Factors that regulate dry matter intake (DMI) by ruminants are complex and not understood fully. Nevertheless, accurate estimates of feed intake are vital to predicting rate of gain and to the application of equations for predicting nutrient requirements of beef cattle, as provided in *Predicting Feed Intake for Food-Producing Animals* (National Research Council, 1987). Previous research has established relationships between dietary energy concentration and DMI by beef cattle based on the concept that consumption of less digestible, low-energy (often high-fiber) diets is controlled by physical factors such as ruminal fill and digesta passage, whereas consumption of highly digestible, high-energy (often low-fiber, high-concentrate) diets is controlled by the animal’s energy demands and by metabolic factors (National Research Council, 1987). This model of intake regulation, however, is not fully compatible with existing data. Ketelaars and Tolkamp (1992a) used data from voluntary intake and digestibility of 831 roughages to evaluate the relationship between organic matter digestibility (OMD) and organic matter intake (OMI). Across a range of 30 to 84 percent OMD, OMI and OMD were related linearly. If intake of highly digestible feeds is regulated by energy demand, OMI (or digestible OMI) would be expected to plateau with increasing OMD. Also difficult to reconcile with the theory that ruminal fill of indigestible residues controls intake are large increases in intake during lactation periods and cold stress and decreases often observed with advancing pregnancy (Ketelaars and Tolkamp, 1992a). This disparity led Tolkamp and Ketelaars (1992) to hypothesize that ruminants do not simply eat as much as they can, but rather eat an amount that will optimize the cost and benefits of oxygen consumption; in effect, ad libitum intake in the model corresponds to the point at which net energy (NE) intake per unit of oxygen consumption is maximized. The approach of Tolkamp and Ketelaars (1992) resulted in accurate predictions of ad libitum intake by roughage-fed sheep. These authors further hypothesized that optimum intake was linked to an optimum metabolic acid load (Ketelaars and Tolkamp, 1992b). Additional research will be needed to develop this hypothesis fully; however, for further discussion of intake regulation theories and comparisons of intake predicted from various models, readers are referred to the thorough review by Mertens (1994).

Because the factors regulating intake by ruminants are not completely understood, models for predicting intake are empirical by nature. Intake prediction equations given in the preceding edition of *Nutrient Requirements of Beef Cattle* (National Research Council, 1984) and in *The Nutrient Requirements of Ruminant Livestock* (Agricultural Research Council, 1980) relate feed intake to dietary energy concentration (NE_{m} and ME, respectively). Based on such equations, energy concentration probably accounts, in part, for effects on feed intake attributed to gastrointestinal fill, energy demands, and potential effects of absorbed nutrients. These equations, however, do not account directly for the numerous physiological, environmental, and management factors that alter feed intake. Clearly, the methods of predicting feed intake described are intended to provide general guidelines. No single, general equation applies in all production situations. Optimally, beef cattle producers should develop intake prediction equations specific to given production situations; such equations should account for a greater percentage of the variation in intake than would be possible with a generalized equation.

Body composition, especially percentage of body fat, seems to affect feed intake (National Research Council, 1987). As animals mature, adipose tissue may, in some way,

Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.

Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter.
Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 85

Nutrient Requirements of Beef Cattle: Seventh Revised Edition, 1996
7
Feed Intake
FACTORS AFFECTING FEED INTAKE
Factors that regulate dry matter intake (DMI) by ruminants are complex and not understood fully. Nevertheless, accurate estimates of feed intake are vital to predicting rate of gain and to the application of equations for predicting nutrient requirements of beef cattle, as provided in Predicting Feed Intake for Food-Producing Animals (National Research Council, 1987). Previous research has established relationships between dietary energy concentration and DMI by beef cattle based on the concept that consumption of less digestible, low-energy (often high-fiber) diets is controlled by physical factors such as ruminal fill and digesta passage, whereas consumption of highly digestible, high-energy (often low-fiber, high-concentrate) diets is controlled by the animal’s energy demands and by metabolic factors (National Research Council, 1987). This model of intake regulation, however, is not fully compatible with existing data. Ketelaars and Tolkamp (1992a) used data from voluntary intake and digestibility of 831 roughages to evaluate the relationship between organic matter digestibility (OMD) and organic matter intake (OMI). Across a range of 30 to 84 percent OMD, OMI and OMD were related linearly. If intake of highly digestible feeds is regulated by energy demand, OMI (or digestible OMI) would be expected to plateau with increasing OMD. Also difficult to reconcile with the theory that ruminal fill of indigestible residues controls intake are large increases in intake during lactation periods and cold stress and decreases often observed with advancing pregnancy (Ketelaars and Tolkamp, 1992a). This disparity led Tolkamp and Ketelaars (1992) to hypothesize that ruminants do not simply eat as much as they can, but rather eat an amount that will optimize the cost and benefits of oxygen consumption; in effect, ad libitum intake in the model corresponds to the point at which net energy (NE) intake per unit of oxygen consumption is maximized. The approach of Tolkamp and Ketelaars (1992) resulted in accurate predictions of ad libitum intake by roughage-fed sheep. These authors further hypothesized that optimum intake was linked to an optimum metabolic acid load (Ketelaars and Tolkamp, 1992b). Additional research will be needed to develop this hypothesis fully; however, for further discussion of intake regulation theories and comparisons of intake predicted from various models, readers are referred to the thorough review by Mertens (1994).
Because the factors regulating intake by ruminants are not completely understood, models for predicting intake are empirical by nature. Intake prediction equations given in the preceding edition of Nutrient Requirements of Beef Cattle (National Research Council, 1984) and in The Nutrient Requirements of Ruminant Livestock (Agricultural Research Council, 1980) relate feed intake to dietary energy concentration (NEm and ME, respectively). Based on such equations, energy concentration probably accounts, in part, for effects on feed intake attributed to gastrointestinal fill, energy demands, and potential effects of absorbed nutrients. These equations, however, do not account directly for the numerous physiological, environmental, and management factors that alter feed intake. Clearly, the methods of predicting feed intake described are intended to provide general guidelines. No single, general equation applies in all production situations. Optimally, beef cattle producers should develop intake prediction equations specific to given production situations; such equations should account for a greater percentage of the variation in intake than would be possible with a generalized equation.
Physiological Factors Affecting Feed Intake
Body composition, especially percentage of body fat, seems to affect feed intake (National Research Council, 1987). As animals mature, adipose tissue may, in some way,

OCR for page 85

Nutrient Requirements of Beef Cattle: Seventh Revised Edition, 1996
have a feedback role in controlling feed intake (National Research Council, 1987). Regardless of the mechanism, the percentage of body fat is often considered in equations to predict feed intake by beef cattle. Fox et al. (1988) suggested that DMI decreases 2.7 percent for each 1 percent increase in body fat over the range of 21.3 to 31.5 percent body fat. As a result of the relationship between feed intake and body fat, careful monitoring of feed intake can be a useful management tool to determine when cattle have reached appropriate slaughter condition.
Sex (steer vs heifer) seems to have limited effects on feed intake (Agricultural Research Council, 1980; National Research Council, 1987). Intake differences attributable to sex may be evident at certain times; Ingvartsen et al. (1992a) reported that at body weights (BW) less than 250 kg, heifers had greater intake capacity than steers or bulls. In the previous edition of Nutrient Requirements of Beef Cattle (National Research Council, 1984), the Subcommittee on Beef Cattle Nutrition suggested that predicted DMI should be decreased by 10 percent for medium-framed heifers. At a given BW, heifers are proportionally more mature (fatter) than steers; hence, Fox et al. (1988) in their equation for predicting DMI use a frame-equivalent weight adjustment instead of a direct adjustment for sex.
The age of an animal when it is placed on feed can affect feed intake. Older animals (e.g., yearlings vs calves) typically consume more feed per unit BW than younger ones. Presumably, the greater ratio of age to body weight (age relative to proportion of mature body composition) for yearling cattle prompts greater feed intake. This effect has been likened to increased feed intake by cattle experiencing compensatory growth (National Research Council, 1987). Assuming that cattle started on feed at heavier BW are generally older cattle, age-related effects on feed intake are partly responsible for the positive relationship between initial weight on feed and DMI (National Research Council, 1987). The 1984 subcommittee (National Research Council, 1984) and Fox et al. (1988) suggested a 10 percent increase in predicted DMI by cattle started on feed as yearlings compared with cattle started on feed as calves. Before more accurate predictions of feed intake are possible, designed studies are needed in which independent effects of age and body weight or body composition on feed intake can be quantified.
The animal’s physiological state can markedly alter feed intake. Lactating animals can increase feed intake by 35 to 50 percent compared with that of nonlactating animals of the same BW fed the same diet (Agricultural Research Council, 1980). For forages, Minson (1990) reported a mean increase in DMI of 30 percent during lactation. Based on data from dairy cows, the Agricultural Research Council (ARC) (1980) and National Research Council (NRC) (1987) reports suggested that DMI increases by 0.2 kg/kg fat-corrected milk. Hence, beef cows bred for greater milk-producing ability would be expected to have greater feed intakes per unit BW during lactation. Advancing pregnancy has an adverse affect on feed intake, most notably during the last month (Agricultural Research Council, 1980; National Research Council, 1987). Ingvartsen et al. (1992a) noted a 1.5 percent decrease per week during the last 14 weeks of pregnancy in Danish Black and White heifers fed diets predominantly of roughage; this value agrees fairly well with the decrease of 2 percent per week during the last month of pregnancy suggested in the NRC (1987) report.
Frame size varies considerably in beef cattle. The 1984 NRC subcommittee (National Research Council, 1984) factored frame size into intake predictions, whereas Fox et al. (1988) suggested predictions could be adjusted by scaling frame sizes to an equivalent mature weight (frame-equivalent weight). However, Holstein and Holstein×beef crosses may consume more feed relative to body weight than beef breeds (National Research Council, 1987). Fox et al. (1988) suggested that intake predictions should be increased 8 percent for Holsteins and 4 percent for Holstein×British breed crosses relative to British-breed cattle. In addition to possible breed-specific effects, in the NRC (1987) report it was noted that genetic selection for feed efficiency could produce animals with increased feed intake potential, suggesting that genetic potential for growth (or increased production demands) may affect feed intake.
Environmental Factors Affecting Feed Intake
Considerable research has been conducted to evaluate effects of ambient temperature on feed intake and digestive function, and the topic has been reviewed extensively (Kennedy et al., 1986; Minton, 1986; Young, 1986; Young et al., 1989). In experimental situations, feed intake has been shown to increase as the temperature falls below the thermoneutral zone and decrease above that zone. With cold stress, ruminal motility and digesta passage increase before changes in intake occur, prompting Kennedy et al. (1986) to conclude that the digestive tract response may be essential to accommodating greater feed intake. As noted by Young (1986), however, this general response to temperature can vary with thermal susceptibility of the animal, acclimation, and diet. Behavioral responses to thermal stress (e.g., decreased grazing time) are restricted by some experimental conditions that could heighten the effects of thermal stress on feed intake. For example, acute cold stress decreased forage intake by as much as 47 percent in grazing cattle (Adams, 1987); however, for thermally adapted grazing cows, Beverlin et al. (1989) reported only small changes in forage intake with temperature deviations of 8° to -16° C. Similarly, feed intake by confined beef cattle fed finishing diets did not generally increase during

OCR for page 85

Nutrient Requirements of Beef Cattle: Seventh Revised Edition, 1996
cold stress and was often less during winter than during other seasons (Stanton, 1995).
Other adverse environmental conditions (wind, precipitation, mud, and so on) can accentuate the effects of ambient temperature. Fox et al. (1988) suggested multiplicative correction factors to adjust intake predictions for various environmental effects. Duration of adverse conditions seems important, and because effects caused by environmental conditions are variable, feed intake in a variable environment is difficult to predict (National Research Council, 1987). Regardless of the variable nature of its effects, thermal stress can markedly alter energetic efficiency of ruminants as evidenced by the effects of cold stress on energy utilization by beef cattle (Delfino and Mathison, 1991).
Seasonal or photoperiod (day length) effects on feed intake are understood less fully than are thermal effects, and photoperiod has been suggested as a potentially important factor influencing feed intake by beef cattle (National Research Council, 1987). Ingvartsen et al. (1992b) evaluated effects of day length on voluntary DMI capacity of Danish Black and White bulls, steers, and heifers. Voluntary DMI increased 0.32 percent per hour increase in day length; the range in the literature reviewed by the authors was -0.6 to 1.5 percent. Based on the deviation from the voluntary intake at 12 hours of daylight, voluntary intake would be expected to be 1.5 to 2 percent greater in long-day months (July in the northern hemisphere) and 1.5 to 2 percent less in short-day months (January). Hicks et al. (1990) grouped intake data into four seasons and thereby accounted for much of the seasonal pattern in feed intake. Nevertheless, temperature, photoperiod, animal, and perhaps management differences contribute to seasonal patterns, and separate effects are difficult to delineate.
Management and Dietary Factors Affecting Feed Intake
With grazing cattle, quantity of forage available can affect feed intake. The authors of the NRC (1987) report reviewed data summarized by Rayburn (1986) and concluded that grazed forage intake was maximized when forage availability was approximately 2,250 kg dry matter/ha or a forage allowance of 40 g organic matter/kg BW. Intake decreased rapidly to 60 percent of maximum when forage allowance was 20 g organic matter/kg BW (450 kg/ha; National Research Council, 1987). Minson (1990) noted that bite size decreased with forage mass of less than 2,000 kg dry matter/ha; this decrease was only partially compensated for by increased grazing time, resulting in decreased forage intake. The break point at which intake of grazed forage was decreased with decreasing forage allowance seemed to lie between 30 and 50 g dry matter/kg BW. Relationships may vary with forage type and sward structure. McCollum et al. (1992) evaluated effects of forage availability on cattle grazing annual winter wheat pasture and noted that peak intake of digestible organic matter was predicted at 1,247 kg dry matter/ha or an allowance of approximately 300 g dry matter/kg BW. The data base for determining the relationship between forage availability and forage intake is derived largely from studies with actively growing pastures. As noted by Minson (1990), gain by sheep is related more closely to green (growing) forage allowance than to total forage dry matter offered. Similarly, Bird et al. (1989) reported that body weight gain by grazing cattle could be modeled more effectively from green pasture mass than from total pasture mass. Selective grazing of growing forage may increase in pastures with both growing and senescent material. Cattle eat only small amounts of senescent forage when some growing forage is available (Minson, 1990). Hence, effects of forage availability on intake should be considered in light of pasture composition and the potential for selective grazing.
Growth-promoting implants tend to increase feed intake. In two trials with beef steers fed a 60 percent concentrate diet, administering an estradiol benzoate/progesterone implant increased DMI from 4 to 16 percent, depending on when the implant was administered relative to slaughter (Rumsey et al., 1992). Fox et al. (1988) suggested that predicted feed intake should be decreased 8 percent for nonimplanted cattle.
Monensin, the ionophore feed additive, typically decreases feed intake. Fox et al. (1988) suggested that feed intake decreases by 10 and 6 percent with 33 and 22 mg monensin/kg diet respectively. With beef steers fed a 90 percent concentrate diet, Galyean et al. (1992) noted a 4 percent decrease in feed intake when animals were fed 31 mg monensin/kg dietary dry matter. Lasalocid, another ionophore approved for use in beef cattle, seems to have limited effects on feed intake. Fox et al. (1988) suggested that feed intake is decreased 2 percent by lasalocid, regardless of dietary concentration. Malcolm et al. (1992) found that feed intake increased approximately 4 percent with 85 percent concentrate diets that contained 33 mg lasalocid/kg diet compared with a nonionophore, control diet. Fewer data are available regarding effects of laidlomycin propionate, an ionophore approved for confined growing and finishing cattle, on feed intake. However, a summary of available data (Vogel, 1995) suggests that laidlomycin propionate has minimal effect on feed intake.
A dietary nutrient deficiency, particularly protein, can decrease feed intake. With low-nitrogen, high-fiber forage, nitrogen deficiency is common, and provision of supplemental nitrogen often increases DMI substantially (Galyean and Goetsch, 1993). Forage intake responses to protein are most typical when forage crude protein content is less than 6 to 8 percent (National Research Council, 1987). Supplementing forages with grain-based concentrates often decreases forage intake, such effects typically being

OCR for page 85

Nutrient Requirements of Beef Cattle: Seventh Revised Edition, 1996
greater with high- than with low-quality forages (Galyean and Goetsch, 1993).
Grinding feeds can affect intake, but effects depend on the type of feed. With forages, fine grinding can increase intake, presumably through effects on digesta passage (Galyean and Goetsch, 1993). With concentrates, fine grinding often decreases feed intake. Adjustments to intake predictions for finely processed diets as a function of dietary NEm concentration have been suggested (National Research Council, 1987). Fermentation of feeds by ensiling generally has little effect on DMI unless the silage is unusually wet or dry and undesirable fermentation has occurred (National Research Council, 1987). Intake of wilted grass silages is usually greater than that of direct-cut silage, but reasons for the decrease with direct-cut silages are not fully understood (Minson, 1990).
PREDICTION OF FEED INTAKE BY BEEF CATTLE
The approach used to develop prediction equations for feed intake involved reevaluating relationships suggested in the previous edition of Nutrient Requirements of Beef Cattle (National Research Council, 1984). Equations presented in the previous edition have been used extensively in practice; however, description of the data base used and statistical validation of the equations were inadequate. Hence, efforts will be made to fully describe the approach used to develop prediction equations for growing and finishing cattle and beef cows. No attempt was made to develop prediction equations for intake by nursing calves; readers are referred to Predicting Feed Intake for Food-Producing Animals (National Research Council, 1987) for a proposed equation. It also should be noted that the focus of prediction in each case was average DMI over an extended feeding period. Although prediction of feed intake for shorter periods is highly desirable, no data base exists from which to develop such prediction equations for the wide variety of production situations and feeds available to beef cattle producers.
Growing and Finishing Cattle: Dietary Energy Concentration
As noted previously, the Nutrient Requirements of Beef Cattle (National Research Council, 1984) provided an equation to predict DMI by growing and finishing beef cattle. This equation describes DMI as a function of dietary NEm concentration, with adjustments for frame size or sex. The base NRC 1984 equation is
Eq. 7-a
where DMI is expressed in kg/day, SBW is expressed in kg, and NEm concentration is expressed as Mcal/kg dietary dry matter. Data from the published literature were used to reevaluate the relationship between dietary NEm concentration and DMI by growing and finishing beef cattle (Figure 7–1).
Data were obtained from experiments conducted with growing and finishing beef cattle and published in the Journal of Animal Science from 1980 to 1992. Each of 185 data points extracted from the literature represented a treatment mean for average DMI throughout a feeding period. Feeding periods varied from 56 to 212 days. Approximately 48 percent of the cattle were implanted with a growth-promoting implant, and approximately 50 percent were fed an ionophore. Information on frame size (small, medium, or large), sex (steer, heifer, or bull), age (calf or yearling), and initial and final SBW was recorded. Because this data contained a mix of full and shrunk body weights, the subcommittee assumed SBW in developing these equations. Dietary NEm concentration was calculated from tabular values (National Research Council, 1984); however, actually determined NEm values were used, when available. Because of the limited number of observations, bulls were classed as large-frame steers and large-frame heifers were classed as medium-framed yearling heifers. Total NEm intake was calculated as the product of DMI and dietary NEm concentration. Total NEm intake was then divided by average metabolic body weight (average SBW0.75 in kg). The intake of NEm per unit SBW0.75 was analyzed by stepwise regression procedures (SAS Institute, Inc., 1987) with dietary NEm concentration, NEm2, length of the feeding period, and dummy variables used to account for effects of sex and frame classes as possible independent selections.
The relationship between NEm intake per unit SBW0.75
FIGURE 7–1 Relationship of dietary NEm concentration to NEm intake by beef cattle. Data points were obtained from published literature and represent treatment means for average intake during a feeding period.

OCR for page 85

Nutrient Requirements of Beef Cattle: Seventh Revised Edition, 1996
and dietary NEm concentration is shown in Figure 7–1. A regression equation that included NEm, NEm2, and an intercept adjustment for yearling cattle accounted for 69.87 percent of the variation in NEm intake per unit SBW0.75. Expressed as total NEm intake (Mcal/day), this equation is
Eq. 7–1
The intercept adjustment terms for medium-framed yearling steers and medium-framed yearling heifers differed slightly, but the standard errors of these adjustments overlapped. Hence, the mean value of the two intercept adjustments was used, resulting in one intercept term for both yearling steers and heifers of -0.0869 instead of -0.1128. DMI (kg/day) can be calculated from Eq. 7–1 by dividing total NEm intake (Mcal/day) by dietary NEm concentration (Mcal/kg).
DMI predicted from Eq. 7–1 and from Eq. 7-a were regressed on actual DMI for the 185 data points. The intake predicted from Eq. 7-a accounted for 62.35 percent of the variation in DMI, with a bias of - 2.2 percent (under prediction). DMI predicted from Eq. 7–1 accounted for 72.85 percent of the variation in actual DMI, with a bias of -1.86 percent.
A comparison of the DMI predicted from the Eq. 7-a (with adjustments for frame size) and Eq. 7–1 is shown in Figure 7–2. In this example, DMI was predicted for a 410-kg average SBW, medium-frame steer (300 and 520 kg initial and final SBW, respectively) over a range in NEm concentrations of 1 to 2.35 Mcal/kg. At low dietary NEm concentrations, both equations yielded similar estimates of DMI. Eq. 7–1 predicted lesser intakes in the middle of the energy range and greater intakes at the upper end of the energy range than did Eq. 7-a.
FIGURE 7–2 Comparison of dry matter intake predictions for a medium-frame (410-kg average BW) steer using Eq. 7-a (National Research Council, 1984) and Eq. 7–1, the equation developed from a literature data set.
As noted previously, Eq. 7–1 was developed to predict average DMI throughout a feeding period. Hence, the SBW term in the equation would be calculated as the average of initial and final SBW for a feeding period on a given diet. In practice, one would generally know the initial SBW and project the final SBW for the feeding period (e.g., estimated SBW at low-choice grade).
Because feed intake can vary greatly with environmental conditions, management factors, cattle type, and dietary factors, any equation should be viewed as providing a guideline rather than an absolute prediction of intake. Feedlot managers, nutritionists, and beef producers should combine such guidelines with their own data bases to develop more accurate predictions for specific situations. Hicks et al. (1990) reported that inclusion of feed intake data from the early portion of the feeding period (days 8 to 28) increased the coefficient of determination for prediction of mean DMI. Similarly, Oltjen and Owens (1987) used a statistical technique to adjust subsequent intake predictions for intake earlier in the feeding period. As noted by Hicks et al. (1990), it may be possible to use intake data obtained early in the feeding period of cattle to detect groups of cattle with particularly low or high feed intakes and thereby initiate appropriate management actions.
Growing and Finishing Cattle: Initial Weight on Feed
As discussed earlier, several factors other than dietary energy concentration can affect feed intake. Combined with data from cattle fed mostly high-energy diets, initial weight on feed seems to have predictive value (National Research Council, 1987). Hence, the relationship between initial body weight and DMI was evaluated in data obtained from the published literature. In addition, data from commercial feedlots were used to evaluate the relationship within a narrower range of dietary energy concentrations.
The data used in a preliminary analysis were the 185 data points described in the preceding section on dietary energy concentration. Dietary NEm concentration ranged from slightly less than 1.0 to approximately 2.4 Mcal/kg. DMI (kg/day) was analyzed by stepwise regression procedures (SAS, Institute, Inc., 1987) with initial BW and dummy variables used to adjust the intercept and slope for effects of sex and frame classes as possible independent selections. The relationship between initial BW (kg) and DMI (kg/day) for the 185 data points taken from the literature is shown in Figure 7–3. Initial weight, with adjustments to the intercept for certain frame size/sex/age classes accounted for 59.78 percent of the variation in DMI. The equation is
Eq. 7–2
where iBW is initial BW in kg. For large-frame steer calves,

OCR for page 85

Nutrient Requirements of Beef Cattle: Seventh Revised Edition, 1996
FIGURE 7–3 Relationship between initial BW and DMI for growing and finishing cattle. Data points were obtained from published literature and represent treatment means for average intake during a feeding period.
the intercept was 2.477, whereas for large-frame heifer calves and medium-frame yearling heifers, the intercept was 3.212. For medium-frame yearling steers, the intercept was 3.616. These equations were used to predict DMI, and predicted intake was regressed on actual intake for the 185 data points. Predicted intake accounted for 57.82 percent of the variation in actual DMI, with a bias of -2.1 percent (under prediction). As noted earlier, the intake equation in the preceding edition of this report (National Research Council, 1984) accounted for 62.35 percent of the variation in actual intake, with a bias of -2.2 percent.
The published data used to evaluate the relationship of initial BW to DMI covered a wide range in NEm concentrations. In an effort to examine the relationship within a narrower range of energy concentrations that might be typical of beef feedlots, the subcommittee used initial weight and DMI data obtained from commercial feedlots. The first set of commercial data was collected from feedlots in Texas, Arizona, and California and included 929 pen means for DMI by crossbred steers and heifers. Average initial weight of cattle in this data set ranged from approximately 76 to 454 kg. Most cattle in this data set had some degree of Brahman breeding. The second data set included 732 pen means for DMI by crossbred steers collected from one feedlot in Kansas. Initial weight of cattle for this second data set ranged from to 201 to 528 kg. The degree of Brahman breeding was minimal in this second data set. Diets fed in both data sets were typical growing and finishing diets (NEg ranged from approximately 1.1 to 1.59 Mcal/kg). Cattle in the first data set were typically on feed longer than those in the Kansas data set, and, as a result, lower energy growing diets made up a greater proportion of the DMI than in the Kansas data set. For both commercial data sets, simple linear regression equations were developed with initial BW as an independent variable to predict DMI. Results are shown in Table 7–1. For the first data set, which included both steer and heifer data, sex was not a significant factor, so the overall equation is presented.
The similarity of the relationship between initial weight and DMI in these two sets of commercial feedlot data are somewhat remarkable. The slope of both equations in Table 7–1 differs somewhat from the slope derived from the preliminary analysis of the literature data set, which might reflect the narrower range in dietary energy concentrations in the commercial data sets. For simplicity, the average values for the two equations shown in Table 7–1 can be used for a general prediction equation based on initial BW. Hence, Eq. 7–2 is revised as
Eq. 7–2
where iBW is initial BW in kg. As with Eq. 7–1 described previously, it should be noted that Eq. 7–2 is designed to predict average feed intake throughout a feeding period.
These results suggest that initial BW when cattle are started on feed is related linearly to average DMI during a feeding period. This finding confirms previous research (National Research Council, 1987); thus feedlot managers and nutritionists should be able to use their own data bases to derive equations to predict DMI from initial BW. Other factors, like management, environment, and cattle type could be factored into such equations for individual production situations. Although Eq. 7–2 could be useful in practice, as noted previously, no single equation is likely to be effective in all production situations.
Validation of Prediction Equations
Three data sets were used as independent tests of Eq. 7–1, Eq. 7–2, and Eq. 7-a (with frame size adjustments). The first data set came from Cornell University (D.G. Fox, Cornell University, personal communication, 1995) (54 data points; average DMI by small-, medium-, and large-framed steers and heifers; NEm [Mcal/kg] ranged from approximately 1.4 to 2.1; length of the feeding period was 100 days or longer). This data set was used to test the equations with diets in the middle-to-upper range of dietary NEm concentrations. The second data set came from the University of Guelph, Ontario, Canada (J.G.
TABLE 7–1 Relationship Between Initial Weight on Feed and Dry Matter Intake by Beef Cattle in Two Sets of Commercial Feedlot Data
Data Set
Intercept
Slope
Sy·x
r2
A
4.4498
0.01081
0.6217
0.571
B
4.6346
0.01422
0.6048
0.452
NOTE: Data set A was collected from commercial feedlots in Texas, Arizona, and California, and included both steer and heifer data (n=929). Data set B was collected from one feedlot in Kansas and included only steer data (n=732).

OCR for page 85

Nutrient Requirements of Beef Cattle: Seventh Revised Edition, 1996
Buchanan-Smith, personal communication, 1995) (38 data points; average DMI by medium- and large-frame steers and heifers fed mostly alfalfa/grass silage-based diets; length of the feeding period ranged from 16 to 24 weeks; NEm [Mcal/kg] ranged from 1.12 to 1.95). This second data set was used to evaluate the equations in the lower-to-middle range of dietary NEm concentrations. The third data set was taken from a summary of intake and digestion data compiled at the University of Alberta (Mathison et al., 1986). This data set included 139 observations with beef cattle fed all-forage diets. Dietary NEm concentrations were calculated from the reported ME values of the diet (range in NEm of 0.69 to 1.71 Mcal/kg). Grasses, legumes, and grass-legume mixtures, as well as crop residues, were included in the data set.
For each of the data sets, DMI predicted from the three equations was regressed on actual DMI. The r2, Sy·x, and bias for each prediction equation are shown in Table 7–2. Bias was also calculated by fitting the model with a forced intercept of 0 and expressing the deviation of the slope from this model as a percentage change from an ideal value of 1.0. Initial SBW data were not available for the Alberta data set.
For the Cornell data set, Eq. 7–1 accounted for approximately the same percentage of variation in actual DMI as Eq. 7-a but had less overprediction bias (Table 7–2). The rather simple Eq. 7–2 accounted for approximately 55 percent of the variation in actual DMI but tended to underestimate intake. With the Guelph data set, Eq. 7-a (with adjustments for frame size) and Eq. 7–1 yielded similar results (Table 7–2). Once again, however, the over-prediction bias of the NRC 1984 equation, Eq. 7-a, was corrected by Eq. 7–1. If frame adjustments were not made to the NRC 1984 predictions, the r2 was 80.1 percent with a bias of 0.1 percent. Hence, the tendency for overprediction noted in this data set with Eq. 7-a was most likely a function of
TABLE 7–2 Results of Regressing Predicted Dry Matter Intake on Actual Dry Matter Intake by Growing and Finishing Beef Cattle for Three Validation Data Sets
Data Seta
Equationb
Observations, n
r2
Sy·x
Bias, %c
Cornell
7–1
54
0.7647
0.3431
+0.16
7–2
54
0.5481
0.3559
-6.49
7-a (NRC, 1984)
54
0.7624
0.5498
+5.88
Guelph
7–1
38
0.7930
0.3731
-0.49
7–2
38
0.3529
0.3330
+4.54
7-a (NRC, 1984)
38
0.7827
0.5581
+8.34
Alberta
7–1
139
0.3078
0.7144
-8.40
7-a (NRC, 1984)
139
0.3102
0.7028
-7.90
aSee text for description of the data sets.
bEq. 7–1=NEm intake=BW0.75 * (0.02435 * NEm-0.0466 * NEm2-0.1128; Eq. 7–2=DMI=4.54+0.0125 * iBW; and Eq. 7-a=DMI=BW0.75 * (0.1493 * NEm-0.046 * NEm2-0.0196).
cBias was calculated as the percentage deviation of the slope from a theoretical value of 1.0 when the predicted DMI was regressed on actual DMI with a zero-intercept model.
use of the frame size adjustments. Eq. 7–2 accounted for approximately 35 percent of the variation in DMI and, in contrast to results with the Cornell data set, tended to overpredict DMI. Both Eq. 7-a and Eq. 7–1 yielded similar results when applied to the Alberta data set, accounting for approximately 30 percent of the variation in actual DMI and underpredicting DMI by approximately 8 percent.
Results of these independent tests were in agreement with the comparison of Eq. 7-a and Eq. 7–1 shown in Figure 7–2. The Guelph and Alberta data sets represented a range in dietary NEm concentrations for which both equations predict similar DMI, whereas the Cornell data set included NEm concentrations in the range for which predictions from the two equations are most divergent. Further testing of Eq. 7–1 with independent data sets will be required to determine whether it is a superior predictive tool than Eq. 7-a. For the three independent data sets evaluated, Eq. 7–1 seemed to decrease slightly the overprediction bias of Eq. 7-a. Evaluation of these data sets affirms the validity of the intake prediction equation, but also raises questions about the value of the suggested frame-size adjustments, in the National Research Council (1984) report.
The failure of both Eq. 7–1 and Eq. 7-a to accurately predict DMI of beef cattle fed all-forage diets (Alberta data set) raises some concerns. Specific considerations for all-forage diets will be dealt with in a subsequent section.
Adjustments to Predictions
Fox et al. (1988, 1992) reported on various factors that can affect feed intake, factors that can be used to adjust feed intake predictions of Eqs. 7–1 and 7–2 and Eq. 7-a. Some caution should be applied in making these adjustments, however, because of the possibility of double accounting. For example, the data base used to derive equations to predict intake includes intake data from cattle under a variety of management systems and an array of environmental conditions. Hence, the equations derived from the data base developed by this subcommittee may reflect partial adjustments for many of the factors suggested by Fox et al. (1988, 1992).
Three specific adjustments need to be addressed. First, as noted previously, approximately 50 percent of the 185 data points used to develop Eq. 7–1 represented cases in which cattle were fed an ionophore. Statistical evaluation of these data, however, suggested no basis for adjustments to intake predictions as a result of ionophore use. Nonetheless, based on field experience, this subcommittee believes considerable evidence suggests that monensin will typically decrease feed intake, whereas lasalocid and laidlomycin propionate have little effect on feed intake. As a result, the subcommittee suggests that predicted DMI be decreased by 4 percent if monensin is fed at concentrations

OCR for page 85

Nutrient Requirements of Beef Cattle: Seventh Revised Edition, 1996
of 27.5 to 33 mg/kg dietary dry matter and that predicted DMI not be adjusted when lasalocid or laidlomycin propionate are added to the diet.
The second case relates to adjustments for use or nonuse of growth-promoting implants. As with ionophores, statistical evaluation of the 185 data points indicated no basis for adjustments to predicted DMI if growth-promoting implants were used. On the other hand, considerable research and field evidence suggest that such implants increase feed intake. Hence, the subcommittee suggests that values suggested by Fox et al. (1992) be used as a guideline for adjustments to predicted DMI in cases where implants are not used (e.g., 6 percent decrease in predicted DMI when implants are not used).
The third case deals with effects of forage allowance. Data presented in Predicting Feed Intake of Food-Producing Animals (National Research Council, 1987) relative to forage availability were reevaluated by Rayburn (1992). He constructed a quadratic regression of relative DMI on available forage mass. The resulting regression equation was
where FM is available forage mass =1,150 kg/ha. The FM value of 1,150 kg/ha represents the maxima of the quadratic equation (first derivative), and relative DMI is assumed to be 100 percent for FM greater than this maxima. For application to grazing situations, the subcommittee suggests that this relationship be used in two steps. First, the daily forage allowance (FA) should be determined
where grazing unit is the pasture size in hectares and SBW is in kg. If FM is =1,150 kg/ha, or FA is four times the predicted DMI (expressed as g/kg SBW), no adjustment should be made to the predicted DMI. If neither of these conditions is true, relative DMI should be calculated from the equation shown above, and the predicted DMI should be multiplied by the relative DMI (expressed as a decimal) to adjust predicted DMI for the effects of limited FM. This adjustment procedure should be applied to all types of grazing systems; however, rotational or other intensive grazing systems with heavy stocking rates will result in more rapid changes in FM than continuous systems with lower stocking rates. This necessitates careful attention to FM in intensive grazing systems and frequent reevaluation of relative DMI.
BEEF COWS: DIETARY ENERGY CONCENTRATION
The preceding edition of Nutrient Requirements of Beef Cattle (National Research Council, 1984) includes an equation for feed intake by breeding beef females similar in form to an equation for growing and finishing beef cattle; DMI is described as a function of SBW0.75, and linear and quadratic effects of dietary NEm concentration (DMI, kg/day=SBW0.75 * [0.1462 * NEm-0.0517 * NEm2-0.0074]). As with the growing and finishing equation, the description of how this equation was developed was inadequate in that publication. Predicting Feed Intake of Food-Producing Animals (National Research Council, 1987) provides an alternative equation for beef cows that described DMI as a linear function of dietary NEm concentration:
Eq. 7-b
To further evaluate the relationship between dietary NEm concentration and intake by beef cows, an approach similar to that described previously for growing and finishing cattle was used. Treatment means for DMI were obtained from a variety of sources. Data were obtained from articles in the Journal of Animal Science (1979 through 1993), unpublished theses, and unpublished data that were solicited from individual scientists. The 153 data points used in the analysis represented treatment or breed×year means for DMI by nonpregnant beef cows or by cows during the middle and last one-third of pregnancy. As with growing and finishing beef cattle data, the beef cow data base contained a mix of full and shrunk body weights; the subcommittee assumed SBW in developing these equations. The data base was not sufficiently detailed to allow incorporation of information about body condition scores or frame sizes of the cows; and for some data points, only information on dietary NEm concentration and DMI per unit SBW0.75 was available. Dietary NEm concentration (range=0.76 to 2.08 Mcal/kg) was taken from the data source or calculated based on tabular values (National Research Council, 1984) for feeds. Total NEm intake was calculated as the product of DMI and dietary NEm concentration and expressed per unit SBW0.75 (average SBW0.75 during the intake measurement period). Data were then subjected to stepwise regression analysis (SAS Institute, Inc., 1987), with dummy variables included to account for the specific physiological stage of the cow.
It should be noted that data points were not included in the regression analysis when an obvious nutrient deficiency existed. This exclusion primarily impacted data points from beef cows fed low-quality forages that were deficient in crude protein. In such cases, only data from protein-supplemented cows were included in the data set. Hence, the resulting equation would not be applicable when the user wants to predict intake of a protein-deficient forage. Alternatively, the resulting equation would be applicable when the user wants to estimate total intake (e.g., forage plus supplement).
The relationship between dietary NEm concentration

OCR for page 85

Nutrient Requirements of Beef Cattle: Seventh Revised Edition, 1996
and NEm intake is depicted in Figure 7–4. In contrast to the quadratic relationship noted for growing and finishing beef cattle (Figure 7–1), intake of NEm by beef cows was relatively linear with dietary NEm concentration. The regression equation that provided the best fit to the data included NEm2 and an intercept adjustment for nonpregnant cows. For pregnant cows,
Eq. 7–3
the intercept for nonpregnant cows=0.03840. Eq. 7–3 accounted for 75.94 percent of the variation in NEm intake. When Eq. 7–3 was used to predict DMI per unit SBW0.75 for the 153 data points, the r2 was 15.47 percent with a prediction bias of -2.2 percent. The relatively low r2 resulted from the fact that a large proportion of the data points for middle-to-late pregnancy (breed×year means obtained from Pfennig, 1992) were for cows fed diets with a narrow range in dietary NEm concentration (approximately 1.15 to 1.4 Mcal/kg). Compared with Eq. 7–3, Eq. 7-a for breeding females accounted for only 0.99 percent of the variation in actual DMI with a bias of -10 percent. Intake predicted from the NRC 1987 equation, Eq. 7-b, for breeding females accounted for 12.06 percent of the variation in actual DMI with a bias of -10.3 percent. The greater similarity in predictions between Eq. 7–3 and Eq. 7-b vs Eq. 7-a may reflect the fact that data points used to construct Eq. 7-b were included in the data set used to derive Eq. 7–3. Overall, these results seem to indicate that Eq. 7–3 provided a superior fit to these data than either Eq. 7-a or Eq. 7-b.
Predicted DMI by a 500-kg cow fed diets with varying NEm concentration for Eq. 7–3, Eq. 7-a, and Eq. 7-b is
FIGURE 7–4 Relationship of dietary NEm concentration to NEm intake by beef cows (nonpregnant, and middle and last third of pregnancy). Data points were obtained from published and unpublished literature and represent treatment and breed×year means for average intake during a feeding period.
shown in Figure 7–5. Compared with Eq. 7-b, Eq. 7–3 predicted greater intakes at lower NEm concentrations and lesser intakes at higher NEm concentrations.
As with Eq. 7–1 for growing and finishing beef cattle, DMI is calculated from Eq. 7–3 by dividing the predicted NEm intake by dietary NEm concentration. Because of the mathematical form of this equation, predicted DMI will increase substantially for NEm values less than approximately 0.95 Mcal/kg. This increase in predicted DMI results from division by a fraction and is not biologically realistic. Based on results that will be described in a subsequent section on all-forage diets, the subcommittee recommends that for feeds with NEm concentrations less than 1.0 Mcal/kg, the user apply Eqs. 7-a or 7-b for breeding females, or, with Eq. 7–3, use a constant value of 0.95 for the NEm concentration of the diet. The subcommittee further suggests that adjustments to predicted intake for effects of ionophores, implants, available forage mass, and other adjustments suggested by Fox et al. (1992) for growing and finishing cattle also be applied to intake predictions for beef cows.
VALIDATION OF THE BEEF COW EQUATION
Beef cows are not typically given ad libitum access to feed in production situations. As a result, obtaining data for both development and validation of intake prediction equations is difficult. In contrast to the equations derived for growing and finishing beef cattle, only one fully independent data set was available for validation of the beef cow equation. This data set, supplied by R.H.Pritchard (South Dakota State University, personal communication, 1995) included 36 pen observations of DMI by nonpregnant beef cows fed a high-concentrate diet (NEm=2.06 Mcal/kg). Cows were either implanted (Finaplix-H) or not
FIGURE 7–5 Comparison of dry matter intake (DMI) by a 500-kg, pregnant beef cow predicted from Eqs. 7-a (National Research Council, 1984) and 7-b (National Research Council, 1987) with that predicted from Eq. 7–3.

OCR for page 85

Nutrient Requirements of Beef Cattle: Seventh Revised Edition, 1996
implanted and were classed in thin or nonthin body condition categories. Monensin was fed at 29.5 mg/kg diet. Cows were serially slaughtered, such that length of the feeding period ranged from 44 to 100 days.
Eq. 7–3 for nonpregnant beef cows, and Eqs. 7-a and 7-b for breeding females, were used to predict DMI. Predicted DMI was then regressed on actual DMI for all three equations. The r2 values for Eq. 7–3 and Eqs. 7-a and 7-b were identical (36.25 percent), reflecting the fact that all cows were fed the same diet; only average SBW differed among observations. Bias was +7.57 percent for Eq. 7–3, -34.3 percent for Eq. 7-a, and +16.49 percent for Eq. 7-b. If predicted intakes for Eq. 7–3 were decreased by 4 percent for feeding monensin, the bias was +3.26 percent. Hence, in this particular set of validation data, Eq. 7-a for breeding females grossly underpredicted DMI, whereas Eq. 7-b overpredicted DMI, a situation that was partially corrected for use of Eq. 7–3. Further testing of these equations with independent data sets is desirable and is required to determine their relative predictive value.
SPECIAL CONSIDERATIONS FOR ALL-FORAGE DIETS
As noted previously, validation tests of Eq. 7–1 and Eq. 7-a, intake prediction equations for growing and finishing beef cattle, indicated that neither equation yielded accurate predictions of DMI by beef cattle fed all-forage diets in the Alberta (Mathison et al., 1986) data set. Because forages constitute all or most of the diet in many production situations, an accurate prediction equation for all-forage diets is critical to practical application of nutrient requirement data. Consequently, the Alberta data set was used to determine whether a specific equation for all-forage diets could be developed that would provide more accurate predictions of DMI by growing and finishing cattle and beef cows than either Eq. 7–1, Eq. 7–3, or Eqs. 7-a and 7-b. The Alberta data set consisted of 139 observations of ad libitum DMI by beef cattle consuming forages in three classes: grasses (65), legumes (39), and grass/legume mixtures (35). After first determining that SBW0.75 accounted for significant (P <0.0001) variation in DMI, DMI expressed as kg/kg of SBW0.75 was evaluated by stepwise regression with dietary crude protein (CP), neutral detergent fiber (NDF), and acid detergent fiber (ADF) concentrations, dietary ME concentration, and intercept adjustment terms for forage class. The resulting best-fit equation included terms for CP and ADF and an intercept adjustment term for grass/legume mixtures. Because the intercept adjustment term resulted in only a slight increase in r2 and a slight decrease in Sy·x, this term was deleted from the model, yielding the following equation (hereafter referred to as CP__ADF):
Eq. CP__ADF
where percentages of CP and ADF in the forage are expressed on a DMI basis. The r2 for this equation was 39.31 percent with an Sy·x of 0.0149. When this equation was used to predict DMI (kg/day) in the Alberta data set, and predicted intake was regressed on actual intake for the 139 data points, r2 was 57.13 percent, Sy·x was 0.7501, and the bias was -2.93 percent.
The predicted vs actual DMI values for Eq. CP__ADF derived from the Alberta data set represent a considerable improvement relative to Eq. 7-a and Eq. 7–1 for this data set (see Table 7–2). Nonetheless, conclusions relative to the merit of various equations must be based on independent validation tests. Hence, two validation data sets were used to compare predicted vs actual intake among equations. The first data set was obtained from experiments with grazing beef steers and heifers. Data from Funk et al. (1987), Krysl et al. (1987), Pordomingo et al. (1991), Gunter (1993), and Gunter et al. (1993) were compiled and used to test Eq. CP__ADF developed from the Alberta data set, Eq. 7–1, and Eq. 7-a. Cattle in these experiments freely grazed native rangelands, and intake of organic matter was determined by marker-based methods. Dietary in vitro organic matter disappearance (IVOMD) was used to calculate dietary NEm concentration, assuming that IVOMD was equal to total digestible nutrient TDN and that 1 kg TDN was equal to 3.62 Mcal ME (National Research Council, 1984). Calculated NEm values for this data set ranged from 0.88 to 1.74 Mcal/kg.
The second data set was derived from the experiment of Vona et al. (1984), in which beef cows were fed warm-season grass hays. Dietary NEm concentration was calculated from in vivo DM digestibility in the same manner as described for IVOMD in the first data set. One data point with an extremely low DM digestibility (30.3 percent) relative to DMI was deleted from this data set. Calculated NEm concentrations for this data set ranged from 0.76 to 1.78 Mcal/kg. Because two of the predicted NEm values were less than 1.0 Mcal/kg, DMI was calculated from Eq. 7–3 by using a constant divisor of 0.95 for those two data points. Eqs. CP__ADF, 7-a, and 7-b for breeding females also were used to predict DMI for this data set.
Results for the two validation data sets are shown in Table 7–3. For the grazing steer and heifer data, Eq. CP__ADF accounted for less variation than either Eq. 7-a or Eq. 7–1. All three equations underpredicted DMI; however, underprediction bias was lowest for Eq. 7–1 and higher for Eq. CP__ADF. Standard errors of prediction were least for Eq. 7-a and Eq. CP__ADF and greatest for Eq. 7–1. For the beef cow data of Vona et al. (1984), Eq. 7-a accounted for considerably less variation in DMI than the other equations. Prediction bias was greatest for Eq. CP__ADF, and similar among Eqs. 7-a and 7-b and Eq.

OCR for page 85

Nutrient Requirements of Beef Cattle: Seventh Revised Edition, 1996
TABLE 7–3 Results of Regressing Predicted Dry Matter Intake on Actual Dry Matter Intake for Two Validation Data Sets with Growing Beef Steers and Heifers and Beef Cows Fed All-Forage Diets
Data Seta
Equation
Observations, n
r2
Sy·x
Bias, %b
Steers/heifers
CP__ADFc
38
0.4750
1.384
-9.71
7–1d
38
0.5997
1.673
-0.93
7-a (NRC, 1984)e
38
0.6800
1.322
-5.49
Cows
CP__ADFc
34
0.4357
0.6721
-10.83
7–3f
34
0.5049
0.8005
+2.12
7-a (NRC, 1984)e
34
0.1101
0.6496
-3.65
7-b (NRC, 1987)g
34
0.6203
1.0536
-0.75
aSee text for description of the data sets.
bBias was calculated as the percentage deviation of the slope from a theorectical value of 1.0 when the predicted DMI was regressed on actual DMI with a zero-intercept model.
cCP__ADF=Mathison et al. (1986).
dEq. 7–1=NEm intake (Mcal/day)=BW0.75 * (0.02435 * NEm-0.0466 * NEm2-0.1128.
eEq. 7-a=the equation for either growing and finishing beef cattle or breeding females (National Research Council, 1984).
fEq. 7–3 and for forages with calculated NEm values of less than 0.95 Mcal/kg, a constant value of .95 was used as the divisor to calulate DMI from NEm intake predicted by Eq. 7–3.
gEq. 7-b=the equation for breeding females (National Research Council, 1987).
7–3. Standard error of prediction was least for Eq. 7-a and greatest for Eq. 7-b.
Neither of these two validation data sets is optimal. For the steer and heifer data set, the use of marker-based estimates of intake, and organic matter intake and digestibility rather than DM-based values, no doubt introduced some bias. For the cow data set, some caution should be used in interpreting the validation tests because Eq. 7-b was actually derived from this data set, and all the observations from this data set were included in the 153 data points used to develop Eq. 7–3. Despite these caveats, the validation tests indicate that Eqs. 7–1 and 7–3 and Eq. 7-a for growing and finishing cattle and breeding females generally yield estimates of DMI that are similar to those predicted from an empirical equation based on CP and ADF concentrations of the forage. Perhaps the similarity in predictions from these different approaches reflects the fairly high correlation between dietary energy metabolizability and fiber (NDF) concentration (Mertens, 1994). Further research is needed to develop more accurate means of predicting intake by beef cattle fed all-forage diets; but until such equations or models are developed, this subcommittee concludes that reasonable estimates of DMI can be obtained from Eqs. 7–1 and 7–3, as well as Eqs. 7-a and 7-b.
REFERENCES
Adams, D.C. 1987. Influence of winter weather on range livestock. Pp. 23–29 in Proceedings of the Grazing Livestock Nutrition Conference. Laramie: University of Wyoming.
Agricultural Research Council. 1980. The Nutrient Requirements of Ruminant Livestock . Technical Review. Farnham Royal, U.K.: Commonwealth Agricultural Bureaux.
Beverlin, S.K., K.M.Havstad, E.L.Ayers, and M.K.Petersen. 1989. Forage intake responses to winter cold exposure of free-ranging beef cows. Appl. Anim. Behav. Sci. 23:75–85.
Bird, P.R., M.J.Watson, and J.W.D.Cayley. 1989. Effect of stocking rate, season and pasture characteristics on liveweight gain of beef steers grazing perennial pastures. Aust. J. Agric. Res. 40:1277–1291.
Delfino, J.G., and G.W.Mathison. 1991. Effects of cold environment and intake level on the energetic efficiency of feedlot steers. J. Anim. Sci. 69:4577–4587.
Fox, D.G., C.J.Sniffen, and J.D.O’Connor. 1988. Adjusting nutrient requirements of beef cattle for animal and environmental variations. J. Anim. Sci. 66:1475–1495.
Fox, D.G., C.J.Sniffen, J.D.O’Connor, J.B.Russell, and P.J.Van Soest. 1992. A net carbohydrate and protein system for evaluating cattle diets. III. Cattle requirements and diet adequacy. J. Anim. Sci. 70:3578–3596.
Funk, M.A., M.L.Galyean, M.E.Branine, and L.J.Krysl. 1987. Steers grazing blue grama rangeland throughout the growing season. I. Dietary composition, intake, digesta kinetics and ruminal fermentation. J. Anim. Sci. 65:1342–1353.
Galyean, M.L., and A.L.Goetsch. 1993. Utilization of forage fiber by ruminants. Pp. 33–71 in Forage Cell Wall Structure and Digestibility, H.G.Jung, D.R.Buxton, R.D.Hatfield, and J.Ralph, eds. Madison, Wis.: ASA-CSSA-SSSA.
Galyean, M.L., K.J.Malcolm, and G.C.Duff. 1992. Performance of feedlot steers fed diets containing laidlomycin propionate or monensin plus tylosin, and effects of laidlomycin propionate concentration on intake patterns and ruminal fermentation in beef steers during adaptation to a high-concentrate diet. J. Anim. Sci. 70:2950–2958.
Gunter, S.A. 1993. Nutrient Intake and Digestion by Cattle Grazing Midgrass Prairie Rangeland and Plains Bluestem Pasture. Ph.D. dissertation, Oklahoma State University, Stillwater, Oklahoma.
Gunter, S.A., F.T.McCollum III, R.L.Gillen, and L.J.Krysl. 1993. Forage intake and digestion by cattle grazing midgrass prairie rangeland or sideoats grama/sweetclover pasture. J. Anim. Sci. 71:3432–3441.
Hicks, R.B., F.N.Owens, D R.Gill, J.W.Oltjen, and R.P.Lake. 1990. Dry matter intake by feedlot beef steers: Influence of initial weight, time on feed and season of the year received in yard. J. Anim. Sci. 68:254–265.
Ingvartsen, K.L., H.R.Andersen, and J.Foldager. 1992a. Effect of sex and pregnancy on feed intake capacity of growing cattle. Acta Agric. Scand. (Sect. A). 42:40–46.
Ingvartsen, K.L., H.R.Andersen, and J.Foldager. 1992b. Random variation in voluntary dry matter intake and the effect of day length on feed intake capacity in growing cattle. Acta Agric. Scand. (Sect. A). 42:121–126.
Kennedy, P.M., R.J.Christopherson, and L.P.Milligan. 1986. Digestive responses to cold. Pp. 285–306 in Control of Digestion and Metabolism in Ruminants, L.P.Milligan, W.L.Grovum, and A.Dobson, eds. Englewood Cliffs, N.J.: Prentice-Hall.
Ketelaars, J.J.M.H., and B.J.Tolkamp. 1992a. Toward a new theory of feed intake regulation in ruminants 1. Causes of differences in voluntary feed intake: Critique of current views. Livestock Prod. Sci. 30:269–296.
Ketelaars, J.J.M.H., and B.J.Tolkamp. 1992b. Toward a new theory of feed intake regulation in ruminants 3. Optimum feed intake: In search of a physiological background. Livestock Prod. Sci. 31:235–258.
Krysl, L.J., M.L.Galyean, M.B.Judkins, M.E.Branine, and R.E.Estell. 1987. Digestive physiology of steers grazing fertilized and nonfertilized blue grama rangeland. J. Range Manage. 40:493–501.
Malcolm, K.J., M.E.Branine, and M.L.Galyean. 1992. Effects of

OCR for page 85

Nutrient Requirements of Beef Cattle: Seventh Revised Edition, 1996
ionophore management programs on performance of feedlot cattle. Agri-Practice 13:7–16.
Mathison, G.W., L.P.Milligan, and R.D.Weisenburger. 1986. Ruminant feed evaluation unit: Nutritive value of some Alberta feeds for cattle and sheep. Pp. 55–57 in The 65th Annual Feeders’ Day Report: Agriculture and Forestry Bulletin, Special Issue. Edmonton: University of Alberta, Canada.
McCollum, F.T., M.D.Cravey, S.A.Gunter, J.M.Mieres, P.B.Beck, R.San Julian, and G.W.Horn. 1992. Forage availability affects wheat forage intake by stocker cattle. Oklahoma Agric. Exp. Sta. MP-136:312–318.
Mertens, D.R. 1994. Regulation of forage intake. Pp. 450–493 in Forage Quality, Evaluation, and Utilization, G.C.Fahey, ed. Madison, Wis.: ASA-CSSA-SSSA.
Minson, D.J. 1990. Forage in Ruminant Nutrition. San Diego: Academic.
Minton, J.E. 1986. Effects of heat stress on feed intake of beef cattle. Pp. 325–327 in Symposium Proceedings: Feed Intake by Beef Cattle, MP-121, F.N.Owens, ed. Stillwater, Okla.: Oklahoma Agricultural Experiment Station.
National Research Council. 1984. Nutrient Requirements of Beef Cattle, Sixth Rev. Ed. Washington, D.C.: National Academy Press.
National Research Council. 1987. Predicting Feed Intake of Food-Producing Animals. Washington, D.C.: National Academy Press.
Oltjen, J.W., and F.N.Owens. 1987. Beef cattle feed intake and growth: Empirical Bayes derivation of the Kalman filter applied to a nonlinear dynamic model. J. Anim. Sci. 65:1362–1370.
Pfennig, M. 1992. Factors influencing feed intake in beef cow-calf pairs. M.S. thesis, Purdue University, West Lafayette, Indiana.
Pordomingo, A.J., J.D.Wallace, A.S.Freeman, and M.L.Galyean. 1991. Supplemental corn grain for steers grazing native rangeland during summer. J. Anim. Sci. 69:1678–1687.
Rayburn, E.B. 1986. Quantitative aspects of pasture management. Seneca Trail RC&D Technical Manual. Franklinville, N.Y.: Seneca Trail RC&D.
Rayburn, E.B. 1992. Modeling the effect of forage availability on the forage intake of grazing cattle. Paper presented at the American Society of Agronomy Northeastern Branch Meetings, University of Connecticut, Storrs, Conn., June 28–July 1, 1992.
Rumsey, T.S., A.C.Hammond, and J.P.McMurtry. 1992. Response to reimplanting beef steers with estradiol benzoate and progesterone: Performance, implant absorption pattern, and thyroxine status. J. Anim. Sci. 70:995–1001.
SAS Institute, Inc. 1987. Procedures Guide for Personal Computers, Version 6 Ed. Cary, N.C.: SAS Institute Inc.
Stanton, T.L. 1995. Damage control strategies for cattle exposed to cold stress. Pp. 289–298 in Symposium: Intake by Feedlot Cattle, P-942 F.N.Owens, ed. Stillwater: Oklahoma Agric. Exp. Sta.
Tolkamp, B.J., and J.J.M.H.Ketelaars. 1992. Toward a new theory of feed intake regulation in ruminants 2. Costs and benefits of feed consumption: An optimization approach. Livestock Prod. Sci. 30:297–317.
Vogel, G. 1995. The effect of ionophores on feed intake by feedlot cattle. Pp. 281–288 in Symposium: Intake by Feedlot Cattle, F.N.Owens, ed. Stillwater: Oklahoma Agricultural Experimental Station.
Vona, L.C., G.A.Jung, R.L.Reid, and W.C.Sharp. 1984. Nutritive value of warm-season grass hays for beef cattle and sheep: Digestiblity, intake and mineral utilization. J. Anim. Sci. 59:1582–1593.
Young, B.A. 1986. Food intake of cattle in cold climates. Pp. 328–340 in Symposium Proceedings: Feed Intake by Beef Cattle, MP-121, F.N.Owen, ed. Stillwater: Oklahoma Agricultural Experiment Station.
Young, B.A., B.Walker, A.E.Dixon, and V.A.Walker. 1989. Physiological adaptation to the environment. J. Anim. Sci. 67:2426–2432.