Abstract: In order to provide guidance for selection of replacement gilts,this study investigated the effect of partial non-genetic factors (parity,birth month and pure/cross-breeding) on the farrowing performance (total number born,number born alive and birth weight) of Landrace,Yorkshire and Duroc sows.The results showed that the birth weight (BW) from 2nd to 5th parities was heavier,and the BW of 2nd and 3rd parities was extremely significantly higher than the 1st and after 6th parities' (P<0.01).The total number born (TNB),number born alive (NBA) and BW in March were extremely significantly higher than other months (P<0.01),and the BW in November was significantly lighter than other months (P<0.05).The NBA of pure-breeding Landrace was extremely significantly higher than pure-breeding Yorkshire and Landrace×Yorkshire (P<0.01).The BW of Yorkshire×Landrace was extremely significantly higher than pure-breeding Yorkshire and Landrace×Yorkshire (P<0.01).There was extremely significant positive correlated with 2nd and 3rd to 4th parities in TNB,and so was 3rd and 4th parities (P<0.01).In conclusion,the farrowing performance of 3rd to 6th parities and January to April were the best;Pure-breeding Landrace and Landrace×Yorkshire were the advantageous ways of hybridization;The farrowing performance of 2nd parity could be regarded as an important basis for replacement gilts selection.

Key words: parity; birth month; pure/cross-breeding; farrowing performance