Ce. Dewey et al., ASSOCIATIONS BETWEEN LITTER SIZE AND SPECIFIC SOW-LEVEL MANAGEMENT FACTORS IN ONTARIO SWINE, Preventive veterinary medicine, 23(1-2), 1995, pp. 101-110
This retrospective study examined associations between sow-level manag
ement factors and litter size in a target population of Ontario swine
farms that used computerized recording systems between 1987 and 1991.
The data were from 112 herds with 66 525 individual sow records. The a
verage litter sizes were ten pigs (SD = 2.7) for primiparous sows and
11.4 pigs (SD = 3.1) for multiparous sows, respectively. The total bor
n litter size (including pigs born alive, stillborn and mummified) was
regressed on the sow-level variables of interest, Multiple regression
models were built using a backward elimination approach and variables
significant at P<0.05 were retained. There was a curvilinear relation
ship between litter size and parity; the largest litters were from sow
s in parities three through ten. Natural matings resulted in one more
pig per litter than artificial insemination. Lactation lengths of 27-3
2 days and 33-40 days resulted in progressively larger litter sizes in
subsequent farrowings than shorter lactation lengths. Litter size was
highest at weaning-to-concepti on intervals of up to 4 days, decrease
d daily from 5 to 7 days, at which point it reached a plateau, until D
ay 11 when it increased again. The sow's current litter size was posit
ively associated with her previous litter size. The model explained on
ly 18% of the variation in litter size. As the age of the gilt at firs
t conception increased, the litter size increased.