Brazil is the largest beef exporter in the world with an annual export of 2.49 million tons of carcass weight in 2019 (ABIEC, 2020; USDA, 2020). In addition, beef production is an important driver of the economy of Brazil, accounting for 8.5% of total gross domestic product of Brazil in 2018 (ABIEC, 2020). The predominant breed found in more than 90% of commercial beef herds in Brazil is the Nelore (ABIEC, 2020). This breed is resilient in the Brazilian environment, and most importantly, it is well adapted to the traditional pasture-based beef production system. However, in recent years, the adoption of intensive and semi-intensive production systems has increased, mainly due to pressure to scale up the production in response to the growing protein demand for human consumption and secondarily to the decline pasture land in Brazil (Maciel et al., 2019; Vale et al., 2019). The major difference in management practices between traditional and semi-intensive production systems is that large amounts of concentrates and feed supplements are used in semi-intensive systems. In addition, pasture is well managed in semi-intensive systems to provide cattle access to high-quality forages. Many studies have explored the relationship between cattle production system, dietary components, and the resident gut microbiota (Mu et al., 2019; Qiu et al., 2019; Andrade et al., 2020; Freetly et al., 2020). Most studies focused on the ruminal microbiota, which is understandable considering their substantial contribution to the generation of substrates for energy and protein supply to the host. However, the lower gut microbiota play an important role in animal health and immunity and contribute, to a lesser extent, to energy supply to the host (Zhang et al., 2019; O’Hara et al., 2020; Valerio et al., 2020). Therefore, it is critical to characterize the fecal microbiota to understand their contribution to host health and nutrition and to understand the effect of management system on the fecal microbiota composition.

In the recent years, there has been an increasing acknowledgment of the need to examine host-microbiome interactions in the rumen and in the lower gut to fully exploit the potential of the gut microbiome for sustainable ruminant production (Hagey et al., 2019; O’Hara et al., 2020). In addition, feces has been used as a proxy of the gut microbiota in animals because of the non-invasive nature of fecal sampling (Muiños-Bühl et al., 2018; Yan et al., 2019; Andrade et al., 2020). A few studies have investigated the association between fecal microbiota and animal phenotypes in Brazilian beef cattle (de Oliveira et al., 2013; Bessegatto et al., 2017; Lopes et al., 2019; Andrade et al., 2020), but there has been little investigation of changes in the microbiota in feces of beef cattle raised under different production systems.

Although investigation of the microbiota provides useful information about the composition of the microbial community, this does not of itself provide a meaningful insight into functional changes in the microbiota. Therefore, we investigated the fecal microbiome by characterizing the functional metagenome and the composition of microbial communities in the feces of beef cattle raised under semi-intensive (fed concentrate supplement) or traditional (without concentrate supplement feeding) management systems. Because the diet plays a key role in manipulating the composition and metabolism of the gut microbiota and the Brazilian beef production systems are distinguished primarily by their difference in dietary management (Myer et al., 2017; Vale et al., 2019), we hypothesized that beef cattle raised under semi-intensive and traditional production systems will show differences in their fecal microbiome. The objective of this study was to determine the taxonomic and functional variations in the fecal microbial communities seen across the beef production systems using a metagenomic approach. Such investigation will provide further insight into the association between management practices and ruminant production and health mediated by changes in the composition and function of the gut microbiota.

The present study, to our knowledge, was the first to compare the cattle fecal microbiome from Brazilian traditional and semi-intensive beef production systems. The bovine fecal microbiota is complex, and management practices play an integral role in shaping the establishment of the fecal microbial community (Shanks et al., 2011). The data from the current study showed that the management system had a significant impact on shaping the beef cattle fecal microbiome diversity and structure. Unlike other studies, age of animals did not influence the fecal microbiome in the current study. This discrepancy was likely because young animals recruited in the current study were relatively more mature (12 months old) compared to other studies that reported age-driven changes in the fecal microbiome (Oikonomou et al., 2013; Klein-Jöbstl et al., 2014). In the current study, I farms clustered distinctly from T farms for both taxonomic and functional structure of the fecal microbial communities, showing the difference in fecal microbiome between the production systems differed in feeding management. Difference in the structure of bacterial communities from feces of cattle fed different diets has been reported before by Kim et al. (2014) and Azad et al. (2019). In the current study, relatively low taxonomic diversity (Shannon diversity index) in the microbial community of fecal samples from I beef farms was consistent with a recent finding that fecal bacterial diversity in growing cattle decreased as dietary concentrate and energy increased (Zhang et al., 2018; Qiu et al., 2019). Similarly, in a recent study, the diversity and structure of the rumen and fecal microbiota in cattle responded differently to dietary treatments (Azad et al., 2019). Alpha-diversity indices of rumen microbial communities have also been reported to contribute to the variation in feed efficiency of cattle, where inefficient animals possessed more diverse microbial communities (Li et al., 2019a). Interestingly, in the current study, relatively less diverse bacterial diversity in the feces from I farms had highly diverse functional profile, suggesting more efficient microbiome in the gut of cattle reared in I management systems. This conclusion could be supported by recent findings that the rumen microbiome of more efficient cattle had more diverse functional potential and higher activities compared with that of less efficient animals (Shabat et al., 2016; Li and Guan, 2017). Similarly, Li et al. (2019a) reported that rumen microorganisms of inefficient cattle produce more products that can be harmful or of little use, whereas relatively simple and less active microbiota in the rumen of efficient animals produced fewer but more useful fermentation products that were absorbed more efficiently by the host.

In addition, the correlation of the fecal microbiome with CP, NDF, and %N on I farms and with %C on T farms suggested a strong influence of the feed composition on the fecal microbiome. Relatively higher concentration of CP and NDF in forages and supplements used on I farms compared with that on T farms (63 and 788 g/kg DM vs. 51 and 737 g/kg DM, CP, and NDF, respectively) indicates higher nutrient intake by the cattle reared in I systems, with potentially better synchronization of energy and N availability for the fecal microbiota that are essential for the fermentation process.

In the current study, relatively higher abundance of the phylum Firmicutes but lower abundance of the phylum Bacteroidetes in the beef cattle from I farms suggested the possibility of beef cattle reared in I systems being more efficient in utilizing nutrients. Díaz Carrasco et al. (2017) suggested that increased F/B can improve performance in bovines, citing previous studies that reported a correlation between F/B and phenotypic parameters of performance such as body weight gain and milk fat yield (Jami et al., 2014; Myer et al., 2015). In addition, increased abundance of the genus Succinivibrio in the gut of possibly more efficient beef cattle reared in I system is consistent with another report evaluating the gut microbiota and feed efficiency in cattle (Li et al., 2019b). Accompanied with a shift in the taxonomic composition, the functional profile of the fecal microbiota in beef cattle from I farms differed from that T farms with greater abundance of gene coding for energy production and nutrient metabolism including biosynthesis of amino acids, ubiquinone, and other terpenoid-quinone. Nutritionally essential amino acids synthesized de novo by the gut microbiota work as potential regulatory factors in amino acid homeostasis and gut health (Lin et al., 2017), suggesting that the gut of beef cattle reared in I systems harbored a relatively more efficient microbiome characterized by enriched gene coding for amino acid synthesis. Increased abundance of gene coding for ubiquinone and other terpenoid-quinone in beef cattle feces from I farms further suggested that beef cattle reared on I farms were possibly more efficient in utilizing energy because high producing cows have been previously found to have upregulated metabolic pathways of ubiquinone and other terpenoid-quinone biosynthesis (Mu et al., 2019).

In addition to the dietary factors mentioned above, it is likely that the fecal microbiome in beef cattle was influenced by dietary mineral supply, as indicated by a significant correlation between the microbial diversity and fecal mineral concentrations observed in the current study. Considering that feces can be a bioindicator of dietary nutrient supply, which has proven to be a cost-effective and reliable method in studies of ruminants (Showers et al., 2006; Fredin et al., 2014; Yang et al., 2020), relatively higher fecal concentrations of P and Zn in I farms and As and Pb in T farms suggested that the dietary supply of these minerals to beef cattle was higher in respective farms, which might have contributed to the difference in fecal microbial community structure between I and T management systems. Indeed, it is widely known that minerals and trace elements are essential micronutrients and an important driver of gut microbiota modulation (Skrypnik and Suliburska, 2018; Chang et al., 2020; Yang et al., 2020). In addition, hematological and biochemical parameters from the same animals used in the current study were evaluated by Jimenez et al. (Unpublished¹). Although hematological parameters were not influenced by the production system, blood urea concentration was higher in beef cattle reared on I farms, and blood glucose concentration was higher in beef cattle reared on T farms. This indicates that differences that were observed in the microbiome were not reflected in the hematological response of the animals, but there were differences in blood biochemistry, which would be expected as a result of the differences in the diet.

Our analysis also showed that specific fecal nutrients were correlated with bacterial diversity. In the current study, bacterial diversity in the feces from I farms was negatively correlated with fecal P and Zn concentrations. However, this reduced diversity could be explained by increased abundance of specific groups of beneficial bacteria (family Lachnospiraceae and genera Anaerostipes, Blautia, Butyrivibrio, Eubacterium, Roseburia, and Ruminococcus) in I farms. All these bacteria are known to produce butyrate in the gut (Li et al., 2012; Levy et al., 2016; Geirnaert et al., 2017), providing a source of energy for the mucosa and hence promoting epithelial health (Kim et al., 2014). Increased dietary phosphorus supply has been related to an increase of butyrate-producing bacteria in the cecal digesta of broiler chickens (Borda-Molina et al., 2016). Similarly, Zn is an important micronutrient that promotes the growth of beneficial gastrointestinal bacteria responsible for maintaining epithelial integrity (Chang et al., 2020; Yang et al., 2020). In addition, the higher abundance of Bifidobacterium in I samples suggested a healthier microbiota in the gut of beef cattle reared in I systems, because this microorganism has been considered both a probiotic and symbiont (Feng et al., 2018) and was associated with healthy fecal microbiota in dairy calves (Gomez et al., 2017).

Interestingly, the most abundant class abovementioned belongs to the phylum Firmicutes. An increase in the abundance of Firmicutes in the gut of semi-intensively managed beef cattle with possibly high dietary supply of Zn was in accordance with the finding by Chang et al. (2020), who reported a higher abundance of Firmicutes in the fecal microbiota of newborn dairy calves supplemented with Zn. In addition, it has been reported that Zn is a gatekeeper of immune function (Wessels et al., 2017) and gut microbiota play a key role in host immunostasis (Lazar et al., 2018). Interestingly, functional analysis showed that the fecal bacterial community in beef cattle reared on I farms had a higher abundance of sequences related to the biosynthesis of peptidoglycan and immune system, which are important for generating innate immune memory and the defense mechanism in bacteria (Horvath and Barrangou, 2010; Negi et al., 2019).

In the current study, the microbiota in beef cattle feces from T farms was characterized by an increased abundance of the phylum Proteobacteria, which has been associated with diet change, gut dysbiosis, or enrichment of stress-response genes in the gut microbiota of cattle (Auffret et al., 2017). Increased abundance of Proteobacteria in parallel to reduced abundance of Firmicutes in the feces from T farms could be attributed to lower fecal concentration of Zn and hence lower dietary supply of Zn to beef cattle on T farms that are used as indicator of gut dysbiosis (Reed et al., 2015; Wu et al., 2019). Therefore, a lower dietary supply of Zn to animals, which was possibly the case on T farms, may contribute to the disruption of gut microbiota homeostasis.

In addition to the abundance of Proteobacteria, genes related to virulence, disease and defense, LPS biosynthesis, sphingolipid metabolism, cell growth and death, and environmental adaptation were more abundant in the fecal microbiome of beef cattle from T farms. Genes related to virulence, disease and defense, or stress response have been found to be associated with an increased abundance of Proteobacteria (Auffret et al., 2017). An enrichment of genes related to the biosynthesis of LPSs in feces from T farms can be explained by an increased abundance of Proteobacteria because all Proteobacteria are Gram-negative bacteria and Gram-negative bacteria contain LPS in their outer membrane (Cani et al., 2007). LPS is a potent endotoxin and has been associated with gut disorders and disease in the host (Shin et al., 2015; Plaizier et al., 2018).

Relatively high fecal concentrations of As and Pb in beef cattle reared on T farms suggested that the gut microbiota in these cattle were exposed to high concentrations of these heavy metals. Although oral administration of a high concentration of As resulted in the enrichment of genes related to LPS biosynthesis in the gut microbiome of mice (Chi et al., 2017), the exposure to a high concentration of Pb leads to a perturbation in the gut microbiota characterized by a reduced abundance of beneficial bacterial genera (Gao et al., 2017). Therefore, in beef cattle reared on T farms, the exposure of the gut microbiota to high concentrations of heavy metals might have contributed to relatively poor gut health or development of a microbiota that is prone to dysbiosis. However, it is difficult to attribute the difference in fecal concentrations of As and Pb between I and T systems to different dietary management practices because the exposure of beef cattle to these heavy metals can be through the consumption of soil, water, and air (Johnsen and Aaneby, 2019). In addition, the concentrations of heavy metals reported in the current study are very low compared to their toxic doses, and therefore, these heavy metals might not have contributed to the gut microbiota modulation as much as they would at higher doses (Reis et al., 2010).

The effect of the I and T production systems on microbiome dynamics was assessed using the co-occurrence network analysis, which has been reported to be useful to visualize the impact of different environmental factors on the adaptability of the microbiome (Wang et al., 2019). A comparison of the complexity of interactions between microbes within each production system revealed that the microbial community in the gut of beef cattle reared in I systems had a greater number of microbial groups interacting, but with less connectivity among the microbes (Table 3). Our observation is based on the higher number of nodes and edges found in I farms compared with that in T farms. However, the microbial networks of the I farms had a lower clustering coefficient, which is a measure of the degree to which nodes tend to cluster together, revealing a closer network. This lower connectivity can also be observed in the higher number of communities in the I farm and higher modularity. A more modular structure is characterized by the presence of different groups of nodes (i.e., communities) with some independencies between groups (Newman, 2006). The higher number of communities suggests an increased diversity in species roles and functionality, increasing niche overlap (Poudel et al., 2016). Besides, we observed a lower average path length in T farms. This network property is defined as the average number of steps along the shortest paths between each node, being a measure of efficiency of a network (Zhou et al., 2010; Mendes et al., 2018). Our finding was consistent with the finding of a previous work (Wolff et al., 2017) that reported complex and variable changes in the metabolic networks of the ruminal microbiome as a result of dietary changes. Another study with cows fed corn silage, grass silage, or grass hay diet also provided a deeper insight into how different diets and microenvironments influence the complicated network of bacterial interactions and adaptations (Deusch et al., 2017). In our study, our network analysis revealed that the microbial community was affected by semi-intensive system, increasing the number of groups interacting within the community but with less connectivity among them. This shift in the community dynamics between the two production systems has a potential effect on metabolic pathways within this microbiome.

Metagenomic investigation of the fecal microbiome in Brazilian beef cattle revealed that some microbial taxa and their genes differ substantially between semi-intensive and traditional production systems with lower taxonomic but greater functional diversity in the fecal microbiome of beef cattle reared in the semi-intensive system. In intensively managed beef cattle, the fecal microbiome was characterized by an increased abundance of beneficial bacteria including butyrate-producing bacteria and microbial genes related to nutrient metabolism, energy production, and immune system. Many dietary and fecal parameters were associated with both taxonomic and functional diversity of the fecal microbiome, but further work is needed to evaluate causative relationships between them. Together, our study findings suggest that the Brazilian semi-intensive management system could promotes a healthier and more efficient gut microbiome than the traditional system, which was accompanied by alterations in fecal microbial community via driving an increase in the abundance of beneficial bacteria and functional genes and supporting the development of a complex microbiome.

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.mg-rast.org/mgmain.html?mgpage=search&search=mgp89663, mgp89663.

The animal study was reviewed and approved by Ethics Committee on the Use of Animals (CEUA) of the Center for Nuclear Energy in Agriculture, University of São Paulo (CENA/USP), Piracicaba, Brazil. Written informed consent was obtained from the owners for the participation of their animals in this study.

HL, CR, PR, and AA: conceptualization, validation, supervision, and funding acquisition. PC: performed the data analysis and writing original draft. CJ: performed the laboratory work and writing. LM: performed bioinformatics analysis. LG and VS: collected the samples. ED: contributed to lab analysis. All authors read and approved the final version.

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.