Multifactorial models to assess responses to sorghum proportion, molasses and bacterial inoculant on in vitro quality of sorghum–soybean silages

Cuban climatological conditions allow culture of both soybean and sorghum. Combined ensiling of these crops might provide an excellent feed for ruminants during the dry season of forage shortage. This study aimed to assess the optimal proportions of both crops to ensure good quality silage using two approaches, being on in vitro fermentation test (Rostock fermentation test; RFT) and lab scale silages. Effects in the RFT of sorghum proportion (SGP), molasses and bacterial inoculant on silage quality characteristics were assessed by multifactorial response models. Two sorghum varieties (i.e., CIAP 2E-95, SG1 and CIAP 49V-95, SG2) and one soybean (SB) variety (i.e., INCASOY-35) were sown, harvested, chopped and ensiled. RFT was run with 72 treatments, including the pure forages (SB, SG1 and SG2) or either SG1 or SG2 in combination with SB in three proportions (SGP; 0.40, 0.60 and 0.80), with or without water soluble carbohydrates (WSC; 10, 20 and 30 g/kg fresh material) from molasses (17, 35 and 52 g/kg of fresh material) and with or without Lactobacillus plantarum as the bacterial inoculant. Quadratic response models were fitted and are presented as contour plots. Responses of pH, lactic acid and NH3–N/N to addition of WSC and SGP when bacterial inoculant was added or not were quantified. Models allowed prediction of minimum SGP or WSC contents to reach target values of desired parameters. Moreover, quantification of equivalent responses to WSC and SGP allowed assessment at their exchange rates. RFT results (i.e., pH and lactic acid) were used to select acceptable quality treatments for laboratory silages. Additionally, pure sorghum or soybean lab scale silages were prepared to enlarge the distribution of the silage quality characteristics. In both procedures (i.e., RFT and lab scale silos), fermentation of good quality was produced with both sorghum varieties alone, but combined silages had improved quality compared to soybean silage (P<0.05). Addition of WSC and bacterial inoculant further improved silage quality (P<0.05). High correlations between RFT and laboratory silage parameters suggest RFT to be an acceptable alternative to laboratory silage calculation before upscaling treatments.