Effect of amylases and storage length on losses, nutritional value, fermentation, and microbiology of silages of corn and sorghum kernels

This study evaluated the effect of two commercial amylases with known activity at low pH and storage length on fermentation profile, microbial diversity, and effective rumen degradation (ERD) of silages of rehydrated kernels of sorghum (RES) or flint corn (REC). Mature kernels were ground through a 1 mm diameter mesh screen with a stationary hammer mill and manually rehydrated to achieve a similar dry matter (DM) concentration (650 g/kg of fresh) for ensiling (4.74 ± 0.30 kg/silo). Treatments were formed using a 2 × 3 × 2 factorial combination of grain type (RES and REC), enzyme [control and amylases AMG (300 AGU/mL) and GAM (138 AGU/mL), both at 0.35 mL/kg of fresh kernels], and storage length (30 and 180 d), with six replicates. The ERD of DM was evaluated in situ with a 2-pool model (A and B fractions) and incubations for 0, 3, 6, 12, 18, and 48 h. The proportional change in ERD mass during silage storage (% of ensiled) was calculated. Lactic acid bacteria (LAB) and spore-forming aerobic bacteria (SAB) were quantified and identified. The loss of DM (6.8 vs. 5.3% of ensiled) and starch (23.9 vs. 13.6% of ensiled) were higher (P < 0.01) for REC than RES. The RES had higher (P < 0.01) DM fraction A (398 vs. 325 g/kg of DM) and ERD (593 vs. 554 g/kg of DM) and lower (P < 0.01) kd of DM fraction B (3.14 vs. 3.33%/h) than REC. Ensiling increased ERD mass of both kernels (+6.8% of ensiled). The RES showed a faster increase (P = 0.04) in ERD mass in response to ensiling than REC. Both amylases and a longer duration of storage reduced (P < 0.01) the starch content and increased (P < 0.01) DM loss. Enzymes induced a faster increase (P < 0.01) in ERD mass relative to control but reduced the gain in ERD mass in extended storage. Longer storage increased (P < 0.01) the ERD content and the pool size of fraction A of both RES and REC. Lactic acid was the primary organic acid in silage, followed by propionic acid. Long storage of sorghum increased (P < 0.01) butyrate and there was no effect of grain type and enzyme on butyrate at 30 days of storage. The RES ensiled for 180 days without enzyme had the highest (P < 0.01) butyrate. The species Levilactobacillus brevis, Lentilactobacillus buchneri, Lactiplantibacillus plantarum, and Pediococcus acidilactici were the most frequent, but in general, amylases reduced numerically their population. The SAB had a greater diversity and lower population than LAB. We concluded that RES was more degradable in the rumen and had a lower loss during storage than REC. The highest ERD mass recovery was observed for long storage without amylase. The amylases seem to be an option only when storage length is short. The LAB population dominated sorghum and corn kernels silage fermentation. Most microorganisms identified were involved in organic acids and alcohol metabolism.