Date of this Version
The objective was to estimate the effect of traits recorded in females and in boars and their semen on farrowing rate (FR), total number (TB) and number of stillborn pigs (SB) at birth. Results of 20,569 inseminations in 4,468 sows on 4 farms with semen from 856 boars in 2 AI centers were analyzed. Records on sows included farm, dline parity and brdeeding interval (Brdint). Records on boars included number of days rest between collections, and 26 characteristics (e.g., volume, sperm concentration, motility, abnormal heads and tails plus 16 traits that described velocity and path of sperm cell movement). At first time, we were trying to use whole boar semen traits for our analysis; however the attempt was not competitive enough to reveal which semen characteristics had been far more deeply involved in FR, TB and SB. Thus, we used STEPWISE, MAXR and R-square were used for choosing statistically best semen characteristics. Data were analyzed with SAS PROC MIXED in models accounting for fixed effects of farm dam line of sow (Dline) and parity, random effects of sow and boar, and regressions of sow reproductive traits on sow, boar, and semen traits. Models were first fitted with only linear regressions; if important (P < 0.10), 2nd models including quadratic effects were fitted. Parity and the interval from 1st insemination (1st estrous during breeding period in gilts, and 1st post-weaning estrus in sows) to the insemination that resulted in a litter affected (P < 0.01) FR, and SB (P<0.1); parity also affected FR, TB and SB (P < 0.01). Average FR declined in a quadratic manner by 0.15 as the interval from 1st insemination to insemination of conception increased from 0 to 65 days. Sow reproductive traits were not affected (P>0.10) by number of days between collections (all boars had at least 3 d rest) or sperm concentration. Ten traits (Tmot, Vol, Proximal, Distal, Compos, Head, Tail, VAP, DSL and AOC) describing semen traits affected sow reproduction (P < 0.10), but differences across the range of variation were relatively small.
Advisor: Rodger K. Johnson