteroidetes, and Firmicutes were the main phyla in all types of samples, although the percentage of Firmicutes was slightly lower than that of Actinobacteria in the soil samples. The three phyla were distributed at approximately similar ratios in the two manure samples , while Proteobacteria dominated in one manure sample . In contrast to the fecal and manure samples, Proteobacteria dominated the bacterial communities in all of the soil samples . In fresh feces, the phylum Firmicutes was predominantly composed of the three classes Clostridia, Bacilli, and Erysipelotrichi, the phylum Proteobacteria was comprised mainly of the classes Gammaproteobacteria, Betaproteobacteria, Alphaproteobacteria, Deltaproteobacteria, and Epsilonproteobacteria, the phylum Bacteroidetes was mainly composed of the classes Bacteroidia, Cytophagia, Sphingobacteriia, Saprospirae, Rhodothermi, and Flavobacteria, and the phylum Spirochaetes was mainly composed of the classes Spirochaetes, MVP-15, Brevinematae, and Leptospirae . In manure, the compositions of the classes in the phyla Proteobacteria, Firmicutes, pruning cannabis and Bacteroidetes were similar to those in fresh feces. However, the phylum Firmicutes in manure consisted of an additional class called AHT28. In addition, the phylum Actinobacteria in manure was mainly composed of the classes Actinobacteria, Acidimicrobiia, Thermoleophilia, Rubrobacteria, Nitriliruptoria, and Coriobacteriia . The data within the OTU table was compared between 5,000 randomly selected samples each at a 97% nucleotide identity level. To investigate variations in the distributions within microbial communities, all OTUs were assigned taxonomically using the RDP classifier. Among a total of 8819 OTUs, most bacteria were concentrated into five phyla, namely, Firmicutes, Bacteroidetes, Proteobacteria, Spirochaetes, and Actinobacteria. The heat map results showed that the compositions of the microbes in the same types of samples appeared to be similar .PCA profiles indicated that microbial communities varied depending on the type of sample. Principal components 1 and 2 demonstrated 14.11% and 11.91% of the total variance, respectively . PCA profiles showed significant separations between F1, F2, and F3 and M1, M2, and M3 treatments at three sites, especially for S1, S2, and S3. Treatments M1–M3 had higher scores than the corresponding treatments F1–F3 along the PC1 axis. According to the PC2 axis, PCA profile scores for treatment S3 were higher than those for both of the corresponding treatments S1 and S2.
To determine the similarity between different samples, a clustering tree was constructed using UPGMA , which is a commonly used method for cluster analysis. Interestingly, the same type of samples from different farms clustered within the same branches . Further, beta analysis of microbial diversity showed that the diversity of species within the fecal, manure, and soil groups was rather small, in contrast to the large differences among the fecal and soil groups . In the analysis of human pathogenic bacteria, this trend was also detected. The boxplot showing the phylum level classification in terms of both bacterial diversity and the diversity of zoonotic pathogens revealed that the fresh feces contained a great abundance of Firmicutes, but a low diversity of zoonotic pathogens. By contrast, little change in diversity was observed in the accumulated manure or soil samples .The diversity and relative abundance of zoonotic pathogens in the manure and soil samples are shown in Figure 8. In total, 32 species of pathogenic bacteria were found in feces, manure, and soils. Acinetobacter calcoaceticus and Bacillus cereus were the dominant zoonotic species in feces, followed by Enterococcus faecalis, Streptococcus uberis 0140J, Escherichia coli O26: H11, Corynebacterium diphtheria, Staphylococcus aureus C0673, and Pseudomonas aeruginosa. The relative abundance of B. cereus in feces was 3 to 17 times higher, respectively, compared with that in manure and soil. Among the 32 pathogens, the relative abundance of 20 pathogens varied significantly in feces and manure, with the relative abundance in manure 2 to 19 times higher than that in feces. The number of Actinomycetes in manure was much higher than that in fresh feces. Further, at the genus level, although the number of genera in manure decreased, the number of bacteria and Pseudomonas increased significantly. This phenomenon was closely correlated with the abundances in feces and manure. In a more detailed classification order, the first dairy farm was detected having 27 zoonotic species, of which five species were increased in abundance in manure,including S. aureus M0406, Clostridium perfringens B str. ATCC 3626, A. calcoaceticus, Bacteroides fragilis NCTC 9343, and Bacteroides vulgatus CL09T03C04. In the second dairy farm, 34 zoonotic species were found, of which 14 species were increased in abundance in manure, namely, B. cereus, Listeria monocytogenes FSL R2-503, S. aureus M0406 and C0673, E. faecalis, S. uberis 0140J, Streptococcus dysgalactiae, Clostridium botulinum, A. calcoaceticus, Acinetobacter baumanii BIDMC 57, Proteus mirabilis BB2000, and Vibrio cholera VCC19. Finally, in the third dairy farm, 29 zoonotic species were found, of which 15 species showed increased abundance in manure, namely, B. cereus, L. monocytogenes FSL R2-503, S. aureus M0406 and C0673, B. cereus, S. uberis, S. dysgalactiae, E. faecalis, P. aeruginosa, Klebsiella pneumoniae, C. diphtheria, Yersinia pestis biovar Antiqua B42003004, V. cholera VCC19, and B. fragilis 3725-D9-ii. Statistical analysis of the OTU at the order level shows that the number of zoonotic bacteria in dairy farm 1 and dairy farm 2 was significantly higher in the manure than in the fresh fecal samples .In the Ningxia region, manure from most of the dairy farms is used as an organic fertilizer by local farmers without proper treatment. However, to our knowledge, the composition, diversity, and abundance of bacterial communities in the manure that has not been properly treated in this region are poorly understood, but this manure is being used as a fertilizer. Livestock and poultry manure contains feces, urine, litter, nose stains, blood stains, shed skin, hair, and placental material.