Abstract
Populations of lactic acid bacteria (LAB) on lucerne before and after havesting were extensively monitored over a two‐year period and correlated with weather data, various wilting parameters and equipment used for harvest. The LAB on standing lucerne were below detectable levels (−1). Immediately after mowing, low levels of LAB (51 CFU g−1) were found on over half the samples. During wilting, the population of LAB generally increased. The amount of increase was affected primarily by the average air temperature during wilting, the length of wilting and the drying rate. The first two variables were positively correlated with bacterial growth whereas increased drying rate adversely affected LAB numbers. If the LAB numbers in the swath were below 100 CFU g−1 prior to chopping, the forage harvester inoculated the chopped lucerne an level of 102 to 104 CFU g−1, depending on air temperature. Considering these factors, LAB counts on chopped lucerne could be predicted within one order of magnitude approximately 95% of the time.