Effects of antibiotic-free pig rearing on ammonia emissions from five pairs of swine rooms in a wean-to-finish experiment.

Antibiotic use and ammonia (NH) emissions during animal production are two environmental issues of worldwide concern. However, the role of antibiotics on NH emissions is still unknown. This study evaluated the effects of rearing pigs without antibiotics on NH emissions from a swine experimental building starting with 657 piglets during a wean-to-finish production cycle of 154 days. Pigs were reared in two groups of 10 rooms that were divided into five 2-room pairs (P1-P5) and fed in nine dietary phases. Each pair consisted of one room without antibiotics (no antibiotics in the diet, water, or injectable) and another room as a positive control. Control animals were fed diets containing carbadox-10 (phases 1-4), chlortetracycline (CTC, phase 5), lincomix (phases 6-7), and tylan 40 (phases 8-9). Temperatures in the pig living space and the under-floor manure pit headspace were continuously measured. Ventilation rates at all wall fans and pit fans were obtained by continuous monitoring. Ammonia concentrations in the wall and pit fan exhaust air, and in room inlet air were measured with two multi-gas monitors. Only days that contained at least 18 h of data each day were validated and used. The study generated 1337 room-days of valid data of NH emission rates, with a data completeness of 88.6%. Daily mean NH emission patterns demonstrated large variations between the paired rooms and among different pairs. Within the individual 2-room pairs, no NH emission differences were found in P1 (rooms 1 and 2, p = 0.34) and P2 (rooms 3 and 4, p = 0.44). Significant differences were found in P3-P5 (p < 0.01). The antibiotic-free rooms emitted more NH from P3 and P4, but less NH from P5. However, the combined cycle mean NH emissions from the group of five antibiotic-free rooms and the group of five control rooms were 41.6 ± 10.5 and 39.4 ± 10.6 g d AU (mean ± standard deviation. AU = 500 kg live body weight), respectively. Therefore, there was no statistical difference in combined cycle mean NH emissions from rearing pigs with or without antibiotics (p = 0.78). This study also revealed that experiments with multiple replicates and long NH monitoring durations were necessary to avoid potential misinterpretation of experimental results.