Visible fungal growth on baled grass silage during the winter feeding season in Ireland and silage characteristics associated with the occurrence of fungi

The extent of visible fungal growth and the identity of fungi causing spoilage of baled grass silage were recorded regularly through a winter feeding period in Ireland. The study was conducted from November 2003 to March 2004 on 50 farms differing in farming systems and geographical locations in the Irish Midlands and two bales were examined in detail on each farm (n = 100 bales). Visible fungal growth was present on 90 of the 100 bales examined. There were more fungal colonies visible on bales in March (mean, 8) than in November (mean, 4) (P<0.05), but the number of fungal colonies on bales did not differ between the very youngest and oldest bales (P>0.05). Fungal contamination was higher (P<0.001) in bales where the surrounding polythene stretch-film was visibly damaged compared to bales where the film appeared intact. No other bale production or storage characteristic significantly predisposed bales to increased fungal spoilage. The most prevalent fungus on bales throughout the feeding season was Penicillium roqueforti which represented 0.43 of all fungal colonies observed and was present on 0.78 of bales. The proportion of bale surface areas visibly colonised by P. roqueforti increased (P<0.05) monthly from November to March. Other fungi frequently isolated throughout the winter included Schizophyllum commune, mucoraceous moulds and Pichia fermentans. Redundancy analysis showed that the distribution of fungi on bales was best explained by the concentrations of ethanol and lactic acid, dry matter content, bale tying, month of bale feed-out, age of bales, polythene film damage, ryegrass dominance, bale storage location and volatile fatty acid concentrations. Forward selection of these silage characteristics showed that butyric acid concentration and age of bales contributed significantly (P<0.05) to the variance found in the species data and these two variables could explain 0.19 of the explained variance. P. roqueforti occurrence was positively correlated to higher concentrations of propionic and butyric acids, to month of bale feed-out and to bale age. S. commune was more commonly isolated from bales made from ryegrass dominant swards, that had visibly damaged polythene film, that were stored in farmyards and from silage with a lower lactic acid concentration. Overall, a high incidence of visible fungal growth was recorded on bales throughout the winter season and the extent of colonisation and fungal species occurring were not random. To prevent fungal spoilage of baled silage, the focus of best farm practice should be to provide low oxygen conditions in well wrapped and carefully managed bales.