|dc.description.abstract||The impact of within-litter weight variation on the productivity of pig enterprises is poorly understood. The objective of the study was to determine the effect of within-litter birth weight variation on litter performance at three weeks of age and at weaning. The study was conducted using records from 1 788 litters, collected between January 1998 and September 2010, from a pig herd at the Agricultural Research Council (ARC), Irene. The records consisted of piglet identity, breed of sow, breed of boar, parity number, date of farrowing, number of piglets born alive (NBA), individual piglet weight at birth, three weeks and at weaning. From these records, mean birth weight (MBWT), litter weight at birth (TBWT), within-litter birth weight coefficient of variation (CVB), minimum birth weight (MinB) and maximum birth weight (MaxB) were calculated. Mean weight at three weeks (MWTT), litter weight at three weeks (LWTT), within-litter weight coefficient of variation at three weeks (CVT), percent survival to three weeks (SURVT), mean litter weaning weight (MWWT), litter weight at weaning (LWWT), within-litter weaning weight coefficient of variation (CVW) and percent survival at weaning (SURVW) were computed as derivatives. The factors affecting CVB were analysed using the General Linear Model procedures (SAS, 2008). For the relationships between CVB and litter performance at three weeks and weaning, PROC STEPWISE was used. The PROC REG (SAS, 2008) was then used to test whether the relationships between CVB and CVT, SURVT, MWTT, LWTT, CVW, SURVW, MWWT, LWWT and LWWT.
Multiparous sows farrowed litters with higher (P<0.05) CVB than gilts. The litter weight (TBWT) and NBA, fitted as covariates, also affected (P<0.05) CVB. The correlation between CVB and NBA was 0.30. The CVB had a linear relationship (P<0.05) with SURVT (SURVT = 83.21 - 0.20 CVB), CVT (CVT = 16.71 + 0.50 CVB), SURV (SURW = 87.9 – 0.04CVB) and CVW (CVW= 15.8 + 0.5CVB). An increase of CVT with CVB depended on parity (P<0.05). The rate of increase of CVT with CVB was highest in Parity 1 (b=0.41) followed by Parity 2 (b=0.36) then middle aged (Parity 3-5) sows (b=0.32). The CVB had no effect on MWTT, LWTT, MWWT and LWWT (P>0.05). The CVB was shown to be an important determinant of SURVT and SURVW. A uniform litter at birth is likely to lead to a homogenous litter at three weeks and weaning, thereby reducing costs of production. Pig producers should, therefore aim at producing homogenous litters at birth.||en