Animal experiments were performed under the institutional guidelines of Ecological Chicken Farm in Jianshi County of Hubei Province and Laboratory of Poultry Genetics, College of Animal Science and Veterinary Medicine, Huazhong Agricultural University, China. The experiments were performed according to recommendations proposed by the European Commission (1997) to minimize the suffering of animals.

Jingyang chickens were reared in the Fire Phoenix Ecological Breeding Cooperative of Jianshi County (Hubei, China). A total of 320 laying hens (body weight, 1.70±0.15 kg; 47 weeks old), derived from silver feather (S-line = 160 hens) and gold feather (G-line = 160 hens), under the two husbandry systems of enrichment cages (EC) and free range (FR) were randomly divided into four groups (80 hens per group): silver-feathered hens in enrichment cages (SEC), silver-feathered hens in free range (SFR), gold-feathered hens in enrichment cages (GEC), and gold-feathered hens in free range (GFR). For EC husbandry system, each hen was reared in an individual cage (70×60×75 cm) with controlled conditions of 18°C to 22°C, 16L/8D photoperiod/day. For FR husbandry system, chickens were raised in mountain areas characterized by arbor evergreen coniferous in the visible small stones and grains of sand, with a stocking density of 1.5 chickens per 10 square meters. Chickens in EC and FR were fed the same local standard diet (2,700 kcal/kg, 13% protein, 1% calcium, 0.45% phosphorous) (DB 42/T 686-2011) and drinking water ad libitum during the experiment period.

At the end of the 60-day experiment, 5 chickens were randomly selected from each treatment (a total of 4 treatments × 5 repeat = 20) and the cecal samples were collected in a cryopreservation tubes and then mixed and stored in liquid nitrogen at −70°C.

Microbial genome DNA was extracted from cecal samples by using QIAamp DNA stool mini kit (Tiangen Biotech, Beijing, China) according to the manufacturer’s instructions. PCR amplification of the V3–V4 region of the 16S ribosomal RNA (rRNA) gene was performed using the 341F/802R primer set (341F: CCTAYGGGRBGCASCAG; 802R: GGACTAC NNGGGTATCTAAT) as previously reported [17]. Only the products without primer dimers and contaminant bands were used for 16S rDNA high throughput sequencing at Shanghai Personal Biotechnology Co., Ltd (Shanghai, China).

The screened and obtained high-quality sequences uploaded to quantitative insights into microbial ecology (QIIME, v1.8.0), were clustered into operational taxonomic units (OTUs) with a sequence similarity of >97% using the USEARCH software.

The OTU analysis results of cecum microbes are presented in Table 1 and 2. A total of 22,660 OTUs consisted of 5,880, 6,245, 6,316, and 4,219 OTUs in SEC, GEC, SFR, and GFR, respectively (Table 1), and the annotation results are shown at various classification levels and the number of annotations in species is far less than that in genus level (Table 2). Moreover, the rarefaction curve (Figure 1) tends to be flat as sequencing depth increases and the number of OTUs reaches a plateau, with a sample coverage rate of over 0.96 (Table 3). These results confirmed that our sequencing has covered most of the sample species, and the sequencing results fully reflect the microbial diversity of the samples.

The results of abundance and diversity analysis are presented in Table 3. In silver-feathered chickens, Shannon index and Simpson index of SFR were higher than those of SEC. Nevertheless, compared with GFR, GEC showed higher values of chao1, Shannon index, and Simpson index. Bailey et al [18] found that exposure to prolonged restraint pressure (a stress method used to induce physiological responses in an animal by restricting its free movement) results in a decrease in the abundance and diversity of intestinal flora in mice. Thus, some of the above results in the present study could be explained by a prolonged restraint pressure-induced decrease in the relative abundance of bacteria of chickens in cages. However, Chao1, a species richness index, was lower in free-range groups than in cage groups. In our former research, a higher level of lactobacilli was observed in free range group, which could inhibit the proliferation of other bacteria and promote gut health [19]. GFR showed the maximum content of Lactobacillus, but the minimum content of chao1, Shannon index, and Simpson index in our present study. These data demonstrated that Lactobacillus did inhibit the growth of some bacteria.

The core microbiota of the 4 groups were counted at the phylum level (Table 4) and genus level (Table 5). At the phylum level, core members are microbes belonging to Bacteroidetes (49% to 60%), Firmicutes (21% to 32%), and Proteobacteria (2% to 4%), which are similar to the results reported in broilers [20], Dagu chickens [21], and Tibetan Chickens [22]. It is reported that the total proportion of Firmicutes and Bacteroides is even more than 90% in some mammalian species [23,24], and the Firmicutes/Bacteroides ratio can directly reflect the capacity of energy absorption and storage of host, such as the increase of the proportion of Firmicutes due to the deposition of fat [25], or the increase of the proportion of Bacteroides with the decrease of the Firmicutes/Bacteroides ratio due to enormous absorption of cellulose food [23,26]. In our study, the total proportion of Bacteroidetes and Firmicutes was more than 75% in Jingyang chickens, which was higher in free-range groups (SFR 81.61%; GFR 82.12%) than in enrichment-cage groups (SEC 76.38%; GEC 75.33%). The ratio of Firmicutes/Bacteroides was lower in GFR (23.00%/59.12%) than in GEC (21.20%/54.13%), whereas opposite results were obtained for silver-feather hens (SFR vs SEC). The reduced ratio of Firmicutes/Bacteroides in GFR group may be attributed to the higher cellulose food intakes in free range environment.

The core bacteria were Bacteroides (26% to 31%), Rikenellaceae (9% to 16%), Parabacteroides (2% to 5%), and Lachnoclostridium (2% to 6%) at the genus level. Ruminococcaceae was higher in SFR (2.25%) than in SEC (1.23%) as well as higher in GFR (1.4%) than in GEC (0.78%). Combined with our previous research results on cecum index, we speculate that the increased of free-range husbandry in the accumulation of microorganisms is related to cellulose degradation, which can be supported by the studies of Saengkerdsub et al [27], Lu et al [28], and Latham et al [29], who reported that ruminococcaceae was responsible for cellulose digestion.

A further analysis was conducted to discover the relative bacterial community abundance at the phylum level, and the results are shown in Figure 2. The phyla were divided into four groups. Group 1 (Synergistes, Phascolarctobacterium, Megasphaera, Megamonas, Sutterella, and Parabacteroides) was predominant in SEC. Group 2 (Ruminococcaceae, Lachnoclostridium, Alloprevotella, Christensenellaceae, Oribacterium, and Anaerotruncus) was typical in SFR. Group 3 (Treponema, Odoribacter, Prevotellaceae, Sphaerochaeta, Desulfovibrio, and Ruminococceae) and Group 4 (Bacteroides, Subdoligranulum, Eubacterium, Saccharimonas, Rikenellaceae, and Lachnoclostridium) were observed in GEC and GFR, respectively. The clustering was dependent on hen strains (gold vs silver).

The clustering result was consistent with the strain classification: the silver-feather strain and the gold-feather strain. Actually, factors affecting the intestinal microbiotas are host and environment, with the former mainly referring to heredity, age and endocrinology, and the latter involving diet, geographical environment, lifestyle, and the use of antibiotics [16,30]. Diet has been reported as the dominant factor for the intestinal microbial community in the formation of host genotype [16]. In this study, we explored other factors affecting cecum microbial diversity under the same dietary conditions. Our study proved that the genetic factor has a greater impact than husbandry system on cecum microbial diversity in Jingyang chickens.