Effects of Direct Additive and Direct Dominance on Litter Traits in Crossbred Sows between Danish Yorkshire and Danish Landrace Pigs in Vietnam

Reproductive improvement of sows depends on suitable selection programs and crossbreeding strategy. Thanks to the heterosis, the diallel crossing system between Yorkshire and Landrace pigs is currently applied to increase reproductive efficiency of crossbred sows. However, to identify an efficient breeding scheme, crossbreeding parameters need to be estimated (Ibáñez-Escriche et al., 2014), especially for additive and dominant effects which are considered as the genetic basis of heterosis (Li et al., 2008). The genetic models based on direct and maternal additive and dominance effects were developed for generating crossbreeding parameters (Eisen et al., 1983). Moreover, these models were used as genetic tools to predict performance of un-tested crossbred groups in practical. In addition, combination effects generated by genetic differences among various breeds provides essential information to guide efficient use of genetic resources in pig crossbreeding systems (Cassady et al., 2002).
 
Effects of direct and maternal dominant genetic on growth, carcass quality traits in the crossbreeding system of Hampshire, Landrace and Large White have been documented (Baas et al., 1992; Bittante et al., 1993). However, these effects were only observed in crossbred terminal sires Duroc and Pietrain (Tinh et al., 2015), Pietrain and Landrace (Hop ​et al., 2015) and between Duroc and Landrace (Hào et al., 2015). Additionally, heterosis and combination effects on growth, carcass quality and reproduction in pig crossbreeding system were also reported in pig breeding scheme (Cassady, et al., 2002). Thus, this study is aimed at estimating the effects of direct additive and dominance on litter traits and predicting the reproduction of un-tested crossbred sows between Danish Yorkshire and Landrace pigs in Vietnam.

Materials
 
To estimate the crossbreeding effects, from 2014 and 2017, 1979 litters including 1,308 litters of purebred Danish Yorkshire and Landrace sows and 671 litters of crossbred sows, including YL (Yorkshire-Sire × Landrace-Dam), LY (Landrace-Sire × Yorkshire-Dam), L(YL) and L(LY) were obtained between in two pig breeding farms: Binh Thang and Khang Minh An. Three groups of un-tested crossbred sows were also included in analysis for predicting litter performance based on estimates of crossbreeding effects including Y (YL), Y (LY) and (YL)(LY). Analyzed traits in this study were total number born (TNB), number born alive (NBA), number weaned (NW) and average piglet weight at weaning (AWW).
 
Statistical analysis
 
Heterosis in percentage was identified by formula: Where,
H  = Heterosis (%).
XF=Average performance in progenies.
Xp = Average performance in parents.
 
Statistical analysis on phenotype of crossbred groups for litter traits was performed using SAS software (version 9.3.1) by GLM (General Linear Model) procedure and statistical model:
  Where,
Yij = Measurements of tested traits.
µ = Average value of observations.
αi = Genetic group effect.
eij = Random error.
 
Crossbreeding effects (direct additive, direct dominance, maternal additive, maternal dominance) and predicted performance of un-tested crossbred groups were estimated using software CBE (Wolf, 1996) by Jakubec statistical model:
  Where,
G = Average phenotype of crossbred groups.
µ = Average phenotype of crossbred groups.
Ad  = Direct additive effect.
Dd = Direct dominant effect.
Am=  Maternal additive effect.
Dm= Maternal dominant effect.
e   =  Random error.

Litter performance and heterosis in crossbred sows
 
In inbreeding systems, increased proportion of homozygosity leads to inbreeding depression, particularly for traits related to reproduction, early growth and survival. However, the level of inbreeding in off spring will be diminished and traits of interest will be improved when two genetically different populations are crossed (McLaren, et al., 1987). The value of litter traits at birth and at weaning period was presented in Table 1 and Table 2, respectively. Our results showed that for all litter traits of TNB, NBA, NW and AWW, value differences between purebred and crossbred groups were statistically significant. In F1 crossbred sows’ generation, heterosis effect was observed in all studied traits and insignificant difference was observed between YL and LY crossbred groups including 6.7-6.4% for TNB, 5.4-6.2% for NBA, 4.4-4.8% for NW and 5.7-6.0% for AWW traits. This indicated that there was not any remarkable distinction in selecting Yorkshire or Landrace as sire or dam when targeting for reproduction traits.




 
Previous studies indicated that in crossbred sows, the manifestation of heterosis effect can be observed on litter traits including litter size, litter weaning weight and pig survival (Lukač et al., 2012; McLaren et al., 1987). However, there was a noticeable difference between YL and LY groups in diallel crossing system for litter traits was reported (Dragomir 2013; Kantanamalakul et al., 2007). In this study, insignificant difference between YL and LY crossbred sows for litter traits could be explained by closer genetic structure of pure Danish Yorkshire and Landrace, which could be also one of the causes for moderate heterosis displayed in current study as compared with reported studies. Additionally, environment can be a have remarkable influences to the level of heterosis present in a crossbred population (Sheridan 1986). Therefore, genetic evaluation program and environmental improvement should be carried out simultaneously in these genetic studies under practical conditions of Vietnam.
 
As crossbred female individuals from F1 generation were used as dams to mate with pure parent for the next generation, the rotational crossbreeding system substantially reduces the proportion of pure line parental stocks. The level of heterosis in the crosses of rotational crossbreeding system was lower as compared to a terminal crossing system (Sheridan 1986). This phenomenon was also observed in this study. For crossbred sows of L(YL) and L(LY), a substantial decreasing trend was found for all studied litter traits as compared to F1 crossbred groups. However, the performance of these traits was slightly higher than that in purebred sows by 1.4-3.0% for these traits.
 
Estimates of crossbreeding effects for litter traits in crossbred sows
 
In crossbreds, dominance effect and the combination of dominance effect and over-dominance effect are the main components influencing heterosis (Li et al., 2008). Over-dominance existence has also been observed for many traits in animal breeding programs based on genetic pleiotropism (Boysen et al., 2010; Dagnachew et al., 2011; Ishikawa 2009). In our study, over-dominance effect was not separately analyzed. The effect of direct dominance was the most to studied traits of crossbred sows YL and LY in relation to other effects, such as 0.92 piglets for TNB; 0.73 piglets for NBA; 0.55 piglets for NW and 0.37 kg for AWW (Table 3). However, thanks to the controlled inbreeding level in pure-breeding programs, heterosis in current crossbred population was at moderate level, even for low heritability traits as litter size, number born alive, number piglet and weight at weaning. In addition, the epistatic interaction, which could be more important as some genes linked closely together, was not investigated in this study (Abasht and Lamont 2007). Therefore, further studies on epistatic interaction will be needed for litter traits in Danish genetics of Y and L to improve the efficiency of breeding scheme.


 
Predicted litter performance of un-tested crossbred sows
 
Beside tested crossbred groups, we also predicted the value of TNB, NBA, NW and AWW traits in three un-tested groups based on estimated components of genetic effects from tested crossbred groups. The predicted phenotypes of litter traits in Y(YL), Y(LY) and (YL)(LY) sows were presented in Table 4 (at birth) and Table 5 (at weaning).




       
Table 4 showed that deviation between observed and estimated values in tested crossbred groups was remarkably low for TNB trait (0.00 to 0.06 piglets) and NBA trait (0.00 to -0.11 piglets). This indicated the high accuracy of predicted means for both TNB and NBA trait in un-tested crossbred groups. Predicted values for TNB and NBA trait in Y(YL), Y(LY) and (YL)(LY) were lower than that of F1 crossbred groups YL and LY, but equal to that of rotational crossbred groups L(YL) and L(LY). Similarly, for the NW and AWW traits, the accuracy of predicted values in un-tested crossbred groups was considerably high, which reflected in the low deviations between observed and estimated values, from 0.00 to -0.06 piglets for NW and from 0.00 to -0.12 kg for AWW. As a result, in un-tested crossbred groups of Y(YL), Y(LY) and (YL)(LY), predicted reproduction means were equal to rotational crossbred groups of L(YL) and L(LY), but lower than F1 crossbred groups of YL and LY.

Conclusively, our data in current research suggested that there was not any difference between Danish Y and L purebred as well as between LY and LY crossbred sows for studied traits (TNB, NBA, NW and AWW). This means using of purebred Y or L individuals as sire or as dam to produce crossbred sows in dialed crossing system did not affect to reproduction traits of F1 crossbred progenies under practical conditions in Vietnam. As for rotational crossbreeding system, such as, L (YL), Y(YL), L(LY) and L(YL), the proportion of Y or L purebred parent was reduced in progeny generation and lower heterosis for litter traits was observed as compared with F1 crossbred sows. It resulted in reproduction dropped down in rotational crossbred sows but was still higher than purebred parents for those traits. Therefore, as mentioned to biological performance, the diallel crossing system between Danish Yorkshire and Landrace to produce F1 crossbred sows should be recommended. But economically, the rotational crossing system among these purebreds should be considered to apply for specific production conditions to continue using the advantage of crossbred dams.

Under practical in Vietnam, direct dominance component was the most positive effect on litter traits of total number born, number born alive, number weaned and average piglet weight at weaning in crossbred sows produced from Danish purebred Yorkshire and Landrace.
 
The manifestation of heterosis in F1 crossbred sows was at moderate level for studied litter traits, respectively; 6.7-6.4% for total number born; 5.4-6.2% for number born alive; 4.4-4.8% for number weaned and 5.7-6.0% for average piglet weight at weaning. As applying the rotational crossing system, reproduction dropped down as compared with F1 diallel crossbred sows, but was higher by 1.4-3% than purebred parents for litter traits. Economically, the rotational crossing system among purebreds of Danish genetics of Yorkshire and Landrace should be considered to apply for specific production conditions to continue using the advantage of crossbred dams; and also a further study on epistatic interaction should be done for litter traits in to improve the efficiency of breeding scheme.

Samples collected were used only for routine diagnostic purpose of the breeding programs and not specifically for the purpose of this project. Therefore, approval of an ethics committee was not mandatory. Sample collection and data recording were conducted strictly according to the Vietnamese law on animal protection and welfare.