| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
ANIMAL NUTRITION |
* Texas A&M University, College Station, TX 77843; and
Universidade Federal de Viçosa, Viçosa, MG 36571, Brazil
 |
Abstract |
|---|
 Top Abstract INTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONLITERATURE CITED |
|---|
Key Words: Bos indicus • comparative slaughter • growth • Nellore • net energy • net protein
|
INTRODUCTION |
|---|
|
TopAbstractINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONLITERATURE CITED |
|---|
).
The National Research Council guidelines for beef cattle production (NRC, 2000) are widely adopted to formulate diets around the world; however, energy and nutrient requirements are based on Bos taurus data. The NRC (2000) indicates that B. indicus breeds require about 10% less NEm than beef breeds of B. taurus. Tedeschi et al. (2002) indicated that steers and bulls of Nellore, a B. indicus breed, had NEm requirements similar to that adopted by the NRC (2000). Additionally, it has been recognized that sex and castrate status (castrate or intact male or female) influences growth of body tissues, affecting carcass composition and feed efficiency (Berg and Butterfield, 1976) and the NEm requirement (ARC, 1980; NRC, 2000). Therefore, improvements in the beef cattle production in tropical and subtropical regions require an accurate assessment of energy and protein requirements of cattle.
The metabolism of fat and protein involves contributions of various organs that are highly dynamic in terms of energy expenditure; consequently, they should be evaluated together. Several studies have been independently conducted to determine energy and protein requirements of B. indicus purebreds and their crosses with B. taurus. A meta-analysis of this data is necessary to provide an overall summary, and its interpretation can provide directions for future experiments. Therefore, the objective of this study was to perform a meta-analysis to determine energy and protein requirements for maintenance and growth of bulls, steers, and heifers of Nellore and Nellore x B. taurus crossbreds from independent studies that used the comparative slaughter technique to measure energy and protein balances.
|
MATERIALS AND METHODS |
|---|
|
TopAbstractINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONLITERATURE CITED |
|---|
Data Collection
A database that included general information (e.g., title, author name, date of publication), qualifying (e.g., breed, sex and castrate status, treatment), and necessary quantitative data were gathered for this study. Quantitative information included days on feed, mean intake of ME and N, initial and final shrunk BW (SBW), empty BW (EBW), and content of ether extract and CP of the EBW for each animal. Studies that provided enough information to compute these variables were included. Data from individual animals were obtained from 16 studies [Ferreira et al., 1999; Paulino et al., 1999, 2004; Martins, 2001; Véras et al., 2001; Putrino et al., 2006; Silva et al., 2002; Tedeschi et al., 2002 (3 studies); Veloso et al., 2002; Freitas, 2004; Backes et al., 2005; Paulino, 2006; Chizzotti et al., 2007; Marcondes, 2007], resulting in 389 animal records. Animals from all studies were individually fed, and no implants were used. Breeds were coded as Nellore purebred and Nellore crossbred with Angus, Red Angus, Simmental, Limousin, or Brangus. Table 1 contains the descriptive statistics of the database classified by breed and sex and castrate status.
|
The procedures used to compute energy requirements for maintenance and growth were similar to those described by Lofgreen and Garrett (1968), except we used data from the baseline animals (within each study) to develop linear regression equations, by study, to determine the initial EBW and initial composition of the EBW of growing animals rather than using the mean SBW and body composition, as described by Tedeschi et al. (2002). The initial EBW was computed from SBW using the appropriate equation for each trial, and then initial empty body fat and protein were estimated for each animal from their predicted initial EBW using the regression equations of the empty body fat or empty body protein on the EBW of the baseline animals. Empty BW gain (EWG) and empty gains of body fat and protein were calculated as the difference between final and initial BW of the respective body components. The caloric values of retained fat and protein were assumed to be 9.367 (Blaxter and Rook, 1953) and 5.686 (Garrett, 1958) Mcal/kg, respectively.
Linear and nonlinear regressions were used to estimate energy and protein utilization. The models that best described the data with a minimal possible number of parameter estimates were determined based on Akaike’s information criteria (AIC) and Schwartz’s Bayesian information criteria (BIC; Mills and Prasad, 1992).
Heat production (HP, kcal/kg0.75 of EBW) was calculated as the difference between ME intake (MEI, kcal/ kg0.75 of EBW) and retained energy (RE, kcal/kg0.75 of EBW). The average of the antilog of the intercept confidence interval (95%) of the linear regression between the log of HP and MEI was used to estimate the requirement for NEm (kcal/kg0.75 of EBW; Lofgreen and Garrett, 1968). The ME required for maintenance (MEm) was calculated by iteration, assuming that the maintenance requirement is the value at which HP is equal to MEI (kcal·kg–0.75 of EBW·d–1), as shown in the Eq. [1]:
 | [1] |
where HP = heat production (Mcal/d); β0 and β1 = coefficients; e = the natural log; and MEI is ME intake (Mcal/d).
The efficiency of ME utilization for maintenance (km) was calculated as the NEm requirement divided by the MEm requirement. The slope of the regression of RE on MEI was assumed to be the efficiency of energy utilization for growth (Kg). An alternative method, the intercept divided by the slope (Kg), was used to compute the MEm requirement, which was then multiplied by the km to estimate the NEm requirement. This second approach of calculating the NEm requirement was compared with the NEm requirement estimated using the regression of the log of HP on MEI.
The net requirement of protein for maintenance (NPm, g·kg–0.75 of EBW·d–1) was assumed to be the intercept of the linear regression of the retained N (g·kg–0.75 of EBW·d–1) on N intake (g·kg–0.75 of EBW·d–1), multiplied by 6.25 to convert to protein.
Animals fed at the maintenance level were not utilized in the calculations of the growth requirement. The NEg requirement (Mcal/d) was calculated as shown in Eq. [2], and the net protein requirement for gain (NPg, g of NPg/d) was calculated as shown in Eq. [3]:
 | [2] |
 | [3] |
where a and b = the antilog of the intercept and the slope of the linear regression of the logarithm of the RE (Mcal·kg–0.75 of EBW·d–1) on the logarithm of the EWG, respectively; and c, d, and e = the intercept and slopes of the multiple regression of the retained protein (g/d) on EWG, kilograms per day, and RE (Mcal/d), respectively.
The partitioning of MEI to RE as fat and protein was computed using a multiple regression, as shown in Eq. [4]:
 | [4] |
where MEI = ME intake (Mcal·kg–0.75 of EBW·d–1); REf and REp = the RE (Mcal·kg–0.75 of EBW·d–1) as fat and protein, respectively.
The intercept (β0) of Eq. [4] was assumed to be the estimate of the MEm, and the coefficients β1 and β2 represented the amounts of ME required to deposit 1 Mcal of ME as fat or protein, respectively. The efficiencies of RE as fat and protein (Kf and Kp, respectively) were calculated as the inverse of the coefficients β1 and β2, respectively.
Statistical Analysis
A random coefficients model was used assuming a random variation for the effect of study (Littell et al., 2006). The general statistical model used is shown in Eq. [5]:
 | [5] |
where Yij = the dependent variable Y at level j of the independent variable X in the study I; β0 = the overall intercept with fixed effect; β1 = the overall slope that results from regressing Y on X across all studies with fixed effect; Xij = the observed value j of the independent variable X in the study I; β2 = the effect of study i (Si) on the intercept; β3 = the effect of study i on the slope of the regression of Y on X in study I; and 
ij = the random, unexplained error.
An initial analysis was conducted assuming random slope and intercept effects, including a possible covariance between the slope and intercept using an unstructured variance-covariance matrix. The MIXED procedure (SAS Inst. Inc., Cary, NC) was used in the random coefficients model analysis. The covariance parameter was considered different from zero if the P-value was less than 0.10. A greater P-value than the traditional P = 0.05 was used, because accurate estimations of variances and covariances require a considerable number of observations. In instances in which the covariance parameters were not different from zero, a variance components structure of the variance-covariance matrix was used. The NLMIXED procedure of SAS was used to fit nonlinear models, and the Gauss method was selected for convergence. Outliers were identified and removed if the studentized residue was greater than 2.5 or less than –2.5 (Neter et al., 1996).
|
RESULTS AND DISCUSSION |
|---|
|
TopAbstractINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONLITERATURE CITED |
|---|
Regression equations have been developed to estimate EBW from BW (NRC, 1984, 2000). According to Owens et al. (1995), EBW is the most precise index of energy and nutrient content of the body, because the digesta are totally washed out from the gastrointestinal tract after animals are slaughtered, and only tissues are weighed and chemically analyzed. Neither effects of breeds (P = 0.30) nor sex and castrate status (P = 0.22) in the prediction of EBW from SBW values were found. The overall equation is shown in Eq. [6] (n = 385). This equation resulted in predictions of EBW within the range reported by Owens et al. (1995) and NRC (2000), in which the EBW was about 85 to 95% of the SBW:
 | [6] |
where EBW = empty BW (kg) and SBW = shrunk BW (kg).
Similarly, the intercept was not different from zero (P = 0.23), and there were no effects of sex and castrate status (P = 0.24) or breed (P = 0.24) on the estimates of the EWG from the ADG. The overall equation is shown in Eq. [7] (n = 385). The NRC (2000) assumes that EWG is 0.951 x ADG, which is very close to our findings:
 | [7] |
where EWG = empty BW gain (kg/d).
Energy Requirement for Maintenance
The nonlinear regression indicated that HP increased exponentially as MEI increased. Ferrell (1988) reported that energy intake affects HP due to an increase on mass and metabolic activity of visceral organs. The intercept and the slope of the regression of the log of HP on MEI and NEm requirements are shown in Table 2. There were no differences in NEm requirements between breeds (P = 0.82). The combined data indicated a NEm requirement of 75 kcal·kg–0.75 of EBW·d–1, which is slightly lower than the NEm of 77 kcal·kg–0.75 of EB-W·d–1 reported by Lofgreen and Garrett (1968) using data from 5 studies involving 208 B. taurus heifers and steers. Freitas et al. (2006) found no differences in NEm requirement among Nellore purebred and Nellore x Angus, Nellore x Brown Swiss, and Nellore x Simmental growing bulls, which was on average 79 kcal·kg–0.75 of EBW·d–1. Ferrell and Jenkins (1998) reported an kg–0.75 average value of NEm requirement of 74.5 kcal· of EBW·d–1 for B. indicus crossbred steers. There were no differences in NEm requirements among sex and castrate status (P = 0.73). The NRC (2000) assumed that steers have NEm requirements 15% less than bulls. Because animals from our database were individually fed in stalls, the absence of difference among sex and castrate status may be attributed to physical activity and other interactions among bulls when fed in groups.
|
). The overall estimate of MEm requirement was smaller than the exponential relationship between HP and MEI (107 vs 112 kcal·kg–0.75 of EBW·d–1, respectively). Similarly, Ferrell and Jenkins (1998) reported that the estimates of NEm requirement, calculated as the MEI at which energy gain equals to zero, resulted in the same ranking among sire breeds but lower values than those estimated from the regression of the log of HP on MEI.
|
. This method tends to yield lower values for NEm because of the positive linear relationship between RE and MEI, whereas the log of HP on MEI has a nonlinear, ever-decreasing HP value as MEI diminishes. Sainz et al. (2005) also reported a lesser NEm requirement of Nellore bulls than that of B. taurus breeds and suggested that the lower maintenance expenditure might partially be explained by the lower protein turnover (one of the major contributors to endogenous energy expenditures) reported in Nellore cattle. Ferrell and Jenkins (1998) found that NEm requirement differed (P < 0.05) among B. taurus and B. indicus sire breeds. The authors also found an effect of breed sire on weight of the liver and visceral organs in which the liver and the gastrointestinal compartments of Tuli-sired steers fed for ad libitum intake represented 1.16 and 5.64% of the EBW vs. 1.27 and 6.1% of the EBW of Angus-sired steers, respectively. Although visceral tissues represent approximately 6% of EBW, their energy expenditure can account for more than 40% of total energy utilized for maintenance (Ferrell, 1988).
The partial efficiency of conversion of ME to NEm was similar among sex and castrate status and breeds with an average value of 67% (Table 2
). Similar to these results, Freitas et al. (2006) reported a km of 67% and found no differences between Nellore purebred and Nellore x B. taurus bulls. Ferrell and Jenkins (1998) reported similar values of km (ranging from 65 to 69%) in crossbreds of B. indicus x B. taurus and B. taurus x B. taurus steers.
We concluded the use of log of HP on MEI to determine NEm is likely to be more adequate than using a relationship between RE and MEI because of the nonlinear relationship between these variables. Furthermore, the regression of the logarithm of HP on MEI resulted in an intercept of 1.87 ± 0.01 (0.5% of the intercept estimate; Table 2), whereas the prediction of MEm based on the regression of RE on MEI was calculated using an intercept of –4.75 ± 0.89 (18.7% of the intercept estimate) and a slope of 0.44 ± 0.03 (6.8% of the slope estimate; Table 3). Therefore, the prediction of NEm using the nonlinear regression of HP on MEI might be less variable and more precise. In addition, our results indicated no differences among sex and castrate status and between Nellore purebred and Nellore x B. taurus on NEm and energy efficiency for maintenance.
Protein Requirement for Maintenance
The NPm requirement is usually assumed to be the sum of endogenous urinary N, metabolic fecal N, and dermal (scurf and hair) N losses, multiplied by the factor 6.25 (NRC, 1985). Alternatively, the NPm requirement might be estimated as the intercept of the regression of retained N on N intake, which should be similar to the N losses at zero N intake. The main difference is that retained N is the amount of N retained by the animal and does not account for losses such as hair and scurf.
Our results indicated an overall NPm requirement of 1.74 ± 0.41 g of NP·kg–0.75 of EBW·d–1 (R2 = 0.35, n = 142) with no differences in NPm requirement among sex and castrate status (P = 0.59) and breeds (P = 0.92). Assuming an efficiency of use of MP to NPm of 0.67 (NRC, 2000) and converting the EBW to SBW according to Eq. [6], the overall MP requirement for maintenance was 2.30 g of MP·kg–0.75 of SBW·d–1. This value is lower than the recommendation of NRC (2000) of 3.8 g of MP·kg–0.75 of SBW·d–1. The Institute National de la Recherche Agronomique (INRA, 1988) used N balance studies to determine the maintenance requirement of 3.25 g of MP·kg–0.75 of SBW·d–1. Similarly, Smuts (1935) determined a value of 3.52 g of MP·kg–0.75 of SBW·d–1.
Energy Requirement for Growth
The coefficients to predict the NEg requirement from the RE are listed in Table 4. Rates of protein deposition increase at decreasing rates, whereas rates of fat deposition increase at increasing rates with the rate of gain (Byers, 1982). Consequently, the NEg requirement was exponentially related to the EWG. The regression of the logarithm of RE on the logarithm of EWG indicated a similar slope (P = 0.92) but a different intercept (P = 0.01) among sex and castrate status. The NEg requirement calculated assuming a common slope and different intercept among sex and castrate status for bulls, steers, and heifers is shown in Eq. [8].
|
 | [8] |
where NEg = net energy for gain (Mcal/d); EBW = empty BW (kg); EWG = empty BW gain (kg/d); and a = either 0.0514, 0.0700, or 0.0771 for bulls, steers, or heifers, respectively.
According to the NRC (2000), heifers and bulls have 18% greater and lesser NEg requirements than steers at the same BW, respectively. Our data indicated that the NEg requirement for bulls was 27% less than that for heifers and 10% greater than that for steers. These findings are in agreement with the differences in carcass composition and rates of fat accretion among sex and castrate status reported by Berg and Butterfield (1976). These authors indicated that fat deposition was greater (in a decreasing order) for heifers, steers, and bulls.
There was a tendency (P = 0.06) of difference in the NEg requirement between breeds. The regression of the logarithm of RE on the logarithm of EWG using a common intercept of 0.059 indicated the slope of the Nellore data set was 9% greater than that of Nellore crossbreds (1.03 vs. 1.13, respectively), indicating that when EWG was greater than 1 kg/d, the RE in the EWG was greater in Nellore crossbreds than in Nellore purebreds. This is confirmed by the greater daily RE (Mcal/d) in the Nellore crossbred data set (Table 1). Solis et al. (1988) reported a greater fat content in the empty body in Angus x Brahman than in Brahman purebreds. Most of the difference among breeds on the NEg requirement might be accounted for by differences in fat distribution and by differences in the mature BW among breeds, indicating that Nellore purebred might have a greater BW than Nellore crossbred with early maturing B. taurus to reach the same body fat content.
Protein Requirement for Growth
Protein deposition in the empty body has been estimated using the rate of ADG and the composition of the gain (NRC, 2000; Table 5). The composition of the gain depends on physiological maturity of the animal, which is affected by sex and castrate status and breed of the animal (NRC, 1984). The NPg requirement was not different among sex and castrate status (P = 0.59) and breeds (P = 0.14), suggesting that energy retained and EWG can account for a significant amount of the variation in the protein retention due to breeds and sex and castrate status (Eq. [9]):
|
 | [9] |
where NPg = net protein requirement for growth (g/d) and EWG = empty BW gain (kg/d).
The percentage of RE deposited as protein (%REp) decreased exponentially as the content of RE in the gain (Mcal/kg of EWG) increased (Figure 1). The %REp can be used to compute the partial efficiency of ME to NE for growth (Williams and Jenkins, 2003; Tedeschi et al., 2004). Because no study effect was observed, we pooled the data across studies to develop Eq. [10].
|
 | [10] |
where REc = the concentration of RE in the empty BW gain (Mcal/kg) and e = the natural log.
Our findings are not in agreement with those reported by Geay (1984), who reported that %REp was greater for bulls than for heifers likely because of greater protein content of protein in the gain. This is possible when animals have similar maturity degree, in which the composition of gain is identical and the RE concentration is comparable.
Efficiencies of Fat and Protein Accretion
The analysis of the partition of MEI above maintenance to protein and fat deposition based on Eq. [4] is shown in Table 6. Based on this multilinear approach, the energetic efficiencies of protein and fat deposition were not different among breeds and sex and castrate status and were on average 0.34 and 0.79, respectively. Owens et al. (1995) reported that protein accretion had an average efficiency of 0.47 and the efficiency of fat accretion was 0.79, which is identical to our finding. Geay (1984) reported efficiencies of protein and fat accretion of 0.20 and 0.75, respectively, using data from 52 experiments. The efficiency of protein accretion is highly variable and depends on physiological stage, nutritional status, and protein turnover. Unlike protein, fat accretion efficiency is thought to be less variable (CSIRO, 1990). According to Tedeschi et al. (2004), the ME for fat deposition has a greater efficiency than for protein likely due to the cost of protein turnover.
|
). Our efficiency of protein accretion was greater than that observed by Geay (1984) (0.34 vs. 0.20, respectively); this was likely because of lower rates of protein turnover of Nellore cattle in comparison with B. taurus breeds (Sainz et al., 2005). Efficiencies of Retained Energy
The partial efficiency of use of ME to NEg estimated as the slope of the linear regression of RE on MEI was not different among sex and castrate status (P = 0.33) and breeds (P = 0.20) and averaged 0.44 (Table 3). Ferrell and Jenkins (1998) reported a nonlinear regression in which energy gain increased asymptotically as MEI increased. However, the linear regression of the RE on MEI fit better in our data set than the exponential regression, because it had smaller values of AIC and Schwartz’s BIC (Mills and Prasad, 1992).
Even though Ferrell (2003) reported that variations in body composition and composition of the gain can affect the partial efficiency of energy utilization, the ARC (1980), CSIRO (1990), AFRC (1993), and NRC (2000) compute Kg based on biological values of the diet. Because the energy content of the fat and lean tissue differs, Tedeschi et al. (2004) suggested that Kg depends on the composition of EWG. Assuming the equation proposed by Tedeschi et al. (2004, 2005) to compute partial efficiency of NE for growth {Kg = (Kf x Kp)/[Kp + (%REp/100) x (Kf – Kp)]} and the Kf and Kp values found in this study, the Kg can be estimated from Eq. [11].
 | [11] |
where %REp = the percentage of protein energy in the retained energy (Mcal/Mcal). The use of Kg based on the EWG composition might be preferable to a single efficiency to compute ME to NEg based on diet ME content. As expected, Eq. [11] will yield greater Kg than the equation proposed by Tedeschi et al. (2004) at the same REp. In addition, as REp increases, the difference between Kg predicted by Eq. [11] and the equation of Tedeschi et al. (2004) increases.
Energy Partition of the MEI
The most efficient animal is the one that converts ME to RE more efficiently by expending less energy for maintenance (Herd et al., 2004). The percentage of the MEI utilized for HP and RE was plotted against MEI to evaluate differences in efficiency among breeds and sex and castrate status (Figure 2).
|
) indicated the maximum RE was 27.1% and the minimum HP was 72.9% of the MEI (Mcal·kg–0.75 of EBW·d–1). However, the quadratic model fit slightly better than the nonlinear model to describe the relationships between the percentage of MEI used as HP or RE and MEI. The quadratic equations were 1.15 (±0.06) – 2.67 (±0.52) x MEI + 4.48 (±1.16) x MEI2 for HP (AIC = –831.7 and BIC = –816.6) and –0.15 (±0.06) + 2.67 (±0.52) x MEI – 4.48 (±1.16) x MEI2 for RE (AIC = –831.7 and BIC = –816.6).
Our data suggested that HP increases exponentially with MEI (Table 2), and, although RE increases with increased MEI (Table 3), it happens at a lesser incremental rate than that with HP. This occurs because as MEI increases, the energy utilized for feed intake, digestion, absorption, and metabolism of nutrients also increases. In addition, the metabolic activity of the visceral organs increases with MEI, resulting in a greater HP. Ferrell and Jenkins (1998) reported that HP increased exponentially and energy gain increased asymptotically as DMI increased above maintenance, supporting the quadratic behavior when HP and RE are expressed as percentage of MEI. Assuming the quadratic model, the maximum proportion of the MEI deposited in the body (maximum growth efficiency) does not occur at the maximum energy intake level; the quadratic equation suggested that at the MEI of 0.298 Mcal·kg–0.75 of EBW·d–1, animals retained the greatest portion of the ME consumed (or lost the least portion of MEI as heat).
In conclusion, our results do not support the hypothesis that bulls have greater NEm requirements than steers and heifers. Similarly, no differences in the NPm among bulls, steers, and heifers were detected. Nonetheless, the NEg requirement of steers was greater than bulls and lower than heifers. The %REp was negatively, nonlinearly correlated with the concentration of energy in the EWG, but our findings indicated no differences in NPg among bulls, steers, and heifers. The greatest energy efficiency did not occur at the maximum MEI.
|
Footnotes |
|---|
2 Corresponding author: luis.tedeschi{at}tamu.edu
Received for publication May 30, 2007. Accepted for publication March 14, 2008.
|
LITERATURE CITED |
|---|
|
TopAbstractINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONLITERATURE CITED |
|---|
This article has been cited by other articles:
|
|
 |
|
  D. D. Millen, R. D. L. Pacheco, M. D. B. Arrigoni, M. L. Galyean, and J. T. Vasconcelos A snapshot of management practices and nutritional recommendations used by feedlot nutritionists in Brazil J Anim Sci, October 1, 2009; 87(10): 3427 - 3439. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
and 
), steers (
and •), and heifers (
and 
). Opens symbols are data from Nellore purebreds, and solid symbols are data from Nellore x Bos taurus crossbreds. The solid line is the predicted values using the following equation: %REp = 10.1 (±0.1) + {166.7 [± 15.8] x e[–0.660 (±0.045) x REc]}.