Knowledge Gaps in the Nutrient Requirements of Beef Cattle

Simple Summary

In light of research and practical applications of the National Academies of Science, Education and Medicine (NASEM) nutrient requirement recommendations for beef cattle, we reviewed published literature and identified knowledge gaps and research needs related to beef cattle nutrient requirements. Key areas identified for energy requirements include the effects of environmental temperature and grazing activity on the maintenance requirements of beef cattle, as well as the need for an improved understanding of the relationship between retained energy and protein. More accurate predictions of dry matter intake, particularly for beef cows, and microbial protein synthesis for all classes of beef cattle are needed. Continued research on trace mineral requirements, notably zinc and chromium, is needed, and updated information on composition of feedstuffs, especially byproduct feeds, is an important practical issue. Nutritional models for beef cattle need to evolve to a mechanistic approach that optimizes the use of artificial intelligence and precision livestock farming technologies.

Abstract

The 8th revised edition of the Nutrient Requirements of Beef Cattle was released in 2016, with the recommendations provided in the publication being used extensively in both research and production settings. In the context of research needs identified in that publication, our objective was to review research on beef cattle nutrient requirements published since 2016 and identify knowledge gaps that should be addressed. Relative to energy requirements, the effects of environmental temperature and grazing activity, along with stress and disease, on maintenance requirements are inadequately characterized or defined. In addition, relationships between retained energy and protein should be more fully elucidated, and additional guidance on body weight at a target compositional endpoint is needed. Areas of continuing concern include accurately and precisely predicting microbial protein supply, predicting N recycling, and the metabolizable protein requirements for maintenance. Mineral and vitamin requirements are often challenging because of a lack of consistency in models used to determine requirements and potential effects of unique production settings on requirements. Based on recent research with feedlot cattle, zinc and chromium requirements should be examined more closely. Because predictions of dry matter intake are critical to supplying nutrients, additional development of prediction equations is needed, especially for beef cows and grazing beef cattle in general. Given considerable research in prediction of greenhouse gases, reevaluation of 2016 recommendations is warranted, along with a need for the updating of equations to predict excretions of N and P. Composition of feeds, particularly byproducts from ethanol production or other industrial streams, represents a knowledge gap, with obtaining reliable energy values of these feeds being a notable challenge. Nutritional models provide the means to integrate nutrient requirement recommendations into practice, and moving towards mechanistic models that take advantage of artificial intelligence and precision livestock farming technologies will be critical to developing future modeling systems.

1. Introduction

Global livestock farming was estimated to generate an output value of approximately USD $1.72 trillion in 2018 [1], with cattle accounting for 34% of the total. Regarding asset value, cattle represent 74% of the total value and approximately 69% of farmed animal liveweight mass [1]. Optimizing inputs, while concurrently minimizing the environmental impact, is critical to sustainable cattle production, with feed inputs consistently recognized as the largest cost of beef production [2]. Systems that accurately define nutrient requirements of cattle are critical to efficient use of feed resources and diminished environmental impacts of cattle production. In the United States, nutrient requirements of beef cattle are recommended by committees that are appointed by and operate under the guidance of the National Academies of Sciences, Engineering, and Medicine (NASEM). The most recent revision of the nutrient requirements for beef cattle [3] was published almost a decade ago.

As former members of the most recent NASEM Committee on Nutrient Requirements of Beef Cattle, our objective for this paper is to identify knowledge gaps in the nutrient requirements of beef cattle. We offer these suggestions on our own, not with the authority or approval of the NASEM, but as scientists who have dedicated their careers to issues related to beef cattle nutrient requirements. We hope that this effort will stimulate conversations and ultimately peer-reviewed research findings that will aid the work of future NASEM nutrient requirement committees, as well as provide funding agencies with information on which to base research priorities.

2. Knowledge Gaps

2.1. General

The Committee on Nutrient Requirements of Beef Cattle summarized research needs (Chapter 21) in the NASEM [3] publication. Several of the issues raised in this summary have been addressed, but many remain to be fully elucidated. Our current objective in identifying knowledge gaps was not to reiterate the items mentioned in the 2016 summary, but rather to emphasize knowledge gaps based on the application of nutrient requirement recommendations and models in practice, with a focus on what the future might hold. Nonetheless, when relevant, the previously identified research needs will be highlighted as they relate to knowledge gaps identified by the authors in this current effort. Moreover, knowledge gaps will generally be delineated in the context of the various chapters of the NASEM [3] publication.

2.2. Energy, Carbohydrates, and Lipids

2.2.1. Calculation of Energy Units

In the NASEM publication [3], metabolizable energy (ME) is calculated as digestible energy (DE) × 0.82, but the committee noted that the efficiency of conversion of DE to ME was likely not constant and needed additional definition. Galyean et al. [4] suggested a linear equation (ME = DE × 0.9611 − 0.2999) to calculate ME from DE. Subsequently, additional studies were added to the database and the ME/DE relationship was reevaluated [5], with a recommendation that ME be calculated as DE − 0.39. Across a wide range of diets in terms of DE and ME concentrations, the latter equation [5] seems to agree well with observations in the literature, particularly those that relate to the conversion of DE to ME in high-concentrate, feedlot-type diets, which have an efficiency of conversion much greater than 0.82. Likewise, the equation [5] seems to be equally effective when applied to diets with lower DE concentrations such as those often consumed by grazing beef cattle.

On a related issue, the NASEM [3] publication assumes that 1 kg of total digestible nutrients (TDN) is equal to 4.4 Mcal of DE. This conversion value was originally derived from the relationship between ME and TDN, with the 4.4 Mcal/kg calculated using an assumed conversion efficiency of DE to ME of 0.82. Applying the DE − 0.39 conversion of DE to ME [5] could yield a different relationship between TDN and DE, resulting in the need to reevaluate this relationship.

Calculating DE from ME by these newer methods also raises questions about the subsequent use of ME to calculate dietary net energy values (NEm and NEg). The cubic equations recommended by NASEM [3] to calculate NEm and NEg from ME were largely derived from an assumed conversion of DE to ME of 0.82, leading to the suggestion that these equations would need to be adjusted for newly derived DE:ME relationships [4]. Although the cubic equations were recalculated [4] to force agreement between published NEm and NEg values with the ME values calculated from the linear relationships reported previously [4,5], further research is needed to verify that these adjusted equations are valid when applied to independently derived data.

2.2.2. Indigestible Neutral Detergent Fiber and Calculation of Total Digestible Nutrients

Both empirical (ELS) and mechanistic (MLS) levels of solution for nutrient requirements were provided in the Beef Cattle Nutrient Requirements Model (BCNRM; version 1.0.37.22; https://nutritionmodels.com/beef.html) by the NASEM [3] committee. Estimates of the protein and carbohydrate fractions in feedstuffs are important to the application of these models, particularly the MLS. As noted in the research needs suggested by NASEM [3], however, data related to the composition of these fractions are limited. One example of this concern is the use of indigestible neutral detergent fiber (iNDF) to calculate TDN in the MLS. The iNDF fraction (defined as the CC fraction in the MLS) is assumed to be a constant value that is estimated as 2.4 × the lignin fraction. In addition, it is not possible to override this value with user input from a laboratory measurement of indigestible NDF (e.g., iNDF values from laboratories based on extended in vitro or in situ fermentations). Because it affects the calculated TDN value of the feed, improved methods to predict the CC value are needed, as well as evaluation of the potential to input a CC value by the user. Relative to improved predictions, Raffrenato et al. [6] analyzed >200 forage species from Australia and South Africa for NDF, acid detergent fiber, lignin (ADL), and iNDF. Ratios of iNDF to ADL ranged from 1.6 to 8.0, suggesting that a single constant value of 2.4 × lignin does not likely reflect the biological relationship. Predictions of iNDF based on the relationship between ADL/NDF and the iNDF/lignin ratio resulted in equations with R² values between 0.82 and 0.95, indicating superior predictions are possible, particularly when individual forage species and environmental conditions are considered.

2.2.3. Effective Fiber and Ruminal pH

Another fiber-related concern in the MLS is the significance of physically effective NDF (peNDF) and associated recommendations for high-starch, feedlot diets. The peNDF concept does not work well for feedlot diets that are mainly grain because the grain, byproduct ingredients, and pelleted supplements are retained on the sieves and thereby used to calculate peNDF content, even though they do not contribute to promoting rumination or “fibrousness.” Moreover, models like the Cornell Net Carbohydrate and Protein System (CNCPS; https://cals.cornell.edu/animal-science/outreach-extension/publications-resources-software/cncps) and Ruminant Nutrition System (RNS; https://nutritionmodels.com/rns.html) acknowledge that peNDF alone is insufficient to predict ruminal pH or microbial efficiency because other dietary and animal factors, including carbohydrate source, intake level, and digestion rates, also modulate fermentation and acid production. Thus, there is a need for the MLS to refine fiber effectiveness estimations, especially for non-forage byproducts, and to link fiber characteristics more directly with fermentation dynamics and pH depression risk. Prediction of ruminal pH needs to be improved in the MLS, particularly in terms of how various types of carbohydrates affect mean ruminal pH and the time that ruminal pH is less than a given benchmark. In addition to potentially improving degradation rate and microbial protein supply estimates in the MLS, more accurate estimates of ruminal pH would help beef cattle producers make management decisions (e.g., choice of ingredients, grain processing, roughage concentrations) related to feeding high-grain diets.

2.3. Protein

2.3.1. Microbial Protein Synthesis

The supply of microbial crude protein (MCP) is vital to predicting the total metabolizable protein (MP) supply and affects calculations of MP required for maintenance and gain [7]. Moreover, in terms of animal production, estimates of the MCP supply affect decisions about supplementation of protein, one of the largest out-of-pocket costs to beef producers. Prediction of MCP synthesis in the NASEM [3] publication is accomplished using regression equations based on either TDN or fat-free TDN [8]. After adding an additional 50 observations to the original database of 285 observations, Galyean and Tedeschi [9] reevaluated their previous equations and concluded that other measures of energy intake besides TDN were effective in predicting MCP and that adding crude protein (CP) to equations improved model precision. In addition, fat-free TDN was not selected as a dependent variable to predict MCP in their initial stepwise-regression analyses, suggesting that adjustment for the ether extract content of the diet recommended by NASEM [3] was not necessary. More recently, Galyean and Tedeschi [10] compared the TDN equation of NASEM [3] with two additional equations they developed [9] and the BR-CORTE [11] equation. They [10] also included a simple approach of estimating MCP as 10% of the TDN intake (10% being the mean value of MCP relative to TDN intake in their database). Regression fit statistics indicated that all equations yielded similar accuracy and precision, including the simple 10% calculation. The root mean square error of prediction across equations averaged 28.7% of the mean MCP concentration, implying the need for improved precision. Thus, although research since the release of the NASEM [3] publication has yielded additional options for prediction equations, these new options might not significantly improve the accuracy and precision of MCP predictions, and further research is needed in this area.

Challenges with MCP prediction, including the limited data for building databases and the inherent variability in estimates of MCP, reflect the indirect and often imprecise methods used for generating MCP estimates [10]. Moving forward, improved digesta sampling methods, prediction of digesta flow, and microbial marker techniques need to be a focus of research. It has been suggested [12] that metagenomics, metaproteomics, metataxonomics, and metatranscriptomics methodologies might allow us to better define and predict MCP synthesis in ruminants. Moreover, adding key predictors of microbial growth, like ruminal pH dynamics and fermentation profiles, along with microbial community composition, might enhance model development. At present, such data are limited and infrequently combined with estimates of MCP synthesis, leading to the suggestion [10] that effectively using these tools will require coordinated and intense research efforts at research locations around the world. Whether associated funding commitments for such research would be made by government, industry, and private foundations remains to be determined.

2.3.2. Metabolizable Protein Requirements

The need for additional data regarding the validity of the MP requirements for maintenance was noted in the research needs of NASEM [3], which is consistent with the concerns expressed [7] about the relationship between predicted MCP flow and its connection to the calculation of MP requirements for maintenance and gain. The NASEM publication [3] defined the MP requirement for maintenance as 3.8 g of MP/kg of shrunk body weight (BW)^0.75, whereas the efficiency of MP use for growth was defined as 49.2% for cattle with a shrunk BW greater than 300 kg. Using Holstein x Gyr crossbred bulls [13], MP for maintenance was reported to be 3.05 g/kg of BW^0.75, with an efficiency of conversion of MP to net protein in gain of 35.7%. More recently, in a meta-analysis involving 385 treatment means with beef cattle [14], a MP requirement for maintenance of 4.31 g/kg of BW^0.75, with an efficiency of converting MP to retained protein of 32%, was reported. Given the inconsistency of data relating to MP requirements for maintenance and conversion of MP to retained protein, additional research is needed.

2.3.3. Nitrogen Recycling

Nitrogen recycling is not a component of the ruminally degraded protein (RDP) supply in the ELS and MLS models [3], but the committee summarized the meta-analysis of urea kinetics studies in growing cattle [15] to provide a framework for predicting urea N recycling and its anabolic utilization. Results indicated that approximately 75% of N intake was converted to urea, and an average of 41% of N intake was recycled to the gastrointestinal tract. Nonetheless, the fraction of the urea entry rate that was recycled, as well as the recycled N used for anabolic purposes, decreased as N intake or dietary CP increased, reflecting diminished reliance on recycled N when dietary N is sufficient. Equations were developed [15] to predict the proportion of recycled urea N used for anabolic purposes, which were included in the NASEM [3] discussion of N recycling but not included in the ELS or MLS. In a recent review [16], it was suggested that regulation of N recycling could be an important component of efforts to increase the efficiency of ruminal N utilization, but for effective regulation strategies to be implemented, accurate predictions of recycled N across a wide range of beef cattle diets are needed. Based on compiling 107 treatment means from the literature (beef and dairy cattle and sheep), it was concluded that the gut entry rate of urea relative to urea entry into the blood (i.e., N recycled) reaches a plateau at a dietary CP concentration of 17.6% [17]. In a revised representation of the Molly cow model [18], urea entry into the blood was about 64% of intake N, of which 64% was recycled to the gut (i.e., 41% of N intake is recycled). In cattle fed a protein-restricted diet [19], N recycling to the gut supplied from 50% to 200% as much N available for microbial growth as the diet. Several regression equations were provided to predict recycled N from dietary CP concentration and the ratio of digestible organic matter intake to CP [19].

We conclude that the prediction of recycled N continues to be a knowledge gap in beef cattle nutrient requirements. Previous work [15] suggests that more mechanistic representations of N fluxes, incorporating urea kinetics and dynamic microbial use, could improve current models, especially under variable protein intake, but additional data are needed for high-concentrate diets and data under varied feeding and physiological conditions. Until such datasets and modeling approaches are integrated, the omission of N recycling in current models will remain a significant limitation to accurately predicting MCP synthesis and overall N efficiency in beef cattle.

2.3.4. Ruminally Degraded and Undegraded Protein Values of Feedstuffs

Effective application of the MP system depends on the ability to accurately define the RDP and ruminally undegraded protein (RUP) concentrations in feedstuffs, as well as estimates of degradation rates and intestinal digestibility of protein and amino acids in the MLS model [3]. The RUP and RDP values are typically predicted from protein solubility estimates combined with degradation and passage rate estimates for the insoluble fraction [20]. Using dairy cattle data from the literature, Hanigan et al. [20] concluded that protein degradation rates derived from in situ techniques and marker-based passage rates for protein could not be used to accurately predict RUP outflow from the rumen. These authors further suggested adjustment factors for degradation rates and feed class adjustments that improved agreement between model-predicted and observed RUP flow values. Using ruminally cannulated beef cattle, in situ degradation rates of grains, protein sources, and various roughages were compared [21]. The RDP and RUP values of the feedstuffs were calculated using an assumed passage rate, and the intestinal digestibility of RUP was estimated from in vitro pepsin–pancreatin digestion. The RUP values for protein ingredients were 37% less on average than values from NASEM [3], but grain and roughage RUP estimates did not differ from the NASEM values, which is consistent with conclusions regarding in situ values [20]. Reported intestinal digestion for protein and roughage, but not for grain [21], differed from values of NASEM [3]. The authors further reported that models based on NDF and CP concentrations of the feed ingredients were effective for predicting RDP, whereas ADF, NDF, and CP were found useful for predicting intestinal digestion of RUP. Thus, knowledge gaps exist relative to the RDP and RUP estimates for common feeds used in beef cattle diets. In addition, assumed values in NASEM [3] for intestinal digestibility of RUP in various classes of feeds (60% for forages and 80% for other feedstuffs) should be reevaluated.

2.4. Minerals and Vitamins

In terms of field applications of the NASEM [3] recommendations, fortification of minerals and vitamins in practical diets often shows considerable variance with recommendations. This is particularly true for trace minerals and vitamins. For example, a survey [22] reported that dietary concentrations of trace minerals recommended by consulting feedlot nutritionists, except for iron, were 1.5 to 5.5 times the NASEM [3] recommended concentrations. Some research studies have shown limited benefits in terms of cattle growth performance and carcass characteristics to trace mineral supplementation beyond NASEM recommendations [23,24], but others [25] reported increased ADG and hot carcass weight when finishing beef steers were provided Cu, Zn, Mn, Se, Co, and I (all from inorganic sources) at concentrations recommended by feedlot nutritionists vs. NASEM [3] recommendations or no supplemental trace minerals. Presumably, nutritionists discount trace minerals supplied by dietary ingredients, choosing to add NASEM-recommended concentrations (or greater) on top of basal concentrations. This practice seems to reflect concerns about the availability of trace minerals in the basal diet, although measurements of mineral solubility using techniques that mimic ruminal and abomasal conditions suggest that minerals in basal ingredients should be highly available [26].

Another concern that might explain why nutritionists often supply trace minerals in practical diets at concentrations that exceed NASEM [3] recommendations relates to the systems used to determine mineral requirements. The factorial method was used by the NRC [27] and NASEM [3] committees for determining Ca and P requirements. This method sums up the needs for bodily functions (i.e., maintenance, growth, pregnancy, and lactation), which are then divided by the absorption coefficient for the mineral to determine dietary requirements. A recent example [28] used the retention and absorption coefficient approach to determine the major and trace mineral requirements of Nellore cattle in Brazil. Although the method works well for Ca and P, where reliable absorption coefficient estimates are available, retention data and absorption coefficients are much less available for trace minerals [29]. Thus, for trace minerals, response data (e.g., growth performance or metabolic responses) often form the basis for recommended dietary concentrations. Moreover, interactions among minerals often limit the inference space for a given study, requiring different experimental conditions to identify consistent recommendations for fortification [29]. Accordingly, developing more accurate systems for determining mineral requirements in beef cattle would prove helpful to future NASEM committees.

2.4.1. Zinc and Growth-Enhancement Technologies

Whether Zn requirements are affected by growth-enhancement technologies (e.g., implants and beta-agonists) has been a notable subject of research since the NASEM [3] publication was released. Feeding crossbred steers 1 g/d of Zn from Zn propionate with or without 3 mg/d of Cr from Cr propionate did not affect growth performance, hot carcass weight, or carcass quality when the beta-agonist ractopamine was included in the diet [30]. Carmichael et al. [31] compared no supplemental Zn (32 mg/kg of dietary DM) or supra-nutritional Zn supplementation (145 mg/kg of dietary DM) in Angus crossbred steers with or without ractopamine hydrochloride for the final 30 d before harvest. Zinc supplementation did not interact with the beta-agonist, but both supra-nutritional Zn and ractopamine increased N retention as a percentage of intake. In further research by the same group [32], Zn was supplemented at 30 or 100 mg/kg of the dietary DM, which was factored with either a single extended-release implant or a re-implant strategy that provided twice the estradiol and trenbolone acetate as the extended-release implant. Although some differences were noted during various segments of the experiment, no Zn concentration x implant strategy interactions were observed, and neither factor affected overall ADG, DMI, or hot carcass weight. Supplemental Zn concentrations of 0, 100, 150, or 180 mg/kg of the dietary DM supplied by ZnSO₄ did not affect final BW, dressing percent, 12th rib fat, or marbling score, but hot carcass weight tended to be greater for Zn supplementation vs. no supplemental Zn [33]. In Angus crossbred steers [34], dietary Zn supplementation (from ZnSO₄; 0, 30, or 100 mg/kg of dietary DM) was factored with implant treatments (no implant or implanted with 20 mg of estradiol plus 200 mg of trenbolone acetate). Neither final BW nor ADG was affected by Zn concentration, and there was no Zn x implant interaction. Nonetheless, metabolite and mineral concentration data taken throughout the study led the authors to conclude that demand for Zn is increased immediately after the administration of a steroidal implant [34]. Overall, the experimental evidence for an increase in the NASEM [3] recommendation for dietary Zn in growing and finishing cattle is ambiguous. Supplying Zn at concentrations above the recommendation does not seem harmful to animal health and performance, and sometimes yields improvements in these variables, but it also increases fecal Zn excretion, suggesting further research is warranted. Although research on trace mineral supplementation of grazing beef cattle, particularly replacement heifers and beef cows, is more limited than in feedlot cattle, the potential for oversupplying trace minerals like Zn exists in those production settings as well, necessitating the need for additional research.

2.4.2. Chromium

The NASEM [3] publication discussed the effects of Cr on beef cattle health and growth performance, but the committee did not make formal recommendations regarding dietary supplementation of Cr. The only permitted source of supplemental Cr for cattle in the U.S. is Cr propionate [35]. Considerable research with growing and finishing beef cattle fed supplemental Cr was conducted in the last decade, with a typical target dose of Cr of 3 mg/animal daily. Effects of supplemental Cr in growing and finishing cattle are mixed, with several studies showing limited effects on growth performance and carcass measurements [36,37,38,39], whereas others have noted increased ADG, improved gain:feed ratio, and increased hot carcass weights [40,41]. Comparisons often include other factors (e.g., trace mineral sources and concentrations, or the inclusion of other supplements (e.g., yeast or other additives), which makes it challenging to get a clear picture of the effect of Cr alone. Metabolically, supplemental Cr can affect glucose concentrations and clearance rates and the ratio of glucose to insulin, at least at certain times during a typical feeding period [36,41,42]. Additional research on the feeding of Cr alone that better defines potential interactions with growth-enhancing technologies and metabolic effects of Cr in growing and finishing beef cattle would help fill knowledge gaps in this area. A recent review [35] addresses the use of Cr in all major livestock species.

2.4.3. Vitamins

The role of vitamin A in the development of intramuscular adipose tissue (marbling) has been the subject of several studies in the past few years. Given that retinol can inhibit adipogenesis, a focus of this research has been to determine whether limiting supplemental vitamin A in finishing beef cattle might positively affect marbling and thereby improve the quality grade. For example, feeding a diet with no supplemental vitamin A for 10 mo to Angus steers resulted in a 46% increase in the percentage of intramuscular fat, but marbling score only tended to differ between treatments [43]. Providing no supplemental vitamin A to Wagyu crossbred steers [44] that were fed for 320 d increased the percentage of intramuscular fat by more than 1.8 times compared with steers supplemented with vitamin A, with effects attributed to altered energy metabolism of muscle. Wellman et al. [45] evaluated supplemental vitamin A concentrations of 0, 2200, or 11,000 IU/kg of dietary DM in finishing beef steers following depletion. Vitamin A concentration did not affect marbling or other carcass characteristics, but DMI was decreased in cattle fed the 0 IU/kg diet. The authors concluded that the NASEM [3] recommendations for finishing beef cattle were adequate to replete vitamin A stores in the liver. Results of a meta-analysis based on seven experiments [46] indicated that neither intramuscular fat percent nor score was significantly affected by vitamin A, but the overall effect of additional vitamin A was slightly negative. Although the results of experiments with no vitamin A supplementation help clarify effects of vitamin A on marbling, they do not represent a practical approach to altering intramuscular fat deposition because of the potential negative effects of vitamin A-deficient diets on cattle health and performance.

Other research studies on vitamin A and marbling have evaluated how early life exposure to high levels of supplemental vitamin A affects intramuscular adipose tissue. Angus steer calves were given 0, 150,000, or 300,000 IU of vitamin A at birth and again at 1 mo of age [47]. Weaning weight was increased by vitamin A supplementation, and the marbling score of the carcass responded quadratically, with an increase for the 150,000 IU treatment. Subsequent work [48] showed that 150,000 IU given at birth and 1 mo of age activated satellite cells in muscle and shifted muscle fibers to oxidative types. Further, it was reported [49] that a single dose of 300,000 IU of vitamin A given at birth increased intramuscular fat in longissimus muscle, and the effect was associated with increased expression of genes involved in angiogenesis, adipogenesis, and lipogenesis. Administering vitamin A at or near the time of birth could have other production benefits, as serum concentrations of vitamin A and E in beef calves less than 1 mo of age were associated with health outcomes [50]. Calves with low serum vitamin A and E concentrations had a greater risk of death and enteritis, respectively.

Further studies are needed to clarify the ambiguous response of intramuscular adipose tissue to vitamin A supplementation depending on stage of growth and dose and duration of supplementation. More generally, additional research with both fat soluble and B vitamins should prove useful to more clearly define the relationship between neonatal status and supplementation and subsequent health, growth performance, and carcass merit.

2.4.4. Effects of Stress

In the chapter on the effects of stress on beef cattle nutrient requirements, the NASEM [3] publication identified several trace minerals (Cr, Cu, Se, and Zn) and vitamin E as key nutrients for which additional dietary fortification (i.e., beyond compensation for decreased DMI) might be beneficial before or after post-weaning periods of transportation and marketing stress. Considerable research was conducted in this area over the past decade, and recent reviews [51,52] summarize this work. Suggested research needs [52] include additional studies to identify how mineral and vitamin status before stress alters the need for, or benefits from, supplementation, as well as how supplementation, particularly of trace minerals, affects the response to respiratory vaccines.

2.5. Feed Intake

The case can be made that feed intake is the most important component of the supply side of nutrient requirement systems, which makes accurate predictions of feed intake critical to the application of the NASEM [3] model. Intake by growing and finishing cattle, particularly feedlot cattle, was an emphasis of data analyses by the NASEM [3] committee, resulting in further evaluation of the NRC [27] feed intake equations and the development of new equations. For growing and finishing cattle, the committee recommended continued use of the NRC [27] equations for calves and yearlings. In addition, the committee recommended the use of a new equation to predict DMI as a percentage of BW using linear and quadratic coefficients for dietary NEm concentration [53] and equations for feedlot steers and heifers fed high-grain diets that were based on initial BW [54]. Recently, data from 479 pens of feedlot steers in Brazil were used to develop an equation to predict DMI based on metabolic BW and linear and quadratic components for average daily gain (ADG) [55]. In a subsequent evaluation with independent feedlot data [55], it was concluded that the equation was more accurate and precise when applied to Nellore cattle than the equations recommended by NASEM [3]. Differences among equations likely reflect the databases used for development (e.g., Zebu vs. non-Zebu cattle), but also reflect a difference in approaches to predicting DMI. The NASEM [3] approaches focus on initial shrunk BW or average shrunk metabolic BW with the potential to include dietary NEm concentration in the prediction equations, whereas the other equation [55] includes average metabolic BW and ADG. Including ADG is likely to yield superior predictions, as it is probably a more accurate reflection of the utilization of dietary energy than NEm concentration alone. Indeed, using ADG in equations would be akin to the DMI required (DMIR) approach suggested and recommended previously by NASEM [3]. The challenge, however, of using the DMIR method or equations that include ADG is that one must accurately predict ADG to apply the equation in practice or only use the equation for post-hoc analysis of observed vs. predicted DMI.

Additional research that is focused on refining DMI predictions, particularly the possibility of combining traditional approaches with data obtained from precision livestock farming (PLF) technologies to predict individual animal intake, could yield improvements. In this regard, feeding-behavior data obtained with electronic feeders and various modeling approaches were used to predict intake by cattle fed diets that varied in forage-to-concentrate ratio at individual and group levels [56]. Machine learning and regression approaches for individual data improved prediction errors, but group-based models yielded greater repeated measures correlations than models based on individual data. In all cases, the authors concluded that prediction errors relative to the mean (approximately 10%) were too high for practical application. Machine learning approaches were also used to predict DMI by bulls and steers fed corn silage-based diets [57]. An in-pen electronic system collected individual feed and water intake and BW data, which were combined with various climatic variables to predict individual DMI. Body weight and ADG were key elements of predictions, and the approach resulted in predictions that were within approximately 9% of the mean. Although more research is needed, these early results indicate the potential value of PLF technology and machine learning predictive approaches.

The validity of DMI predictions for beef cows has been a concern expressed since the NRC [27] recommended a DMI equation based on a quadratic function of dietary NEm concentration and average shrunk metabolic BW. For example, the NRC [27] equation was criticized [58,59], because it lacked both precision and accuracy, particularly when applied to cows fed primarily forage and mixed forage:concentrate diets. Nonetheless, it was concluded [60] that the equations that included BW and digestibility [58] were highly correlated with predictions from the NRC [27] beef cow equation, although the NRC [27] predictions were consistently higher or lower depending on the physiological state of the cow. Other equations [59] were developed from a much narrower range of dietary NEm concentrations than the database used to develop the NRC [27] equation, which likely contributed to the lack of accuracy and precision in the NRC-based predictions. It should also be noted that the use of an incorrect adjustment [59] to DMI for milk yield in the NRC [27] equation likely affected comparisons, and the authors [59] also incorrectly stated that the NRC [27] database included protein-deficient diets. Despite these issues, it is clear that the NRC [27] beef cow equation often fails to provide accurate predictions of DMI by beef cows.

Although beef cows consume predominantly high-forage diets during their lifetime, forage quality and energy concentrations vary considerably over time. In addition, cull cows are often fed high-grain diets at the upper end of the energy spectrum. Thus, the challenge for NASEM committees is to provide DMI prediction equations that can be applied over a wide range of diet quality (i.e., energy concentrations). The NASEM [3] committee followed the lead of the NRC [27] committee and chose an approach of recommending one equation that would be applicable over a wide range of diets; however, this one-size-fits-all approach might not be the best in practice. Indeed, the equations reported by others [58,59] are more diet-specific than the NRC [27] equation, which likely explains their improved fit to the databases used by these authors. Perhaps an alternative approach is needed by the next NASEM committee, one in which multiple equations specific to narrower energy ranges (or other dietary components) could be provided for the prediction of DMI by beef cows.

In addition to a NEm-based prediction equation for beef cows, the NASEM [3] committee also suggested the potential for using NDF concentration to predict the intake of all-forage diets. This concept was derived from previous work [61], which indicated that dairy cows consumed approximately 1.2% of BW as NDF. This concept was extended to include models based on either physical regulation of intake (i.e., NDF) alone or in combination with physiological (i.e., ME) regulation of intake that were evaluated using literature data collected from cattle in the tropics and sub-tropics [62]. The authors concluded that such models could provide adequate predictions of DMI by stall-fed cattle. Data from growing dairy and beef bulls fed mixed forage:concentrate diets were used to develop DMI prediction equations [63]. A non-linear equation that included BW alone was improved when NDF concentration of the diet was included in both the constant and exponent terms, with a resulting prediction error of 5% of the mean total DMI. Further research on using NDF to predict intake, especially for beef cows fed predominantly forage-based diets, seems warranted.

Predicting intake in grazing cattle is particularly difficult. The unique challenges of the grazing environment relative to intake measurements (selective grazing; changing forage mass, structure, and composition; climatic and environmental factors; social influences) were discussed previously [60]. Practically, however, the greatest challenges relate to the fact that intake and grazed forage composition cannot be measured directly, necessitating indirect measurement and specialized sampling methods. Including PLF technologies and other digital and sensor-based approaches can provide novel tools to potentially yield more accurate estimates of grazed forage intake [64], but the science in this area is in the early stages of development.

2.6. Maintenance, Reproduction, Growth, Stress, and Disease

2.6.1. Environmental Temperature and Grazing Activity

Several research needs related to the maintenance requirements of various classes of beef cattle were identified by the NASEM [3] committee. Among these, estimating the effects of environmental temperature and humidity (e.g., hot and cold environments) and effects of physical activity associated with grazing are key issues. Concerns about the effects of climate change on beef production, particularly increased heat stress, have been noted in several recent reviews [65,66,67,68], with research needed on both animal (e.g., genetic and physiological adaptations) and physical (e.g., shade, sprinkling, air movement) mitigation strategies. Cold stress is also a concern, with demonstration that long-term cold stress during the winter increased DMI but decreased ADG and nutrient digestibility in Simmental crossbred bulls compared with bulls fed in the autumn when temperatures were generally in the thermoneutral zone [69]. Cold stress increased lying time, and several blood metabolites were altered. Short-term (14 d) mild or extreme cold stress was evaluated in bull calves and fattening steers [70]. Calves exposed to extreme cold stress (mean ambient temperature = −4.33 °C) had lower DMI, with increased heart rate and rectal temperature compared with calves exposed to temperatures near the threshold of cold stress (mean ambient temperature = 4.66 °C). Readers are referred to a recent, practically focused review of cold stress in cattle for additional information [71].

Effects of both heat and cold stress on production by various classes of beef cattle are well documented in the literature, but translating production responses to equations that define energetics in nutritional models is not an exact science. Concerns about how to describe short-term environmental stress using models that are generally developed from long-term studies, where effects of these stressors are moderated over time, is challenging. Defining short-term nutritional modifications during heat or cold stress that could help maintain production levels might be a logical focus of future research in this area, with the information generated contributing to the refinement of nutritional models.

The energy cost of grazing can be substantial, with the NASEM [3] publication citing earlier work that indicated a 10 to 20% increase in maintenance requirements under optimal grazing conditions, with a potential increase up to 50% under extensive, hilly pastures. An equation previously developed by the Commonwealth Scientific and Industrial Research Organisation (CSIRO) was recommended by NASEM [3] to estimate the energy costs of grazing, which includes DMI, digestibility, a factor for terrain, BW, and green forage availability. Nonetheless, because of uncertainty about the accuracy and precision of this and other potential equations, the BCNRM did not include an adjustment to the maintenance requirement of grazing cattle.

Advances in PLF technology might be a means by which the BCNRM and other nutritional models could be updated to estimate the effects of grazing activity on maintenance requirements more accurately. Various electronic movement detection systems, combined with positioning system data to identify location and terrain variables, could be applied to an entire herd or to a limited number of “sentinel” animals with a larger group, with real-time or integrated locomotion and terrain data being fed back to nutritional models to update estimates of energetic costs. This use of PLF technology could concurrently aid in identifying animal welfare issues and forage availability across the grazed landscape. Several recent reviews are available on the potential value of PLF technologies for grazing animals [64,72,73,74].

2.6.2. Beef × Dairy Crossbreds

The increased importance of beef × dairy crossbred cattle in the U.S. feedlot industry [75] necessitates evaluating whether current NASEM [3] nutrient requirement recommendations are applicable to these cattle. In the BCNRM, dairy breeds are assumed to have a 20% increase in maintenance energy requirements, with a beef × dairy crossbred assumed to have a 10% increase over the beef breed in the cross. Nonetheless, in practice, Holstein steers are often assumed to have a 12% increase in their NEm requirements compared with beef breeds [76]. Assuming the 12% increase, Angus × Holstein crossbreds had a calculated maintenance requirement that was midway between the breeds in the cross [76]. Additional research based on feeding studies and direct measurement of maintenance requirements using indirect calorimetry or comparative slaughter would be useful to more fully characterize the energy needs of these cattle. More generally, the assumption of a 20% increase in NEm requirements for all classes of dairy cattle and some dual-purpose breeds of cattle (e.g., Simmental) deserves additional consideration.

2.6.3. Developmental Programming

Nutritional aspects of developmental programming, particularly the extent to which maternal deficiencies or excesses of nutrients affect metabolism, growth, and body composition of the offspring is an important topic that was addressed by the NASEM [3] committee. Nonetheless, the committee did not make specific nutritional recommendations related to developmental programming. Considerable research has been conducted on this topic over the last decade, with a recent review [77] providing an excellent summary of the findings relevant to nutrient requirements and strategic supplementation of nutrients. Given that developmental programming responses are driven by epigenetic changes, and these changes depend on one-carbon metabolism (i.e., methyl donors; [78]), additional research related to strategic supplementation of nutrients involved in one-carbon metabolism (e.g., methionine, folic acid, betaine, choline, and vitamin B₁₂) is needed to clarify appropriate timing and potential effects. The ways in which other nutrients, particularly trace minerals (e.g., Se, S, Co, and Zn), are involved in epigenetic processes and one-carbon metabolism also needs to be considered.

2.6.4. Retained Nutrients and Body Composition Endpoints

Existing methodologies used in the NASEM [3], such as those in the California Net Energy System, rely heavily on interdependent predictions. Although this approach ensures consistency, it can mask inaccuracies in subcomponents. A good example of these inaccuracies is the prediction of retained energy (RE) and RP in current NASEM [3] models, which rely on linear relationships that fail to capture the nonlinear dynamics of fat and protein deposition, leading to inherent assumption mismatches. Many models assume a fixed linear relationship between RE and RP; however, the exponential nature of fat deposition and the plateauing of protein accretion [3] challenge this assumption, particularly as animals mature. Tedeschi [79] highlighted the interdependence between RE and RP predictions, noting that offsetting errors decrease the reliability of the model. Indeed, predictions of RP are often more accurate when based on predicted RE rather than observed RE, further illustrating the instability caused by internal dependencies. The limitations of empirical equations in capturing these dynamics, highlights the need for nonlinear mechanistic approaches [80].

An accurate estimate of BW at a compositional endpoint is vital to applying the NASEM [3] models for predicting beef cattle performance. A previous equation [81] has been used extensively to calculate empty body fat and thereby determine the final BW adjusted to a common empty body fat (e.g., 28%). This equation was based on data from cattle fed to a lower empty body fat than is current practice in the U.S. feedlot industry. Thus, it might be beneficial to reevaluate this equation with body composition data more consistent with current feeding practices and final BW at harvest.

A significant challenge related to improving predictions of retained nutrients and compositional endpoints is that traditional methods for assessing body composition, such as comparative slaughter, are labor-intensive and imprecise, leading to substantial variability in predicted outcomes. Thus, the disconnection between RE and RP is, at least in part, a methodological problem, and very few improvements can be made at the modeling backend. Developing mechanistic sub-models to capture the nonlinear interactions between fat and protein deposition could be an option. These sub-models should reflect exponential fat deposition and plateauing of protein accretion as animals mature. Moreover, integrating advanced imaging technologies, such as dual-energy X-ray absorptiometry and computerized tomography to improve the accuracy of body composition data, could be an effective approach moving forward to address such problems.

2.6.5. Effects of Stress and Disease on Energetics and Nutrient Requirements

It is generally accepted that infection and inflammatory processes increase energy requirements. Moreover, stress associated with marketing and transportation, particularly in recently weaned cattle, elicits various metabolic responses that undoubtedly affect energy requirements. Although these events can have long-term consequences for animal production [82], because the duration of the effects associated with inflammation, disease, and stress is generally limited, it is challenging to model changes in maintenance energy requirements. Decreased feed intake is a common response to stress and disease, which creates a practical dilemma as to how to best fortify diets during periods of stress and disease to offset lower intake. In this regard, Gouvêa et al. [83] concluded that stress-induced inflammatory responses result in physiological responses that disrupt the normal processes that control feed intake by ruminants, with proinflammatory cytokines being identified as potential mediators of stress-induced decreases in feed intake. Because the metabolic and physiological changes that occur during stress and disease are many, vary among individuals, and are not fully elucidated, incorporating these changes into nutritional models is virtually impossible.

What can we do to address changes in nutrient requirements, specifically energetics, associated with disease and stress? Direct measurements of energy expenditure in animals undergoing inflammatory and disease stress would help to quantify the whole-animal effects. This will most likely require model systems that apply well-defined and reproducible inflammatory and/or disease responses (e.g., lipopolysaccharide and/or specific respiratory pathogen challenges). From a practical perspective, a recent review of nutrition and management research on the health of newly received cattle [52] led to the suggestion that additional research is needed to improve performance and health outcomes by identifying targeted nutritional supplementation programs that could offset the negative effects of inflammation, disease, and stress.

2.7. Environmental Issues

The Statement of Task to the NASEM Committee on Nutrient Requirements of Beef Cattle included a mandate to “review nutritional and feeding strategies to minimize nutrient losses in manure and reduce greenhouse gas production” [3]. As a result, a new chapter entitled “Environment” was added to the document, which focused on factors affecting nutrient excretion, primarily N and P, and prediction of greenhouse gas emissions. These areas have been the subject of extensive research over the past decade, necessitating summarization and updating of the NASEM [3] recommendations.

2.7.1. Greenhouse Gas Emissions

Beef cattle production systems in the U.S. need to adopt CH₄ prediction equations to use with tier 3 methodologies [84,85], as well as develop applied tools and computer models that can be used by cattle producers to evaluate how management changes affect their carbon footprint, potentially leading to the development of viable carbon markets [86]. Accurate predictive models can be used by researchers to estimate greenhouse gas emissions in cases where the research facilities are not adequate to measure emissions. Continued research focused on developing accurate methods to predict CH₄ emission from beef cattle under varying production practices is needed, but considerable progress has been made in recent years. The NASEM [3] publication identified several equations to predict CH₄, but specific equations were not recommended. Newer equations developed from a large international database [87] have been published and need to be evaluated by future NASEM committees. Moreover, because DMI and various measures of digestibility are key components of most equations, accurate estimates of these variables are also critical to practical applications of CH₄ prediction equations.

In addition to the development of prediction equations, considerable research has been conducted in the past decade to evaluate potential CH₄ mitigation strategies. Among the feed additives that either are currently available or could be soon, 3-nitrooxypropanol can significantly decrease enteric CH₄ production. Although only approved in the U.S. for lactating dairy cows, it would seem to have potential for use in beef production systems. A recent comprehensive review [88] includes information on 3-nitrooxypropanol and other potential feed additives. An approach to mitigation strategies that focuses on capturing the energy typically lost as CH₄ in animal products would benefit both the climate and the efficiency of beef production.

2.7.2. Manure Fertilizer Value and Nitrogen and Phosphorus Losses

The NASEM [3] publication provided information on manure output and suggested equations that could be used to predict urine and fecal N and total P excretion. In addition, an equation was provided to estimate the fraction of total N excreted in the urine, which is an important component of ammonia emissions. Further evaluation and refinement of these equations is needed, as well as additional research that would help describe the fertilizer value of manure, thereby assisting producers in developing comprehensive nutrient management plans. Because housing systems for cattle can affect nutrients removed in manure, their effects should be considered in future research and equation development.

Management of P excretion to avoid contamination of surface waters has long been recognized as a concern by the beef feedlot industry. As noted by Warner et al. [89], most feedlot nutritionists do not supply P beyond that provided by the dietary ingredients, which generally results in P concentrations that exceed animal requirements. For the cow–calf sector, however, supplementation of P is often practiced, at least during forage dormancy, leading the authors to suggest approaches by which producers could avoid watershed contamination [89]. Research conducted in different cow–calf production systems relative to production responses and environmental effects of P supplementation seems warranted.

2.8. Feed Composition

Providing users with nutrient composition data of commonly used feed ingredients has been a longstanding practice of the NASEM nutrient requirement publications. Many nutritional components are consistent across livestock species (e.g., protein, fat, various measurements of fiber, vitamins, and minerals). In addition to tabular values provided in NASEM publications, many excellent feed composition databases are available online. Examples include the U.S. National Animal Nutrition Program feed composition database (https://animalnutrition.org) and the French/European INRA-CIRAD-AFZ feed tables (https://www.feedtables.com), along with the related Feedipedia database (https://www.feedipedia.org).

One challenge with compiling data for beef cattle (and ruminants in general) is the wide variety of feed ingredients incorporated in their diets, including many byproducts from agricultural industries (e.g., byproducts of ethanol production, plant-based oil processing byproducts, and grain milling byproducts), as well as a wide variety of residues from crop and vegetable production. These byproducts can vary considerably in composition depending on agronomic practices and the processes used to generate them. A second challenge with beef cattle relates to the need for data beyond the basic chemical components (e.g., protein, fat, and minerals). For example, both the ELS and MLS models of the BCNRM require reliable estimates of RDP and RUP fractions of ingredients. Similarly, degradation and passage rates for carbohydrate and protein fractions are needed for the application of the MLS model. Moreover, given the importance of the environmental impact of livestock feeding, incorporating sustainability factors that describe potential effects of feed ingredients on greenhouse gas and N emissions could be a useful addition to future feed composition databases. Continued research in this area will provide useful information for subsequent revisions of the NASEM [3] beef cattle requirements.

Energy values of feed ingredients are not consistent across livestock species, requiring values to be developed for beef cattle, and potentially for various classes of beef cattle. Whether the application is to complete mixed diets in confined cattle feeding operations or to supplements fed to grazing cattle, accurately defining NEm and NEg values for feeds, especially novel feed ingredients, is critical to the application of the energetic components of the BCNRM. For example, with the growth of ethanol production in the U.S., the availability of wet distillers grains plus solubles (WDGS) required the NASEM [3] committee to reevaluate the NEm and NEg values of this ingredient. Based primarily on growth performance studies, the committee recommended NEm and NEg values of 2.47 and 1.74 Mcal/kg, respectively, compared with the values of 2.24 and 1.55 Mcal/kg, respectively, suggested by NRC [27]. Subsequent research using indirect respiration calorimetry to determine energy values in finishing beef steers [90,91] indicated that the NASEM [3] values overestimated net energy concentration of WDGS by approximately 15%, resulting in values much closer to those suggested by NRC [27]. Further calorimetry work or reevaluation of growth performance data seems warranted for this important ingredient and other byproduct feeds. Because the nutrient composition of byproduct feeds, particularly the fat content of distillers grains, has changed over time, consideration of complete nutrient profiles is needed when evaluating growth performance data. Moreover, defining experimental designs and calculation methods for estimating net energy values from growth performance data might prove valuable for future NASEM committees.

2.9. Models for Nutrient Requirements

Accurately modeling the nutrient requirements of beef cattle is foundational to sustainable livestock production, with global significance for meeting the rising demand for using environmentally sustainable practices. Current models, including the BCNRM of NASEM [3], provide valuable frameworks but exhibit restrictions that limit their broader applicability. Critical challenges include the disconnection between RE and RP discussed previously, oversimplifications in CH₄ and energy partitioning, and the lack of readiness for including inputs from PLF technologies, among many others. Addressing these challenges requires integrating dynamic mechanistic models, leveraging artificial intelligence (AI), and incorporating real-time data capabilities.

2.9.1. Integrating Mechanistic and AI-Driven Models for Precision Predictions: Enhancing Individual Precision and Precision Livestock Farming Readiness

Current models rely heavily on group-level data, limiting their precision for individual animals. This is particularly problematic in optimizing feeding strategies in systems aiming for individual animal management [92,93]. Artificial intelligence-driven pattern recognition can enhance the scalability and adaptability of mechanistic frameworks by identifying nonlinear relationships in large datasets. Precision livestock farming technologies, like wearable sensors, enable continuous monitoring of individual variables, including feeding behavior, body temperature, and activity. These sensors, ranging from accelerometers and gyroscopes to temperature and pH and pressure sensors, provide real-time data critical for decision-making [94]. For example, accelerometers placed on collars or legs can monitor grazing and rumination, whereas temperature sensors can detect physiological changes indicative of stress or illness. Pressure sensors near the jaw can evaluate feeding patterns and chewing efficiency, offering insights into nutrient intake, and pH sensors could be used to monitor the potential for acidosis with changing feed resources, weather patterns, and management activities.

By combining mechanistic principles with AI, the BCNRM could dynamically adjust model parameters based on real-time data, significantly improving its utility in diverse production environments. For example, AI algorithms can integrate data from wearable sensors to predict the onset of heat stress, allowing producers to adjust feeding regimens preemptively. Moreover, real-time analysis of feeding behavior using AI can identify inefficiencies in nutrient delivery, thereby optimizing feed conversion rates and decreasing waste. In dairy systems, machine-learning models have been successfully used to forecast milk yield [95], showcasing the potential of AI-driven insights across livestock operations. Machine learning algorithms further refine these predictions by identifying emergent patterns and providing actionable insights for producers [96]. For instance, hybrid systems that integrate accelerometers with location-based technologies, such as Ultra-WideBand, have demonstrated increased sensitivity in detecting behavioral changes, decreasing false alarms, and improving overall precision in livestock management [94]. Integrating advanced communication technologies, such as the NarrowBand Internet of Things (NB-IoT) and long-range (LoRa) wireless technology, can enhance data transmission across extensive grazing areas, addressing limitations in connectivity [94]. These advancements enable seamless integration of wearable devices with centralized databases, allowing real-time animal-specific and environmental data synthesis. As PLF methods advance, such systems are pivotal in ensuring that nutrient requirement models remain adaptable and relevant in the context of modern livestock production.

2.9.2. Dynamic Energy and Methane Partitioning Models: Unifying Energy Loss Frameworks

The calculation of CH₄ emissions in current models—whether empirical or mechanistic—does not adequately integrate these losses into energy-partitioning frameworks [97,98]. This omission creates inconsistencies in calculating the efficiency of converting DE to ME, which undermines the internal consistency of the models. As discussed previously, empirical equations have been developed to predict the DE-to-ME conversion efficiency, but integrating CH₄ losses into energy partitioning dynamically, using feedback loops to adjust DE-to-ME conversion efficiencies, would be a superior approach. For example, incorporating volatile fatty acid (VFA)-driven CH₄ calculations into broader energy loss frameworks could ensure a consistent and accurate representation of energy dynamics. These enhancements would harmonize empirical and mechanistic approaches, bridging gaps in CH₄ and energy modeling.

2.9.3. Mechanistic Models for Dynamic Nutrient Partitioning: Addressing Environmental and Biological Variability

Mechanistic models offer a deeper understanding of nutrient metabolism than empirical approaches by explicitly considering interactions with environmental stressors, variability in feed composition and quality, and individual animal differences. These models can dynamically account for protein and energy metabolism, providing scenario-based predictions for grazing vs. confined systems. An approach that includes variability aligns with the demands of PLF, where adaptability and precision are paramount [92].

2.9.4. Leveraging Genomics and Omics Technologies: Improving Model Predictability with Advanced Data

Advances in genomic and omics technologies have provided important insights into ruminal microbial populations and their interactions with the host. As noted previously, including omics technologies has been suggested as an approach to improve the accuracy and precision of MCP predictions. Practical applications in commercial livestock production include leveraging metagenomics to identify microbial communities linked to greater feed efficiency and decreased CH₄ emissions. For example, sequencing data have been used to develop targeted probiotics that enhance fiber digestion. Similarly, metabolomics has facilitated the identification of biomarkers that can predict nutrient utilization efficiency, enabling producers to tailor feed formulations for optimal performance. These technologies are increasingly integrated into management practices, offering producers actionable insights to improve sustainability and productivity. Integrating meta-omics approaches can reveal microbial features associated with feed efficiency, offering a deeper understanding of ruminal microbial functions critical for accurate nutrient modeling [99]. A “rumen microbiome nutriomics” framework, which combines metagenomics, metabolomics, and advanced data analyses to elucidate the complex relationships between dietary inputs and microbial activity has considerable potential [100]. By incorporating these datasets into mechanistic models, predictions of MCP synthesis and nutrient utilization efficiency could become more precise and accessible. Furthermore, AI-driven regression models can complement this approach by identifying key biomarkers that predict CH₄ production and nutrient metabolism. These innovations make it possible to directly link MCP synthesis to modeling frameworks, streamlining the integration of microbial insights into practical decision-support tools for ruminant nutrition.

2.9.5. Enhancing Fiber and Nutrient Digestibility: Advanced Mechanisms and Methods

Improving fiber digestibility is critical for enhancing nutrient utilization efficiency [101] in ruminants. Mechanistic models should incorporate advanced imaging technologies and in vitro methodologies to better represent complex interactions within the rumen. These models could simulate fiber degradation pathways, emphasizing the dynamics of microbial attachment and enzymatic hydrolysis under varying grazing conditions. Integrating dynamic models to simulate the interplay between fiber quality (e.g., lignin content) and microbial activity offers a pathway to optimize nutrient availability. For example, using mathematical models to predict fiber digestibility improved voluntary feed intake predictions in grazing cattle [102]. Similarly, real-world applications of these models have shown success in adapting feeding strategies to account for varying forage quality across seasons, optimizing nutrient supply in extensive grazing systems. These examples underscore the potential of dynamic models to enhance decision-making and nutrient efficiency in diverse production settings. Models can incorporate feedback mechanisms to refine predictions, including the effects of grazing behaviors, forage selection, and environmental factors. Furthermore, the application of sensor technologies to measure real-time grazing behavior and intake can significantly enhance the precision of nutrient utilization models [102]. Advanced mechanistic frameworks should also address the synchronization of carbohydrate and protein degradation in the rumen to maximize fermentation efficiency and MCP synthesis.

2.9.6. Unified Model Components for Consistency: Harmonizing Predictive Frameworks

Inconsistencies in current models often arise from disconnected components, with CH₄ estimation and energy efficiency, as noted previously, being a good example. A unified model approach where changes in one component seamlessly adjust others is essential for the success of nutrient requirement modeling efforts. As noted above, feedback loops within a unified framework could dynamically adjust DE and ME calculations based on CH₄ emissions. If CH₄ output increases, the model will recalibrate the energy available for production, thereby maintaining consistency across all energy partitioning components. Similarly, when protein utilization efficiency changes because of dietary modifications, the feedback loop adjusts related nutrient requirements, ensuring the model remains robust and responsive to real-time data inputs. These mechanisms enhance accuracy and internal harmony within the model, as described for feedback loops that integrate VFA-driven CH₄ calculations with energy partitioning [97,98,103].

2.9.7. Implications for the Future of Beef Cattle Nutrition Models

By integrating dynamic mechanistic models, AI-driven insights, and real-time data from PLF technologies, the next generation of the BCNRM can bridge critical gaps in precision and applicability, addressing limitations in energy partitioning, CH₄ modeling, and individual animal predictions. These advancements will align the model with the evolving demands of the livestock industry, promoting sustainable intensification of livestock production and precision-driven nutrition management, as well as enhancing nutrient utilization efficiency and decreasing environmental impacts. Furthermore, harmonizing model components will ensure robustness and adaptability, positioning the BCNRM as a cornerstone of modern animal nutrition.

3. Summary and Recommendations for Moving Forward

Key knowledge gaps in beef cattle nutrient requirements identified by the authors are summarized in

Table 1. Summary of knowledge gaps and research needs in beef cattle nutrition identified by the authors.

Category	Knowledge Gaps and Research Needs 1
Energy, carbohydrates, and lipids	Verify relationships between TDN, DE, ME, NEm, and NEg concentrations of feedstuffs Better characterize effective fiber in feedlot diets and its relationship to ruminal pH
Protein	Improve methodology to measure digesta flow and MCP synthesis Better define MP requirements for maintenance and efficiency of converting MP to retained protein Develop accurate and precise prediction equations for recycled N Improve estimates of RDP and RUP in common feedstuffs
Minerals and vitamins	Develop and define standards for determining mineral requirements Define relationships between fetal/neonatal vitamin and mineral status, and subsequent growth performance and carcass merit Determine whether supplementation of Zn beyond recommendations is beneficial and define specific recommendations Clearly define conditions under which Cr supplementation is beneficial
Feed intake	Continue to refine DMI predictions, particularly for grazing beef cows, and consider inclusion of PLF technologies in prediction systems
Maintenance, growth, reproduction, stress, and disease	Define both short- and long-term effects of cold and heat stress on maintenance requirements Improve methods for predicting the energy cost of grazing, with incorporation of PLF technologies as appropriate Define maintenance requirements of dairy x beef crossbred steers and heifers Better define the role of developmental programming in offspring growth and body composition and identify key nutrients involved in these processes Reevaluate the relationship between RE and RP and BW at compositional endpoints Define both short- and long-term effects of inflammation, disease, and stress on maintenance requirements
Environmental issues	Refine recommendations for prediction of CH4 Refine equations for prediction of N and P excretion
Feed composition	Better characterize the nutrient composition, particularly energy values, of byproduct feedstuffs
Models for nutrient requirements	Incorporate PLF and AI technologies into nutritional models to better assess biological variability Increase the use of dynamic feedback loops and dynamic mechanistic frameworks to improve consistency of model predictions

¹ AI = artificial intelligence; DMI = dry matter intake; MCP = microbial crude protein; ME = metabolizable energy; MP = metabolizable protein; NEm and NEg = net energy for maintenance and gain, respectively; PLF = precision livestock farming; RE = retained energy; RDP = ruminally degraded protein; RP = retained protein; RUP = ruminally undegraded protein.

. As noted earlier, many of these recommendations were identified in the research needs section of the NASEM [3] publication. Where possible, we have updated these previously identified areas based on recently published data and have added new areas based on our experience in applying the NASEM [3] nutrient requirements in research and production settings. It is our sincere hope that these recommendations will be a valuable resource for future NASEM committees as they face the challenge of updating and revising nutritional recommendations for beef cattle. In addition, our findings could provide a guide for funding agencies in setting priorities for future beef cattle research.