It is widely accepted that robust and accurate in vitro techniques are required to predict the proportion of food nitrogen (N) degraded in the rumen. One such technique is to estimate the solubility of food N. In these experiments, relationships between solvent soluble N and in situ rumen degradability of forage N were investigated.

Samples of 11 fresh grasses (FG) (mainly perennial ryegrass) and their corresponding silages (GS) were used. GS was prepared from material ensiled in laboratory scale silos for 90 days. Prior to the experiments, FG and GS samples were initially hand chopped to approximately 1 cm lengths. In the in situ study a fresh sample equivalent to 0-5 g DM was weighed into polyester bags (pore size 43 μ 200 X 90 mm internal diameter). Duplicate bags for each of FG or GS were incubated in the rumen of three wethers for 0, 3, 8, 16, 24, 45 and 72 h. The incubated residues including the 0 h samples were washed in a washing machine and freeze-dried for 48 h. Rumen degradability characteristics and effective degradability (ED, at rumen outflow rate of 0.08 per h) of N were calculated using the exponential model of Ørskov and McDonald (1979). In vitro solubility of N (S) was determined by incubating for 1 h (at room temperature) the fresh sample (0.5 g on dry matter basis) in each of the four solvents: Borate phosphate buffer (BFB), Durand's buffer (DB), clarified rumen fluid (CRF) and distilled water within a balanced three way factorial design (three operators; four solvents; 11 forages; Deaville et al., 1997). Residues from S were filtered under vacuum and the filter paper plus residue were oven dried for 18 h at 100°C. All samples and residues were analysed for total N using Kjeldahl method (Ministry of Agriculture, Fisheries and Food, 1986). Factorial analysis on the general linear model (Minitab®, 1994) was used in the analysis of variance(ANOVA) for in vitro data and regression analyses of in situ and in vitro data were performed (Minitab®, 1994). Only the regression results are reported here.