Previous studies proposed that the feasibility of rumen fistula in animal experiments is based on the assumption that rumen fistula surgery does not impact the rumen function (1). Rumen fistula installation is a challenge for ruminants and requires a slit through the abdominal cavity and rumen, causing serious trauma (2, 3). As we all know, the rumen microbial community and protozoa undertake the digestion and metabolism of fiber, starch, protein, and other substances in an anaerobic environment (4). The destruction of the anaerobic environment is a disaster for the microorganisms exposed to the oxygen stress in the rumen after fistula surgery (5). Compared with the intact rumen, the rumen microbial diversity of the fistulated is reduced (6–8), and the total volatile fat acid (TVFA) increased and the pH decreased (9). The fistulated abdominal cavity owned a lower proportion of CH₄, H₂, and CO₂, and a higher proportion of O₂ and N₂ in the rumen headspace gasses, which results in a decrease in energy utilization (8, 10).

On the other hand, the presence or absence of fistula surgery directly determines the sampling methods. Svga et al. (11) considered that the oral tube sampling is not a feasible alternative to fistula sampling to research the gas production and composition when comparing oral tube sampling to fistula sampling. In addition, there is little difference in profile of VFA (12) and microbial community because of the influence of saliva (13–15). However, some scientists concluded that there was no difference between oral tube sampling and fistula sampling, and it can adequately represent rumen fermentation parameters such as the pH, VFA profile, and mineral concentrations (16, 17).

The contribution of fistula animals cannot be denied, and they had been applied in ruminants for many decades (17). However, there is still controversy surrounding the question of whether the sample of rumen fistula animals really representatively reflects the purpose of the experiment for animal experiments. In practice, many animal trials and sampling methods leave these changes of microbial community wholly out of account. In some cases, comparing the sample of the oral tube and fistula on fistulated animals was equivalent to comparing with the sample of the fistulated or intact (15, 17, 18), which lacks sufficient evidence to provide proof. At present, figuring out the impact of fistula needs to unify three key variations: whether fistulated or not; whether it is the same group of animal before and after the fistula; whether a unified sampling method has been selected (oral tube or fistula). In addition, the different fistula methods (the one-step and two-step surgery) (19), individual differences, and species genetic backgrounds (15), may also be considered, but the impact of fistulation on destruction of rumen environment may be consistent.

Therefore, it is necessary to study the impact of rumen fistulation on the microbial community and the digestive physiology of ruminants. In this experiment, six sheep from the same batch were used to monitor the change before and after the fistulation, which could systematically reveal the effect of rumen fistulation on the microbial community and the digestive physiology of ruminants.

In this experiment, the DMI of sheep was not affected by fistula surgery, but the output of dry matter feces increased gradually, which resulted in a gradual decrease in DM digestibility until the 12th day (Figure 1). The output of dry matter feces decreased 14 days after fistula surgery. There was no difference in DM, OM, NDF, and ADF between before and 14 days after surgery (Table 1). Nevertheless, in terms of rumen microbes, we observed that the diversity of rumen microbes was increased and the relative abundance of Prevotella was decreased. The contradictory result may be attributed to two factors. The fistula surgery destroyed the pressure environment of the rumen and slowed down the peristalsis, to which led an increase in the retention time of the diet in the rumen (28), and it was conducive to rumen digestion. Likewise, the change in the composition of rumen microorganisms, the decrease in the relative abundance of Prevotella, and the increase in the unclassified microorganisms were observed, which influence the digestion activity of rumen microorganisms. From the results of previous studies, MacRae et al. (29) and Moate et al. (10) also observed no difference in the digestibility of DMI and DM on sheep and Holstein dairy cow in the ARS14. Furthermore, Hirayama et al. (28) observed that the fistula surgery reduced rumen pressure and rumen content passage rate on goats, thereby increasing the cumulative number of recovered particles in feces. From a numerical view in this experiment, the apparent digestibility of DM, OM, NDF, and ADF in the ARS14 were lower than those of the DRS1, but there was no difference. The phenomenon indicated that the influence of rumen microorganisms may be slightly greater than the slowing down of rumen peristalsis, which contributed to the conclusion that rumen fluid of rumen fistula animals cannot completely reflect the rumen fermentation function (11, 30). In summary, the rumen fistula surgery has no effect on the apparent digestibility of DM, OM, NDF, and ADF, and its rationalities are acceptable by judging from present results and previous studies.

The fermentation parameters are important indicators for determining the impact of fistula surgery on the rumen microbial function of sheep. In this experiment (Table 2), compared with the ARS14 and DRS1, the value of rumen pH and the concentration of butyrate were decreased (P < 0.05), the concentration of TVFA and acetate were increased (P < 0.05), as observed in the ARS3. However, there were no difference in rumen pH, the concentration of TVFA, and acetate between the ARS14 and DRS1. Consistent with our results, Wang et al. (7) also reported that no difference in rumen pH was observed at 14 days between the fistulated and the intact goat (P > 0.05). However, Rumsey et al. (6) reported that the rumen pH of the fistulated cattle was lower than that of the intact, and the TVFA was higher than that of the intact after 20 days. Wang et al. (8) compared the fermentation parameters of fistulated and intact cows, and the results were consistent with Rumsey et al. (6). The rumen peristalsis was decreased after fistula surgery, leading to an increase in the retention time of feed and a decrease in the pass rate of rumen (28). A study reported fully verified this results that considering the degree of rumen acidosis in fistula cow, the intact cows have a characteristic difference in oxygen utilization and nitrogen metabolism and suggested that the results of rumen fistula cows cannot reflect the normal metabolic process (30). Judging from VFA results in this experiment, it objectively reflects that the metabolism of rumen microbes is indeed affected by rumen fistula surgery.

Considering the result of rumen microbial diversity, the Shannon-wiener and Simpson indexes are higher in the ARS14 than those in the ARS3 and DRS1 (P < 0.05, Table 3) and the results of PcoA analysis suggested that there are differences in the diversity of rumen microbes in the three stages after surgery (Figure 2). Wang et al. (7) and Wang et al. (8) also reported that the ruminal microbial diversity of goats and cows received fistula surgery was lower than that of the intact goats and cows. The fistula surgery usually led considerable amounts of air to flow into the rumen headspace (10) and introduce the unclassified microbes into the rumen during the surgery, which could lead to the increase of microbial diversity. Moreover, in accordance with Wang et al. (8) who studied the gas compositions in the rumen by comparing the intact and fistulated cow and observed that the concentration of CO₂, H₂, and CH₄ in fistula cattle are reduced and N₂ and O₂ are increased. Comprehensively, although the plug can prevent the unimpeded exchange of gas inside and outside of the rumen, it is inevitable that rumen fistula cause gas leakage. The destruction of the gas environment and the increase in the unclassified may be the reason for the change in diversity.

The different microbial diversity among the three stages may provide the evidence that the fistula surgery could change the rumen microbial composition in different taxonomic levels. In comparison with the DRS1, decrease in the relative abundance of Bacteroidetes was observed in the ARS3 and ARS14 (P < 0.05, Table 4). In comparison with the DRS1 and ARS3, the relative abundance of Firmicutes in the ARS14 increased to 44.58% (P < 0.05). The relative abundance of Proteobacteria in the ARS3 was higher than the ARS14 and DRS1 (P < 0.05). However, Wang et al. (7) reported that no differences in the relative abundance of Firmicutes and Bacteroidetes between the intact and fistulated goat after 2 weeks, but an increase in the relative abundance of Proteobacteria was observed (P < 0.001) (7). The similar relative abundance of Firmicutes and Bacteroidetes between the treatments mainly attributed to the different genotypes of rumen microorganisms carried by different species (8). Shin et al. (31) suggested that the increase of Proteobacteria is a distinctive signal for the occurrence of animal intestinal disorder and imbalance of dietary nutrition, which may indicate that a disorder of rumen microbes were caused by rumen fistula surgery.

Compared with the DRS1, the relative abundance of the Prevotella decreased (P < 0.05) nearly by 50% in the ARS3 and ARS14 (Table 5). However, Wang et al. (7) found no difference in the relative abundance of the Prevotella compared with the intact and the fistulated goat after 2 weeks. A study indicated that the increase by 5 fold in the level of recA mRNA was observed in response to conditions of oxidative stress when Prevotella was exposed to oxygen by removing the culture from the anaerobic box, and the recA genes are related to the stress response of cells when chromosomal DNA was severely damaged (32). The destruction of the anaerobic environment is the direct cause of the decrease of Prevotella in the present study. In addition, the DMI of sheep may be also the cause of the decrease in Prevotella. However, the relative abundance of Prevotella did not return to that before the operation in the ARS14, which may lead to the decrease in digestive activity of microorganisms in the ARS14.

Compared with the DRS1, the relative abundance of the Succinivibrio was increased in the ARS3 (P < 0.05) and then decreased in the ARS14 (P < 0.05, Table 5). The first reason was the decline in Prevotella, which caused the relative abundance of the Succinivibrio to increase. The second reason was the optimal carbon dioxide concentration and CO₂ balance of the Succinivibrio. A study indicated that Succinivibrio dextrinosolvens grow well in 5% CO₂ environment compared with 0% CO₂, and the 5% CO₂ flowing environment was contributed to the fixation of CO₂ (33). The ratio of CO₂ in the rumen headspace gas is 49.8%, and the dissolved CO₂ was 69.7 mM before rumen fistula surgery, but the ratio of CO₂ in the rumen headspace gasses was 13.4% after fistula surgery and the dissolved CO₂ was 43.6 mM (8, 10). The high concentration of CO₂ before the fistula surgery may inhibit the metabolic activity of the Succinivibrio (Succinivibrio produce CO₂ as anaerobic metabolism). Moreover, the CO₂ concentration after fistula surgery was closer to the optimum CO₂ concentration of the Succinivibrio and the increases of air circulation rate promote the CO₂ circulation. The third factor, and possibly the most likely factor, is that Succinivibrio has a wide selection of substrates, which can ferment dextrin, disaccharides and monosaccharides, and fructans (33), which makes it more energy-supplying than other microorganisms and better adapted to the rumen environment in the early stage of fistula surgery.

Furthermore, it should be noted that the relative abundance of un-classified microorganisms has increased. This result indicates that there was introduction of non-rumen microorganisms due to fistula surgery, but whether these non-rumen microorganisms are related to apparent digestibility, it cannot be determined due to the limitations of the sequencing data. According to microbial structure and apparent digestibility, this experiment could be read that the main factors affecting rumen digestibility are still concentrated in the obligated anaerobic (Firmicutes), such as the decrease of Prevotella (participating in synergistically digestion of fiber and starch) caused the continued impact on rumen digestive function in the ARS14. Considering the structure and diversity of microorganisms, a series of studies agreed that oral tube sampling and fistula sampling (fistula and non-fistula) can be substituted for each other in the composition and diversity of rumen microbes, and the representative samples can also be obtained (12, 17, 34). However, some scientists have made completely the opposite conclusions (11, 30), and suggested that the rumen fluid of fistula animals can not reflect the condition of intact animals. In summary, whether the fistula and intact sampling can replace each other may require reconsideration of the impact of rumen fistula surgery.

Through the analysis of the BugBase Feature Prediction, the changes in rumen microbial function have been further verified. In this trial, the microorganism abundance of aerobic and facultative anaerobic increased, and the microorganism abundance of anaerobic phenotypes decreased in the ARS3 (Figure 3). The study showed that the fistula surgery destroys the rumen anaerobic environment, resulting in the air enter the rumen along with the fistula moving, the activity of obligate anaerobic was impacted by the imbalance of rumen anaerobic environment (8, 10, 14). In addition, the changes in the microorganism abundance of contains-mobile element, form-biofilm and potential-pathogenic phenotypes increased in the ARS3 and then declined in the ARS14, which demonstrated that some rumen microbial community and function trend to become active after fistula surgery, then gradually forms a new balance as the internal environment becomes stable. The microorganism abundance of gram-negative and gram-positive phenotypes showed completely opposite results after fistula surgery. In addition, the microorganism abundance of stress-tolerant phenotype showed that rumen fistula surgery has caused stress in the ARS3. A study conducted by Dahl et al. (35) illustrated that the outer membrane-lipopolysaccharide portion of the gram-negative cell wall initially protects the bacteria from extracellular singlet oxygen, which further verified the existence of oxygen stress in the rumen of a fistulated animal. The microorganism abundance of stress-tolerant phenotype in the ARS14 was declined in those before the surgery, but the results of aerobic phenotype, anaerobic phenotype, and facultative anaerobic phenotype did not seem to be considered as recovery, or there was no difference. The acid-base balance of experimental acidosis given in the fistulated and intact illustrated that the intact cows have more pronounced and longer duration of acidosis caused by the administration of sucrose to the rumen and the trial of the fistulated cows do not reflect the normal metabolic process and characteristic differences were observed in oxygen utilization and nitrogen metabolism (30). All in all, the results of the BugBase feature prediction turn out that the function of rumen microorganism has experienced fluctuations due to the damage of rumen anaerobic environment, and then rebalance existed in this process, but whether the new balance can be equivalent to the balance of the intact needs to be further verified by experiments.

The present results illustrated that the rumen fistula surgery increased microbial diversity, and led to the decrease of Prevotella (Bacteroidota) 3 and 14 days after fistula surgery and the increase of Succinivibrio (Proteobacteria) 3 days after fistula surgery. In addition, the rumen fistula surgery leads to an increase of the microorganism abundance of aerobic, stress-tolerance, facultative anaerobic phenotype, and a decrease in anaerobic phenotype. However, there is no difference in DMI, apparent digestibility of DM, OM, NDF, ADF, pH, and TVFA both before and 14 days after surgery. The rumen fluid of fistula animals in microbiota research may not be representative of the intact animals, which requires further detailed functional analysis and identification.

The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found below: https://www.ncbi.nlm.nih.gov/, SUB9200751.

The animal study was reviewed and approved by the Institutional Animal Care and Use Committee of Lanzhou University. Written informed consent was obtained from the owners for the participation of their animals in this study.

YC and FeL: conceptualization, investigation, and writing—review and editing. BZ, DZ, and TG: validation. GY: resources. YC: data curation and writing—original draft preparation. FaL: supervision. FaL and FeL: project administration and funding acquisition. All authors have read and agreed to the published version of the manuscript.

The research was financially supported by the National Natural Science Foundation of China, grant number: 32072754, and the Program for Changjiang Scholars and Innovative Research Team in University, grant number: IRT_17R50.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.