PREDICTION OF LACTIC-ACID BACTERIAL NUMBERS ON LUCERNE

The numbers of lactic acid bacteria (LAB), as measured by Rogosa SL agar, on lucerne (Medicago sativa) chopped for ensiling were predicted using two methods. The first prediction method consisted of regression equations developed from two previous harvesting seasons and required inputs of wilting time between mowing and chopping, average wilting temperature and moisture content of the lucerne at chopping. The second method simulated the growth of LAB on the lucerne during wilting and needed similar inputs. The former method predicted LAB counts within one order of magnitude more than 95% of the time except for lucerne samples drier than 60% dry matter. The prediction of the simulation model had more variation than those of the regression equations, but the simulation model was applicable over a wider range of conditions. These results, together with earlier studies, provide a preliminary basis for determining when a silage inoculant containing LAB will be most beneficial. Copyright © 1990, Wiley Blackwell. All rights reserved