J. Anim Sci. 2008. 86:2680-2689. doi:10.2527/jas.2007-0561
© 2008 American Society of Animal Science
Performance by feedlot steers and heifers: Daily gain, mature body weight, dry matter intake, and dietary energetics
R. A. Zinn1,
A. Barreras2,
F. N. Owens3 and
A. Plascencia2
University of California, Davis 95616
 |
Abstract
|
|---|
Performance, DMI, diet composition, and slaughter data from 9,683 pens of steers and 5,009 pens of heifers that were fed high-concentrate diets for 90 d or more were obtained from 15 feedlots from the western United States and Canada. The data set included pen means for more than 3.1 million cattle fed between 1998 and 2004. Performance measurements assessed included ADG, DMI, dietary NE, shrunk initial weight (SIW), and shrunk final weight. Mature final weight (MFW) for cattle in each pen was estimated based on regression of slaughter weight against SIW and ADG across all pens. Equations were developed to standardize performance projections (ADG, MFW, and break-even values) and analyze feedlot cattle close-outs. Generally, as diet NE concentration increased, DMI was decreased but G:F, dressing percentage, and yield grade all increased. Pens of cattle with greater SIW had greater ADG, DMI, and shrunk final weight but a lower G:F and dressing percentage. Dressing percentage and yield grade were correlated positively. Equations of the NRC relating gain to NE intake explained 85 and 80% of the variation in DMI of steers and heifers, respectively, with mean ratios of predicted to observed DMI (DMIratio) at 1.000 ± 0.0506 and 0.974 ± 0.0490. However, a significant (P < 0.001) bias in the NRC estimate of DMI was detected (r2 = 0.10 and 0.05, for steers and heifers) between the DMIratio and ADG in which DMIratio increased as ADG increased. This was due to inherent confounding of ADG and MFW in the original NE equation of Lofgreen and Garrett. Based on iterative optimization to minimize the difference between expected and observed DMI, revised equations for retained energy (RE, Mcal/kg) were developed for steers and for heifers: REsteer = 0.0606 x (LW x 478/MFWsteer)0.75ADG0.905; REheifer = 0.0618 x (LW x 478/MFWheifer)0.75ADG0.905, where LW = mean shrunk live weight. The revised equations decreased the SD of the DMIratio by 5.4% (from 0.0496 to 0.0469) and eliminated the bias in DMIratio that was related to ADG (r2 = 0.0006). The similarity between the 2 equations derived for steers and for heifers for estimation of RE from ADG supports the concept that scaling by MFW accounts for energy utilization differences between sexes.
Key Words: energetics • feedlot • heifer • performance • steer
|
INTRODUCTION
|
|---|
The California Net Energy System (Lofgreen and Garrett, 1968
) partitions energy requirements into 2 factors, viz., maintenance and growth, yielding a relationship of the form: energy intake, Mcal/d = a x W0.75 + b x W0.75 x ADGc, where W = shrunk BW (full BW x 0.96; NRC, 1984) and ADG = daily shrunk BW gain. As discussed by NRC (1996), the growth coefficient b is discrete based on the sex of an animal as well as body composition that varies with age, frame size, or mature BW. To linearize the growth coefficient across BW and frame sizes, Fox and Black (1984) proposed the concept of equivalent BW (i.e., the BW in which cattle that differ in frame size have similar body composition). Tylutki et al. (1994) expanded this concept to include both age and sex effects; equivalent shrunk weight (EQSW) was considered equal to shrunk weight times shrunk reference weight divided by mature final weight (MFW), shrunk reference weight being an arbitrary value (467 kg, Tylutki et al., 1994; 478 kg, NRC, 1996) and MFW being the weight of the animal at chemical maturity (28% empty body fat). This concept condensed multiple equations for estimating energy retention of feedlot cattle into a single equation: 0.0557 x EQSW0.75 x ADG1.097, where EQSW = shrunk weight x 478/MFW. Unfortunately, MFW is not readily predicted a priori (e.g., at arrival of cattle to a feedlot). Consequently, the first objective of this study was to predict slaughter weight and other feedlot performance measurements from initial shrunk weight of arriving cattle based on thousands of pen records from commercial feedlots. Second, because a bias in gain predicted from NE intake was detected, the relationship of feedlot gain to NE intake was employed to derive revised coefficients for the NE equations for steers and heifers. Precision and bias of the revised equations were appraised using an independent data set.
|
MATERIALS AND METHODS
|
|---|
Animal Care and Use Committee approval was not obtained for this study because no animals were used.
Net energy equations serve as the critical lynchpin between intake of feed (or NE) and retained energy (RE, performance). When NE intake is known, performance can be predicted accurately from NE equations. Conversely, if performance is known, the amount of NE required (and feed intake for a diet with any specified NE content) can be calculated from NE equations. The validity and bias of any NE equation can be determined and appraised using feedlot performance data. The predicted DMI for a pen of cattle must precisely match the observed DMI if the NE equation is accurate. Basic to such an evaluation, certain points must be fixed. These include 1) NE values for diets, 2) the relationship between NEg and NEm, 3) the NEm requirement for maintenance, and 4) the relationship of performance (rate of gain) to retained energy. For our evaluation, we employed the NEg values for feeds specified by NRC (1996) that were based largely upon measurements of TDN. We used the equation NEg = 0.877 NEm – 0.41 (Zinn et al., 1996) to relate NEm to NEg, we assumed that the NE requirement (EM, Mcal/d) was 0.077 LW0.75 (Lofgreen and Garrett, 1968), and we related RE to rate of gain based on the NRC (1996) equations. Alteration in any of these factors will alter the veracity of a NE equation. Accordingly, performance, DMI, diet composition, and slaughter data from 9,683 pens of steers and 5,009 pens of heifers that housed 30 or more cattle and were fed for 90 d or more were obtained from 15 feedlots from various locations in the western United States and Canada (Alberta, California, Colorado, Idaho, Kansas, Nebraska, and Texas). The data set consisted of pen means from a total of more than 3.1 million cattle fed between 1998 and 2004. All cattle were fed diets rich in concentrates. Consequently, derived relationships may not apply to cattle fed diets containing more roughage. All steers had received a growth-stimulating ear implant, and all diets contained an ionophore.
Initial live purchase BW and BW at slaughter both were decreased by 4% (pencil shrink) to account for digestive tract fill; these BW were employed as initial shrunk weight (SIW) and final shrunk weight (SFW) for all calculations except in the very rare instances when arrival BW exceeded 96% of purchase BW. In those cases, the arrival (off-truck) BW rather than 96% of purchase BW was assumed to equal SIW. Because MFW (e.g., weight at 28% empty body fat) is not available for commercial feedlot cattle, we assumed that MFW for each set of feedlot cattle was equivalent to SFW. The decision to market each pen of cattle is an independent decision targeted to optimize economic return. If marketing is current, the percentage of empty body fat will be relatively constant to maximize carcass value. When cattle are slaughtered before maturity or when MFW is not known, accuracy of predicting SFW (and thereby MFW) from SIW was appraised as will be discussed later. Mean feeding weight (LW) was calculated as the mean of SIW and SFW. Dietary NEm estimates for cattle in each individual pen was based on diet composition and tabular NEm values (NRC, 1996) for individual feed ingredients in the diet. The amount of diet DM (kg/d) required for maintenance for each pen of cattle was calculated as the sum of the average energy need of an animal for maintenance divided by dietary NEm. The amount of diet DM (kg/d) required for gain was the average energy requirement of an animal for growth divided by dietary NEg. The requirement for gain was calculated from equations proposed by various NRC committees (NRC, 1984, 1996). To match the NRC (1984) estimates, the NEg need for steers was calculated from the NRC (1984) equation for medium frame steers (NEg Calf = 0.0557 LW0.75ADG1.097, NEg Yearling = 0.0493 LW0.75ADG1.097); steers with SIW less than 273 kg were classified arbitrarily as calves. The NEg need for heifers was calculated from the NRC (1984) equation for medium frame heifers (NEg Calf = 0.0686 LW0.75ADG1.119, NEg Yearling = 0.0608 LW0.75ADG1.119); heifers with SIW less than 250 kg were classified arbitrarily as calves. To match the NRC (1996) estimates, NEg need was calculated from the NRC (1996) equation [NEg = 0.0557 (LW x 478/MFW)0.75ADG1.097] assuming that MFW was equivalent to SFW. The total DMI (kg/d) required for each pen of cattle then was calculated as the sum of the daily DM required for maintenance plus the daily DM required for growth by these 2 different NE systems; calculated DMI was compared with the observed DMI for each pen of cattle. If calculated DMI fails to match observed DMI across BW, sexes, and rates of gain, this indicates that some segment of the NE relationship is erroneous. Relative to devising energy values for feeds or altering maintenance energy requirements, adjusting the equation relating energy intake to rate of gain is the most direct and applicable solution.
The precision of predictions was surprisingly high, with the ratio of expected to observed DMI (DMIratio) averaging 1.011 (SD = 0.064) and 0.993 (SD = 0.051) for NRC (1984) and NRC (1996), respectively. For individual feedlots, this ratio ranged from 0.97 to 1.03 for NRC (1984) and 0.97 to 1.02 for NRC (1996). Close average agreement of observed with expected DMI confirms that NE equations when combined with NE values for feeds will predict gain accurately, for the average animal at the average performance level. This is not a check on the accuracy of NRC equations for sets of cattle that deviate from the mean. Because ratios differed among feedlots for unidentified reasons, DMI for each pen of cattle in each feedlot was adjusted (using a single multiplier for each feedlot) so that the ratio of observed:expected DMI for each feedlot averaged 1.00 based on NRC (1996). For example, if the ratio of observed:expected DMI for a specific feedlot averaged 0.9800, DMI for each pen in that feedlot were multiplied by 1.02. Performing this adjustment decreased the CV for DMIratio based on NRC (1996) by 5.5%.
Statistical Analysis
For analytical purposes, each individual pen was considered to be an independent observation. Data were analyzed by means of the following linear mixed model: Yij = B0 +B1 Xij + B2X2ij + ti + bi1 Xij + bi2 X2ij + eij, where: i = 1, 2,..., 15 feedlot; j = 1, 2,..., ni observations within trial i; Yij = observed Y values being modeled (final BW or ADG variables). The equation B0 +B1 Xij + B2X2ij is the fixed effect component of the model; ti + bi1 Xij + bi2 X2ij + eij is the random effect component of the model, where [ti bi1 bi2]' ~iid N [(0 0 0) ', their variance- covariance matrix], eij ~iid N (0, 
). The feedlot was entered as a random effect in the linear mixed model. Full models containing linear and quadratic terms were decreased by removing terms that did not contribute significantly to the model. Dependent variables were not weighted. Estimation for all models was carried out using the method of REML, assuming a variance-covariance structure among the regressors. The MIXED procedure (SAS Inst. Inc., Cary, NC) was used in the analysis. The variance-covariance components were subjected to a test of hypothesis using the option COVTEST. The resulting covariances between random parameters were different from zero (P < 0.01). Therefore, different variance-covariance structures were used in the TYPE option of the RANDOM statement of PROC MIXED. Final model was selected based on Akaike’s and Schwarz’s criteria (Littell et al., 1996). Adjusted Y observations (predicted Y values + residuals) for mixed model effects were used to calculate coefficients of determination (R2) by regression of Y-adjusted dependent variables against the respective observed independent variables (St-Pierre, 2001). Statistical computation was carried out using the REG procedure of SAS. Performance measurements of greatest interest were MFW, ADG, DMI, and dietary NE. Using the above model, prediction equations were developed for performance projections (ADG, MFW, and break-even) and to analyze feedlot cattle pen closeouts.
|
RESULTS AND DISCUSSION
|
|---|
Means and ranges of measurements for pens of steers and heifers are presented in Tables 1 and 2, respectively. Simple Pearson correlations between individual measurements for these pens are presented in Tables 3 and 4, respectively. Generally, as NE content of the diet increased, DMI decreased but G:F, dressing percentage, and yield grade all increased. Pens of cattle with greater initial BW had greater ADG, DMI, and final weight but a lower dressing percentage. As DMI of the pen increased, G:F and dressing percentage decreased; this presumably reflects the expected difference between calves and yearlings. Dressing percentage and yield grade were positively correlated. As will be shown presently, the effect of initial BW on performance, including slaughter weight, DMI, ADG, and G:F, is sizable, giving emphasis to the importance of stratifying cattle used in performance trials based on initial BW. This illustrates one complexity for correctly interpreting results of progeny tests when cattle differ in SIW.
View this table:
[in this window]
[in a new window]
|
Table 1. Pen means and SD for feedlot steer closeouts used for generation of growth-performance and NE evaluations
|
|
Steer Data and Relationships
The NRC (1984, 1996) equations are based on cattle performance, and these specific equations explained 81.8 and 85.1% of the variation in DMI of steers, respectively. Corresponding ratios of predicted vs. observed DMI were 1.020 ± 0.0631 and 1.000 ± 0.0506, respectively. Based on NRC (1984) equations, a small but appreciable (r2 = 0.0263; P < 0.001) bias in prediction of DMI was associated with SIW; SIW did not affect (r2 < 0.001) DMIratio calculated from NRC (1996) equations. However, both NRC (1984) and NRC (1996) had a pronounced bias (r2 = 0.269 and 0.103, respectively; P < 0.001) in the relationship between ADG and DMIratio. In the case of NRC (1984), this bias was expected due to the arbitrary application of 2 discrete growth coefficients (medium frame calf or yearling). However, in the case of NRC (1996), a significant bias also was detected (DMIratio = 0.882 + 0.0760 ADG, P < 0.001). This bias, indicating that predicted divided by observed DMI varied significantly with ADG, shows either 1) that pens of cattle gaining slowly are obtaining more NE value from their feed than pens of cattle gaining rapidly, an unwieldy premise for predicting performance from NE intake, or 2) that the NE requirement for either maintenance or gain predicted by the NRC (1996) is excessive for pens of commercial feedlot cattle gaining slowly. In testing the NRC (1996) predictions of DMI, we generalized that final slaughter weight was equivalent to MFW. Had cattle feeders marketed their cattle based on target final BW rather than standard maturity or fatness (i.e., 1.3-cm backfat or yield grade 2.8), then cattle that gain faster could be slaughtered less chemically mature (leaner) than average, and vice versa.
Any underestimation of MFW increases apparent energy retention and hence the DMIratio. To test this consideration, the data were filtered to remove all pens where average yield grade was less than 2.7 or greater than 2.9. In the resulting data set (1,270 pen closeouts; average yield grade 2.81 ± 0.06), the observed DMIratio for NRC (1984) and NRC (1996) were 1.027 ± 0.068 and 1.005 ± 0.051, respectively. Again, for both NRC (1984) and NRC (1996), a pronounced bias (r2 = 0.383 and 0.222, respectively) based on DMIratio associated with ADG was evident. Consequently, even when cattle are slaughtered at a similar degree of finish, these equations apparently were not adjusting appropriately for ADG.
When dietary energy density is not limiting and feed access is not limited, rate of BW gain should increase as mature BW increases (Taylor, 1980; Emmans, 1997). In developing the NE system, Lofgreen and Garrett (1968) did not include this principle. Their equation was based on the simple relationship between ADG and RE per unit point-in-time metabolic size (LW0.75). Thus, ADG and MFW were confounded. To address the dilemma, Fox and Black (1984), and later Tylutki et al. (1994), proposed the concept of equivalent BW (i.e., the BW at which cattle with different MFW have similar body composition). However, their scaling factor (equivalent BW adjustment) was applied as a multiplier to the original Lofgreen and Garrett (1968) steer calf equation. In so doing, they assumed that the MFW of all cattle used for deriving the coefficient and exponent in the original data set (Lofgreen and Garrett, 1968) was constant (viz., 478 kg).
A more appropriate coefficient and exponent for RE unbiased by the correlation between ADG and MFW was derived using our data set based on iterative optimization: iteration of the coefficient and exponent values to obtain the lowest SD for DMIratio subject to the constraint that DMIratio = 1.000. The equation derived for steers is shown below:
 | [1] |
Compared with the NRC (1996) equation, the SD of the estimate was decreased by 5.4% (from 0.0496 to 0.0469). Furthermore, the significant bias in DMIratio associated with ADG was eliminated (r2 = 0.0006). Consistent with Ferrell and Jenkins (1998), the above equation predicts that RE increases, but at a slightly decreasing rate, as ADG increases. In contrast, NRC equations for NEg (NRC, 1984, 1996) indicate that RE increases at an increasing rate as ADG increases. Conversely, shrunk ADG (kg) for steers can be calculated from equivalent BW and RE from Eq. [1]:
Heifer Data and Relationships
The NRC (1984, 1996; using slaughter weight as MFW) NE equations explained 80.0 and 80.3%, respectively, of the variation in DMI of heifers with mean DMIratio for the 2 equations being 1.069 ± 0.0641 and 0.974 ± 0.0478, respectively. Thus, for feedlot heifers, NRC (1984) overestimated, whereas NRC (1996) underestimated DMI (by 6.9 and 2.6%, respectively; P < 0.0001). No bias was apparent in the DMIratio (r2 < 0.001) associated with SIW using NRC (1996) equations. However, with NRC (1984), a small but statistically appreciable (r2 = 0.0023; P < 0.001) bias in DMI-ratio associated with SIW was detected. For both the NRC (1984, 1996) approaches, the bias in DMIratio associated with ADG was less pronounced for heifers than it was for steers (r2 = 0.0151 and 0.0477, respectively; P < 0.001).
Using iterative optimization as described above, the ADG exponent for RE for heifers in our data set was equal to that for steers, but the slope was slightly (1.8%) greater:
 | [2] |
Compared with NRC (1996), the SD for DMIratio was 2.9% less (decreased from 0.0490 to 0.0476) for the new equation, and the bias in DMIratio associated with ADG was eliminated (r2 = 0.0005). The similarity between the derived steer and heifer equations for estimating RE (Eq. [1] and [2]) from ADG supports the concept that scaling on the basis of MFW satisfactorily accounts for differences attributed to sex (Taylor, 1980). Again, as for steers, shrunk ADG (kg) for heifers can be calculated from equivalent weight and RE (or NEg intake) from Eq. [2]:
Tests of Derived Equations
As a test of their applicability, Eq. [1] and [2] were used to predict DMI for an independent data set consisting of 1,880 pens of steers and 420 pens of heifers, respectively (representing a total of 357,173 steers and 78,098 heifers). Summary information from this independent data set is shown in Table 5. The corresponding DMIratio for steers using Eq. [1] was 0.996 ± 0.0450, and variation in DMIratio was not explained by either SIW (r2 = 0.001) or ADG (r2 = 0.014). For heifers, the corresponding DMIratio using Eq. [2] was 1.007 ± 0.0469, and likewise, the variation in DMIratio was not explained by either SIW (r2 = 0.001) or ADG (r2 = 0.014).
View this table:
[in this window]
[in a new window]
|
Table 5. Pen means and SD for 1,880 feedlot steer and 420 feedlot heifer closeouts used for testing applicability of revised retained energy equations developed using data set in Tables 1 and 2
|
|
Effect of Intake on Efficiency
For feedlot cattle, average NEm intake (the multiple of NEm of the diet and DMI per kilogram of shrunk BW) seems amazingly constant. Indeed, the ratio of NEm intake to BW in the original data set (Tables 1 and 2) was remarkably similar for steers and heifers averaging 0.0416 ± 0.0028 and 0.0424 ± 0.0027 Mcal/kg of average shrunk BW, respectively. Furthermore, the NEm intakes of steers and heifers within the smaller independent data set (Table 5) also were in extremely close agreement with those of the original data set (Tables 1 and 2), averaging 0.0418 ± 0.0021 and 0.0425 ± 0.0025 Mcal/kg of average BW, respectively.
As expected, when observed NEm intake per kilogram of average shrunk BW increased, the DMIratio decreased (r2 = 0.42, P < 0.001). What is not known is how much of the variance is due to level of intake, per se. Does this imply that feed conversion is poorer at greater intakes of energy? The effect of level of intake on dietary energetic efficiency remains a topic of debate. One fundamental premise of the NE system proposed by Lofgreen and Garrett (1968) was that the partial efficiency of utilization of ME for growth was constant. Indeed, comparative slaughter studies (Lofgreen et al., 1963; Lofgreen and Garrett, 1968) revealed that as ME intake increased, RE increased linearly, indicating that the ME value of feed ingredients did not decrease as intake increased. However, when evaluated directly, particularly with greater roughage and less processed grains, level of feed intake altered ME of the diet (Moe et al., 1972). Zinn et al. (1995) reported that increasing intake from 1.6 to 2.4% of LW daily decreased the DE and calculated ME value of finishing diets based on either dry-rolled or steam-flaked corn. Although adjustments in dietary energy availability based on feed processing and intake above maintenance currently are employed in formulation of diets for lactating cows (NRC, 2001), feed intakes and dietary forage levels are much greater for lactating cows than for feedlot cattle, and the extent of grain processing generally is less. Nevertheless, greater-roughage diets and less extensive grain processing, both associated with less energy digestibility, would be expected to increase rate of passage and may decrease digestibility.
MFW Prediction
Feedlot application of equivalent BW scaling of LW to estimate RE requires that MFW be specified. According to NRC (1996), MFW can be predicted based on cattle frame size (i.e., MFWsteers = 366.52 + 33.35 frame; MFWheifers = 293.2 + 26.7 frame). However, the term frame size is not readily quantified by scientific standards. Lack of information about the background of an animal renders frame size estimation excessively subjective for reliably predicting MFW. Frame size might be assigned reliably if information about hip height, age, and dam and sire is available (BIF, 2002), but such information is not available for most cattle entering the feedlot. However, even when specified information (BIF, 2002) is available for determining frame size, few studies have tested its value for predicting MFW. Grona et al. (2002) studied the relationships between frame size, muscling, and condition score assessments of cattle of known age and background to measures of carcass traits; subjective frame score assessments were affected by both condition score and muscularity. Most notably in these studies, frame score assessment was confounded with initial BW. The NRC (1984) noted this interaction between initial BW and frame size on MFW, but they grouped cattle into only 2 initial BW classes: calves and yearling. No such differentiation based on initial BW is apparent in NRC (1996). Indeed, when feedlots have assigned individual frame scores to cattle upon arrival into the yard based on visual appraisal and physical measurements, this index had limited utility. One feedlot in the basic data set visually appraised the frame size for all cattle that entered their feedyard during 1999 and 2000 (405 pen closeouts). Their frame size assessment explained only 13% of the variation in final slaughter weight and only 10% of the variation in ADG. Across our entire data set, frame scores were reported for 4,809 pens of steers (43% of total). Frame scores were not associated (r2 < 0.0001) with slaughter weight and explained only 1.1% (P < 0.001) of the variation in ADG. In multiple-regression analysis in which both initial weight and frame score are included to predict final slaughter weight, the subjectively assessed frame score dropped out of the model (P = 0.27 for the frame score variable) when initial BW was included in the model.
Mature BW of feedlot cattle previously have been associated with differences in sex, SIW, and ADG (Emmans, 1997). In our data set, regression analysis (r2 values) indicated that SIW could explain 39 and 50% of the variation in SFW and 25 and 21% of the variation in ADG for steers and heifers, respectively. When the data set was subdivided into 10-kg increments of SIW (resulting in 28 subclasses of SIW ranging from 199 to 468 kg), SIW of the average animal, presumably being a medium frame or average steer or heifer, was closely associated with both SFW and ADG:
 | [3] |
 | [4] |
 | [5] |
 | [6] |
These equations imply that mean performance and final BW for a pen of cattle within an initial BW group could be predicted accurately from SIW alone. Imprecision of prediction by linear regression alone, as noted above, can be attributed to the curvilinearity present in these equations, as depicted in Figures 1 and 2. Shrunk final weight increased linearly (P < 0.001) with increasing SIW for steers but exponentially (P < 0.001) with increasing SIW for heifers. Assuming that SFW for the average steer or heifer in our data set is equivalent to MFW (mature weight at which cattle would be expected to contain 28% empty body fat), the difference in MFW of steers and MFW of heifers decreased as SIW increased (Figure 1).
View larger version (16K):
[in this window]
[in a new window]
|
Figure 1. Final shrunk weight (FSW) predicted from initial shrunk weight (SIW) of steers (SFW = 421.16 + 0.4565 SIW; r2 = 0.39; Sy,x = 25.37; P < 0.001) and heifers (SFW = 470.0 – 0.2371 SIW + 0.00117 SIW2; r2 = 0.50; Sy,x = 20.06; P < 0.001) at commercial feedlots (the + symbol identifies the sex mean).
|
|
View larger version (15K):
[in this window]
[in a new window]
|
Figure 2. Average daily gain predicted from initial shrunk weight (SIW) of steers (ADG = 0.72 + 0.0029 SIW – 0.00000142 SIW2; r2 = 0.25; Sy,x = 0.147; P < 0.001) and heifers (ADG = 0.5015 + 0.004027 SIW – 0.00000397 SIW2; r2 = 0.22; Sy,x = 0.127; P < 0.001) at commercial feedlots (the + symbol identifies the sex mean).
|
|
Daily BW gain of steers and heifers both increased quadratically (P < 0.001; Figure 2) with SIW. However, as SIW approached 350 kg, ADG tended to plateau for heifers, whereas for steers, ADG continued to increase linearly. Assuming that cattle with a SIW under 350 kg probably represent calves rather than yearling cattle, this would suggest that for calves, the difference in ADG between steers and heifers at a given SIW was relatively constant; in contrast, for yearlings, the advantage of steers over heifers in ADG increased as SIW increased.
Quality Assessment of Cattle
Considering again the original data set, SFW (an assumed proxy for MFW) can be estimated from SIW alone, as noted above. However, the accuracy of predicting SFW was enhanced additionally by including ADG in the model:
 | [7] |
 | [8] |
Any deviation in ADG from that of the average steer or heifer in the above equations might be designated as a measure of performance quality for a pen of cattle. We use the general term quality instead of frame size, because this measure would include environmental, management, nutritional, and health effects in addition to simple physiological differences. Based on the degree of deviation in ADG of a given pen from that of the average pen with specified mean initial BW, one could arbitrarily assign performance quality score (PQS) to an individual pen of cattle on a continuous scale from 1 to 3, where 2 is average quality, 1 is exceptionally high quality (much faster rate of gain than typical), and 3 is exceptionally poor quality (unusually low rate of gain):
 | [9] |
where ADGmean = the expected gain for average cattle based on initial BW (Eq. [5] and [6]). The intercept of 2 represents the average steer or heifer; the coefficient 0.30 was included as a divisor so that 99% of the pens of cattle in our data set would have a PQS between 1.0 and 3.0. Accordingly, ADG can be calculated directly from measures of SIW and PQS:
 | [10] |
 | [11] |
Predicted final BW for heifers and steers with PQS of 1, 2, and 3 are shown in Figure 3.
View larger version (26K):
[in this window]
[in a new window]
|
Figure 3. Predicted final BW of steers and heifers with performance quality scores (PQS) of 1, 2, and 3 from commercial feedlots (the + symbol identifies the mean of each sex).
|
|
Although ADG of feedlot cattle is not known a priori, one could estimate PQS visually or by observed BW gain or possibly by feed intake during the first 30 to 60 d in the feedlot (Strasia et al., 1988). Each group of feedlot cattle on arrival also could be classified based on their expected performance quality [i.e., whether the cattle within a pen at any given BW are equal (2.0), better (less than 2.0), or poorer (more than 2.0) than the average of all pens]. For example, if a pen of cattle were classified to have a PQS of 1.75, one would expect ADG to be faster than average for cattle with similar SIW. Substituting PQS for ADG, MFW can be calculated:
 | [12] |
 | [13] |
Accordingly, for a pen of steers with a SIW of 300 kg and a visual PQS of 1.75, the expected ADG (Eq. [10]), MFW (Eq. [12]), and feeding interval [(MFW – SIW)/ADG] to achieve MFW would be 1.57 kg, 569 kg, and 171 d, respectively. The MFW also could be adjusted for a target degree of fatness according to the equation: MFWfat adjusted = [(28 – EBF) x 14.26]/0.891, where EBF = the target empty body fat (Guiroy et al., 2002).
If MFW can be predicted from SIW and either PQS or ADG as discussed above, then energy retention can be determined independently of frame score. The practicality of this approach was assessed by comparing the predicted MFW based on SIW and PQS (Eq. [10] and [11]) versus using SFW as MFW in Eq. [1] and [2] for estimation of RE in steers and heifers, respectively. For steers, using predicted MFW for estimating RE explained 83.7% of the variation in DMI. Using SFW as MFW for estimating RE explained 85.9% of the variation. The corresponding DMIratio were 1.000 ± 0.0503 and 1.000 ± 0.0469, respectively. For heifers, using predicted MFW for estimation of RE explained 80.5% of the variation in DMI, whereas using SFW as MFW for estimation of RE explained 81.04% of the variation. The corresponding DMIratio were 1.000 ± 0.0482 and 1.000 ± 0.0476, respectively. For both steers and heifers, there was no apparent bias (r2 < 0.001) between SIW or ADG and DMIratio using the predicted MFW to estimate RE.
When our data set (Tables 1 and 2) was used to derive estimates of RE based on an equation proposed to predict MFW from feedlot-assigned frame sizes as stipulated by NRC (1996; MFW = 33.35 x frame + 366.52), the CV of DMIratio was 53% greater (SD = 0.0746 vs. 0.0488) than when predicted from SIW alone. Differences in ADG explained 17.8% of the variation in DMIratio when slaughter weight was used as MFW; estimating MFW from feedlot-assigned frame score increased this bias (DMIratio = 0.667 + 0.232 ADG; r2 = 0.35). Including feedlot-assigned frame size measurements to estimate MFW also caused an appreciable bias between SIW and DMIratio (DMIR = 0.834 + 0.000524 IW; r2 = 0.143).
Placing expectations (standards of performance) on growth performance of feedlot cattle based on sex, initial BW, and expected performance can prove useful for retrospective analysis of closeout data. A PQS analysis based on pen closeouts will reveal how each set of cattle has performed relative to other cattle independent of SIW. Compared with visual appraisal of the cattle, a PQS can aid in fine-tuning management. For example, after visual review of a feedlot X, a consulting nutritionist could classify all pens of cattle with an overall PQS of 2.0 (average quality). If closeout analysis indicated that PQS for an individual pen or period was poorer (e.g., 2.5), one should investigate management and environmental factors that might explain the subnormal ADG. Such factors could include animal health or background problems, marketing and transportation stress, the implant program, diet composition (protein level, ionophore, too little or too much forage NDF, grain processing adequacy), animal measurements (unusual initial shrink; i.e., a large difference between purchase BW and off-truck arrival BW), inadequate pen space (i.e., <12 m2/head), insufficient shade (i.e., <1.86 m2), mud, climate, etc. The ultimate goal of PQS evaluation is to set a standard and determine why an individual pen of feedlot cattle or the entire feedlot failed to achieve their PQS potential independent of SIW. On the other hand, if projected and observed (calculated) PQS are similar, efforts to enhance ADG should not be ascribed to animal management but instead could be altered by improved selection for cattle type and altering initial purchase BW. Assigning a PQS as well as SIW can help explain performance changes from season to season or year to year. For example, when the supply of feeder cattle is limited, feedlots often purchase cattle with less SIW. Subsequently, cattle feeders often express their concern about suboptimal ADG. Yet, based on the relationship described above, a decreased ADG should be expected whenever SIW is decreased.
Additional Relationships
Some additional tools may have value for evaluating nutrition-management of feedlot cattle. These include the estimation of dietary NE from cattle performance and assessment of factors that can change the maintenance requirements (e.g., mud, cold exposure). From the derived estimates of RE and EM (as indicated previously; Table 6), the NEm value of a diets can be calculated from animal performance information using the quadratic formula:
where a = –0.41EM; b = 0.877 EM + 0.41 DMI + RE; and c = –0.877 DMI (Zinn and Shen, 1998): NEm, Mcal/kg = {– [0.877 EM + 0.41 DMI + RE] – [(0.877 EM + 0.41 DMI + RE)2 – (1.438 x EM x DMI)]0.5}/–0.82 EM; NEg, Mcal/kg = 0.877NEm – 0.41. Accordingly, differences among diets in observed versus expected DMI can be ascribed directly to the concentration of NE of the diet. Alternatively, one might attribute differences in the DMIratio to an altered maintenance coefficient (maintenance requirement). The standard maintenance coefficient for beef cattle is 0.077 (NRC, 1996). The effect of variation in DMI with respect the maintenance coefficient (Q) could be calculated as: Q = [NEm (DMI – (RE/NEg)]/W0.75.
Revised estimates of the NEg required for cattle fed feedlot diets were developed and tested based on minimizing the error in matching energy intake to cattle performance, using data from 14,692 pens of cattle in 15 commercial feedlots. In previous equations, rate of gain had been confounded with slaughter weight, whereas the revised equations were not biased by rate of gain. Equations to predict slaughter weight and daily gain from the initial shrunk BW of cattle entering a feedlot also were devised. A quality appraisal system based on feedlot performance of cattle was developed to evaluate whether performance of a pen of cattle does or does not meet expectations based on initial BW, sex, and feed intake.
|
Footnotes
|
|---|
2 Present address: Instituto de Investigaciones en Ciencias Veterinarias, UABC, Mexicali, México 21386. 
3 Present address: Pioneer Hi-Breed, A DuPont Business, Johnston, IA 50131.
1 Corresponding author: razinn{at}ucdavis.edu
Received for publication September 4, 2007.
Accepted for publication May 27, 2008.
|
LITERATURE CITED
|
|---|
BIF. 2002. Guidelines for Uniform Beef Improvement Programs. 8th ed. http://www.beefimprovement.org/library/06guidelines.pdf Accessed Jan 25, 2008.
Emmans, G. C. 1997. A method to predict the food intake of domestic animals from birth to maturity as a function of time. J. Theor. Biol. 186:189–200.[CrossRef]
Ferrell, L., and T. G. Jenkins. 1998. Body composition and energy utilization by steers of diverse genotypes fed a high-concentrate diet during the finishing period: I. Angus, Belgian Blue, Hereford, and Piedmontese sires. J. Anim. Sci. 76: 637–646.[Abstract/Free Full Text]
Fox, D. G., and J. R. Black. 1984. A system for predicting body composition and performance of growing cattle. J. Anim. Sci. 58:725–739.[Abstract/Free Full Text]
Grona, A. D., J. D. Tatum, K. E. Belk, G. C. Smith, and F. L. Williams. 2002. An evaluation of USDA standards for feeder cattle frame size and muscle thickness. J. Anim. Sci. 80:560–567.[Abstract/Free Full Text]
Guiroy, P. J., L. O. Tedeschi, D. G. Fox, and J. P. Hutcheson. 2002. The effects of implant strategy on finished body weight of beef cattle. J. Anim. Sci. 80:1791–1800.[Abstract/Free Full Text]
Littell, R. C., G. A. Milliken, W. W. Straub, and R. D. Wolfinger. 1996. SAS System for Mixed Models. SAS Inst. Inc., Cary, NC.
Lofgreen, G. P., D. L. Bath, and H. T. Strong. 1963. Net energy of successive increments of feed above maintenance for beef cattle. J. Anim. Sci. 22:598–603.[Abstract/Free Full Text]
Lofgreen, G. P., and W. N. Garrett. 1968. A system for expressing net energy requirements and feed values for growing and finishing beef cattle. J. Anim. Sci. 27:793–806.[Abstract/Free Full Text]
Moe, P. W., W. P. Flatt, and H. F. Tyrrell. 1972. Net energy value of feeds for lactation. J. Dairy Sci. 55:945–958.[Abstract/Free Full Text]
NRC. 1984. Nutrient Requirements of Beef Cattle. 6th rev. ed. Natl. Acad. Press, Washington, DC.
NRC. 1996. Nutrient Requirements of Beef Cattle. 7th rev. ed. Natl. Acad. Press, Washington, DC.
NRC. 2001. Nutrient Requirements of Dairy Cattle. 7th rev. ed. Natl. Acad. Press, Washington, DC.
St-Pierre, N. R. 2001. Invited review: Integrating quantitative findings from multiple studies using mixed model methodology. J. Dairy Sci. 84:741–755.[Abstract]
Strasia, C. A., S. L. Harp, B. J. Skaggs, R. L. Schemm, D. L. Deen, C. L. Schultz, D. R. Gill, R. B. Hicks, and S. D. Kraich. 1988. The effect of sorting on feedlot animal performance. Okla. Agric. Exp. Stn. Rep. MP-125:161–163.
Taylor, St. C. S. 1980. Live-weight growth from embryo to adult in domesticated mammals. Anim. Prod. 31:223–235.
Tylutki, T. P., D. G. Fox, and R. G. Anrique. 1994. Predicting net energy and protein requirements for growth of implanted and nonimplanted heifers and steers and nonimplanted bulls varying in body size. J. Anim. Sci. 72:1806–1813.[Abstract]
Zinn, R. A., C. F. Adam, and M. S. Tamayo. 1995. Interaction of feed intake level on comparative ruminal and total tract digestion of dry-rolled and steam-flaked corn. J. Anim. Sci. 73:1239–1245.[Abstract]
Zinn, R. A., M. Montano, and Y. Shen. 1996. Comparative feeding value of hulless vs. covered barley for feedlot cattle. J. Anim. Sci. 74:1187–1193.[Abstract]
Zinn, R. A., and Y. Shen. 1998. An evaluation of ruminally degradable intake protein and metabolizable amino acid requirements of feedlot calves. J. Anim. Sci. 76:1280–1289.[Abstract/Free Full Text]
This article has been cited by other articles:
|
|
|
|
 
J. Salinas-Chavira, J. Lenin, E. Ponce, U. Sanchez, N. Torrentera, and R. A. Zinn
Comparative effects of virginiamycin supplementation on characteristics of growth-performance, dietary energetics, and digestion of calf-fed Holstein steers
J Anim Sci,
December 1, 2009;
87(12):
4101 - 4108.
[Abstract]
[Full Text]
[PDF]
|
|
|
Top
Abstract
INTRODUCTION
