The study aimed to explore whether an increase in bunker silage density obtained by turning to a heavier packing machine than a farm size tractor would reduce losses and improve grass silage quality and aerobic stability. At each of three harvests, two bunkers were packed with either a 14.5 t wheel loader (WL) or an 8.3 t tractor (T). For comparison with the bunker silages, silage was produced simultaneously in round bales with high and low chamber pressure and wrapped immediately or after delay, and in laboratory silos.Compaction with WL increased silage dry matter (DM) density by 9 % compared with T, from 204 to 222 kg DM/m3. On average for three harvests, DM recovered as silage, or lost, was almost identical for the two packing treatments, with 870 g/kg of harvested DM recovered as feed offered to animals, 55 g/kg as wasted silage, and 75 g/kg as invisible losses due to respiration, effluent, fermentation and aerobic deterioration. However, in the harvest with lowest crop DM content, 266 g/kg, invisible DM losses with WL exceeded losses with T by 46 g/kg, of which the main portion was assumed to be caused by more effluent squeezed out by the WL. In the harvest with highest crop DM, 332 g/kg, invisible DM losses with T exceeded losses with WL by 43 g/kg, of which the main portion was assumed to be caused by poorer compaction with T, and therefore higher respiration and aerobic deterioration losses. Wasted silage DM was lower in bales than in bunkers (P = 0.004). The proportion of offered silage DM from poorly compacted bales sealed after delay (867 g/kg) was similar to that of bunkers, whereas the proportion of offered silage DM from well compacted and immediately sealed bales (963 g/kg) was similar to that of laboratory silos.Significant increases in protein bound in the neutral detergent and acid detergent fiber fractions were found in bales sealed after delay where temperatures had rised to 47 °C at wrapping. Similar levels of fiber bound protein were found in bunker silage, suggesting that they were also heated during filling. Spot samples from bunker silo shoulders were more infected by yeasts, moulds and Clostridium tyrobutyricum than samples from mid in bunkers and from bales. No differences in losses, silage composition or aerobic stability were observed between bunker silo packing with WL or T on average over three harvests.
Read full abstract