Effects of inoculation with lactic acid bacteria on the bacterial communities of Italian ryegrass, whole crop maize, guinea grass and rhodes grass silages

Abstract

Italian ryegrass (IR), whole crop maize (WM), guinea grass (GG) and rhodes grass (RG) were stored in laboratory silos for 30 d with and without Lactobacillus plantarum or Lactobacillus brevis inoculation. The grasses were wilted to low moisture levels (about 500 g/kg) prior to ensiling, while WM was ensiled directly after harvest. Microbial counts and bacterial community analyses were undertaken by plate-culture and denaturing gradient gel electrophoresis (DGGE), respectively, on pre- and post-ensiled crops. A total of 31 DGGE bands were retrieved, cloned, and sequenced to identify bacterial species according to their 16S rRNA gene sequences. Inoculation with L. plantarum and L. brevis increased (P<0.05) lactic acid and decreased (P<0.05) 2,3-butanediol contents in wilted grass silages, with higher (P<0.05) contents of lactic acid observed in L. plantarum-inoculated silages compared to L. brevis-inoculated silages. Inoculation with L. brevis increased (P<0.05) both lactic and acetic acid contents in IR silage and intensified lactic acid production in GG and RG silages. Inoculation with either L. plantarum or L. brevis did not affect fermentation of WM silage, except that the lactic acid to acetic acid ratio increased (P<0.05) in L. plantarum-inoculated silage. Differences in inoculation effects could be accounted for by changes in bacterial communities present in the silages. Interestingly, in wilted grass silages, the inoculation nearly eliminated the indigenous communities, leaving the inoculated species as the principal bacteria. In maize silage, desirable bacterial communities were found in untreated control, and the inoculated lactic acid bacteria (LAB) were additive to the uninoculated communities. Combination of fermentation product analysis and bacterial community analysis may describe the effects of LAB inoculation on the ensiling process.