ABSTRACT Traits associated with meat performance are characterized by a complex inheritance model, including the effects of non-allelic interactions. The aim of this study was to estimate the significance of non-allelic interactions between pairs of loci in each of two alternative groups based on four important performance traits of ducks: body weight (BW), carcass yield calculated as the ratio of carcass weight to body weight (CY), breast and leg muscle yield as a fraction of carcass weight (BL-CW), and skin weight and abdominal fat, again as a fraction of carcass weight (SF-CW). The experimental material was based on an F2 cross between two parental lines: A-55 (Polish Peking) and GL-30 (French Peking). In total, 368 individuals were recorded. All of the birds (including parents and F1) were genotyped for 23 polymorphic microsatellite loci. Entropy was taken as a measure of interaction between pairs of loci, and the chi-square test was used to verify significance. In total, 253 pairs of loci were analyzed in both categories of each of the four traits. Statistically significant interactions were obtained for each trait. However, it was observed that some of the loci showed a greater tendency to have significant opposite interactions for alternative categories of the same trait. The results indicate the dependence of the recombination rate within pairs of loci on the level of each of the four traits.

ABSTRACT The objective of this study was to identify genome regions determining duck meat performance traits with possible small variation. In total, 368 crossbred ducks of F2 generation obtained from two parental lines: Pekin-type ducks of Polish origin (A55) and Pekin-type ducks of French origin (GL-30) were recorded. The following seven traits were analyzed: body weight, breast muscle weight, leg muscle weight, water holding capacity in the breast and leg muscles, and color lightness L* of the breast and leg muscles. All birds (including parental and F1 generations) were genotyped (29 microsatellite markers). Means and coefficients of variation (CV) were calculated for 28 full-sibs (four sires by six dams and one sire by four dams). Number of progeny per full-sib group ranged from 7 to 17. The multivariate cluster analysis using grouping by k-means algorithm was used on transformed data. The multivariate cluster analysis gave two clusters: first group with 10 full-sibs and second one with 18 families. Differences among half-sibs in the CV of the recorded traits were determined. It should be noted that one out of five sire groups showed statistically significant differences from the other ones. Moreover, the CVs in this group were smaller. The analysis of microsatellite markers indicated three alleles from three loci were present only in the “superior” sire group. The obtained results indicate a promising opportunity of effective selection for improving carcass technological quality using molecular markers.