Precision Determination of Energy and Protein Requirements of Grazing and Feedlot Animals

Introduction

The United Nation’s sustainable development goals aim to “…achieve a better and more sustainable future for all” by addressing “…the global challenges we face, including poverty, inequality, climate change, environmental degradation, peace and justice.” These challenges are interconnected and depend on integrating innovative ideas of several science fields to achieve sustainability, especially in livestock production systems. The perfect match between what animals require to perform as-desired and what dietary ingredients supply to them is a sine qua non condition to sustainably meet animal products’ demand by the human population worldwide. It is a wicked problem to be solved in many ways, and solving it requires a systems approach (Tedeschi et al., 2015). Grazing and feedlot systems are two completely different production conditions for ruminants, and they are not necessarily mutually exclusive. Sometimes the grazing phase precedes the feedlot phase, and, for some regions in the world, it can be one or the other for the complete production duration (Cottle and Kahn, 2014).

The majority of recommendations for grazing animals’ energy and protein requirements have been determined on confined animals that are much closer to animals’ requirements raised under feedlot conditions. In part, the determination of energy and protein, as well as other nutrients, for grazing animals require special equipment and methodology that makes it more challenging, expensive, and laborious. On top of that, such estimates’ variability is much greater for grazing than confined animals because grazing animals are prone to the impact of additional environmental factors that confined animals are not. The requirements of energy and protein for confined animals have been extensively studied and disseminated for over 116 years (Kellner, 1905), using concepts devised in the late 1700s by many illustrious scientists, including Leonardo da Vinci^1452-1519, Joseph Priestley^1733-1804, Carl Wilhelm Scheele^1742-1786, Antoine-Laurent Lavoisier^1743-1794, Pierre-Simon Laplace^1749-1827, and Justus Freiherr von Liebig^1803-1873 (Tedeschi and Fox, 2020a, Ch. 4). Digestion trials have been implemented for a bit longer: since 1860 at the Weende Experiment Station at the University of Goettingen in Germany, and since 1884 at the University of Wisconsin Agricultural Experiment State in the United States (Schneider and Flatt, 1975).

Many classical books (Baldwin, 1995; Blaxter, 1962; Brody, 1945; Kleiber, 1961) and literature reviews (Ferrell and Oltjen, 2008; Garrett and Johnson, 1983; Johnson et al., 2003; Tedeschi, 2019a) have comprehensively discussed the transactions of animal bioenergetics. For beef cattle, currently recommended energy and protein requirements stemmed from pen-fed animal studies using the comparative slaughter technique and the California Net Energy (NE) System methodology (Oltjen, 2019), though discrepancies might exist (Tedeschi et al., 2017; Tedeschi, 2019b). For dairy cattle, current recommendations for energy and protein requirements are based on the indirect calorimetry using open-circuit respiration chambers (Flatt et al., 1965a, b; Moe et al., 1972; Moe, 1981) at the Energy Metabolism Laboratory at the Dairy Cattle Research Branch, United States Department of Agriculture, Agricultural Research Center, Beltsville, Maryland, US (Flatt et al., 1958). Most modern requirement studies were initiated in the 1960s worldwide.

Grazing animals have an additional energy requirement associated with the grazing activity compared to confined animals’ requirements. It comprises the additional energy needed for body movements (i.e., locomotion) and forage browsing, selection, and prehension. The non-activity maintenance requirement of energy between grazing and confined growing or finishing animals might be identical on a metabolic weight basis, as long as animals are at the same maturity degree (i.e., same composition of gain) (Tedeschi and Fox, 2015). However, because the diet consumed by grazing animals (i.e., essentially forage) have a lower partial efficiency of energy use for growth (kg), grazing animals would require a greater dry matter intake (DMI) to achieve the same average daily gain (ADG). This fact becomes a significant limitation for grazing animals for two main reasons: 1) the distance traveled to reach maximum voluntary intake within a 24-h period (daily basis) may exceed animal’s locomotion capacity, worsening its energy balance; and 2) because DMI is also a function of rumen size (i.e., volume/space), low-quality forages, i.e., forage containing a higher proportion of fibrous material, may further restrict intake by triggering the negative effect of rumen fill on voluntary feed intake (Tedeschi and Fox, 2015; Tedeschi and Fox, 2020a, Ch. 10). Both reasons concurrently impede grazing animals to have the same ADG as confined animals. Although the energy cost of physical activities in cattle, sheep, and buffaloes has been extensively documented worldwide, a comprehensive physical activity calculation logic does not exist because the lack of information on the energy required for consuming and processing feed by the animal is considerably larger for grazing animals than for confined animals (Tedeschi and Fox, 2015).

This review’s main objectives are to 1) illustrate different methods to estimate feeds’ nutritive value, 2) briefly discuss existing techniques and methods to assess energy expenditure (EE) in grazing and confined animals, and 3) provide a different calculation logic in predicting energy requirement for grazing animals.

Determination of Feed Nutritive Value

An essential step in determining energy and protein requirements is to have an accurate assessment of the nutritional value of the feeds consumed by the animals. This is not a trivial step. Based on wet gravimetric chemistry methodology, chemometric techniques are the preferred methods to determine nutrient contents in feeds (Faithfull, 2002; Moughan and Hendriks, 2018; Van Soest, 2015), and when associated with animal’s in vivo digestibility trials (Schneider and Flatt, 1975), an assessment of digestible energy (DE) and total digestible nutrients (TDN) becomes possible. In essence, chemometric techniques are used for critical nutrients (e.g., protein, ether extract, and carbohydrate) in determining the “energy content” of feeds through DE or TDN. For practical purposes, in ruminants, 1 kilogram of TDN is assumed to contain 4.4 kcal of DE (Swift, 1957). Although energy is defined as “the ability to perform work,” different “forms of energy” have been defined, depending on how it is used (or lack of) by the animal. The energy content of a feed (or feed’s nutrient) is determined using adiabatic or ballistic bomb calorimeters (McLean and Tobin, 1987), and it is referred to as gross energy (GE).

Upon consumption (and rumination) by the animal, the digesta undergoes many physicochemical transformations in the gastrointestinal tract, including fermentation in the rumen and hindgut, and enzymatic digestion in the midgut, before being absorbed. The amount of DE is computed by difference based on the GE contents in the feed and feces, DE = (Feed intake x GE_Feed - Feces x GE_Feces)/(Feed intake x GE_Feed). In ruminants, the amount of energy available to the animal is called metabolizable energy (ME), and it is computed after removing the energy lost to gas production in the rumen (methane, CH₄E) and urine substances (UE), 𝑀𝐸 = 𝐷𝐸 − (𝐶𝐻₄𝐸 + 𝑈𝐸). Finally, ME is further adjusted to the heat production (HP) depending on its fate (e.g., maintenance, tissue deposition, milk production) (NASEM, 2016; NRC, 2001). Though this sketch is an oversimplification of factors involved in the animal’s bioenergetics, it still provides a general idea of the feeds’ supply of energy.

Energy is an attribute of the feed that results from the combination of different nutrients and their digestibility and metabolism by the animal. Hence, it is interesting to note that plant defensive compounds (PDC) (Tedeschi and Fox, 2020a), also known as antinutritional factors, are not explicitly considered when determining TDN or DE of feeds, likely because it is assumed that such compounds would exert their negative impact during the in vivo digestibility trial; thus, already decreasing the digestibility. However, the conditions in which animals consume PDC-containing feeds might differ from those used to determine their TDN, especially under grazing conditions when plants might be at various growth stages and contain different types and concentrations of PDC during the growth of new leaves (Harborne, 1993).

Techniques to Determine the Nutritive Value of Feedstuffs

Wet chemistry has been the golden standard for determining the chemical composition of feeds and, ultimately, their nutritive value. Nevertheless, there are limitations such as the cost of the analyses, the time it takes to obtain the results, and the labor for collecting representative samples and processing them for subsequent analyses that prevent broad adoption of wet chemistry. Perhaps the main problem happens with grazing animals. A representative sample of what the animal is consuming is practically impossible to obtain, and the speed the forage changes its physicochemical characteristics is likely faster than the obtention of the laboratory results. Thus, alternative, quicker techniques have to be employed.

Near-infrared spectroscopy (NIRS) resulted from Sir Frederick Herschel’s pioneering work, who studied the infrared radiation while passing sunlight through a prism and measuring its temperature. Spectroscopy and spectrography were born in the 19th century, and circa the 1960s (Shenk and Westerhaus, 1994), scientists started adopting spectrophotometric techniques given their automation in the laboratory and high correlations with some forage components measured via wet chemistry (Norris et al., 1976). For grazing systems, NIRS provided opportunities that did not exist or were very difficult to achieve through wet chemistry, including daily flux of diet quality consumed by grazing animals (Lippke and Barton, 1988; Stuth et al., 1991; Stuth et al., 2003). However, limitations also exist. Quicker methods may be less reliable and prone to the need for constant cross-validation and calibration (Stuth et al., 1991). Real-time NIRS has been used to determine pasture nutritive composition (Bell et al., 2018) and quality (Serrano et al., 2020), and when associated with fecal scans, pasture nutritional profiling for grazing animals can be predicted (Dixon and Coates, 2009; Stuth et al., 2002).

Another novel spectroscopic technique is Raman microscopy (or spectroscopy) with laser excitation (Delhaye and Dhamelincourt, 1975; Landsberg and Imandelstam, 1928; Raman and Krishnan, 1928). It is based on measuring the degree of energy loss due to light scattering associated with different chemical molecules’ fingerprints, allowing for their identification in addition to revealing their spatial distribution within the sample at micrometer or nanometer scales. Raman spectroscopy has a wider spatial resolution than NIRS, and it can operate through water and glass (De Gelder et al., 2007). It also requires fewer preparatory steps of the sample, making it faster and easier to use in field conditions. For example, along with chemometric methods, a handheld Raman spectrometer device was used to determine the contents of carbohydrates, fibers, carotenoids, and protein in corn grain as accurate as NIRS’ predictions (Krimmer et al., 2019). Raman spectroscopy has assisted in genotyping high-starch containing potato (Morey et al., 2020) and peanut (Farber et al., 2020) and in identifying the nutritional quality of sweet cream butter of dairy cows fed different total mixed diets (Gómez-Mascaraque et al., 2020).

The literature seems scarce or inexistent about using Raman spectroscopy to determine the nutrient composition of forages, mostly pasture composed of mixed forages such as grasses and legumes. Similarly, there is no concrete resolution on the use of Raman or NIRS to determine PDC. Regardless of the lack of definitive literature on the application of Raman or NIRS to estimate the chemical composition of forages (or any other feedstuff), it is not clear if the precision and accuracy obtained with these techniques and current devices are sufficient to reliably estimate energy consumption, more specifically DE or TDN, to assess energy requirement of grazing and confined animals.

Both physicochemometric analytical methods and spectroscopic techniques provide a snapshot of the feed’s nutrient content. There have been reports that spectroscopic techniques (e.g., NIRS) can adequately assess nutrient contents, i.e., pool sizes, but they cannot reliably and repeatedly estimate abstract measurements such as digestibility (Coates and Dixon, 2011). Furthermore, there are incoherences in using NIRS to determine digestibility between in vivo and in vitro systems (Mahipala et al., 2010). In part, spectroscopic techniques cannot inform rumen fermentation dynamics because many factors, other than feed composition, are involved in the digestion process. They may correctly appraise the pool size, but they reveal nothing about what happens in the ruminant animal’s gastrointestinal tract; unless multiple snapshots are taken throughout the digestive process, which is nearly impossible. Consequently, feeds with identical physicochemical composition may degrade differently in the rumen for reasons not unveiled by wet chemistry procedures or spectroscopic techniques. Schneider and Flatt (1975) go one step further by affirming that “two feeds may be equal in composition and equally digestible, yet one may be more valuable than the other because its digested matter can be used to better advantage by the body.” Alternatively, researchers need to reliably measure the same pool size in the same animal’s feces to compute DE by difference like digestibility trials. However, researchers rarely can track down the same animal under pasture conditions to collect feces, let alone collecting the feces that represent the consumed forage.

Approaches using in vitro techniques as a proxy for in vivo digestibility have been suggested, such as the original or modified two-stage digestion proposed by Tilley and Terry (1963) or the gas production-based techniques originally proposed by Menke et al. (1979). However, limitations also exist, including the lack of confirmation that and when in vitro can represent in vivo digestibility (Tedeschi and Fox, 2020a).

Finally, given our inabilities to definitively and accurately assess the consumption of DE of animals when a digestibility trial cannot be carried out, some have proposed the use of mathematical modeling or empirical predictions to predict DE or TDN given the chemical composition of the diet in addition to other factors (Tedeschi and Fox, 2020a, b), especially for those under grazing conditions (Tedeschi et al., 2019; Woli et al., 2020). In part, the problem arises not only because the DE content is not known with a high degree of certainty but because of inadequate feed intake predictability even though many models for such prediction abound in the literature (Tedeschi et al., 2019). The question persists even when using mathematical models: can the intake of DE be accurately determined so energy partitioning can be estimated to assess animals’ EE at grazing or confined conditions? In other words, how confident can one be in using wet chemistry, NIRS/Raman, or modeling to estimate DE or TDN? This conundrum applies to grazing and confined animals when wet chemistry and digestibility trial cannot be conducted.

Determination of Energy Expenditure

As indicated above, the comparative slaughter technique has been the preferred method to determine the energy and protein requirements of growing animals, whereas indirect calorimetry has been used for lactating dairy cows. Regardless of the technique used to determine animals’ EE, they must follow the first law of thermodynamics in which energy can neither be created nor destroyed but instead transformed from one form to another. Therefore, based on the first law of thermodynamics, we assume the intake of ME is equal to the sum of heat produced and energy retained (or excreted, e.g., milk) by the animal. This assumption is fundamental to any technique used to determine the EE by animals.

In the United States, comprehensive discussions about energy and protein requirements, mostly based on confined animals (Tedeschi, 2019a), have been published by the National Research Council (NRC) and National Academies of Sciences, Engineering, and Medicine (NASEM) for beef cattle from 1945 (NRC, 1945a) to 2016 (NASEM, 2016), for dairy cattle from 1945 (NRC, 1945b) to 2001 (NRC, 2001), for sheep and goats from 1945 (NRC, 1945c) to 2007 (NRC, 2007), as well as other publications (Cannas et al., 2004; Tedeschi and Fox, 2020a, b). Other countries and regions around the globe have followed suit and devised their own set of recommendations to meet their needs and production conditions. Few publications have addressed the grazing animal requirement meticulously as the Australian Nutrient Requirements of Domesticated Ruminants (CSIRO, 1990; 2007).

Comparative Slaughter Technique

The comparative slaughter technique consists of measuring the body energy at the beginning and the end of a feeding trial (usually longer than 30 days), using precisely identical animals to determine their retained energy (RE) and protein (RP). Because no two animals are precisely identical (except for monozygotic twins, genetically speaking), many animals are needed to achieve adequate representation, but unfortunately, hardly attained. Significant concerns have been raised about 1) incoherent relationships between predicted and observed RE and RP, and 2) the calculation logic for kg derived from diet characteristics rather than carcass composition (Tedeschi et al., 2019) besides other inconsistent-prone ancillary calculations that support the comparative slaughter technique (Tedeschi et al., 2017), including adequate sample size. Although the comparative slaughter technique provides a direct assessment of RE and RP, it is an expensive method due to sampling problems; thus, been recommended for small animals (Blaxter, 1967). In the comparative slaughter technique, ME intake and RE are measured, and HP is computed by difference.

Calorimetry Techniques

On the other hand, calorimetry measures the HP produced by animals and computes RE as the difference between ME intake and HP. Two types of calorimetry exist: direct or indirect.

Direct calorimetry techniques measure the heat emission from the calorimeter and its contents or occupants during a period through changes in the water’s temperature surrounding the calorimeter. It relies on the heat balance equation in which heat elimination is the sum of heat lost by radiation, convection, conduction, and evaporation. Direct calorimetry is hardly used to determine an animal’s energy requirement because of its costs and laborious management, making its use even harder for grazing conditions. In animal science, direct calorimetry finds its way in measuring feed’s GE with adiabatic or ballistic bomb calorimeters (McLean and Tobin, 1987).

Indirect calorimetry techniques determine heat production by quantifying the consumption of O₂ and the production of CO₂ and CH₄ and the excretion of urinary N (Gerrits and Labussière, 2015; McLean and Tobin, 1987). Besides the measurements of HP, indirect calorimetry can also assist in nutrient assimilation, thermogenesis, the pathogenesis of obesity and diabetes, and energetics of physical exercise (Ferrannini, 1988). Indirect calorimetry relies on the principle of conservation of energy (i.e., the first law of thermodynamics), and it can be either closed- or open-circuit indirect calorimetry. In the closed-circuit indirect calorimetry, CO₂ is removed from the air inside the respiration chamber using a potassium hydroxide solution, and O₂ is provided from a volumetric flask (McLean and Tobin, 1987). The open-circuit indirect calorimetry measures the volume and gaseous composition of the air coming in and out of the respiration chamber (McLean and Tobin, 1987), and O₂ consumption and production of CO₂ and CH₄ are estimated by difference. The validity of indirect calorimetry relies on the proven assumption that O₂ update by the animal can be estimated by the difference between the inflow and outflow of O₂ (Nienaber et al., 1993). The open-circuit indirect calorimetry is more amenable to determine the EE of grazing animals.

The investigations on the calculations of energy metabolism by a sub-committee on constants and factors appointed at the First Energy Metabolism Symposium in 1958 led to the development of an empirical equation to estimate HP based on an animal’s gaseous exchange from closed-circuit calorimetry (Brouwer, 1965), as shown in Eq. [1]. For open-circuit calorimetry, McLean (1972) proposed Eq. [2], which is based on Eq. [1] after adjusting for the composition of inlet air (20.95% O₂ and 0.03% CO₂).

Where CH₄ is methane production, L; CO₂is carbon dioxide production, L; HP is heat production, kcal; N is urine-N, g; O₂ is oxygen consumption, L; ΔCH₄ is the difference of CH₄ between the inlet and outlet air compositions, L; ΔCO₂ is the difference of CO₂ between the inlet and outlet air compositions, L; and ΔO₂ is the difference of O₂ between inlet and outlet air compositions, L.

There are several techniques based on the open-circuit indirect calorimetry that can be used in determining the EE of grazing animals, including:

1. The portable mask technique, sometimes referred to as mobile indirect calorimetry (Lachica and Aguilera, 2008), uses an airtight facemask fitted with inlet and outlet valves and a flow meter to determine breath volume and composition, and EE is computed based on the differences of gas volumes after adjusting for temperature, pressure, and humidity (Lawrence et al., 1991). Limitations with this technique include 1) slow or lagging O2 sensor responses, 2) water vapor condensation inside the facemask, and 3) animals do not habituate with the facemask and have undesirable behavior, including animal’s panting. Tracheal cannulas have partially replaced portable facemasks (Whitelaw, 1974). However, problems still exist, such as adequate sample size to adequately determine EE and the impact of the technique on the animals’ normal grazing behavior.

2. Portable troughs containing a headstall unit to restrict and control the atmospheric mixing of gases using a gas manifold and sensors to measure the airflow and its composition (e.g., CO₂ and CH₄) have been used in recent years (Zimmerman et al., 2011). The main application of this partial open-circuit indirect calorimetry has been to estimate CH₄ emission of cattle herd (Alemu et al., 2017; Huhtanen et al., 2019; Waghorn et al., 2013) and, more recently, EE via CO₂ output (Caetano et al., 2018). Pending additional validation of this technique on a large scale, the benefits are incredible because it can measure gaseous exchanges of many animals at different times of the day, overcoming the variations of short-term measurements and mimicking the measurements obtained with respiration chambers (Hegarty, 2013).

3. The labeled (or heavy) water technique proposed by Lifson and McClintock (1966) is an isotopic non-radioactive method that administers deuterium (D = ² H) and oxygen-18 (¹⁸O) into an animal and measures the elimination of D₂¹⁸O (heavy water) and C¹⁸O₂. The rate of CO₂ production can be estimated through isotopic enrichment difference between consecutive measures of urine samples. For instance, an animal given a dose of water (D₂¹⁸O) will reduce the specific activity of the O₂ in the body faster than that of D. The rate of C¹⁸O₂ production is computed as the difference in these two rates multiplied by the volume of the total initial body water content (Lawrence et al., 1991). Several adjustments have to be made before estimating the EE (Fancy et al., 1986; Haggarty, 1991; Nagy, 1980) that make this technique laborious, let alone the expenses involved in producing the isotope-enriched water and highly sophisticated analytical techniques needed to determine the isotope concentration.

4. The carbon dioxide entry rate technique was initially used to estimate the EE of grazing sheep (Young et al., 1969) and cattle (Young, 1970). This is an exciting technique that relies on continuous CO₂ production and excretion, leading to a CO₂ pool in the body, and a constant infusion of ¹⁴CO₂ (e.g., 20 nCi/min of NaH¹⁴CO₃ in an aqueous solution) until an equilibrium concentration in the excreted CO₂ is reached. Such equilibrium depends on the rate of infusion of NaH¹⁴CO₃ and the elimination of CO₂ produced endogenously (Lawrence et al., 1991; Whitelaw, 1974). The CO₂ production is estimated by dividing the infusion rate of ¹⁴CO₂by the concentration of 14CO₂ excreted. This technique was originally developed to determine the specific radioactivity of ¹⁴CO₂in the blood (Young et al., 1969) or urine (Havstad and Malechek, 1982), but later saliva withdrawn from the parotid gland using peristaltic pumps were employed (Sahlu et al., 1988; Sanchez and Morris, 1984). Limitations also exist for this technique (Lawrence et al., 1991; Whitelaw, 1974), including the use of a radioactive element (¹⁴C) that has led to the modification of the technique to use its stable non-radioactive isotope (¹³C). Regardless of which isotope to use, there have been reports of overestimating CO₂ production (Havstad and Malechek, 1982; Prieto et al., 1997; Sahlu et al., 1988).

5. The δPDB ¹³C-to¹²C ratio technique uses stable, non-radioactive C isotopes to estimate EE (Carro et al., 1980; Chevalier et al., 1984; Jones and Lefeuvre, 1989) and as a tracer in metabolism trials to study the nutrient utilization by animals (Tyrrell et al., 1984). The CO₂ is analyzed with a mass spectrometer adapted to stable isotope analysis, and the ratio of the masses of ¹³CO₂ and ¹²CO₂ in the sample is compared to the ratio of a standard gas of known ¹³CO₂-to¹²CO₂ ratio. The isotopic composition of the sample is expressed in terms of relative difference (‰) compared to the Pee Dee Belemnite (PDB) limestone, which contains 1.111% ¹³C) standard (Carro et al., 1980). However, different forages have different natural concentrations of ¹³C that may interfere with the δPDB 13C-to12C ratio, leading to incorrect estimates of CO₂ production.

Alternative techniques, which are calibrated with calorimetry under controlled experimentation, can determine HP through different means and provide more feasible and practical determinations of grazing animals' EE, including: 1. The heart rate technique provides a satisfactory correlation between heart rate and EE in free-ranging and wild ruminants (Chabot, 1993; Lawrence et al., 1991; White, 1993), but its main limitation lies in its very application in field animals because it requires calibration for individual animals. 2. The near-infrared (NIR) laser technique is a noninvasive technique based on the NIRS of oxyhemoglobin (Hb-oxy) and deoxyhemoglobin (Hb-deoxy) of tissues (e.g., brain) in situ according to Lambert-Beers law, which defines the relationship between concentration and infrared absorbance. Studies conducted with Cheviot sheep (Takahashi and Eda, 1997) to monitor changes in Hb-oxy and Hb-deoxy at 775, 808, 825, and 800 nm wavelengths reported metabolic rate of the whole body of 75.7 watts at resting and 132.8 postprandial (about 18 min after feeding), which corresponds to 65.1 and 114.2 kcal/h (1 watt = 0.86 kcal/h), respectively. Assuming a shrunk body weight (SBW) of 55 kg, the average NE requirement would be 77.4 kcal/kg^0.75/d. Others have also determined EE in calves using cerebral hemoglobin oxygenation (Pringle et al., 1998b; Pringle et al., 1998a).

Energy Expenditure for Grazing Animals

Grazing systems are the central system of cattle production in tropical and subtropical ecozones, whereas in temperate ecozones, grazing is explored at different seasons and intensities. As aforementioned, the EE for physical activities and forage browsing of grazing animals is primarily the only additional energy cost compared to confined animals, assuming the dietary ME is similar. Therefore, the correct understanding of EE for physical activities and quantifying the nutrient requirements for such animals under diverse conditions is necessary to develop more efficient feeding systems to optimize energy and protein use and assist with supplementation strategies (Tedeschi et al., 2019).

The ARC (1980) developed a factorial approach to estimate the EE (kcal/d) (Eq. [3]) associated with physical activities by assigning coefficients to the number of hours animals spent standing (h/day), the number of body changes (laying down and standing), and horizontal and vertical (ascent) locomotion (km/day). Appendix 1 has the relationships between these locomotion calculations. Assuming the typical values for feedlot and continuous grazing of 12 and 18 h/d, 6 times/d, 0 and 2 km/d, and 0 km/d; respectively, the EE values for animals' physical activity with 300 kg of body weight (BW) are 471.6 and 1024.6 kcal/d, respectively. If the required NE for maintenance is assumed to be 70 kcal/kg^0.75 BW/d, these EE values for physical activity become an additional 9.35 and 20.3%, respectively. That means, the daily required NE for maintenance becomes 1.0935×(70×BW^0.75) and 1.203×(70×BW^0.75) for animals under feedlot and continuous grazing conditions, respectively.

The CSIRO (2007) devised a different approach to compute the EE of grazing animals by adding an assessment of animals' DMI and grazing density (animals per ha), and dry matter digestibility (DMD) and availability of the forage to the animals’ physical activity (horizontal locomotion). While this is a more mechanistic approach to estimate EE of grazing animals, it requires additional information that might be neither available nor measurable. Novel methods and techniques to assess forage availability and quality (González et al., 2018) are promising, and their adoption in these situations will help our understanding of grazing animals’ EE tremendously.

Table 1 has a revised literature data published by Tedeschi and Fox (2020a), which was initially reported by Tedeschi (2001) and Tedeschi and Fox (2015), who provided a detailed description of the studies. The dataset summarizes EE for different grazing ruminant species, using different techniques from 18 studies, and a summary of Israeli studies for grazing Simmental x Hereford beef cows, using the heart rate technique and global positioning system from four studies. Studies are listed in Table 1, and Figure 1 has the boxplot of EEs for basal metabolism, chewing (i.e., eating, ruminating), and locomotion (horizontal, vertical, and ascent). Assuming an animal walking 1 km/d either on the horizontal, vertical, or ascent, Figure 1 provides a relative comparison of EE (kcal/kg^0.75 BW) for these activities on a daily basis. After the basal metabolism, eating would have the second-largest EE for grazing animals, which is likely to hold for confined animals.

The average values for horizontal and vertical locomotion are nearly identical between the two datasets (Table 1), but the horizontal locomotion differs considerably from the values recommended by ARC (1980) and CSIRO (1990). For instance, for the average values of both datasets versus the ARC (1980) and CSIRO (1990) values, a 300-kg cow would require about 33 (300 × 0.11 kcal/kg BW/km) versus 186 (300 × 0.62 kcal/kg BW/km) for horizontal distance walked. The EE values for standing are very similar.

Table 1. Summary of additional energy expenditure for diverse physical activities (1)

For goats, the EE for horizontal walking varied from 0.8 to 0.87 kcal/kg BW/km, for vertical ascent walking was 7.58 kcal/kg BW/km, for eating was 0.43 kcal/kg BW/kg of dry matter (DM) for pelleted food and grain concentrates and from 1.55 to 2.87 kcal/kg BW/kg DM for chopped and long dried forages, respectively (Lachica and Aguilera, 2005). The EE for horizontal walking and vertical ascent walking was similar to that recommended by the ARC (1980) and CSIRO (1990), but seven times greater than that reported in Table 1 by Tedeschi (2001). Amazingly, the cost of eating (kcal/kg BW/kg DM) in goats was at least 8 and 1.4 times more expensive than that reported for cattle and sheep, respectively, consuming similar diets (Lachica and Aguilera, 2005).

The main disadvantage of different techniques to assess EE of grazing animals is that their results vary tremendously (Figure 1), but this is partially explained by the variations in feed quality and amount used and types of animals under different environments in each experiment. Thus, a direct comparison of the results is complicated.

Tedeschi and Fox (2015) and Tedeschi and Fox (2020a) proposed a holistic approach to predict the increase in the NE required for maintenance due to physical activities, eating, and ruminating forages of different quality based on 1) the energy partitioning proposed by the NRC (1981), 2) the basal metabolism proposed by Lofgreen and Garrett (1968) adjusted for physical activity proposed by Fox and Tylutki (1998), 3) Eq. [3] to estimate physical activity, and 4) the data reported by Susenbeth et al. (1997, 1998) to estimate the EE for chewing (i.e., eating and ruminating). Eq. [4] shows the calculation of ME required for maintenance, but because the EE for chewing requires dietary ME at maintenance to compute ME intake at maintenance, it can only be solved iteratively through optimization. Figure 2 depicts a simulation of EE and energy partitioning using Eq. [4].

ANCC2021 - Precision Determination of Energy and Protein Requirements of Grazing and Feedlot Animals - Image 1

The simulation was conducted for animals at 250 and 450 kg SBW, the horizontal locomotion varied from 1,500 to 0 meters per day, and the vertical locomotion varied from 500 to 0 meters per day for dietary ME varying from 1.5 (low-quality forage) to 3.5 (grain-based diet) Mcal/kg, respectively. The logic behind the variable horizontal and vertical locomotion versus dietary ME was that animals would have to graze farther to meet their energy needs on low-quality forages than those on high-quality forages. It was assumed that horizontal and vertical locomotion would decrease 75 and 25 m for each 0.1 increase in dietary ME. Standing and body position changes were assumed to be 12 h/d and 6 times/d, respectively.

Figure 2B shows that the proportion of MEI (i.e., MEmr) that was lost (i.e., heat increment) was identical between the 250- and 450-kg animals because the partial efficiency of use of ME to NE for maintenance (km) was computed from dietary ME, which was identical for both animals. In reality, both km and kg could change differently between these animals if the body composition was considered (Tedeschi et al., 2004). The heat increment varied from 1.84 to 9.16 Mcal/d for the 250-kg animal and 3.1 to 13.9 Mcal/d for the 450-kg animal. However, the proportion of MEI used for chewing was more remarkable for the 250-kg animal (19% to 22%) than for the 450-kg animal (13% to 14%). Similarly, the EE for chewing was more significant for low-quality forage (1.5 Mcal/kg) than grain-based finishing diets (3.5 Mcal/kg DM). The proportion of physical activity (i.e., movement or locomotion) was similar, ranging between 7% and 14%.

Based on this simulation, animals consuming diets containing 1.5 to 1.9 Mcal ME/kg DM (or less) will likely not meet their daily energy needs and may lose BW. In fact, the 1.9 Mcal ME/kg DM is approximately 53% TDN, which is similar to the 50% suggested by Van Soest (1994; Figures 7.8 and 7.9) for cattle to maintain their energy balance.

Challenges in Determining Voluntary Feed Intake

Although the utilization of technologies in animal nutrition has improved significantly in the last decades, the need for reliable prediction methods to estimate voluntary feed intake (VFI) by grazing ruminants still limits a broader application of nutrition models (Galyean, 2020). The utilization of empirical equations adopted by the NRC (2016; 2000) and NASEM (2016) has shown flaws when predicting DMI for grazing animals (Coleman et al., 2014; Lardy et al., 2004). The primary issue with utilizing empirical equations is that the accuracy pre-established might only be achieved when utilized under similar settings. The data utilized to develop the equations originated, which accentuates the problem that pen setting studies will rarely reflect grazing behavior. Predictors such as NDF (Mertens, 1987) and NEm concentration in the feed (NRC, 2016; NASEM, 2000) are somewhat utilized when predicting DMI, which strongly relies on the physical and metabolic mechanisms that seem to control VFI (Poppi et al., 1989).

Even though the physical mechanisms might fit well as a regulatory means for grazing animals while the metabolic or the chemostatic seems to be more applicable in feedlot conditions, mechanistic models that integrate both approaches appear to be more reasonable in explaining variation in VFI (Detmann et al., 2014). Tedeschi and Fox (2020a) illustrated the variability of VFI in two different grazing scenarios, poor-quality forage and a good-quality forage (40% in vitro DMD, 40% NDF, and 9% CP versus 75% in vitro DMD, 65% NDF, and 10% CP, respectively) based on the mechanistic model proposed by Fisher (1996), whose model adopted both physical distention and chemostatic effects to estimate VFI simultaneously. The DMI estimated were 1.6% and 2.4% of BW for the poor- and good-quality forages, respectively. In addition to the lower intake for the poor-quality forage, the model suggests that animals consuming poor-quality forages would have more significant daily variation in the intake than animals consuming good-quality forages. The NASEM (2016) suggested that forage availability and quality must be incorporated in prediction models to estimate VFI for grazing animals. Several studies have shown that ADG and feed intake are more associated with green pasture (i.e., quality; OM available/kg BW) than total forage mass per se (kg DM/ ha) (Minson, 1990).

The use of new technologies to estimate forage availability has been extensively studied in the last few years (Reinermann et al., 2020; Wigley et al., 2019; Woodward et al., 2019), its applicability in real scenarios is still limited (Kallenbach et al., 2020). The influence of forage availability in the eating behavior alters animal activity (Minson, 1990); thus, increasing animals’ NEm requirements (Fox and Tylutki, 1998). A better understanding of forage intake for grazing ruminants seems pivotal to comprehend grazing ruminants’ behavior and how forage quality affects animal activity, contributing to a better EE account and improving energy requirement estimation.

Conclusions

The EE for physical activity and chewing accounts for nearly all the differences between confined and grazing ruminants, and yet, our incomplete understanding of these components keeps rebounding time after time without a more definitive solution. In part, data collection of plant and animal interaction (forage selection, grazing behavior, pasture growth/regrowth, pasture quality, nutrient digestion and absorption, volatile fatty acids production and profile, energy requirement) remains a critical bottleneck for adequate knowledge of forage intake by ruminants (Tedeschi et al., 2019). The majority of the data on EE for ruminants climaxed in the early 1960s towards the mid-1980s, boosted by many open-circuit, indirect calorimetry apparatuses and methods, but none can be tagged as ideal as limitations exist. New data is needed. The scientific community has been under a spell since then, and only recently, in the last five years, we have been invigorated by the advance of sensors (and artificial intelligence) that has allowed us to embark on a gradual crescendo of excitement. We must continue to boost the investment in noninvasive techniques onwards; otherwise, we will once again paralyze our understanding of energy needs by grazing animals, jeopardizing our search for sustainable livestock production, and upsurge our dependability on feedlots to produce meat to satisfy the population demand.

Figure 1. Boxplot of energy expenditure for chewing and locomotion activities using the datasets described in Table 1. The red asterisks indicate the average.

Figure 2. Amount (A; Mcal/d) and relative proportion (B; percentage) of predicted energy required for basal metabolism (H_eE), physical activity (H_jE_pa), eating and ruminating (H_jE_er), and heat increment (H_iE) for animals at 250 and 450 kg of SBW fed diets containing from 1.5 to 3.5 Mcal of ME/kg of DM at maintenance-level intake. It was modified from Tedeschi and Fox (2020a) with permission.

Appendix 1.

Relationship between energy expenditure for horizontal, vertical, and ascent locomotion.

In practice, an animal’s locomotion is either on leveled terrain (horizontal, flat) or sloped terrain (vertical) with a gradient or an inclination angle (α). The main limitation of measuring the energy expenditure (EE) for sloped terrains is that the inclination angle may differ between studies, making it challenging to compare measured EE needed for vertical locomotions. Thus, the EE of ascent locomotion removes the inclination angle as it assumes the differences in heights only to estimate EE.

ANCC2021 - Precision Determination of Energy and Protein Requirements of Grazing and Feedlot Animals - Image 2

Thus, as shown in the Figure above, an animal can go from point 1 (green star) to point 2 (yellow star) by directly walking vertically from 1→2, or walking horizontally from 1→4 and then ascent from 4→2. Given the laws of thermodynamics of conservation of energy, it is assumed that the a×EE_1→2 is equal to the h×EE_1→4 plus s×EE_4→2, assuming that the EE is expressed as energy per distance unit, and a, h, and s are distances. Hence, we have:

The h and s can be computed from the triangle 1-2-4 using trigonometric relationships, as follows:

cos(α) = h/a∴h = / cos(α) x α

sin(α) = s/a∴s = sin(α)xα

Replacing h and s, simplifying the equation, and re-arranging to estimate EE_4→2, we have:

Because EE_1→4 is the same as EE_1→3 as they are expressed as energy per distance, we can substitute EE_1→4 with EE_1→3 to obtain the final equation to compute EE_4→2 with known measurements expressed as energy per distance. Note that EE_4→2 may not be the same as EE_2→4.