Genetic analysis for farrowing rate and litter size for Landrace and Yorkshire sows in South China

Abstract

Abstract Farrowing rate (FR) and litter size (LS) are economically important traits that directly affect the number of weaned piglets per sow per year. This study aimed to estimate genetic parameters and calculate genetic trends for FR and LS at birth in herds of Landrace and Yorkshire sows in South China. We analyzed 16,451 Landrace sows with 58,091 observations and 33,953 Yorkshire sows with 101,486 observations for FR, and we analyzed 15,079 Landrace sows with 47,831 litters and 31,470 Yorkshire sows with 83,974 litters for LS traits. We used a single-trait threshold model to analyze FR. A joint multi-trait linear-threshold model was used to analyze three LS traits simultaneously: total number of piglets born (NBT), number of piglets born alive (NBA), and number of piglets born dead (NBD). Estimated heritability values for FR were 0.112±0.012 in Landrace and 0.120±0.010 in Yorkshire sows. For the three LS traits, the highest heritability was estimated for NBT, followed by NBA and NBD in both breeds, with values ranging from 0.044±0.004 to 0.109±0.008 in Landrace and from 0.039±0.004 to 0.108±0.007 in Yorkshire. Estimated correlations between NBT and NBA ranged from 0.861±0.018 to 0.949±0.010 in Landrace and from 0.895±0.009 to 0.933±0.009 in Yorkshire for random additive genetic effects, random permanent effects, and random service sire effects. Estimates of correlations between NBT and NBD ranged from −0.010±0.092 to 0.353±0.058 in Landrace and from 0.253±0.065 to 0.516±0.044 in Yorkshire. Estimates of correlations between NBA and NBD were not significantly different from 0 in either population, except for additive genetic effects in Yorkshire sows (0.267±0.059). Among the genetic trends, NBT and NBA showed persistent improvements in both breeds, whereas NBD shows no apparent improvement in either population. Genetic trends for FR showed slight declines over time in both populations. The estimated genetic parameters suggest the possibility of improving these reproductive traits by selection. Our results confirm that the selection criterion for LS should be NBA rather than NBT. Despite its low effect, the effect of service sire should be included as a random effect in the statistical model. Finally, due to the genetic downtrend for FR, this trait should be included in future pig breeding goals in South China.