BALANCE TRIALS AND COMPARATIVE SLAUGHTERING IN BREEDING SOWS - DESCRIPTION OF TECHNIQUES AND OBSERVED ACCURACY

Balance trials and comparative slaughtering with breeding sows were us ed to calculate the standard deviation (sd) of the retention and to ca lculate the contribution of the different error sources to the total v ariance of retention within a method. During pregnancy and lactation b alance trials (nitrogen, carbon, energy, calcium, magnesium and phosph orus) were performed with 60 sows. In lactating sows the retention in the litter was calculated from the chemical analysis of weaned piglets . For the comparative slaughter technique 12 sows were slaughtered at first mating (live weight 125 kg) as a reference, 14 sows at the end o f first pregnancy and 23 sows after the third lactation. In the course of the experiment 11 sows were culled. The sd of retention in the lac tation balance trials was much higher than in pregnancy trials. The ma in source of variance for nitrogen and Ca retention was the intake and for Mg and P retention the faecal excretion and for carbon and energy retention the gaseous exchange. With comparative slaughter the sd of fat and energy retention was affected by the estimate of body composit ion at the start of the experiment. The sd of nitrogen and mineral ret ention was mainly associated with the sampling procedure and analysis of the carcass. From the results it can be concluded that the balance technique resulted in a smaller ad of retention than the comparative s laughter technique when retention is measured over a limited period (< 3 months).