1st Page

Journal of Animal Science - Animal Genetics

Genetics of slaughter precocity, carcass weight, and carcass weight gain in Chianina, Marchigiana, and Romagnola young bulls under protected geographical indication1


This article in JAS

  1. Vol. 91 No. 6, p. 2596-2604
    Received: Jan 04, 2013
    Accepted: Mar 07, 2013
    Published: November 25, 2014

    2 Corresponding author(s):

  1. F. Sbarra*,
  2. R. Mantovani 2,
  3. A. Quaglia* and
  4. G. Bittante
  1. National Breeders Association of Italian Beef Breeds, Via Visciolosa 06132 S. Martino in Colle, Italy
    Department of Agronomy, Food, Natural Resources, Animals and the Environment, University of Padova, Agripolis 35020 Legnaro, Italy


The aim of this study was to estimate the heritability and genetic correlation of age at slaughter (AS), as an indicator of slaughter precocity, carcass weight (CW), and CW gain (CWG = CW × AS–1) obtained from young bulls of 3 Italian autochthonous beef cattle breeds [i.e., Chianina (CHI), Marchigiana (MAR), and Romagnola (ROM)]. In addition, the study aimed at evaluating the effect of corrected or uncorrected CW for AS, and analyzing the relationship between adjusted or unadjusted CW with CWG in term of changes in rank correlation in groups of sires with high accuracy. Data were obtained from the Consortium of protected geographical indication (PGI) “Vitellone Bianco dell’Appennino Centrale” (i.e., white young bull of Central Apennines), approved by the European Union. After editing, 20,872 complete records were retained for subsequent Bayesian analysis. Univariate animal model produced h2 estimates of medium value for AS (i.e., from 0.28 for CHI to 0.39 for the ROM breed). The CW presented h2 estimates less than AS, ranging from 0.13 for CHI to 0.24 for ROM bulls. The adjustment of CW by AS (CWU-AS) increased the h2 values for CW in all breeds (i.e., from 0.20 to 0.29). Point estimate of genetic correlations between AS and CW obtained by a bivariate analysis were moderate to low, and negative in all breeds (from –0.08 to –0.29). Rerankings of sire for univariate CW (CWU) analysis and CWU-AS were more noticeable for CHI and ROM (rank correlation of 0.875 and 0.897, respectively) than for the MAR breed (rank correlation of 0.967). Comparing bivariate EBV for CW with EBV for CWU or CWU-AS increased rank correlation to 0.937 for ROM, but for CHI it remained lower (i.e., 0.861), indicating a possible large reranking of sires by correcting CW for AS in this breed. Daily CWG presented h2 estimates greater than CW and similar or greater than AS. It appears to be a good indicator of instant growth rate capacity of the animal but lacking information on the endpoint of fattening in terms of age and weight. Field slaughter data for the CHI, MAR, and ROM breeds under the PGI labeling indicate that AS is not a mere environmental factor to be corrected for but trait subjected to genetic control. Because of its economic relevance in fattening, age at slaughter, as an indicator of slaughter precocity, could become a trait requiring careful consideration for selection of beef breeds.


Genetic improvement of beef cattle in Europe has been based primarily on growth and in vivo conformation recorded at test stations on young candidate bulls (Andersen et al., 1981; Albera et al., 2001). More recently, large data sets from abattoirs have been used to study the genetic components of carcass (Reverter et al., 2000; Rìos-Utrera and Van Vleck, 2004) and meat quality traits (Johnston et al., 2003; Boukha et al., 2009, 2011; Cecchinato et al., 2011, 2012). This is the result of an increased availability of slaughter data, which, in Europe, depends on the traceability regulations introduced after the bovine spongiform encephalopathy outbreaks and on the increased number of labeled products (i.e., protected geographical indication, PGI; or protected designation of origin) monitored all along the productive chain. However, the use of slaughter data for the genetic improvement of beef cattle is not yet documented at European level. In particular, doubts can arise when modeling carcass traits that consider the individual differences in age at slaughter. Indeed, at population level, the age at slaughter can depend not only on BW or market condition, but also on factors like conformation and fatness. These aspects are part of slaughter precocity (i.e., time to reach finishing status required by the market), which can be a genetically controlled trait (Bittante et al., 2011; Boligon et al., 2011, 2013). The aim of this study was to analyze genetics of the age at slaughter, carcass weight (CW), and CW gain (CWG) in Chianina (CHI), Marchigiana (MAR), and Romagnola (ROM) breeds, using the slaughter database of the Consortium for PGI. In particular, the study aimed at: i) estimating the heritability of and genetic correlations between the age at slaughter and CW; ii) comparing the effect of corrected or uncorrected CW for age at slaughter; and iii) analyzing the relationship between the adjusted or unadjusted CW and CWG.


Animal Care and Use Committee approval was not obtained for this study because the data were extracted from an already existing database. The authors had no direct control over the care of the animals included in this study.

Data Source and Editing Procedure

Data used in this study were obtained from the slaughterhouses that belong to the Consortium “Vitellone Bianco dell’Appennino Centrale” (i.e., white young bull of the central Apennines), a PGI for meat obtained from CHI, MAR, and ROM young bulls and heifers. These 3 autochthonous Italian beef breeds are currently widespread across the country (Bittante, 2011; Maretto et al., 2012), although in the PGI consortium only purebred animals born and raised within the typical area of production located in the Apennines of Central Italy are allowed. The labeled meat is obtained by using specific regulations for management and feeding of the animal. Only animals between 12 and 24 mo of age that respond to specific conformation characteristics can be marked by PGI (Mengoli, 2005). Usually, animals within PGI are reared mainly in small private cow-calf operations, with indoor housing during winter and pasture during summer. Alternatively, the fattening is performed in small units in closed or open barns with straw bedding and a feeding regimen based on hay and concentrates, or maize silage, straw, and concentrates. Usually, when animals are between 18 and 24 mo of age and have achieved a good body condition (i.e., approximately equivalent to a fatness score of 2 to 3 in the EUROP carcass trait grading system; see Department of Agriculture and Food, 2004 for review), they are slaughtered in approved abattoir. All production and slaughtering phases are controlled by the Consortium of PGI, which also provides a sampling scheme to control quality variables in meat (e.g., color, shear force, cholesterol, saturated/unsaturated fatty acid ratio; Mengoli, 2005).

The initial data set obtained from the consortium of PGI consisted in raw data registered over 4 yr (i.e., from January 2004 to September 2007), and reporting for each animal, namely the identification number, code of the approved slaughterhouse, slaughter date, birth date, gender, farm in which the calf was born, farm in which each animal was fattened, CW, and muscularity and fat grades of the carcass (EUROP grading system; Department of Agriculture and Food, 2004). However, both muscularity and fat grades presented a limited variability, i.e., 44% of grades R and 49% of grades U for conformation, and 58% and 42% of grade 2 and 3 of fatness, respectively. These variables have indeed a reduced variability imposed by the PGI protocol for muscularity and marketability of carcasses for fat covering. Therefore, due to the limited range of variability of these 2 variables, both traits were not considered in the present study. Three traits were retained from raw data: i) age at slaughter (AS, d), calculated as the difference between slaughter date and birth date; ii) CW (kg), registered immediately after harvest; and iii) CWG (g/d), calculated as the ratio between CW and AS. Data editing was performed to retain data set records related only to young bulls (because of the limited number of heifers available) and herd-years and slaughterhouses having at least 2 observations. After editing, 20,872 records (10,658 for CHI, 5,440 for MAR, and 4,774 for ROM breed, respectively; Table 1) were retained for further analysis. Moreover, 50,865 animals in the pedigree files, and particularly 24,070 animals in pedigree for CHI, 13,759 for MAR, and 13,216 for ROM, respectively (Table 1), were considered. Data analyzed were paternal half-sib families. The within breed family structure and size are reported in Table 1 for each breed analyzed. Briefly, the average number of slaughtered sons per sire ranged between 10.2 for ROM to 14.4 for CHI (Table 1); on the other hand, the mean number of sons per dam was similar for all 3 breeds analyzed (Table 1).

View Full Table | Close Full ViewTable 1.

Number of animals with record, data structure, no. animals in pedigree file, no. sires and dams, and average family size within each breed considered in the study

Item Chianina Marchigiana Romagnola
Animals with record, no. 10,658 5,440 4,774
Average record per year, no. 2,665 1,360 1,194
Average record per slaughterhouse, no. 484 247 682
Average record per year × slaughterhouse, no. 121 62 171
Animals in pedigree file, no. 24,070 13,579 13,216
Sires, no. 740 417 467
Dams, no. 8,477 4,532 3,815
Average sons per sire, no. 14.4 13.0 10.2
Average sons per dam, no. 1.3 1.2 1.3

Statistical Analysis

After a preliminary ANOVA (SAS Inst. Inc., Cary, NC), performed only for nongenetic fixed effects, this mixed model was applied separately for each breed and variable analyzed:where yijk is the individual variable analyzed (i.e., AS, CW, and CWG); HYi is the fixed effect of herd-year i (1,356 levels for CHI, 1,120 for MAR, and 357 for ROM); Sj is the fixed effect of slaughterhouse j (22 levels for CHI and MAR, and 7 levels for ROM); uk is the random additive effect of each slaughtered young bull k; and eijk is the random residual term. Alternatively, the CW variable was analyzed with this model:where a regression coefficient b for AS on CW based on the inclusion of ASk (i.e., individual effect of age at slaughter) was added with respect to model [1].

In the matrix notation, the 2 models can be expressed as:where y is an N × 1 vector, β is the vector of systematic effects of order p, u is the vector of animal additive effects with order q, and e is the vector of residual effects. Furthermore, X and Z are the corresponding incidence matrices with the appropriate dimensions. All models were analyzed under a standard Bayesian approach. The join distribution of the parameters was proportional to:where A is the numerator relationship matrix among the animals and σ2u and σ2e are the additive and residual variances, respectively. The a priori distribution of u was assumed multivariate normal, as follows:Priors for β and variance components were assumed to be flat.

A bivariate analysis to estimate correlation between AS and CW was also performed. In this case, model [1] was implemented in a multivariate manner. The traits analyzed were assumed to follow a multivariate normal distribution (MVN), as well as the additive and residual effect. For these effects, the a priori distribution was:where G and R are the (co)variance matrices for the animal and residual effects, respectively; I is an identity matrix; and ⊗ is the Kronecker product operator.

Data were analyzed within breed via a Bayesian approach through the program “gibbs3f90” (Misztal, 2008), whereby the Gibbs sampling algorithm (Geman and Geman, 1984; Misztal, 2008) was applied to run univariate analyses separately for each trait or a bivariate analysis for AS and CW. The Gibbs sampler performed 990,000 iterations with a starting burn-in that discarded 90,000 samples (Raftery and Lewis, 1992). Convergence was screened by graphical inspection (i.e., drawing plots of the sampled values vs. performed iterations; Kass et al., 1998). The posterior mean of 3,000 samples (i.e., 1 of every 300 samples of the remaining 900,000) was used as a point estimate of parameters under investigation. Lower and upper bounds of the 95% highest posterior density intervals (HPD 95%) for additive and residual effects were estimated from the Gibbs samples. The posterior means and corresponding HPD 95% were also computed for all heritability estimates and correlations between AS and CW. The posterior probability of a heritability value for AS, CW, and CWG >0.20, and for a genetic correlation between AS and CW <0 was also calculated from the posterior distribution.

A rank correlation analysis was performed considering EBV derived from BLUP univariate CW analysis (CWU) or the bivariate CW analysis (CWB), and the univariate CW analysis with the age at slaughter as covariate (CWU-AS). Moreover, rank correlations between CWG and the 3 previous CW traits were obtained. Each BLUP run was performed by accounting for the appropriate estimates of (co)variances previously obtained via Bayesian approach (i.e., posterior mean). Rank correlation analyses were performed within breed by considering the first 90 bulls ranked for CWU among bulls with greater accuracy, i.e., sires with at least 12 sons slaughtered. The homogeneity across groups of breed sire were validated by considering the mean accuracy of EBV, calculated per sire with the formula reported in Mrode (2005) for EBV from progeny records.


Table 2 shows some descriptive statistics of all considered traits for each breed. The mean age at slaughter was very similar among breeds, resulting in ∼21 mo, with a CV slightly greater than 10%. More than 80% of animals were harvested within the range of 18 and 24 mo of age (82% for CHI, 81% for MAR, and 87% for ROM). The mean CW was greater for CHI young bulls than for the other 2 breeds (+10.6%). As a result, carcass ADG of CHI young bulls was on average 11% and 14% greater than for the MAR and ROM breeds, respectively (Table 2). The CV were around 13% for CW and 15% for CWG, with the exception of the ROM breed that exhibited a lower CV for CWG (12.6%). All variables analyzed in the study were normally distributed as confirmed by the Kolmogorov-Smirnov normality test (Table 2).

View Full Table | Close Full ViewTable 2.

Descriptive statistics and Kolmogorov-Smirnov normality tests (Norm. K-S; with significance within brackets; SAS Inst. Inc., Cary, NC) for age at slaughter, carcass weight, and carcass weight gain of Chianina, Marchigiana, and Romagnola young bulls retained for the study

Item Chianina Marchigiana Romagnola
Observations, no. 10,658 5,440 4,774
Age at slaughter, d
    Mean ± SD 628.2 ± 66.1 631.1 ± 72.0 645.9 ± 66.5
    Range 451.0 to 780.0 450.0 to 766.0 450.0 to 775.0
    Norm. K-S 0.056 (<0.01) 0.082 (<0.01) 0.103 (<0.01)
Carcass weight, kg
    Mean ± SD 479.0 ± 65.3 432.9 ± 58.0 433.3 ± 52.9
    Range 226.0 to 713.0 203.0 to 645.0 240.0 to 623.0
    Norm. K-S 0.015 (<0.01) 0.015 (<0.01) 0.014 (0.017)
Carcass weight gain, g/d
    Mean ± SD 768 ± 112 693 ± 107 675 ± 85
    Range 342 to 1,183 367 to 1,102 386 to 1,010
    Norm. K-S 0.013 (<0.01) 0.017 (<0.01) 0.015 (0.013)

In Table 3, the point estimates (mean and SD of marginal posterior density of the parameters) for additive genetic and residual variances, and heritability obtained from univariate analysis are summarized. The marginal posterior distributions were approximately normal, with similar mean, median, and mode (data not shown). Therefore, only the posterior mean has been reported in the table. The AS showed a rather high point estimate for additive variance in all 3 breeds, varying from 541 to 882 d2, meaning that heritability also ranged from 0.28 for CHI to 0.39 for the ROM breed. The HPD 95% for heritability of AS indicated asymmetry in the posterior densities measured for the 3 breeds, although in all breeds the probability of a heritability of AS to be >0.20 ranged from 0.98 to 1.00. The CW showed an additive genetic variance of ∼411 kg2 and heritability coefficients were lower than those for AS in all the breeds, varying from 0.13 in CHI to 0.24 in ROM. Also, in this case, the 3 breeds showed asymmetry in the marginal posterior distribution of heritability, with a more dispersed heritability for MAR than for ROM and CHI, respectively (Table 3). In this case, the probability of a heritability of CW to be >0.20 ranged from 0.01 for CHI to 0.81 for ROM. Adjusting CW by AS increased the mean point estimate for heritabilities of CW in all breeds, but nonetheless, the estimates remained less than those obtained for AS (i.e., from 0.20 to 0.30 for CHI and ROM, respectively). Indeed, the posterior means of the additive genetic variance increased adjusting CW by AS by ∼23%, as compared to the values observed for CW. On the other hand, the posterior mean of the residual variances were reduced by the use of AS as covariate on CW of ∼12%, in comparison with the values obtained for CW (Table 3). In general, the use of AS covariate for CW determined a shift of the heritability posterior mean toward greater values (Table 3), but it did not change the posterior density, resulting in similar distributions of heritability when CW was corrected or not by AS. The CWG, conversely, presented greater heritability coefficients than CW, approaching the heritability values estimated for AS, but still with asymmetry for the 3 breeds. The probability of a heritability for CWG >0.20 ranged from 0.98 for CHI to 1.00 for ROM.

View Full Table | Close Full ViewTable 3.

Estimated marginal posterior densities of genetic variance (σ2a), residual variance (σ2e), and h2 for age at slaughter (AS), carcass weight (CW), CW with AS as covariate, and carcass weight gain for Chianina, Marchigiana, and Romagnola young bulls

Item Mean ± SD HDP 95%1 Mean ± SD HDP 95%1 Mean ± SD HDP 95%1
Age at slaughter, d
    σ2a 541 ± 75 393; 689 723 ± 169 391; 1,055 882 ± 133 621; 1,142
    σ2e 1,414 ± 61 1,295; 1,534 1,372 ± 129 1,120; 1,624 1,398 ± 104 1,194; 1,603
    h2 0.276 ± 0.040 0.206; 0.346 0.344 ± 0.073 0.201; 0.486 0.386 ± 0.052 0.284; 0.488
Carcass weight, kg
    σ2a 318 ± 67 187; 449 449 ± 132 190; 708 466 ± 92 286; 647
    σ2e 2,060 ± 61 1,940; 2,180 1,537 ± 105 1,331; 1,743 1,479 ± 78 1,325; 1,633
    h2 0.134 ± 0.030 0.080; 0.187 0.225 ± 0.062 0.104; 0.347 0.239 ± 0.044 0.152; 0.326
CW with AS covariable, kg
    σ2a 440 ± 70 303; 577 507 ± 131 251; 764 548 ± 90 379; 724
    σ2e 1,789 ± 61 1,670; 1,908 1,396 ± 104 1,193; 1,599 1,278 ± 76 1,129; 1,426
    h2 0.197 ± 0.030 0.138; 0.256 0.265 ± 0.064 0.141; 0.390 0.300 ± 0.045 0.211; 0.389
CW gain, g/d
    σ2a 1,853 ± 245 1,373; 2,333 2,448 ± 541 1,388; 3,509 2,262 ± 329 1,617; 2,908
    σ2e 5,067 ± 203 4,669; 5,464 3,861 ± 407 3,065; 4,658 3,170 ± 255 2,669; 3,671
    h2 0.267 ± 0.033 0.203; 0.332 0.386 ± 0.076 0.237; 0.536 0.416 ± 0.054 0.311; 0.521
1lower and upper bounds of the 95% highest posterior density interval

The mean and SD of the posterior additive genetic and residual variances of AS and CW obtained from bivariate analysis were very close to those yielded by univariate models (data not shown). Therefore, the posterior distribution of heritability was unchanged, comparing univariate with bivariate analysis (data not shown). The environmental (i.e., residual) correlations between AS and CW were moderate and positive, and closer among all breeds considered (from 0.33 to 0.41; Fig. 1a). In addition, the posterior density obtained for the environmental correlations were progressively more dispersed (Fig. 1a) from CHI (HPD 95% between 0.33 and 0.43) to MAR (HPD 95% between 0.21 and 0.45). On the other hand, the means of the genetic correlations between AS and CW were all negative, with a point estimate (mean of posterior distribution) varying from moderate values for the CHI breed (–0.29), to low (–0.11) for the MAR breed, and very low for the ROM (–0.08) breed. However, the posterior density for the genetic correlations (Fig. 1b) indicated a wider distribution for MAR than for CHI and ROM breeds. The probability of a negative genetic correlation (i.e., <0) resulted very high for CHI (0.99) and high for the other 2 breeds (0.71 and 0.75 for MAR and ROM, respectively).

Figure 1.
Figure 1.

Estimated marginal posterior densities of environmental (a) and genetic (b) correlations obtained from bivariate animal model analysis for age at slaughter (days) and carcass weight (kg) in Chianina (continuous line), Marchigiana (dashed line), and Romagnola (dotted line) young bulls.


Table 4 reports the rank correlations between EBV for CW obtained from the analysis of CWU, CWU-AS, or CWB, and the rank correlation among these 3 traits and CWG, considering a group of 90 sires per breed, ranked according to CWU. It has to be pointed out that these sires had a different average number of sons within each breed (i.e., from 25 to 45), but depending on the different heritability values, the mean accuracies of EBV were closer among the 3 breeds (i.e., from 0.89 to 0.93). Thus, in such a homogeneous situation, the reranking of EBV of the sire was substantial in the case of CHI (r = 0.875) and ROM bulls (r = 0.897), and imperceptible for MAR (r = 0.967). When the comparison was performed, considering the breeding values obtained from the bivariate analysis of CW and AS, ranking correlations increased in all breeds, although reranking resulted still perceptible in CHI breed (mean r = 0.906) and less important in the other 2 breeds (i.e., correlation coefficient from 0.956 in ROM to 0.990 in MAR breed). As a consequence, MAR showed a less consistent variation in the ranking of the sire than in CHI, where, despite an increase in the rank correlation coefficient, a consistent variation in rankings of the sire was still detectable, indicating that a reranking in EBV estimates when correcting by AS could be greater in this breed than in MAR or ROM breed. Correlation of the 3 different expressions of CW with CWG indicated fewer, but still evident, changes in animal ranking when the correction of CW for AS was applied (mean r = 0.907; Table 4) or when CWB was considered (mean r = 0.875; Table 4).

View Full Table | Close Full ViewTable 4.

Rank correlations (and SE of correlation within brackets) obtained from EBV, produced by univariate BLUP analysis of carcass weight (CWU) or bivariate analysis carcass weight with age at slaughter (CWB), univariate CW analysis with age at slaughter as covariate (CWU-AS), and univariate analysis of carcass weight gain (CWG) in Chianina, Marchigiana, and Romagnola breeds

Rank correlation1 Chianina Marchigiana Romagnola
CWU with:
CWU-AS 0.875 (0.052) 0.967 (0.027) 0.897 (0.047)
CWB 0.861 (0.054) 0.986 (0.018) 0.973 (0.025)
CWU-AS with:
CWB 0.965 (0.028) 0.988 (0.016) 0.956 (0.031)
CWG with:
CWU 0.641 (0.082) 0.781 (0.067) 0.529 (0.090)
CWU-AS 0.913 (0.043) 0.901 (0.046) 0.828 (0.060)
CWB 0.893 (0.048) 0.856 (0.055) 0.676 (0.079)
1Correlation was obtained considering the first 90 ranked animals within breed for CWU chosen among sires with high accuracy (i.e., sires with at least 12 sons slaughtered).


The main results obtained in this study provide evidence that AS is worth being considered as a trait influenced by genetics and that it cannot be regarded solely as an environmental factor. This result was consistent among the 3 breeds analyzed in the study. Therefore, despite observing a different magnitude in the 3 breeds analyzed, a genetic component for market-slaughter precocity was detected in CHI, MAR, and ROM cattle, using a large data set from the approved abattoirs belonging to the Consortium of PGI. The heritability values estimated in the present study for AS were slightly greater than the value reported on Charolais cattle for the number of days from on test to slaughter by Johnston et al. (1992) and much greater than those observed for the age at selling found on Brown Swiss young calves in Bittante et al. (2011) and in Penasa et al. (2012), where CV was greater, but the range of age was less (i.e., 24 ± 9 d) than the AS considered in this study (i.e., 633 ± 68 d). Traditionally, age has been considered to be a totally environmental trait and, therefore, useful for correcting carcass traits to refer them to a common endpoint. Reviewing literature on genetic components for carcass traits, Rìos-Utrera (2004) outlined that a large proportion of the study reported heritability values for slaughter weight adjusted by AS; on the other hand, only a few studies, with variable results, compared estimates of heritability and genetic correlations for carcass traits adjusting to different slaughter endpoints (Lee et al., 2000; Shanks et al., 2001; Rìos-Utrera et al., 2005; Bergen et al., 2006; Rumph et al., 2007; Choy et al., 2008). However, in all literature, age was assumed to be an environmental factor, without considering that it could also reflect different market-slaughter precocity among animals (i.e., the time to reach a finishing status; Boligon et al., 2011, 2013). Nevertheless, it is well known that slaughter precocity is not entirely an environmental factor and depends on a complex of phenotypic and genetic interactions, involving feed intake, feed conversion efficiency, growth potential, and growth composition (i.e., traits for which a genetic component has been reported; Marshall, 1994; Herd and Bishop, 2000; Arthur et al., 2001a,b; Herring and Bertrand, 2002). In particular, the lean:fat deposition ratio, another trait under genetic control as demonstrated in pigs (Hermesch et al., 2000), can influence age at slaughter. Departing from this basis, the estimate of a >0 genetic component for the age at slaughter seems fully justifiable, as is the case of the present study. If the age at slaughter can mirror in some way different animal precocity in terms of body composition, its inclusion as correcting endpoint for carcass trait analysis should be very carefully considered, as observed in the present study for at least 2 of the 3 breeds analyzed (i.e., CHI and ROM). On the other hand, age at slaughter could become a trait to be taken into account to select animals for market-slaughter precocity, particularly if the more frequent fattening endpoint is based on fat amount, as in the Italian market (i.e., grades 2 and 3 of fatness in the EUROP system). It is evident that, in this market condition, different slaughter precocity means different slaughter weight, feed intake and efficiency, and possibility to dilute the cost of the stock calf. Moreover, a different feeding strategy could be required to fatten calves characterized by different slaughter precocity. On the other hand, slaughter precocity could also be of interest when fat deposition is slow because of the reduced amount of energy given to animals. For example, in Brazilian Nelore cattle fattened at pasture, a finishing precocity visual score (i.e., ability of fat deposition) is used and combined with growth rate and conformation score in weaning and yearling indexes used to select animals for meat production (Boligon et al., 2013). A different approach has been followed in the past by the American Gelbvieh Association, which computed the expected progeny deviation as “days to finish” (i.e., time to reach an appropriate amount of back fat thickness), considering this endpoint more closely related to the industry practice of finishing feedlots animals (Willmon, 2013). However, the Gelbvieh Association has also recently moved to an age-adjusted carcass expected progeny deviation to allow an easier across-breed comparison (Willmon, 2013). Therefore, days to finish has been converted to backfat measures obtained on live animals (Willmon, 2013). Considering the reduced variability of the European fatness scoring system, (i.e., only 5 classes; Department of Agriculture and Food, 2004), the backfat measure could also become a worthwhile criterion to discriminate carcass fatness in the European context. However, Continental Europe generally offers lean beef breeds and demands lean carcasses from entire young bulls, and in lean live animals, the use of ultrasound or other indirect measurement of fat are characterized by a low precision, because of the high ratio between measurement error and average fat covering. This is particularly evident in double-muscled breeds characterized by a very low fat deposition ability (homozygote Piemontese or Belgian Blue), where the choice of the optimal age at slaughter is based more on feed efficiency of animals and market trends, rather than fat covering (Schiavon et al., 2010, 2012).

As in the case of BW of calves sold at auction analyzed by Bittante et al. (2011), CW controlled at population level in the present research showed a moderate heritability, with some differences among the 3 breeds. Moreover, when AS was introduced as covariate, heritability estimates of CW remained low, although showing small increases compared with CW. In general, heritability estimates reported in the present study for CW are within the range reported in the review of Rìos-Utrera (2004) or in other studies (Shojo et al., 2006; Hickey et al., 2007).

The results of the present research outlined that, as expected, the environmental correlation between CW and AS was positive and moderate in all the 3 breeds (i.e., older animals at slaughter were characterized by heavier carcasses). However, regression coefficients of CW on AS obtained by phenotypic correlation and raw SD (255, 166, and 201 g/d for CHI, MAR, and ROM, respectively) resulted much less than the average carcass growth rate (768, 693, and 675 g/d, for CHI, MAR, and ROM, respectively) and did not reflect the growth curve of individual animals. Therefore, it seems evident that the young bulls slaughtered at older ages are characterized by a slower growth rate than animals slaughtered at younger ages. By contrast, the genetic correlations between CW and AS were negative in all the breeds, even if their entity was much less than the negative correlation between BW and age obtained on calves by Bittante et al. (2011) and Penasa et al. (2012). This means that animals with greater genetic merit for CW (i.e., fast-growing animals) tended to be slaughtered earlier than animals with a lower genetic merit for CW. However, the choice for the right moment of slaughtering is a complex decision relating to the farmer and it is strongly influenced by market characteristics and seasonality. Therefore, a better understanding of the genetic aspects of slaughter precocity is still necessary and further research on this topic is needed. In particular, investigations into the possible interactions of slaughter precocity with feed efficiency on one hand, and with muscle development and fat deposition on the other could be of great interest. Furthermore, the direct or indirect effect of selection for slaughter precocity should also take into account indirect effects on conformation and reproduction precocity of both males and females (Cammack et al., 2009; Mantovani et al., 2010; Santana et al., 2010). Last, the investigation into possible artifacts due to the nonstable slaughtering criterion seems necessary to clarify the exact magnitude of the negative correlation between age at slaughter and CW observed in this study.

An alternative way to obtain an estimation of growth potential is by using CWG (i.e., dividing CW by AS). Results of this study indicate that CWG is characterized, in all the 3 breeds studied, by heritability values greater than those obtained for CW (adjusted or unadjusted for AS) and very close to values obtained for AS. However, animals with the same genetic merit for CWG can be characterized by very different slaughter precocities and CW that are, therefore, not taken into account by CWG. In other words, CWG alone cannot completely express growth potential in beef cattle, lacking to give information about the endpoint of fattening in terms of AS and/or CW. This was also confirmed in our study by the generally low rank correlations with CW adjusted or unadjusted for AS.

The correction of CW for a given AS endpoint through phenotypic covariance in the univariate analysis in this study produced significant reranking in EBV ranking among sires in 2 of the 3 breeds considered and yielded biased results due to the fact that AS was heritable and genetically correlated with CW and not an environmental factor to consider. The correction for AS tended to yield an underestimation of growth rate of early-maturing animals and an overestimation of late-maturing ones. A better approach was to analyze CW and AS in a bivariate model, simultaneously taking into account their heritability values and correlations, especially in the case of CHI breed, characterized by a negative genetic correlation between AS and CW. Moreover, data obtained from the Consortium of PGI have confirmed that the CHI breed has a superior growth capacity, not only under performance testing at the central station of the National Breeders Association of Italian Beef Breeds (Sbarra et al., 2009), but also in field conditions, according to the traditional systems of rearing.

In conclusion, in this study, AS of the young bulls has shown a rather high coefficient of heritability in all 3 autochthonous Italian beef breeds specialized for beef production, demonstrating that AS is not a mere environmental factor to be corrected for but a trait of the animal undergoing genetic control and related to market-slaughter precocity. Taking into account market needs, in terms of fat covering of the carcasses, AS probably reflects different velocity of fat deposition among animals. Daily CWG, like ADG in performance testing, seems to be a good indicator of instant growth rate capacity of the animal but lacks to give information on the endpoint of fattening in terms of age and/or CW. Therefore, both CW and AS traits seem of interest for a complete animal evaluation all along the fattening cycle, especially considering the almost genetic independence or slightly negative correlation of the 2 traits in MAR and ROM breeds, and the moderate negative correlation in CHI breed.




Be the first to comment.

Please log in to post a comment.
*Society members, certified professionals, and authors are permitted to comment.