|Livestock Research for Rural Development 26 (7) 2014||Guide for preparation of papers||LRRD Newsletter||
Citation of this paper
A high percentage of buffaloes in Colombia are managed under dual-purpose systems, therefore a study was carried out to evaluate feed efficiency traits in growing buffaloes. Two performance tests were conducted, each lasting 112 days, with the first 28 days serving as adaptation to the facilities and diet, and the remaining 84 days used for data collection. Growth, dry matter intake (DMI) and ultrasound were measured in 61 buffaloes, whereas nutrient digestibility (ND) was measured in 33 randomly selected buffaloes. Buffaloes were intact males and had an average initial weight and age of 262±53 kg y 461±50 d, respectively. Residual feed intake (RFI) was computed as the difference between actual and expected feed intake. Buffaloes were assigned to three RFI groups: high (> 0.12 kg DM/d), low (< -0.12 kg DM/d), and medium RFI (between ± 0.12 kg DM/d). A fixed effect model was used to analyze differences between groups for growth traits, intake, ND, efficiency and ultrasound measurements. Stepwise regression analyses were used to evaluate the contribution of ultrasound measurements and ND in explaining the variations in DMI.
No differences were observed for initial and final BW, ADG, relative growth rate (RGR), Kleiber ratio (KR) or ultrasound measurements among RFI groups. RFI was correlated with DMI, DM and CP digestibility (r = 0.53, -0.58 and -0.59, respectively). RFI correlations with initial and final BW, ADG, RGR, and KR were not different from zero. In contrast, high correlations were observed between ADG and RGR, KR and FCR (r = 0.91, 0.96 and -0.93, respectively). The variation in DMI was explained by MBW (58%), DMD (15%), digestibility of NDF (4%) and ADG (4%).
Our results indicate that: digestibility traits were useful in predicting DMI whereas ultrasound measurements were not; RFI can be used to identify the most efficient buffaloes.
Keywords: nutrient digestibility, performance test, residual feed intake, ultrasound
In recent years, the Colombian buffalo herd has had a great expansion, growing at an annual rate of around 10%, three times the growth rate of the Colombian cattle herd. There are two main production systems in which most buffaloes are handled in Colombia: dual purpose and no milking. The main objective in the dual purpose system is to produce milk, since given its quality (high fat, protein and total solids content), buffalo milk is usually sold at a price that is 30% higher than that paid for bovine milk. In dual purpose systems, the production of buffalo calves is a second priority; and animals are weaned at lower live weights and reach harvest weights at later ages than in the no milking system. A high percentage of buffaloes in Colombia are managed under dual-purpose systems.
It has been reported that buffaloes possess the ability to efficiently utilize low quality pastures. High feed prices and the potentially negative environmental impact of meat production imply that the efficiency of feed utilization in any animal production system is an economically important trait. Many indices have been proposed and used to determine the energetic efficiency of livestock. During the last decade, it has been shown that residual feed intake (RFI; Koch et al 1963) has great potential as an index of energetic efficiency for beef cattle (Herd and Bishop 2000; Liu et al 2000; Arthur et al 2001a) because it is relatively independent from production traits (Koch et al 1963), probably reflecting more variation in basic metabolic processes (Richardson et al 2001; Nkrumah et al 2006) than variation due to differences in production levels. However, most RFI research in beef cattle has been conducted in temperate regions of the world.
According to Koch et al (1963), feed intake can be adjusted by metabolic body weight (MBW) and average daily gain (ADG), where the residue can be used to identify deviations from expected feed intake levels, and represent high or low efficiency (negative or positive RFI, respectively) (Montanholi et al 2009). Additionally, the RFI equation can be used to study the impact of variables that represent basic metabolic processes on the explanation of dry matter intake (DMI) variations, such as ultrasound measurements (US) and nutrient digestibility (ND). The objectives of this study were to evaluate feed efficiency traits and their relationship with productive performance of growing buffaloes from dual purpose systems and to test the usefulness of US and ND to explain and model DMI variation.
The study was conducted at the El Progreso Experimental Station of the University of Antioquia, located at the Barbosa municipality (Antioquia, Colombia), at 1300 m above sea level, 23°C of temperature and 1800 mm/yr of rainfall in an area classified as subtropical wet forest. Chemical analyzes were performed at the Integrated Animal Nutrition, Biochemistry, and Forage Laboratory of the University of Antioquia. Determination of in situ indigestible acid detergent fiber (in ADF) was conducted at the Paysandú Production Center of the National University of Colombia, an area classified as lower-mountain rain forest (altitude above sea level: 2300 m, temperature: 16 ° C, rainfall: 1800 mm / year).
Two performance tests (December 2009 to April 2010 and February to May 2011) were conducted using 61 buffaloes with an average initial weight and age of 262 ± 53 kg and 461 ± 50 d, respectively. All buffaloes were at minimum 75% Murrah bulls. The tests lasted 112 d, with the first 28 d serving as adaptation to the facilities and diet, and the remaining 84 d used for data collection. Growth, DM intake (DMI) and ultrasound were measured in 61 buffaloes, whereas apparent digestibility of DM (DMD), crude protein (CPD), neutral detergent fiber (NDFD) and acid detergent fiber (ADFD) were measured in 33 randomly selected buffaloes.
Buffaloes were housed in individual pens (16 m²) with concrete floors, no bedding, and a partial (4 m²) roof, each with a feeding bunk and a water dispenser. The diet consisted of fresh Maralfalfa grass (Pennisetum sp.) offered at libitum and two kg of a concentrate supplement provided daily. Concentrate ingredients were maize (50%), extruded soybeans (15%), soybean meal (10%), extruded maize (10%), homogeneous mixture of extruded maize and soybean meal (10%) and a mixture of minerals containing of 8% phosphorus, calcium carbonate, and a vitamin-micromineral premix (5%). The chemical composition of Maralfalfa grass and concentrate supplement is reported in Table 1.
|Table 1. Composition of experimental diets (on DM basis) used in performance tests of fifteen-month old buffaloes from a dual-purpose system in Colombia¹|
|Nutritional composition||Maralfalfa grass²||Supplement|
|Crude protein (%)||6.74||15.6|
|Gross energy (Kcal/kg)||3916||4132|
|Acid detergent fiber (%)||44.2||8.82|
|Neutral detergent fiber (%)||67.4||16.2|
|Total ash (%)||6.58|
|Ether extract (%)||2.21|
¹Data reportedare the average of
the diets used in the two
²The composition of maralfalfa (Pennisetum sp.) grass is the average of four samples taken each month, which in turn were composed of three sub-samples taken in three consecutive days during each measurement.
Longissimus muscle area (REA; cm²) and rump fat thickness (RFT; mm) ultrasound measurements were taken every 14 using an Akila-Pro equipment (Esaote Europe BV, Maastricht, Netherlands) with a 3.5 MHz (18 cm) transducer. Images were measured with Eview software (Pie Medical, Maastricht, Netherlands). To assess REA, an image was taken between the 12th and 13th rib, perpendicular to the loin muscle. To measure RFT, images were taken from the tip of the hip towards the back region between the iliac and ischial tuberosities (Realini et al 2001; Jorge et al 2005).
Animals were weighed every 14 days after 12 hours of feed withdrawal. Before each weighing, DMI was evaluated for three consecutive d. To determine DMI in a given fourteen-d measurement period, an average DMI estimate was obtained from the previous and the current DMI estimate, which was multiplied by the number of days in the period. In turn, total DMI throughout each performance test was estimated by adding the DMI calculated in all fourteen-d measurement periods included in a given performance test. Daily DMI was calculated by dividing total DMI by the number of days in the test.
Residual feed intake (RFI) for each animal was computed as the difference between actual average and expected DMI (Koch et al 1963) during the 84 d period. Expected DMI was estimated as a linear regression of average DMI on ADG and metabolic mid-body weight:
DMI = β0 + β1(ADG) + β2(MBW) + RFIKoch
Feed conversion ratio (FCR) for each buffalo was computed as the ratio of daily DMI to ADG. The RGR of each buffalo (growth relative to instantaneous body size) was computed as the percentage of the difference between the log of final body weight and initial body weight over the number of days on test (Fitzhugh and Taylor 1971). The Kleiber ratio was computed as the ratio of ADG to MBW at the end of the test (Bergh et al 1992; Arthur et al 2001a).
Apparent digestibility of nutrients (ND) was evaluated during the last four days of each performance test and included the estimation of dry matter (DMD), crude protein (CPD), neutral detergent fiber (NDFD), and acid detergent fiber (ADFD) digestibility. The feed offered and refused was weighed and DM, CP, ADF and NDF contents were determined so as to estimate consumption of these fractions. The apparent digestibility of DM, CP, NDF, and ADF was estimated according to Bondi (1989).
To estimate fecal output indigestible acid detergent fiber (inADF) was used as an internal marker (Khan et al 2003, Rodríguez et al 2007). The inADF was measured in residues recovered after 144 h of in situ ruminal incubation (Berchielli et al 2000) of feces, forage, and feed supplements (Correa et al 2009). Fecal samples were taken directly from the rectum every six h during a four day period (Correa et al 2009), for a total of 16 samples per buffalo. Samples were kept frozen until the end of the collection period. Subsequently, after being dried at 60 °C for 72 h, these samples were mixed to obtain a single sample per buffalo which was stored until chemical analyzes.
Fecal production (F) was estimated using the following equation:
Recovery of internal markers in feces, in this case (inADF), is variable (Berchielli et al 2000). The precision and accuracy in the estimation of digestibility depends on fecal recovery rate (Zeoula et al 2002). Given the above, an initial experiment was conducted to determine percent recovery of inADF. Three 350-kg, 18 months-old male buffaloes were used. The diet used was similar to that offered during the performance tests.
Total fecal collection was conducted during four consecutive d using harnesses and collection bags designed for this purpose, as proposed by Gorski et al (1957) and Border et al (1963) (Figure 1). Animals had an adaptation period to the equipment (10 d) and diet (30 d) and remained with the harness on throughout the experimental period, whereas collection bags were only attached on the day of fecal collections. The fecal collection bags were changed four times a day. At each change, feces were weighed and homogenized individually for each buffalo. Then, a sample was taken and immediately stored at freezing temperature (-20°C). Subsequently, these samples were mixed to obtain a single sample per animal, dried at 60 °C for 72 h, and stored until inADF determination.
Simultaneously, grass and supplement samples were obtained and kept under refrigeration until the end of the experiment, when samples were subsequently dried at 60 °C for 48 h before chemical analysis. Knowing the amount of inADF consumed and excreted, inADF recovery was then estimated. The recovery rate of inADF in feces was calculated from the following equation:
|Figure 1. Buffalo wearing harness and bags for total collection of feces.|
Linear mixed models with first and second order regression polynomials for age were used with each buffalo to determine initial and final weights, REA, RFT, MBW (in the middle and end), and both REA and RFT in the middle of the test. Models used unstructured variance and covariance matrices (different variances and covariances between random parameters), as described by Littell et al (2004). A daily gain estimate was obtained from the difference between initial and final values for each characteristic, divided by the number of days in the test. The model used was the following:
yijk = (β0 + b0i:k) + (β1 + b1i:k)Xj + β2X²j + Pk + eijk
Where Υijk = trait (weight, ultrasound and bovinometric) measured at the j-th age of the i-th buffalo in k-th test; β0, β1 and β2 = intercept, linear, and quadratic regression coefficients for all animals;Χj is the j-th age; Τk fixed effect of test (two performance tests) and eij = residual associated with the individual variability of the observations not explained by the model, where eij ∼ N (0, σ²e); b0i:k and b1i:k = intercept and linear regression coefficient of the i-th animal, representing random deviations from β0 and β1 coefficients, respectively, where:
The b2i:k random effect of the i-th animal was not included in the final model because convergence was not achieved when this effect was included in preliminary runs. Computations were performed with the MIXED procedure of SAS (SAS 2009).
According to residual feed intake (RFIKoch), buffaloes were assigned to three groups (Nkrumah et al 2004; Elzo et al 2009). The RFI groups were high (buffalo RFI > mean + 0.5 SD, less efficient), medium (buffalo RFI between mean ± 0.5 SD), and low (buffalo RFI < mean - 0.5 SD, more efficient).
A fixed-effect model was used to analyze differences among RFI groups (high, medium and low) for growth traits, feed intake, ND, efficiency, and ultrasound measurements. The traits were individually analyzed by the restricted maximum likelihood method (RML) using the MIXED procedure of SAS (SAS 2009). The model included fixed effects of test (two performance tests) and RFI group. Age and initial weight were included in the model as covariates when they were significant. The residual effect was assumed to have mean zero and a common variance (σ²e). The least-square means were compared using the Tukey-Kramer test. The RFI trait was also analyzed with the former effects, except the RFI group effect.
Stepwise regression analysis was performed (PROC REG; SAS 2009) to determine the contribution of ultrasound measurements and ND to explain the variation in DMI. Ultrasound measurements taken were REA and RFT in the middle of the test; gain REA (REAg), and RFT (RFTg). Digestibility measures included in the model were DMD, PCD, NDFD, and ADFD. Contribution of each variable was determined using partial R². The Cp criterion of Mallows (Gilmour 1996) was used to select the optimal model.
The Cp statistic is defined as:
where SSEp = sum of square error of the model (containing p explanatory variables, including the intercept), S² = mean square error of the full model (containing all the explanatory variables of interest), n = number of observations, and p = number of parameters. The selected model had a Cp closest to its number of parameters (Draper and Smith 1981).
The full model (with all ultrasound and digestibility measurements) was as follows: yi = β0 + β1ADGi + β2MWi + βx1Xĳ + βx2Nĳ +ei, is the DMI of the i-th buffalo, Xij is the j-th ultrasound trait of the i-th Buffalo; Nĳ is the j-th digestibility trait of the i-th buffalo; β0 is the intercept; β1 is the ADG regression coefficient; β2 is the MW regression coefficient; βx1 is the regression coefficient of the j-th ultrasound trait (REA, RFT, RFTg and REAg); βx2 is the regression coefficient of the j-th digestibility feature (DMD, CPD, NDFD, and ADFD) and ei the residual of the i-th buffalo, where ei ∼N (0,σ²e).
Initial weight and age of buffaloes after the 28 d adjustment-period were 283 ± 58 kg and 489 ± 50 d, respectively. ADG was 0.542±0.14 kg/ day and feed conversion was 11.8±3.65. DM intake, as a percentage of live weight, was 1.9%, smaller than those reported for growing buffaloes (2.26 and 2.10%) weighing 260 and 320 kg, respectively (Paul and Lal 2010). These results can be explained by the diet used (nutritional value similar to that of a medium quality pasture), high in fiber, and low in CP. Average digestibility (%) of DM, CP, ADF and NDF found in this study were 59.1±2.68, 63.4±3.48, 49.2±3.50 and 47.8±3.54, respectively. Paul and Lal (2010), analyzing the information available from different experiments, found similar average digestibility of DM and CP (56 and 55.8%), and 61.8%.crude fiber digestibility.
Table 2 presents the least-square means for all RFI categories. It was not possible to compare results with other studies because literature reports on RFI were conducted in beef cattle breeds (Basarab et al 2003; Nkrumah et al 2004; Nkrumah et al 2006; Elzo et al 2009; Montanholi et al 2009; Sobrinho et al 2011). Animals from groups with high, medium, and low RFI presented similar initial and final BW and ADG (Table 2). Similar results were reported by Sobrinho et al (2011) in Brazilian Nellore cattle and Montanholi et al (2009) in Angus, Simmental, Gelbvieh and Piedmontese bulls in Canada. Different results were reported by Elzo et al (2009), who found 11.3 ± 5.29 kg difference in postweaning weight gain between groups of high and low RFI, indicating that the less efficient animals gained more weight during the test.
|Table 2. Least squares means of performance traits and efficiency measurements for dual-purpose growing buffaloes according to their residual feed intake (RFI) categories|
|Characteristic||Meanb||Residual feed intake group a|
|N° of animals||61||16||26||19||-|
|Initial weight (kg)||283±57.8||289±11.5||282±9.02||290±10.7||0.842|
|Final weight (kg)||338±49.7||345±11.0||336±8.62||342±10.2||0.815|
|RFI (kg DM/d)||0±0.24||0.31±0.028c||-0.01±0.022d||-0.24±0.026e||<0.0001|
aLeast squares means
and standard error.
bOverall mean and standard deviation.
c,d,eMeans in the same row with different superscripts are statistically different (Tukey-Kramer test).
ADG = average daily gain, DMI = dry matter intake, RFI = residual feed intake, FCR = feed conversion, RGR = relative growth rate, RFT = fat thickness of the hip in the middle of the test, REA = Longissimus dorsi muscle area in the middle of the test, RFTg = average daily hip-fat gain, REAg = average daily Longissimus dorsi.gain.
The RFI averaged 0.00 kg/d (SD = 0.24 kg/d), ranging from -0.52 to 0.68 kg/d, corresponding to 1.12 kg/d difference between the most and least efficient animals. The DMI was different (P<0.001) among the three RFI groups (Table 2). The low RFI group consumed 0.580 kg/d less (10.3 % difference) than the high RFI group for animals. These results agree with those reported in other studies with cattle breeds, but differences between high and low RFI groups, were higher. Nkrumah et al (2004), evaluating crossed cattle in Canada, reported a 1.86 kg/d difference. Similarly, Nkrumah et al (2007), evaluating a larger number of animals, found a 1.96 kg/d difference, and Montanholi et al (2009) a 2.24 kg/d difference. Intermediate differences were reported by Sobrinho et al (2011) in Nelore cattle in Brazil (0.705 kg/d). These results can be largely explained by the quality of the diet used, as it has a significant effect on consumption. The diet used in this study was low in protein and high in fiber, which may have limited intake, preventing the detection of major differences between RFI groups.
Variation in feed intake seems to be mostly, if not completely, explained by differences in energy costs related to maintenance per unit of metabolic size, when differences in gain rate and composition are not considered (Nielsen 2004). Some authors have suggested that, beyond DMI, some intake patterns vary between less and more efficient animals. In this regard, Golden and Kerley (2004) reported that more efficient animals consumed less food and spent less time eating the diet (fewer bites per d) than less efficient animals. Herd et al (2004) concluded that high RFI animals remained 13% longer time at the feeders, than low RFI animals. Due to this increased feeding time, there is a higher energy costs associated with feed apprehension, chewing, and rumination.
The feed conversion did not differ among RFI groups (Table 2). These results disagree with those reported by Nkrumah et al (2004) and Montanholi et al (2009) and Elzo et al (2009), who reported poorer conversion rates for less efficient (high RFI) animals. These results suggest that, compared with RFI, conventional conversion calculations may underestimate differences between animals with regard to energy feed efficiency, probably due to the effects of growth and BW on feed intake (Archer et al 1999).
Performance traits, relative growth rate (RGR), and Kleiber ratio (KR) did not differ among RFI groups or buffaloes (Table 2). This happened because weight, weight gain, and test days were similar between the less and the most efficient animals. These results agree with those reported by Nkrumah et al (2004) and Sobrinho et al (2011), who observed that RGR and KR may not detect obvious differences in energy efficiency among animals. Ultrasound measurements taken in the middle of the test and gain rate were similar in the three RFI groups, suggesting a small association between RFI and body composition. These results are consistent with those reported by other authors (Basarab et al 2003; Baker et al 2006; Montanholi et al 2009). However, other studies have reported differences in lean tissue deposition (Richardson et al 2001; Schenkel et al 2003; Richardson and Herd 2004) and fat tissue deposition (Richardson et al 2001; Nkrumah et al 2004) between beef cattle with different RFI, based on Koch's model. Nkrumah et al (2004), found higher backfat and higher deposition rate in high RFI animals.
Differences were found between RFI groups for DM (P = 0.006) and CP (P = 0.002) digestibility, indicating that more efficient animals (according to RFI values) exhibit improved feed utilization. According to Herd and Arthur (2009), the five most important physiological processes that contribute to RFI variation are associated with feed intake, digestibility, metabolism (anabolism and catabolism), physical activity, and thermoregulation.
Digestibility explained 10% of the variation in RFI in Angus steers product of divergent selection for RFI. However, the physiological mechanisms identified so far are based on few studies, and some studies used a small sample size (Herd and Arthur 2009). Richardson and Herd (2004) reported that animals with low RFI had higher DMD. Richardson et al (1996) found that animals phenotypically classified by high or low RFI tended to differ in their ability to digest DM. This difference in DMD explained 14% of the difference in feed intake between both groups of animals.
|Table 3. Least squares mean of nutrient digestibility for dual-purpose growing buffaloes according to their residual feed intake (RFI) categories.|
|Characteristic||Meanb||Residual feed intake group a|
|N° of animals||33||11||10||12||-|
aLeast squares means and standard error.
bOverall meanand standard deviation.
c,dMeans in the same row with different superscripts are statistically different (Tukey-Kramer test).
DMD = dry matter digestibility, CPD = crude protein digestibility, ADFD = acid detergent fiber digestibility, NDFD = neutral detergent fiber digestibility.
Positive correlations were found in this study between DMI and ADG (r = 0.43; P < 0.0001), initial BW (r = 0.82; P < 0.0001) and final BW (r = 0.84; P < 0.0001). These results are consistent with previously reported phenotypic correlations for growing animals (Basarab et al 2003; Lancaster et al 2009; Sobrinho et al 2011) and also for bulls (Arthur et al 2001a).
The RFI was positively correlated with DMI (r = 0.53; P < 0.0001), indicating that less efficient animals (high RFI) consume more food, coinciding with reports by Lancaster et al 2009 (r = 0.60) and Nkrumah et al 2004 (r = 0.77). No correlation was found (around zero; P > 0.05) between RFI and initial BW, final BW and ADG, indicating that RFI is independent of BW and ADG, which may be explained by the use of these traits in the linear regression to calculate RFI which in turn determines the phenotypic independence between RFI and these two characteristics. It has been shown in several studies that RFI is independent of weight and growth rate in cattle (Arthur et al 2001a; Nkrumah et al 2004).
In contrast to the lack of correlation between RFI and ADG, a high and negative correlation was observed between ADG and FCR (r = -0.93; P < 0.0001). Several studies have reported strong correlations phenotypic and genetic between FCR and ADG (r = -0.42 to -0.72; Lancaster et al 2009; Sobrinho et al 2011). These results suggest that selection for feed conversion can indirectly increase growth rate and size of the adult animal, increasing its nutritional requirements (Lancaster et al 2009). Similarly, the high and positive correlation between ADG and RGR (r = 0.91; P < 0.0001) and between ADG and Kleiber ratio (r = 0.96; P < 0.0001) suggest a desirable phenotypic effect of the Kleiber proportion and CRT on animal growth, but undesirable effects on energy consumption (high maintenance requirements), especially when adult size is reached (Archer et al 1999; Nkrumah et al 2004). High correlations have been reported by Nkrumah et al (2004) and Sobrinho et al (2011) between ADG and RGR (r = 0.48 to 0.72) and between ADG and Kleiber ratio (r = 0.70 to 0.85).
In contrast, low correlations (around zero; P > 0.05) were found between RFI and RGR and between RFI and Kleiber ratio. These results agree with those reported by Nkrumah et al (2004) and Sobrinho et al (2011), indicating that selection for RFI would have no effect on growth. Negative correlations were found between RGR with initial BW (r = -0.85; P < 0.0001) and final BW (r = -0.72; P < 0.01), and between Kleiber ratio and initial BW (r = -0.83; P < 0.0001) and final BW (r = -0.68; P < 0.0001). Therefore, the response to selection by Kleiber and RGR may not be as independent of growth and animal size as observed in selection by RFI.
The significant effect of initial BW on growth, intake, and efficiency measurements (except RFI) is consistent with reports which claim that RFI is generally not affected by differences between animals in terms of growth rates and maturity patterns (Archer et al 1999; Liu et al 2000). This finding also indicates that RFI -compared with other efficiency measures- may not be affected by pre-weaning environment, variations in age, body weight, or age of the mother, thus being a more robust index for data comparisons across environments and contemporary groups (Herd and Bishop 2000).
RFI was correlated with digestibility of DM (r = -0.58, P < 0.0001) and CP (r = -0.59, P < 0.0001). Negative correlations indicate that low RFI was associated with increased digestibility. This is consistent with the differences found for ND among the three RFI groups discussed above. Thereon, Richardson et al (1996) found that heifers and young bulls with low or high RFI tended to differ in their ability to digest DM. Similarly, Richardson and Herd (2004) found a negative correlation (r = -0.44) between RFI and DM digestibility.
No correlation (P > 0.05) was observed between RFI and ultrasound measurements, coinciding with the similarity among the three RFI groups discussed above. Several studies have reported correlations between RFI and backfat (measured between the 12th and 13th rib), but not with hip fat. Thereon, Nkrumah et al (2004) only found correlations between RFI and back fat gain (r = 0.30) and backfat at the middle of the test (0.19). Similarly, Lancaster et al (2009) reported 0.20, 0.30, and 0.17 correlations between RFI and backfat at the end of their test, fat gain, and REA gain, respectively. Arthur et al (2001b) found low correlations between RFI and subcutaneous fat measured on the 12th and 13th rib (r = 0.14), RFI and hip fat (r = 0.11), and RFI and REA (r = 0.06). Schenkel et al (2004) reported low correlations between RFI and back fat (r = 0.17), and RFI with REA (r = -0.14).
The magnitude of the association between body composition and RFI variation is influenced by the age and maturity of the animals tested. Performance tests for beef cattle usually involve the use of growing animals, in which protein synthesis is more efficient than fat deposition, while maintenance requirements for protein are higher than those for fat in adult animals, favoring the association between increased fat deposition and low RFI (Tixier-Boichard et al 2002).
Negative correlations were found between FCR and gains of RFT (r = -0.32, P < 0.01) and REA (r = -0.52, P < 0.0001). This indicates that the most efficient buffaloes deposited muscle and fat more rapidly, which is in agreement with observations by Lancaster et al (2009), who found negative yet lower correlations (r = -0.15 backfat, REA = -0.19). Opposite results were reported by Nkrumah et al (2004), who observed a positive correlation between FCR and fat deposition (r = 0.30).
The Koch model explained 72% of the variation in DMI for buffaloes from a dual-purpose system. This model was: DMI = 2.56 + 0.47ADG + 0.04MW + RFIKoch. Likewise, Arthur et al (2003), Basarab et al (2003) and Schenkel et al (2004) reported that the Koch model explained between 68 to 82% of DMI. However, other authors have found that this model explains in minor proportion the variation of DMI: Koch et al (1963, 48 and 60%), Montanholi et al (2009, 58%) and Lancaster et al (2009, 61.5%). According to Elzo et al (2009), this model explained the DMI change to a lower extent (30%), because the regression included data of all feedlots, years, racial groups, and genders.
Table 4 shows regression equations to predict DMI, generated by stepwise regression. The variables that contributed the most to DMI variation were MW (58%) and DMD (15%), followed by NDFD (4%) and ADG (4%). The other variables that were taken into account in the construction of models by the stepwise procedure were CPD and gain REA. According to the Cp criterion, the optimal model was the one who had MW, DMD and NDFD as independent variables (R² = 0.77).
|Table 4. Regression equations for predicting dry matter intake of growing buffaloes from dual-purpose systems.|
|DMI = 2.85 +0.042MBW + RFI||0.58||24.4|
|DMI = 6.33 + 00.37MBW – 0.054DMD + RFI||0.73||7.41|
|DMI = 6.50 + 0.037MBW – 0.078DMD + 0.027NDFD + RFI||0.77||4.69|
|DMI = 5.71 + 0.709ADG + 0.040MBW- 0.81DMD + 0.033NDFD + RFI||0.81||2.14|
|DMI = 5.76 + 0.810ADG + 0.041MBW – 0.062DMD – 0.018CPD + 0.030NDFD + RFI||0.82||2.17|
|DMI = 5.72 + 0.698ADG + 0.041MBW + 1.027REAg – 0.060DMD – 0.019CPD + 0.028NDFD + RFI||0.83||3.66|
1 Cp Criterion of Mallows
2 DMI = dry matter intake,
MBW = metabolic body weight in the middle of the test, DMD = dry matter digestibility, NDFD = neutral detergent fiber digestibility, ADG = average daily gain, CPD = crude protein digestibility, REAg = Longissimus dorsi area gain (cm²/d), RFI = residual feed intake, corresponds to the error term.
In general, it can be said that the two parameters used to calculate RFI in Koch model (MBW as an indicator of maintenance needs, and ADG as a measure of productivity) largely explain the DMI variation. However, the addition of digestibility measurements -particularly DMD- substantially improved the model, while ultrasound measurements had no significant impact on DMI prediction. However, under typical production conditions, measuring digestibility is not a common practice.
The contribution of ND in explaining DMI variation agrees with the differences found between the three RFI groups. According to Herd and Arthur (2009), digestibility explained 10% of RFI variation. Richardson and Herd (2004) observed that the difference in DM digestibility explained a 14% difference in consumption between two groups of animals (high and low RFI).
The low impact of ultrasound traits for explaining DMI variation agrees with the similarity among the three RFI groups regarding all ultrasound measurements, which was discussed earlier. The muscle and fat deposition is affected by race, fattening period and diet. Moreover, the category of animals may also influence the amount of hip fat, since young buffaloes, as the ones used in this study, have lower fat levels compared to animals in the final fattening stages. Therefore, ultrasound measurements may be relevant for determining DMI when factors that determine muscle and fat deposition have more variation than that observed in this study. Moreover, the time during the production cycle in which the ultrasound measurements are taken can influence the successful application of this technique (Montanholi et al 2009).
Montanholi et al (2009) compared different regression models to determine RFI. They found that mid-test REA and backfat measurements only improved the explanation of the DMI variation in 4 and 1%, relative to the Koch model. However, when total gain or backfat gain rate were included in the model, the explanation of DMI variation improved by 16%. Arthur et al (2003) observed a 5% improvement in DMI prediction after back fat at the end of the test was included in the model. The model that included REA, marbling, and back fat, improved DMI prediction by 9% compared with Koch model (Montanholi et al 2009). Likewise, Arthur et al (2003) and Basarab et al (2003) observed a 6% increase using similar models in young beef cattle. Knott et al (2008) found a 16% increase in prediction when ultrasound measurements were added to Koch model for predicting DMI in meat sheep. Likewise, Lancaster et al (2009) found that backfat and REA accounted for 9% of the DMI variation not explained by MW and ADG.
The residual consumption of food is a useful tool to identify more efficient buffaloes, without altering their productive performance. The two parameters used in the Koch model, metabolic body weight and daily weight gain, largely explained the variation in DM intake, but the addition of DM digestibility measurements improved the model. The ultrasound measurements were similar in the three groups of residual feed intake (high, medium and low) and had no significant impact on DM intake prediction. Additional studies should be conducted involving other ultrasound backfat and marbling measurements, as well as using older animals, to establish the association between these measurements and feed intake.
The authors wish to thank the Colombian Ministry of Agriculture and Rural Development (MADR) for financing the project “Performance tests on baby and dual purpose buffalo to select the best individuals for characteristics related to meat production and yield”, part of the “Genetic improvement of meat type buffaloes in Colombia” program and the Codi Program “E01727 Sustainability 2013-2014". We also thank all the people who contributed with the project, especially undergraduate and MS students from the University of Antioquia for their work at the Experimental Station.
Archer J A, Richardson E C, Herd R M and Arthur P F 1999 Potential for selection to improve efficiency of feed used in beef cattle: A review. Australian Journal of Agricultural Research 50: 147–161
Arthur P F, Renand G and Krauss D 2001a Genetic and phenotypic relationships among different measures of growth and feed efficiency in young Charolais bulls. Livestock Production Science 68: 131–139 http://www.sciencedirect.com/science/article/pii/S0301622600002438
Arthur P F, Archer J A, Johnson D J, Herd R M, Richardson E C and Parnell P F 2001b Genetic and phenotypic variance and covariance components for feed intake, feed efficiency and other postweaning traits in Angus cattle. Journal of Animal Science 79: 2805–2811 http://www.journalofanimalscience.org/content/79/11/2805.full.pdf
Arthur P F, Herd R M and Archer J A 2003 Should measures of body composition be included in the model for residual feed intake in beef cattle? Proceedings of the Association for the Advancement of Animal Breeding and Genetics 15: 306–309 http://www.aaabg.org/livestocklibrary/2003/306-309.pdf
Baker S D, Szasz J I, Klein T A, Kuber P S, Hunt C W, Glaze Jr J B, Falk D, Richard R, Miller J C,Battaglia R A and Hill R A 2006 Residual feed intake of purebred Angus steers: effects on meat quality and palatability. Journal of Animal Science 84: 938–945 http://www.journalofanimalscience.org/content/84/4/938.full.pdf+html
Basarab J A, Price M A, Aalhus J L, Okine E K, Snelling W M and Lyle K L 2003 Residual feed intake and body composition in young growing cattle. Canadian Journal of Animal Science 83: 189–204. http://pubs.aic.ca/doi/pdf/10.4141/A02-065
Berchielli T T, Andrade P and Furlan C L 2000 Avaliação de indicadores internos em ensaios de digestibilidade. Revista Brasileira de Zootecnia 29:830-833. http://www.revista.sbz.org.br/artigo/visualizar.php?artigo=2496
Bergh L, Scholtz M M and Erasmus G J 1992 Identification and assessment of the best animal: The Kleiber ratio (growth/metabolic mass) as a selection criterion for beef cattle. Proceedings of the Association for the Advancement of Animal Breeding and Genetics 10: 338–340. http://www.aaabg.org/proceedings/1992/ab92078.pdf
Bondi A A 1989 Nutrición Animal. Editorial Acribia, S. A. Zaragoza, España.
Border J R, Harris L E and Butcher J E 1963 Apparatus for obtaining sustained quantitative collections of urine from male cattle grazing pasture or range. Journal of Animal. Science. 22: 521-525. http://www.journalofanimalscience.org/content/22/2/521.full.pdf+html
Correa H J, Pabón M L y Carulla J E 2009 Estimación del consumo de materia seca en vacas Holstein bajo pastoreo en el trópico alto de Antioquia. Livestock Research for Rural Development Article 59: 21 (4) http://www.lrrd.org/lrrd21/4/corr21059.htm
Draper N R and Smith H 1981 Applied regression analysis. University of Wisconsin, WI and Proctor & Gamble Company New York: Wiley.
Elzo M A, Riley D G, Hansen G R, Johnson D D, Myer R O, Coleman S W, Chase C C, Wasdin J G and Driver J D 2009 Effect of breed composition on phenotypic residual feed intake and growth in Angus, Brahman, and Angus x Brahman crossbred cattle. Journal of Animal Science 87: 3877-3886. http://www.journalofanimalscience.org/content/87/12/3877.full.pdf+html
Fitzhugh Jr H A and Taylor C S 1971 Genetic analysis of degree of maturity. Journal Animal. Science 33: 717-725. http://www.journalofanimalscience.org/content/33/4/717.full.pdf+html
Gilmour S G 1996 The interpretation of Mallows’s Cp-statistic. The Statistician. 45: 49-56. http://www.jstor.org/stable/2348411
Golden J W and Kerley M S 2004 Relationship of feeding behavior to feed efficiency in crossbred Angus steers. In: Annual Meeting of the American Society of Animal Science, Saint Louis. Proceedings Saint Louis: American Society of Animal Science.
Gorski J, Blosser T H, Murdock F R, Hodgson A S, Sony B K and Erb R E 1957 A urine and feces collection apparatus for heifers and cows. Jounal of Animal Science 16: 100-109. http://www.animal-science.org/content/16/1/100.full.pdf
Herd R M, and Bishop S C 2000 Genetic variation in residual feed intake and its association with other production traits in British Hereford cattle. Livestock Production Science 63: 111–119. http://www.sciencedirect.com/science/article/pii/S0301622699001220#
Herd R M, Oddy V H and Richardson E C 2004 Biological basis for variation in residual feed intake in beef cattle. 1. Review of potential mechanisms. Australian Journal of Experimental Agriculture 44: 423–430.
Herd R M, and Arthur P F 2009 Physiological basis for residual feed intake. Jounal of Animal Science 87(E Suppl): E64–E71. http://www.journalofanimalscience.org/content/87/14_suppl/E64.full.pdf+html
Jorge A M, Andrighetto C, Domingues D M, Golfetto C and Freitas V A D 2005 Características Quantitativas da Carcaça de Bubalinos de Três Grupos Genéticos Terminados em Confinamento e Abatidos em Diferentes Estádios de Maturidade. Revista Brasileira de Zootecnia 34 (Suppl): 2376-2381. http://www.revista.sbz.org.br/artigo/visualizar.php?artigo=4656
Khan M A, Un-Nisa M and Sarwar M 2003 Techniques Measuring Digestibility for the Nutritional Evaluation of Feeds. International Journal of Agriculture and Biology. 5 (1), 91 - 94.
Knott S A, Cummins L J, Dunshea F R and Leury B J 2008 The use of different models for the estimation of residual feed intake (RFI) as a measure of feed efficiency in meat sheep. Animal Feed Science and Technology 143: 242–255 http://ac.els-cdn.com/S037784010700171X/1-s2.0-S037784010700171X-main.pdf?_tid=776b4308-b5fa-11e3-91a9-00000aab0f6c&acdnat=1395957490_d4098a77fc23f9cbd3bc2b28ab7bb21e
Koch R M, Swiger L A, Doyle C and Gregory K E 1963 Efficiency of feed use in beef cattle. Journal of Animal Science 22: 486-494. http://www.journalofanimalscience.org/content/22/2/486.full.pdf+html
Lancaster P A, Carstens G E, Ribeiro F R B, Tedeschi L O, and Crews D H Jr 2009 Characterization of feed efficiency traits and relationships with feeding behavior and ultrasound carcass traits in growing bulls. Journal of Animal Science 87: 1528–1539. http://www.journalofanimalscience.org/content/87/4/1528.full.pdf+html
Littell R C, Pendergast J, and Natarajan R 2004 Modeling covariance structure in the analysis of repeated measures data. In Statistical Modelling of Complex Medical Data in Medicine, R. B. D’Agostino, Editor, USA 2, 159 – 186.
Liu M F, Goonewardene L A, Bailey D R C, Basarab J A, Kemp R A, Arthur P F, Okine E K, and Makarechian M 2000 A study on the variation of feed efficiency in station tested beef bulls. Canadian Journal of Animal Science 80: 435–441. http://pubs.aic.ca/doi/pdf/10.4141/A99-030
Montanholi Y R, Swanson K C, Schenkel F S, McBride B W, Caldwell T R and Miller S P 2009 On the determination of residual feed intake and associations of infrared thermography with efficiency and ultrasound traits in beef bulls. Livestock Science 125 (1): 22-30. http://ac.els-cdn.com/S1871141309000766/1-s2.0-S1871141309000766-main.pdf?_tid=19020578-b5fe-11e3-b163-00000aacb35e&acdnat=1395959049_0f12be0e524b1ca77b9346d8ef6adee0
Nielsen M K 2004. Genetic variation in feed utilization: selection responses in mice. In: Annual Meeting of the American Society of Animal Science, 2004, Saint Louis. Proceedings Saint Louis: American Society of Animal Science. 410 p.
Nkrumah J D, Basarab J A, Price M A, Okine E K, Ammoura A, Guercio S, Hansen C, Li C, Murdoch B and Moore S S 2004 Different measures of energetic efficiency and their phenotypic relationships with growth, feed intake, and ultrasound and carcass merit in hybrid cattle. Journal of Animal Science 82: 2451–2459. http://www.journalofanimalscience.org/content/82/8/2451.full.pdf+html
Nkrumah J D, Okine E K, Mathison G W, Schmid K, Li C, Basarab J A, Price M A, Z Wang and Moore S S 2006 Relationships of feedlot efficiency, performance, and feeding behaviour with metabolic rate, methane production, and energy partitioning in beef cattle. Journal of Animal Science 84: 145–153. http://www.journalofanimalscience.org/content/84/1/145.full.pdf+html
Nkrumah J D, Crews D H Jr, Basarab J A, Price M A, Okine E K, Wang Z, Li C, and Moore S S 2007 Genetic and phenotypic relationships of feeding behavior and temperament with performance, feed efficiency, ultrasound, and carcass merit of beef cattle. Journal of Animal Science 85: 2382–2390. http://www.journalofanimalscience.org/content/85/10/2382.full.pdf+html
Paul S S and Lal D 2010 Nutrient Requirements of Buffaloes. Satish Serial Publishing House.
Realini C E, Williams R E, Pringle T D and Bertrand J K 2001 Gluteus medius and rump fat depths as additional live animal ultrasound measurements for predicting retail product and trimmable fat in beef carcasses. Journal of Animal Science 79: 1378–1385. http://www.journalofanimalscience.org/content/79/6/1378.full.pdf+html
Richardson E C, Herd R M, Arthur P F, Wright J, Xu G, Dibley K and Oddy V H 1996 Possible physiological indicators for net feed conversion efficiency in beef cattle. Proceedings of the Australian Society of Animal Production 21: 103–106. http://livestocklibrary.com.au/bitstream/handle/1234/8783/Richardson96.PDF?sequence=1
Richardson E C, Herd R M, Oddy V H, Thompson J M, Archer JA and Arthur P F 2001 Body composition and implications for heat production of Angus steer progeny of parents selected for and against residual feed intake. Australian Journal of Experimental Agriculture 41(7): 1065–1072. http://www.publish.csiro.au/?act=view_file&file_id=EA00095.pdf
Richardson E C and Herd R M 2004 Biological basis for variation in residual feed intake in beef cattle. 2. Synthesis of results following divergent selection. Australian Journal of Experimental Agriculture 44: 431–440.
Rodríguez N M, Oliveira S E y Guimarães-Júnior R 2007 Uso de indicadores para estimar consumo y digestibilidad de pasto. LIPE, lignina purificada y enriquecida. Revista Colombiana de Ciencias pecuarias 20 (4): 518-525. http://rccp.udea.edu.co/index.php/ojs/article/viewFile/309/306
SAS Institute Inc 2009 The mixed procedure SAS/STAT(R) 9.2 User's Guide, Second edition. http://support.sas.com/documentation/cdl/en/statug/63033/HTML/default/viewer.htm#mixed_toc.htm
Schenkel F S, Miller S P and Wilton J W 2003 Genetic parameters of feed efficiency, growth and carcass traits of beef bulls. Canadian Journal of Animal Science 83: 617–618. http://pubs.aic.ca/doi/pdf/10.4141/A02-abstracts
Schenkel F S, Miller S P and Wilton J W 2004 Genetic parameters and breed differences for feed efficiency, growth, and body composition traits of young beef bulls. Canadian Journal of Animal Science 84: 177–185. http://pubs.aic.ca/doi/pdf/10.4141/A03-085
Sobrinho T L, Branco R H, Bonilha S F M, de Castilhos A M, de Figueiredo L A, Razook A G and Mercadante M E Z 2011 Residual feed intake and relationships with performance of Nellore cattle selected for post weaning weight. Revista Brasileira de Zootecnia 40 (4): 929-937. http://www.revista.sbz.org.br/artigo/visualizar.php?artigo=66155
Tixier-Boichard M, Bordas A, Renand G and Bidanel J P 2002 Residual food consumption as a tool to unravel genetic components of food intake. Proceeding. 7th. World Congress. Genetics Applied to Livestock Production. Institut National de la Recherche. Agronomique, Montpellier, France. Session 10:0–08.
Zeoula L M, Prado I N, Dian P H M, Valério L J, Ferreira S, Midore E, Peron P D P, Araujo J e da Silva A J 2002 Recuperação fecal de indicadores internos avaliados em ruminantes. Revista Brasileira de Zootecnia31 (4): 1865-1874 http://www.revista.sbz.org.br/artigo/visualizar.php?artigo=3530
Received 30 March 2014; Accepted 8 June 2014; Published 1 July 2014
Go to top