Beef cattle production in the Intermountain West region of the United States — an area spanning nearly 500 million acres across states including Utah, Nevada, Idaho, Wyoming, Colorado, Montana, Arizona, New Mexico, and Oregon — relies predominantly on extensive rangeland grazing systems. Each year, more than 5 million beef calves are produced in this region under conditions characterized by seasonal forage shortages caused by low precipitation (125–500 mm/year) and high daytime temperatures (29-34°C; USDA-NASS, 2023). Expanding late-summer access to legumes and forbs could enhance production efficiency by improving nutrient retention in ruminants, increasing soil water- and nutrient-holding capacity, and supporting wildlife habitat (Rotz et al., 2019).

Historically, efforts to enhance rangeland productivity and forage quality have prioritized introducing species such as crested wheatgrass, aiming to extend the grazing season, increase forage yield, and reduce soil erosion (Holechek, 1981; Sharp, 1986). While crested wheatgrass is well-adapted to arid and semi-arid environments, its nutritional quality declines sharply in mid-to-late summer, leading to reduced livestock performance and increased supplementation costs (Savage and Heller, 1974). Furthermore, the reliance on low-diversity, grass-dominated landscapes has created ecological and nutritional bottlenecks that limit the resilience and multifunctionality of these systems (Distel et al. 2020).

To address these challenges, recent work has proposed the development of Smart Foodscapes —small, strategically distributed patches of diverse, high-quality forage species embedded within the broader rangeland matrix (Villalba et al., 2022, 2025). These “islands of diversity” can be more intensively managed (e.g., re-seeded or fertilized) at small spatial scales, allowing animals to graze the surrounding rangeland vegetation while accessing the islands under rotational systems for enhanced nutritional and functional inputs. Although this model requires greater additional labor and an initial investment, it leverages ecological principles of complementarity and diversity to enhance both livestock performance and ecosystem services.

A critical step toward implementing this model is the identification of effective forage mixtures that incorporate multiple functional groups — such as combinations of grasses, legumes, and forbs — to optimize rumen function, nutrient use, and microbial efficiency. Legumes such as sainfoin (Onobrychis viciifolia) and birdsfoot trefoil (Lotus corniculatus) are particularly promising, as they typically contain higher protein levels due to symbiotic nitrogen fixation and may also reduce energy losses via methane emissions through the action of condensed tannins (Mueller-Harvey et al., 2019). Forbs like small burnet (Sanguisorba minor), while less studied, may contribute additional nutritional and phytochemical diversity, including hydrolysable tannins and other plant secondary compounds that influence rumen fermentation and nitrogen dynamics (Stewart et al., 2019). There are known synergistic interactions among sainfoin and small burnet (Lagrange et al., 2020) and between condensed and hydrolysable tannins (Aboagye et al., 2018). Importantly, all of the cited studies were conducted in the Intermountain West, providing region-specific evidence of the potential value of these forage species and their interactions.

Thus, the objective of this study was to evaluate how strategic combinations of crested wheatgrass with selected legume (alfalfa, sainfoin, birdsfoot trefoil) and forb (small burnet) species affect rumen fermentation, nutrient digestibility, nitrogen metabolism, and microbial community. We hypothesized that forage mixtures incorporating multiple functional plant types would enhance fermentation efficiency and fiber degradation by modulating the rumen microbiome and improving microbial protein synthesis.

The nutritive value of the diets was relatively uniform across treatments (Table 2). Although legumes typically contain higher crude protein than other forages, crude protein levels across treatments averaged 14.1 ± 0.34. This uniformity is likely due to the predominance of crested wheatgrass in the diets (75%), which largely dictated the overall protein concentration. In contrast to nutritive value, tannin concentrations varied more noticeably across treatments: condensed tannins averaged 0.59 ± 0.15%, and hydrolysable tannins averaged 0.17 ± 0.09%. The greatest condensed tannin concentration was observed in CW-SA (1.02%), followed by CW-SA-SB (0.72%), CW-SA-SB-BT (0.68%), CW-SB (0.41%), and CW-AL (0.14%). Hydrolysable tannins were only detected in treatments containing small burnet, with the highest concentration in CW-SB (0.46%), followed by CW-SA-SB (0.23%) and CW-SA-SB-BT (0.15%). These results reflect the distinct tannin profiles of the forages used. Sainfoin is well-known for its high levels of condensed tannins, which are polymeric flavonoids synthesized and stored in vacuoles, contributing to plant defense by complexing with proteins and reducing pathogen access (Mueller-Harvey, 2006; Min et al., 2003). Small burnet, a forb adapted to dry, nutrient-poor environments, contained both condensed and hydrolysable tannins, which are esters of phenolic acids such as gallic and ellagic acid. Unlike condensed tannins, hydrolysable tannins are synthesized in the cytoplasm and tend to be more chemically reactive and susceptible to degradation under oxidative stress (Scalbert, 1991). From a plant physiology standpoint, both classes of tannins function in protecting plant tissues from herbivory, pathogens, and environmental stressors (Foley and Moore, 2005; Huang et al., 2018), but they differ in their chemical structure, site of biosynthesis, and mode of action (Naumann et al., 2017). The consistency in basal nutrient composition across diets in this study provides a controlled context in which tannin effects on rumen fermentation can be more clearly assessed.

Dry matter digestibility was higher for the CW-AL treatment compared with all the other treatments (P = 0.01; Table 3). In addition, NDF digestibility was affected by the treatments (P = 0.01). The highest NDF digestibility was observed in CW-AL (32.9%), which was significantly greater than all other treatments. CW-SA (25.2%) and CW-SA-SB-BT (25.0%) showed intermediate digestibility, while the lowest values were recorded for CW-SA-SB (23.0%) and CW-SB (22.1%). Interestingly, a previous study in dairy replacing alfalfa with sainfoin or birdsfoot trefoil reported no differences in dry matter intake or in OM and NDF digestibility among treatments (Grosse Brinkhaus et al., 2016), which contrasts with the treatment effects on digestibility observed in our experiment. To our knowledge, no previous studies have directly compared the digestibility of small burnet with alfalfa. In the current study, since fiber composition (NDF, ADF, hemicellulose) varied little among diets, the results observed for OM and NDF digestibility likely reflect the influence of tannins on microbial activity. Previous research suggests that the effects of tannins on feed intake, nutrient digestibility, rumen fermentation, and animal performance are more influenced by their dosage than by their source or type (Jayanegara et al., 2015). While all polyphenolics interact with both the cell wall and enzymatic secretions of rumen microbes (McSweeney et al., 2001), condensed tannins, typically of higher molecular weight, are known to bind dietary proteins, thereby reducing protein degradation by ruminal microbes, while hydrolysable tannins — typically of lower molecular weight — may affect rumen microbes more directly. There is evidence that hydrolyzable tannins decrease the methanogen population (Jayanegara et al., 2015), serve as toxins that penetrate archaeal cells, or directly interfere with fibrolytic activity (Aboagye and Beauchemin, 2019). In the current experiment, the CW-SA treatment contained only condensed tannins and resulted in reduced fiber digestibility compared with the CW-AL treatment. The CW-SB and CW-SA-SB treatments contained both condensed and hydrolysable tannins (Verma, 2025), and fiber digestibility was further reduced. Substituting small burnet with birdsfoot trefoil in the CW-SA-SB-BT, which reintroduced only condensed tannins, partially improved fiber digestibility.

Total VFA concentration differed among treatments (P< 0.01; Table 3). The CW-AL diet yielded the highest VFA concentration (120.9 mmol/day), significantly greater than all other treatments. Intermediate values were observed for CW-SA (114.5 mmol/day) and CW-SA-SB-BT (113.5 mmol/day), which did not differ from each other. The lowest VFA production occurred with the CW-SB (102.2 mmol/day) and CW-SA-SB (101.0 mmol/day) treatments. Acetate concentration was affected by treatment (P< 0.01). CW-AL resulted in the greatest acetate concentration (80.6 mmol), followed by CW-SA (72.9 mmol) and CW-SA-SB-BT (72.4 mmol), which did not differ from each other. CW-SB (63.4 mmol) and CW-SA-SB (64.6 mmol) produced the lowest acetate concentrations and did not differ significantly from one another. Results for both DM and NDF digestibility, as well as acetate, indicate that the presence of hydrolysable tannins is more influential than the total tannin concentration. Other VFAs, including propionate, butyrate, valerate, isobutyrate, and isovalerate, were not affected by treatments (P > 0.05). The alignment between NDF digestibility and acetate production reinforces the relationship between fiber degradation and acetate formation, as acetate is the primary VFAs generated from the fermentation of fibrous material in the rumen. Once produced, acetate is rapidly absorbed across the rumen wall into the bloodstream, serving as a major energy source for ruminants (Bergman, 1990). Acetate is primarily used in peripheral tissues for oxidative metabolism and lipogenesis. Therefore, treatments that enhance acetate production through improved fiber digestion, such as CW-AL and CW-SA-SB-BT, may support better energy status and productive performance.

The use of different plant combinations in forage mixtures affected ruminal nitrogen metabolism in dual-flow continuous culture fermenters (Table 4). Ammonia-N concentration was highest for CW-AL (28.2 mg/dL) and CW-SA (27.4 mg/dL), intermediate for CW-SA-SB-BT (22.5 mg/dL) and CW-SA-SB (20.8 mg/dL), and lowest for CW-SB (19.3 mg/dL) (P = 0.01). Despite these significant effects on ruminal ammonia levels, there were no significant effects on total bacterial nitrogen flow (P = 0.96), bacterial nitrogen yield per unit of NDF digested (P = 0.98), non-ammonia-N (P = 0.97), or non-ammonia non-bacterial nitrogen flow (P = 0.99). Ammonia-N in the rumen is primarily formed through the microbial deamination of dietary protein and non-protein nitrogen sources. Once released, ammonia serves as a key substrate for microbial protein synthesis, which is the primary source of metabolizable protein for the host animal (Owens and Bergen, 1983; Tan et al., 2021). However, when the rate of protein deamination exceeds the nitrogen requirements of the rumen microbial population, due to the presence of rapidly degradable protein or insufficient fermentable energy, excess ammonia accumulates in the rumen (Calsamiglia et al., 2010). This surplus is absorbed through the rumen wall into the bloodstream and transported to the liver, where it is rapidly converted to urea to prevent toxicity. Although this conversion is necessary for animal health, it is energetically expensive, requiring ATP, and contributes to nitrogen inefficiency and increased nitrogen excretion in the urine (Calsamiglia et al., 2010). Therefore, the higher ammonia-N concentrations observed in the CW-AL and CW-SA treatments may reflect an oversupply of degradable nitrogen relative to microbial requirements, leading to excess ammonia accumulation and reduced nitrogen use efficiency. This pattern may also be explained by the presence of tannins in certain treatments, as tannins are known to bind to proteins and form complexes that are less susceptible to microbial degradation in the rumen (Mueller-Harvey et al., 2019). By reducing the rate of proteolysis, tannins effectively limit the rapid release of amino acids and subsequent deamination, thereby decreasing the production of ammonia. This moderated release of nitrogen can lead to a better alignment between nitrogen availability and microbial energy supply, thereby improving nitrogen utilization efficiency (Aguerre et al., 2016). From a practical perspective, such improvements in ruminal nitrogen efficiency could enhance the supply of metabolizable protein to the animal, supporting animal growth. At the same time, reducing excess ammonia formation and subsequent urinary nitrogen excretion has direct environmental benefits by mitigating nitrogen losses from manure. Thus, the lower ammonia-N concentrations observed in some of the tannin-containing treatments (i.e., those containing birdsfoot trefoil and small burnet) may reflect this protective effect, highlighting the role of plant secondary compounds in modulating rumen nitrogen dynamics but also suggesting potential benefits for both animal performance and environmental sustainability. Interestingly, Rufino-Moya et al. (2019) previously reported that sainfoin decreased ruminal ammonia compared with alfalfa, whereas in the present study, the treatments containing these forages showed similar responses. Together, these findings demonstrate that the type of secondary compounds present in forage mixtures, particularly tannins, plays a key role in regulating ruminal nitrogen metabolism. They also suggest that the specific combinations of forage species further influence these responses. Strategically incorporating tannin-containing plants into forage mixtures may therefore offer a valuable approach to improving nitrogen efficiency in forage-based systems.

The rumen microbial composition was evaluated by enumerating protozoa and analyzing the bacterial community using 16S rRNA gene sequencing. Protozoa numbers did not differ among treatments (P = 0.43; Table 4). The effects of dietary treatment on the bacterial community were evaluated using alpha diversity, beta diversity, and taxonomic composition analyses. To better exemplify these concepts, imagine each rumen microbiome as a basket of fruits. Alpha diversity refers to the number of different fruit types in each basket (richness) and how equally the fruits are distributed (evenness). For example, a basket with three apples, three bananas, and three oranges would have high richness and high evenness. In contrast, a basket with eight apples and one orange has the same richness (three fruit types if a third type is included) but lower evenness. In our study, alpha diversity —assessed using the Shannon and Chao1 indices — showed no significant differences among treatments (P ≥ 0.49; Figure 1). This suggests that each microbial community had a similar number of microbial species (like fruit types) and a similar distribution of those species (like the number of each fruit), regardless of the dietary treatment. In other words, all the treatments had comparable richness and evenness. This stability is consistent with previous findings that the rumen microbiota maintains core diversity even when diet composition changes, likely due to functional redundancy and ecosystem resilience (Henderson et al., 2016). In contrast, beta diversity compares the types of fruits in different baskets — not just how many types or how evenly they are distributed, but whether the actual kinds of fruits differ. For example, one basket might have apples, bananas, and oranges, while another has grapes, mangos, and kiwis. In our study, beta diversity analysis using Bray-Curtis dissimilarity revealed a tendency for microbial communities to cluster by treatment (P = 0.09; Figure 2). Specifically, the CW-SB group tended to differ in beta diversity from all other treatments. This suggests that while overall microbial diversity (i.e., the number and distribution of species) remained relatively similar, the specific microbial taxa present tended to vary with the forage mixture. These shifts indicate that dietary treatments influenced the composition of the rumen microbiome, even though overall diversity was not significantly altered.

The relative abundance of key prokaryotic genera was significantly influenced by dietary treatments (Figure 3). Among the twenty most abundant genera, four showed statistically significant variation across treatments (Prevotella, Succiniclasticum, Prevotellaceae_UCG-001, and NK4A214_group). These changes suggest that differences in forage composition and plant secondary compounds may modulate specific microbial groups within the rumen and, consequently, the type and amount of VFAs produced as well as N metabolism in the rumen. Prevotella was most abundant in the CW-AL, CW-SA-SB, and CW-SA-SB-BT treatments, with CW-SB showing the lowest levels (P ≤ 0.05). Prevotella is known for its role in degrading protein and non-cellulosic polysaccharides, particularly in diets rich in legumes. Its fermentation end products include acetate, propionate, and succinate, the latter often serving as a substrate for other microbes such as Succiniclasticum. The high abundance of Prevotella in the CW-AL treatment is consistent with its dominance in alfalfa-based diets, while its persistence in the more diverse CW-SA-SB-BT mixture suggests that certain combinations of legumes or tannins may create favorable niches for this genus. For instance, some combinations of HT and CT have been found to create synergistic effects in modulating fermentation quality and methanogenesis relative to when just one type of tannin is provided (Chen et al., 2021). The combination of hydrolizable and condensed tannins alleviated the inhibitory impacts of hydrolyzable tannins on Butyrivibrio, Ruminobacter, and Prevotella (Chen et al., 2021). The reduced abundance in CW-SB may indicate that small burnet alone provides fewer or less accessible fermentable substrates for Prevotella or that it contains bioactive compounds that selectively suppress this genus. In support of this, previous studies have found a significant reduction in Prevotella rumen populations after supplementation with hydrolyzable tannins, including chestnut tannins (Díaz Carrasco et al., 2017). The relative abundance of Succiniclasticum was highest in CW-AL and declined in more complex mixtures, with the lowest levels in CW-SA-SB-BT (P ≤ 0.04). This genus specializes in converting succinate to propionate a key short-chain fatty acid that contributes to gluconeogenesis in ruminants (Van Gylswyk, 1995). Although the propionate flow was not affected by the treatments, the higher abundance of Succiniclasticum in CW-AL may be linked to more consistent carbohydrate fermentation patterns, whereas the reduction in more diverse treatments could reflect shifts in fermentation pathways, possibly due to the presence of secondary metabolites that alter succinate production or utilization. Prevotellaceae_UCG-001 showed a similar trend, with the highest abundances in CW-AL, CW-SA, and CW-SA-SB-BT, followed by CW-SA-SB and the lowest levels in CW-SB (P ≤ 0.04). As a member of the Prevotellaceae family, this group likely contributes to acetate, propionate, and succinate production through starch and peptide fermentation. Its variable response across treatments further supports the idea that individual forages influence the rumen environment in ways that can promote or inhibit specific microbial populations. Although Prevotella species are not primarily proteolytic, some members of this genus participate in protein and peptide degradation, which may help explain the higher ammonia-N concentrations observed in the CW-AL and CW-SA treatments. In contrast to Prevotellaceae_UCG-001, NK4A214_group was most abundant in CW-SB, with decreasing abundance in CW-SA-SB, CW-SA-SB-BT, CW-AL, and CW-SA. Members of this group are associated with the degradation of structural carbohydrates, producing primarily acetate and butyrate. Its enrichment in CW-SB may reflect adaptation to the phytochemical profile of small burnet, possibly indicating a microbial shift toward pathways less affected by condensed tannins or other bioactives present in that forage. Taken together, these results demonstrate that forage composition not only alters the relative abundance of specific microbial taxa but also shifts the VFA production profile of the rumen. Shifts in the abundance of key genera such as Prevotella, Succiniclasticum, and NK4A214_group may influence rumen fermentation profiles, nutrient utilization, and ultimately, animal productivity. These findings underscore the potential of using forage diversity and strategic inclusion of plant secondary compounds to modulate rumen microbial ecology. However, since this study relied on 16S rRNA gene sequencing, the analyses are limited to taxonomic resolution and do not provide direct insights into microbial functional activity. Further studies employing metatranscriptomics or other approaches are needed to elucidate the metabolic functions and regulatory pathways involved.

Overall, our findings partially supported the original hypothesis. We proposed that incorporating diverse functional plant groups, particularly tannin-containing legumes and forbs, into crested wheatgrass-based forage mixtures would enhance rumen fermentation, improve fiber degradation, and increase microbial protein synthesis through shifts in the rumen microbiome. While forage diversity and tannin profiles did influence fermentation patterns, ammonia-N dynamics, and the abundance of specific bacterial taxa, these changes did not lead to improvements in fiber digestibility or microbial nitrogen flow beyond the alfalfa-based control. Instead, the CW-AL mixture consistently supported the highest digestibility and VFA production, whereas mixtures containing small burnet, rich in hydrolysable tannins, reduced fiber degradation and altered microbial structure. These results show that although plant diversity does affect rumen function, the benefits depend more on the specific combinations and chemical characteristics of the forages than on diversity alone. Importantly, this information will help refine the design of “forage islands” within the Smart Foodscapes framework, ensuring that plant species and mixtures are selected not only for agronomic resilience but also for their targeted effects on rumen fermentation and nitrogen efficiency.

Including 75% crested wheatgrass with 25% alfalfa resulted in the highest dry matter digestibility, neutral detergent fiber digestibility, total VFA production, and acetate production, making it the most effective treatment for enhancing rumen fermentation under the conditions of this study. The inclusion of diverse forage species significantly influenced rumen fermentation dynamics and microbial ecology. Variations in condensed and hydrolysable tannin concentrations across treatments played a key role in modulating fiber digestibility, ammonia production, and microbial community composition. Mixtures containing predominantly condensed tannins (e.g., sainfoin and birdsfoot trefoil) supported intermediate fiber digestibility and maintained microbial protein synthesis and acetate production with reductions in ruminal ammonia concentrations. In contrast, combinations containing predominantly hydrolyzable tannins (e.g., small burnet) likely impaired fiber degradation. Future research should investigate the long-term effects of these forage mixtures under in vivo conditions, with particular attention to their impact on animal performance, methane emissions, and overall system sustainability.