Investigating maternal effects on production traits in Duroc pigs using animal and sire models

Variance components for production traits were estimated using different models to evaluate maternal effects. Data analysed were records from the South African pig performance testing scheme on 22 224 pigs from 18 herds, tested between 1990 and 2008. The traits analysed were backfat thickness (BFAT), test period weight gain (TPG), lifetime weight gain (LTG), test period feed conversion ratio (FCR) and age at slaughter (AGES). Data analyses were performed by REML procedures in ASREML, where random effects were successively fitted into animal and sire models to produce different models. The first animal model had one random effect, the direct genetic effects, while the additional random effects were maternal genetic and maternal permanent environmental effects. In the sire model, the random effects fitted were sire and maternal grand sire effects. The best model considered the covariance between direct and maternal genetic effects or between sire and maternal grand sire effects. Fitting maternal genetic effects into the animal model reduced total additive variance, while the total additive variance increased when maternal grand sire effects were fitted into the sire model. The correlations between direct and maternal genetic effects were all negative, indicating antagonism between these effects, hence the need to consider both effects in selection programmes. Direct genetic correlations were higher than other correlations, except for maternal genetic correlations of FCR with TPG, LTG and AGES. There has been direct genetic improvement and almost constant maternal ability in production traits as shown by trends for estimated (EBVs) and maternal breeding values (MBVs), while phenotypic trends were similar to those for EBVs. These results suggest that maternal genetic effects should be included in selection programmes for these production traits. Therefore, the animal–maternal model may be the most appropriate model to use when estimating genetic parameters for production traits in this population.     

Estimation of the additive and dominance variances in SA Duroc pigs

The objective of this study was to estimate dominance variance for number born alive (NBA), 21-day litter weight (LWT21) and interval between parities (FI) in South African Duroc pigs. A total of 10,703 NBA, 6883 LWT 21 and 6881 FI records were analysed. Bayesian analysis via Gibbs sampling was used to estimate variance components and genetic parameters. Estimates of additive genetic variance were 0.554, 16.84 and 4.535 for NBA, FI and LWT21, respectively. Corresponding estimates of dominance variance were 0.246, 9.572 and 0.661 respectively. Dominance effects were statistically not significant for all traits studied. Further research utilizing a larger data set is necessary to make concrete conclusions on the importance of dominance genetic effects for the traits studied.     

Predicted response in food conversion ratio for growth by selection on the ratio or on linear component traits,in a (sequential) selection programme

1. Direct versus indirect selection for food conversion ratio of growth (FCR) after selection for live body weight (LWT), a sequential scheme often applied in broilers, was considered. In the present study loss of response in either FCR or aggregate genotype (H) when LWT was included was investigated under selection on a linear index of ratio component traits (and LWT) or FCR (and LWT) by selection index methodology.

2. Relative responses in FCR and H were generally very similar under single‐step and sequential selection. Without LWT in H, selection for linear index or ratio gave similar responses in FCR when heritabilities of components were equal. With large differences in heritabilities (0.2 compared with 0.5) or genetic and environmental correlation (>0.6) significant differences in response (5 to 12%) in FCR emerged. Therefore, whether additional costs are justified for parameter and breeding value estimation when using a linear index in place of selection for the ratio depends on the difference in heritabilities and correlations of the ratio component traits. With LWT in H, loss of response in FCR was partially or entirely offset by response in LWT.

3. The non‐normality of FCR and consequences for (co)variance component estimation were studied in terms of the coefficients of variation of the component traits of FCR. Restricted Maximum Likelihood (REML) estimation of (co)variance components for both FCR and logarithm transformed FCR (closer to normality) showed the robustness of REML to such deviations from normality.     

Genetic, phenotypic and environmental relationships between sow body weight and sow productivity traits

Yorkshire and Duroc litter records were used to estimate genetic, phenotypic and environmental relationships between sow body weight and sow productivity traits. Two data sets with two subsets each were used to complete this study; 663 and 460 records included litter traits only, while 522 and 359 records also contained sow body weight for Yorkshires and Durocs, respectively. Heritability estimates for number born (NB), number born alive (NBA), total birth weight of live pigs (BWLIT), litter weight at 3 wk (WT3WK), sow weight at parturition (WTDAMPAR) and sow weight at weaning (WTDAMWN) were .24 +/- .14, .21 +/- .14, .42 +/- .16, .19 +/- .14, .72 +/- .21 and .42 +/- .18, respectively, for Yorkshires and .05 +/- .10, .04 +/- .10, .21 +/- .14, .25 +/- .15, .85 +/- .25 and .87 +/- .26, respectively, for the Durocs. Repeatability estimates for NB, NBA, BWLIT, WT3WK, WTDAMPAR and WTDAMWN were .13 +/- .06, .17 +/- .06, .27 +/- .06, .13 +/- .06, .64 +/- .05 and .54 +/- .05, respectively, for Yorkshires and .17 +/- .06, .21 +/- .06, .14 +/- .06, .17 +/- .06, .28 +/- .07 and .39 +/- .07, respectively, for Durocs. Genetic correlations among litter traits were high and positive in the Yorkshire data. Genetic correlations between NBA and WTDAMPAR, NBA and WTDAMWN, WT3WK and WTDAMPAR, and WT3WK and WTDAMWN were .37 +/- .25, .18 +/- .34, .60 +/- .29 and .29 +/- .45, respectively, in the Yorkshire data. Genetic correlations among litter traits in the Duroc analysis had large standard errors but were generally similar to the estimates obtained from the Yorkshire data. The genetic correlation between WTDAMPAR and WTDAMWN was .93 +/- .09 for Yorkshire sows. The primary conclusion from this study is that as selection increases sow productivity traits, there will be a positive correlated response in sow body weight.     

杜洛克猪专门化品系选育研究

杜洛克猪专门化品系是以丹系杜洛克为育种素材,采用不完全闭锁的群体继代选育法,运用最佳线性无偏估计(BLUP)法、综合选择指数和分子标记辅助选择(MAS)等育种新方法,主选日增重和活体背膘厚等性状。经5个世代选育,产仔数达10.48头,产活仔数9.85头;肥育期日增重831g,料重比2.55:1,活体背膘厚11.83mm;瘦肉率69.5%,肌内脂肪为2.95%,主选性状全面达到或超过育种目标,成功培育了一个高性能的专门化父本新品系(ZFD系)。     

快速型黄羽肉鸡饲料利用效率性状的基因组选择研究

旨在探究快速型黄羽肉鸡饲料利用效率性状的遗传参数,评估不同方法所得估计育种值的准确性。本研究以自主培育的快速型黄羽肉鸡E系1 923个个体（其中公鸡1 199只,母鸡724只）为研究素材,采用"京芯一号"鸡55K SNP芯片进行基因分型。分别利用传统最佳线性无偏预测（BLUP）、基因组最佳线性无偏预测（GBLUP）和一步法（SSGBLUP）3种方法,基于加性效应模型进行遗传参数估计,通过10倍交叉验证比较3种方法所得估计育种值的准确性。研究性状包括4个生长性状和4个饲料利用效率性状：42日龄体重（BW42D）、56日龄体重（BW56D）、日均增重（ADG）、日均采食量（ADFI）和饲料转化率（FCR）、剩余采食量（RFI）、剩余增长体重（RG）、剩余采食和增长体重（RIG）。结果显示,4个饲料利用效率性状均为低遗传力（0.08~0.20）,其他生长性状为中等偏低遗传力（0.11~0.35）;4个饲料利用效率性状间均为高度遗传相关,RFI、RIG与ADFI间为中度遗传相关,RFI与ADG间无显著相关性,RIG与ADG间为低度遗传相关。本研究在获得SSGBLUP方法的最佳基因型和系谱矩阵权重比基础上,比较8个性状的估计育种值准确性,SSGBLUP方法获得的准确性分别比传统BLUP和GBLUP方法提高3.85%~14.43%和5.21%~17.89%。综上,以RIG为选择指标能够在降低日均采食量的同时保持日均增重,比RFI更适合快速型黄羽肉鸡的选育目标;采用最佳权重比进行SSGBLUP分析,对目标性状估计育种值的预测性能最优,建议作为快速型黄羽肉鸡基因组选择方法。     

Genetic relationship between different measures of feed efficiency and its component traits in Japanese Black (Wagyu) bulls

Genetic parameters of average daily gain (ADG), metabolic body weight (MWT), body weight at finish (BWF), daily feed intake (DFI), feed conversion ratio (FCR), and residual feed intake (RFI) were estimated in 740 Japanese Black bulls. RFI was calculated as the difference between actual and expected feed intake predicted by the residual of multiple regression (RFI_phe) and genetic regression (RFI_gen) from the multivariate analysis for DFI, MWT, and ADG. The estimations were made for the test periods of 140 days (77 bulls) and 112 days (663 bulls). The mean for RFI_phe was close to zero and RFI_gen was negative. Most of the traits studied were moderately heritable (ranging from 0.24 to 0.49), except for ADG and FCR (0.20 and 0.15, respectively). The genetic correlations among growth traits (ADG, MWT and BWF) and between DFI and growth traits were high, while the phenotypic correlations between them were moderate to high. The genetic and phenotypic correlations between RFI_phe and RFI_gen were > 0.95 implying that they are regarded as the same trait and the genetic correlations of RFI (RFI_phe and RFI_gen) with FCR and DFI were favorably high. RFI_phe was phenotypically independent of its component traits, MWT (r_p = − 0.01) and ADG (r_p = 0.01). RFI_gen was genetically independent of MWT (r_g = − 0.07), while there was a weak genetic relationship (r_g = 0.18) between RFI_gen and ADG. These results provide evidence that RFI_gen should be included for genetic improvement of feed efficiency in Japanese Black breeding program.     

Persistency of accuracy of genomic breeding values for different simulated pig breeding programs in developing countries

Genetic improvement of pigs in tropical developing countries has focused on imported exotic populations which have been subjected to intensive selection with attendant high population‐wide linkage disequilibrium (LD). Presently, indigenous pig population with limited selection and low LD are being considered for improvement. Given that the infrastructure for genetic improvement using the conventional BLUP selection methods are lacking, a genome‐wide selection (GS) program was proposed for developing countries. A simulation study was conducted to evaluate the option of using 60 K SNP panel and observed amount of LD in the exotic and indigenous pig populations. Several scenarios were evaluated including different size and structure of training and validation populations, different selection methods and long‐term accuracy of GS in different population/breeding structures and traits. The training set included previously selected exotic population, unselected indigenous population and their crossbreds. Traits studied included number born alive (NBA), average daily gain (ADG) and back fat thickness (BFT). The ridge regression method was used to train the prediction model. The results showed that accuracies of genomic breeding values (GBVs) in the range of 0.30 (NBA) to 0.86 (BFT) in the validation population are expected if high density marker panels are utilized. The GS method improved accuracy of breeding values better than pedigree‐based approach for traits with low heritability and in young animals with no performance data. Crossbred training population performed better than purebreds when validation was in populations with similar or a different structure as in the training set. Genome‐wide selection holds promise for genetic improvement of pigs in the tropics.     

Genetic correlations of pig performance and sow productivity traits

Records for a total of 732 daughter-dam pairs were analyzed to estimate the genetic correlations of pig performance traits with sow productivity traits, with implications to the development of specialized sire and dam lines for use in crossing. Major pig performance traits analyzed included average daily gain from 56 d of age to a final weight of 90.7 kg (ADG), average backfat thickness at 90.7 kg (BF) and a performance index (PI) consisting of ADG and BF. Major sow productivity traits included number of pigs born alive in a litter (NA), litter size (N21) and litter weight (W21) at 21 d of age and two sow productivity indexes, one with NA, N21 and W21 (SPI-3) and one with NA and W21 (SPI-2). All records were expressed as deviations from breed-line-year-season means of this population. Genetic correlations were computed from daughter-dam covariances. The mean genetic correlation of PI with SPI consisted of two correlations, that of daughters' PI with dams' SPI and that of dams' PI with daughters' SPI. The mean genetic correlation of PI with SPI-3 and SPI-2 was .07 +/- .12, suggesting that concurrent improvement in both PI and SPI would not be restricted by selection within a single composite line. The genetic correlation of daughters' PI with dams' SPI (-.18 +/- .13) was appraised as more critical than the reciprocal correlation of dams' PI with daughters' SPI (+.28 +/- .13). This appraisal is based on the fact that only one generation separates a daughter's PI from her dam's SPI, as compared with two generations in the reciprocal covariance. However, the -.18 correlation was not significantly different from zero, indicating that formation of specialized sire and dam lines for use in crossing would be only marginally more effective at best for improving the overall efficiency in pork production than use of a single composite line, aside from the heterosis effects from crossing the lines. Indexes were proposed for combining PI and SPI for use either in specialized sire and dam lines or in a single composite line.     

Genetic parameters and trends for production traits and their relationship with litter traits in Landrace and Yorkshire pigs

Genetic parameters and trends in the average daily gain (ADG), backfat thickness (BF), loin muscle area (LMA), lean percentage (LP), and age at 90 kg (D90) were estimated for populations of Landrace and Yorkshire pigs. Additionally, the correlations between these production traits and litter traits were estimated. Litter traits included total born (TB) and number born alive (NBA). The data used for this study were obtained from eight farms during 1999 to 2016. Analyses were carried out with a multivariate animal model to estimate genetic parameters for production traits while bivariate analyses were performed to estimate the correlations between production and litter traits. The heritability estimates were 0.52 and 0.43 for ADG; 0.54 and 0.45 for BF; 0.25 and 0.26 for LMA; 0.54 and 0.48 for LP; and 0.56 and 0.46 for D90 in the Landrace and Yorkshire breeds, respectively. The ADG and D90 showed low genetic correlation with BF and LP. The LMA had ?0.40, ?0.32, 0.49, and 0.39 genetic correlations with ADG, BF, LP, and D90, respectively. Genetic correlations between production and litter traits were generally low, except for the correlations between LMA and TB (?0.23) in Landrace and ADG and TB (?0.16), ADG and NBA (?0.18), D90 and TB (0.19), and D90 and NBA (0.20) in Yorkshire. Genetic trends in production traits were all favorable except for LMA.     

Inferring phenotypic causal structure among farrowing and weaning traits in pigs

Direct selection for litter size or weight at weaning in pigs is often hindered by external interventions such as cross‐fostering. The objective of this study was to infer the causal structure among phenotypes of reproductive traits in pigs to enable subsequent direct selection for these traits. Examined traits included: number born alive (NBA), litter size on day 21 (LS21), and litter weight on day 21 (LW21). The study included 6,240 litters from 1,673 Landrace dams and 5,393 litters from 1,484 Large White dams. The inductive causation (IC) algorithm was used to infer the causal structure, which was then fitted to a structural equation model (SEM) to estimate causal coefficients and genetic parameters. Based on the IC algorithm and temporal and biological information, the causal structure among traits was identified as: NBA → LS21 → LW21 and NBA → LW21. Owing to the causal effect of NBA on LS21 and LW21, the genetic, permanent environmental, and residual variances of LS21 and LW21were much lower in the SEM than in the multiple‐trait model for both breeds. Given the strong effect of NBA on LS21 and LW21, the SEM and causal information might assist with selective breeding for LS21 and LW21 when cross‐fostering occurs.     

苏博美利奴羊主要经济性状的遗传参数估计

苏博美利奴羊是我国近年来培育出的超细型细毛羊品种,为了解该品种主要经济性状的遗传参数以及群体改变的速度和方向,本研究收集2014—2020年新疆巩乃斯种羊场苏博美利奴羊的周岁母羊鉴定记录共计9140条、剪毛称重记录共计7158条以及初生重记录5540条;利用DMU软件采用AIREML模块结合EM算法配合单性状动物模型估...     

A genetic analysis of sow productivity traits

Genetic and phenotypic parameters for sow productivity traits were derived from analyses of records for 999 purebred Duroc and Yorkshire litters at the Beltsville Agricultural Research Center. Included were 682 spring-farrowed litters and 317 fall-farrowed litters. Primary traits examined were number of pigs born alive in the litter (NA), litter size at 21 d of age (N21), litter weight at 21 d (W21) and an index (I21) combining the three traits. Also analyzed were total number of pigs farrowed alive or stillborn, litter weight of pigs born alive and litter size and litter weight at 42 d of age (weaning). Selection was based mainly on the sow productivity index among litters in the spring season and mainly on performance traits among fall-farrowed pigs. Select and control lines were maintained in both breeds in both seasons. Yorkshires started at higher levels of productivity than Durocs in spring 1975, showed more fluctuation and ended in spring 1982 at about their starting levels. Durocs started at lower productivity levels and recorded generally significant increases by 1982, with comparable selection differentials for the two breeds. Heritabilities, estimated from daughter-biparental (mid-dam, paternal granddam) regressions using records expressed as deviations from contemporary breed-line means, varied widely among the breed-line groups. Suggested composite heritability estimates were 5, 10 and 20% for NA, N21 and W21, respectively. Genetic correlations, computed from daughter-biparental covariances, were high and positive (.7 or above) among all traits. Based on estimated genetic gains, it was proposed that W21, alone or in combination with N21 and(or) NA in an index, be used as the primary selection criterion to bring about genetic improvement in sow productivity traits.     

东北梅花鹿优化育种规划中育种目标的确定

讨论了东北梅花鹿依据经济指标确定育种目标的问题 ,以及在东北梅花鹿育种规划中 ,应如何挑选在综合育种目标中的生产性状及其对应的选择性状。并着重介绍了应用综合育种值作为数量化育种目标的方法。指出在确定以茸用为主的东北梅花鹿育种目标时 ,除了主要考虑茸用性状外 ,还应适当地考虑繁殖和使用寿命等性状 ,同时对于直接影响茸鹿生产效益的直接性状和次级性状不容忽视。     

Effects of incorporating environmental cost and risk aversion on economic values of pig breeding goal traits

Economic values (EVs) of traits, accounting for environmental impacts and risk preferences of farmers, are required to design breeding goals that contribute to both economic and environmental sustainability. The objective of this study was to assess the effects of incorporating environmental costs and the risk preferences of farmers on the EVs of pig breeding goal traits. A breeding goal consisting of both sow efficiency and production traits was defined for a typical Brazilian farrow‐to‐finish pig farm with 1,500 productive sows. A mean‐variance utility function was employed for deriving the EVs at finishing pig level assuming fixed slaughter weight. The inclusion of risk and risk aversion reduces the economic weights of sow efficiency traits (17%) while increasing the importance of production traits (7%). For a risk‐neutral producer, inclusion of environmental cost reduces the economic importance of sow efficiency traits (3%) while increasing the importance of production traits (1%). Genetic changes of breeding goal traits by their genetic standard deviations reduce emissions of greenhouse gases, and excretions of nitrogen and phosphorus per finished pig by up to 6% while increasing farm profit. The estimated EVs could be used to improve selection criteria and thereby contribute to the sustainability of pig production systems.     

Variance components due to direct genetic, maternal genetic and permanent environmental effect for growth and feed-efficiency traits in young male Japanese Black cattle

Variance components and genetic parameters were estimated using data recorded on 740 young male Japanese Black cattle during the period from 1971 to 2003. Traits studied were feed intake (FI), feed‐conversion ratio (FCR), residual feed intake (RFI), average daily gain (ADG), metabolic body weight (MWT) at the mid‐point of the test period and body weight (BWT) at the finish of the test (345 days). Data were analysed using three alternative animal models (direct, direct + maternal environmental, and direct + maternal genetic effects). Comparison of the log likelihood values has shown that the direct genetic effect was significant (p < 0.05) for all traits and that the maternal environmental effects were significant (p < 0.05) for MWT and BWT. The heritability estimates were 0.20 ± 0.12 for FI, 0.14 ± 0.10 for FCR, 0.33 ± 0.14 for RFI, 0.19 ± 0.12 for ADG, 0.30 ± 0.14 for MWT and 0.30 ± 0.13 for BWT. The maternal effects (maternal genetic and maternal environmental) were not important in feed‐efficiency traits. The genetic correlation between RFI and ADG was stronger than the corresponding correlation between FCR and ADG. These results provide evidence that RFI should be included for genetic improvement in feed efficiency in Japanese Black breeding programmes.     

国外生猪育种体系简析及对我国生猪育种的几点思考

选择是生猪育种的核心,而准确地选择和选种依赖于完善和成熟的育种体系。国际养猪发达国家(国际育种企业)对育种体系不断完善,长期进行生长、繁殖和肉品质等各类性状的表型测定,运用多场联合评估和基因组选择等方法不断提高遗传评估的准确性,不断挖掘新的育种性状,并进行测定、评估和选育,实现了种猪群体持续的遗传改良和综合性能的不断提升。本文对欧美发达国家(国际育种企业)的生猪育种技术体系进行了综述,以期为我国的生猪育种工作提供参考,并促进我国生猪种业的创新发展。     

猪PLA2G4A基因单核苷酸多态性及其与饲料报酬性状的关联分析

猪肉是世界第一大肉类消费品,通过提高饲料报酬来降低猪的养殖成本有巨大的经济意义。PLA2G4A基因作为早期脂肪分化的重要基因,已经在多个物种上进行研究。本实验以2016年出生于北京中育种猪有限责任公司的166头大白猪为材料,研究PLA2G4A基因与平均日采食量、饲料转化率、剩余饲料采食量和平均日增重4个饲料报酬性状的关系。结果发现:猪PLA2G4A位于外显子1、12、13、21的4个SNP位点;内含子2、2、11、12、16的5个SNP位点,共计9个SNP位点;对这9个SNP和4个饲料报酬相关性状进行关联分析发现,位于内含子16的SNP位点与猪的平均日增重有极显著相关,而其他SNP位点与这4种饲料报酬性状均无显著相关。     

Population parameter estimates for performance and reproductive traits in Polish Large White nucleus herds

Performance test records from on-farm tests of young Polish Large White boars and reproductive records of Polish Large White sows from 94 nucleus farms during 1978 to 1987 were used to estimate population parameters for the measured traits. The number of boar performance records after editing was 114,347 from 3,932 sires, 21,543 dams, 44,493 litters and 1,075 herd-year-seasons. Reproductive performance records of sows involved 41,080 litters from 2,348 sires, 18,683 dams and 1,520 herd-year-seasons. Both data sets were analyzed by using restricted maximum-likelihood programs. The model used for the performance records included fixed herd-year-seasons, random sires, dams and error effects, and covariances for the year of birth of sire and year of birth of dam. The model used for the reproduction data set was the same as the performance data with parity as an additional fixed effect. Estimated heritabilities were .27, .29, .26, .07, .06, .06 for average daily gain standardized to 180 d (ADG), backfat thickness standardized to 110 kg BW (BF), days to 110 kg (DAYS), litter size at birth born alive (NBA), litter size at 21 d (N21) and litter weight at 21 d (W21), respectively. Estimated common environmental effects for the same traits were .09, .10, .09, .06, .07 and .08, respectively. Genetic correlations were .25 (ADG and BF), -.99 (ADG and DAYS), -.21 (BF and DAYS), .91 (NBA and N21), .68 (NBA and W21) and .80 (N21 and W21). The respective phenotypic correlations were .23, -.99, -.20, .88, .75, .86. These population parameters for Polish Large White pigs are similar to those for breeds in other countries.     

Genetic parameters for measures of the efficiency of gain of boars and the genetic relationships with its component traits in Duroc pigs

Genetic parameters for the efficiency of gain traits on 380 boars and the genetic relationships with component traits were estimated in 1,642 pigs (380 boars, 868 gilts, and 394 barrows) in 7 generations of a Duroc population. The efficiency of gain traits included the feed conversion ratio (FCR) and residual feed intake (RFI) and their component traits, ADG, metabolic BW (MWT), and daily feed intake (FI). The RFI was calculated as the difference between the actual and expected FI. The expected FI was predicted by the nutritional requirement and by the residual of phenotypic (RFI(phe)) and genetic (RFI(gen)) regressions from the multivariate analysis for FI on MWT and ADG. The means for RFI(phe) and RFI(gen) were close to zero, and the mean for nutritional RFI was negative (-0.11 kg/d). The traits studied were moderately heritable (ranging from 0.27 to 0.53). The genetic and phenotypic correlations between ADG and FI were moderate to high, whereas the genetic correlation between MWT and FI was moderate, and the phenotypic correlation between them was low. The corresponding correlations between RFI(phe) and RFI(gen) were > 0.95, implying that they can be regarded as the same trait. The genetic and phenotypic correlations of FCR with measures of RFI were high but lower than unity. The RFI(phe) was phenotypically independent of its component traits, MWT (r(p) = 0.01) and ADG (r(p) = 0.03). The RFI(gen) was genetically independent of MWT (r(g) = -0.04), whereas there was a weak genetic relationship (r(g) = 0.15) between RFI(gen) and ADG. Residual FI was more heritable than FCR, and the genetic and phenotypic correlations of RFI(phe) and RFI(gen) with FI were positive and stronger than that of FCR with FI. These results provide evidence that RFI(phe) or RFI(gen) should be included in breeding programs for Duroc pigs to make genetic improvement in the efficiency of gain.