Variance and covariance estimates for litter size of purebred and crossbred sows

The aim of this study was the estimation of additive genetic variances and heritabilities of purebred and crossbred sows for the trait number of pigle ts born alive as well as the estimation of the genetic correlation between the litter size of purebred and crossbred sows. Two data sets were available for the analysis, which were very different in their genetic and organisational structure. At the one hand there was a da ta set from an Australian 3-way-crossbreeding program with litter performan ces of 8334 Large White sows (LW), 3262 Landrace sows (LR), 10416 LW * LR c rossbred sows and 1733 LR * LW crossbred sows. For the genetic analysis no differences were made between the two reciprocal crosses. The data sets wer e recorded from October 1991 until November 1994 and were produced on the s ame farm, which was seperated in different breeding compartments (modules). On the other hand there was a second data set from the German Bundeshybrid zuchtprogramm (BHZP). Litter preformances from 211006 purebred sows (01-sow s) and 26422 crossbred sows (31-sows) were available. After the pedigree wa s set up the data set was reduced to 57263 purebred sows and 7773 crossbred sows. The data sets were recorded from the second quarter 1991 until third quarter 1996. The litter performances of the purebred sows were recorded o n 181 different farms, the litter performances of the crossbred sows on 61 different farms. The results for the Australian data set showed heritabilities for number of piglets born alive in first, second and third litters of .059, .082 and .0 72 for Large White, .117, .158 and .063 for Landrace and .078, .083 and .09 0 crossbred sows. The genetic correlations between the litter size in first and second, first and third and second and third litter were for Large Whi te .61, .51 and .99, for Landrace .77, .88 and .88 and for crossbred sows . 88, .62 and .81, respectively. The estimated heritabilities for the German data set in first, second and t hird litters were .070, .069 and .099 for purebred and .077, .041 and .083 for crossbred sows. The genetic correlations between the litter size in fir st and second, first and third and second and third litter were for purebre d sows .808, .700 and .957 and for crossbred sows .665, .576 and .98, respe ctively. Estimated coefficients of genetic correlations between purebred and crossbr ed performance were in the Australian data set .99, .69 and .83 (litter 1-3 ) and in the German data set .80, .99 and .81 (litter 1-3), respectively. T he estimated correlations of the two very different structured data sets ar e very high, but they are only different in the second litter and confirm e ach other.