Forage fermentation product measures are related to dry matter loss through meta-analysis

Forage preservation through fermentation can result in substantial DM and economic losses. The objective of this research by the use of meta-analysis was to determine whether fermentation end-product measures and forage parameters were capable of predicting forage DM losses following ensiling. The data set was built by searching a database for “forage,” “fermentation,” and “dry matter loss.” The data set contained 405 means from 43 peer-reviewed research reports. Forage DM losses (% of original forage DM) ranged from 0 to 28.6% with a raw mean of 6.2%. Report, forage biology, fermentation treatment, fermentation length, DM, pH, lactic acid, and acetic acid parameters were related to natural logarithm DM loss using a mixed-model approach. Parameters were evaluated for linear and quadratic effects and linear interactions. Report was classified as a random effect. The resulting model had a 1.403 mean natural logarithm DM, R2 of 0.813, and root mean square error of 0.418. Forage DM (%), acetic and lactic acid, pH, fermentation length, and forage biology were related to losses. Fermentation treatments tended (P < 0.10) to differ. Forage pH × lactic acid, DM × fermentation treatment, DM × forage biology, and lactic acid × forage biology all exhibited interactions. Forage DM, biology, fermentation, and treatment parameters were capable of describing most DM-loss variation across a range of published research reports (P ≤ 0.01). The final model described here has utility to predict forage DM losses due to fermentation and may be useful to diagnose problematic fermentations and assess opportunity costs.