The objectives were to evaluate (1) the use of 2 types of experimental silos (S) to characterize whole-crop oat (Avena sativa L.) silage with or without addition of an inoculant (I), and (2) the effect of inoculation on the microbial community structure of oats ensiled using only plastic bucket silos (BKT). From each of 6 sections in a field, oats were harvested, treated (INO) or not (CON) with inoculant, packed into 19-L BKT or vacuum bags (BG), and ensiled for 217 d. The inoculant added contained Lactobacillus buchneri and Pediococcus pentosaceus (4 × 105 and 1 × 105 cfu/g of fresh oats, respectively). The experimental design was a complete randomized design replicated 6 times. Treatment design was the factorial combination of 2 S × 2 I. Some differences existed between BG versus BKT at silo opening (217 d), including a decreased CP (7.73 vs. 7.04 ± 0.247% of DM) and ethanol (1.93 vs. 1.55 ± 0.155) and increased lactic acid (4.28 vs. 3.65 ± 0.241), respectively. Also, WSC and mold counts were reduced in BG versus BKT for CON (1.78 vs. 2.70 ± 0.162% of DM and 0.8 vs. 2.82 ± 0.409 log cfu/fresh g) but not for INO (∼1.53 and 1.55), respectively. Application of INO increased DM recovery (96.1 vs. 92.9 ± 0.63%), aerobic stability (565 vs. 133 ± 29.2 h), acetic acid (2.38 vs. 1.22 ± 0.116% of DM), and reduced NDF (65.0 vs. 67.0 ± 0.57), ADF (36.7 vs. 38.1 ± 0.60), ethanol (0.63 vs. 2.85 ± 0.155), and yeast counts (1.10 vs. 4.13 ± 0.484 log cfu/fresh g) in INO versus CON, respectively. At d 0, no differences were found for S and I on the nutritional composition and background microbial counts. Leuconostocaceae (82.9 ± 4.27%) and Enterobacteriaceae (15.2 ± 3.52) were the predominant bacterial families and unidentified sequences were predominant for fungi. A higher relative abundance of the Davidiellaceae fungal family (34.3 vs. 19.6 ± 4.47) was observed in INO versus CON. At opening (217 d), INO had a lower relative abundance of Leuconostocaceae (42.3 vs. 95.8 ± 4.64) and higher Lactobacillaceae (57.4 vs. 3.9 ± 4.65) versus CON. Despite several differences were found between BKT and BG, both techniques can be comparable for characterizing effects of INO on the most basic measures used in silage evaluation. The use of inoculant improved oat silage quality partially by a shift in the bacterial community composition during ensiling, which mainly consisted of an increased relative abundance of Lactobacillaceae and reduction of Leuconostocaceae relative to CON.

We evaluated the effects of adding a combination inoculant to 4 corn (Zea mays L.) hybrids harvested at low moisture on the nutritive value, fermentation profile, aerobic stability, bacterial and fungal populations, and community structure. The treatment design was the factorial combination of 4 corn hybrids ensiled with (INO) and without (CON) inoculant. The hybrids were TMF2R737 (MCN), F2F817 (MBR), P2089YHR (PCN), and PI144XR (PBR), ensiled at 44.0, 38.1, 42.0, and 41.3% of dry matter, respectively; MBR and PBR were brown midrib mutants. The inoculant contained Lactobacillus buchneri and Pediococcus pentosaceus (4 × 105 and 1 × 105 cfu/g of fresh corn). The experimental design was a complete randomized design with treatments replicated 6 times. Corn was chopped, treated or not with inoculant, packed into 7.6-L bucket silos, and stored for 100 d. At d 0, we found higher bacterial observed operational taxonomic units in the brown midrib mutants (MBR and PBR) relative to MCN and PCN (654 and 534 vs. 434 and 444 ± 15.5, respectively). The bacterial and fungal families with the highest relative abundance (RA) were Enterobacteriaceae (61.4%) and incertae sedis Tremellales (12.5%). At silo opening, we observed no effects of INO treatment on dry matter recovery (∼94.3 ± 1.07%), but aerobic stability was extended for all INO-treated hybrids (∼217 vs. ∼34.7 h), except for MBR (∼49 ± 38 h), due to a decreased yeast population (3.78 vs. 5.13 ± 0.440 log cfu/g of fresh corn) and increased acetic acid concentration (1.69 vs. 0.51 ± 0.132%) compared with the control. Furthermore, INO treatment reduced bacterial (61.2 vs. 276 ± 8.70) and increased fungal (59.8 vs. 43.6 ± 2.95) observed operational taxonomic units compared with CON. We observed that INO treatment increased the RA of Lactobacillaceae across all hybrids (∼99.1 vs. ∼58.9), and to larger extent MBR (98.3 vs. 34.3 ± 5.29), and decreased Enterobacteriaceae (0.614 vs. 23.5 ± 2.825%) among 4 other bacterial families relative to CON. For fungi, INO treatment increased the RA of Debaryomycetaceae (63.1 vs. 17.3 ± 8.55) and 5 other fungal families and decreased the RA of Pichiaceae (6.47 vs. 47.3 ± 10.95) and incertae sedis Saccharomycetales (8.47 vs. 25.9 ± 5.748) compared with CON. The bacterial and fungal community structures changed, due to ensiling, to a distinct and more stable community dominated by Lactobacillaceae and Debaryomycetaceae, respectively, when INO treatment was applied relative to CON. In conclusion, the INO treatment used in this study improved low-moisture whole-crop corn silage quality because of a shift in the bacterial and fungal community composition during ensiling.