A Random Regression Analysis of Sow Lactation Feed Intake and the Effect of Temperature on Intake

Abstract

Lactation feed intake (LFI) has only recently been considered as a possible selection criterion to assist in tackling the issue of sow longevity. It has been postulated that adequate LFI is needed to maintain sow body condition and development and to provide for suckling offspring. Hermesch and Jones (2007) showed that, on average, primiparous sows eating <3.5kg/day during lactation were significantly less likely to farrow in their second parity. Previously LFI, defined as the daily intake averaged across the lactation, has been shown to be a heritable trait (Bergsma et al. 2008; Hermesch, 2007). However it has also been demonstrated that the heritability estimates differ across the lactation curve (Hermesch, 2007). Considering that LFI is recorded over time and the trait is measured repeatedly on the sow, a more appropriate model to estimate the parameters could be random regression (Schaeffer, 2004). While random regression analyses have been utilized to examine finisher feed intake in pigs (Schnyder et al. 2001; Huisman and van Arendonk 2004) no published work using random regression is available for LFI. To add further complexity to both LFI and longevity, the effect of heat stress has been implicated as a cause for a reduction in LFI and also longevity. This study sought to firstly understand the temporal component of LFI in the random regression context and then further quantify the effect of temperature on LFI.