The effect of PEG addition in vitro on dry matter and nitrogen digestibility of Calliandra calothyrsus and Leucaena leucocephala leaf

A modified two stage in vitro digestion method was used to study the effects of rate of PEG addition (0–1100 mg/g DM) on dry matter (IVDMD) and nitrogen digestibility (IVND) of freeze dried leaf material from actively growing shoots of Calliandra calothyrsus (Calliandra) and Leucaena leucocephala cv. Cunningham (Leucaena). For both species, IVDMD and IVND increased asymptotically with PEG rate; the IVND response being markedly greater for Calliandra. There was a three-fold greater amount of PEG bound to the residue after Stage 1 (rumen fluid/buffer) digestion with Calliandra than with Leucaena (105 versus 33 mg/g DM). The presence of PEG in the residue, which was not reduced after Stage 2 (acid–pepsin) digestion, resulted in a higher residue weight and therefore an underestimated IVDMD. Quantification of the PEG in the residue enabled a corrected IVDMD (CIVDMD) to be estimated. PEG did not bias estimates of nitrogen digestibility in the same way. In a second experiment, estimates of CIVDMD and IVND were made after Stage 1 and after Stage 2 in the presence and absence of PEG at 160 mg/g sample. For Leucaena, there were small, but significant, effects of PEG, whereas with Calliandra there were large responses to PEG and to timing of addition. When measured after Stage 1, IVND in the absence of PEG was not significantly different to zero (compared with 57% for Leucaena), whereas with PEG, IVND was 75% (compared with 68% for Leucaena). The low IVND for Calliandra was associated with low NH4 N levels in the rumen fluid/buffer after Stage 1. mary, about 160 mg PEG/g sample is appropriate for most studies with tropical tanniniferous shrub legumes to estimate any deleterious tannin effect. The use of PEG to estimate this effect on IVDMD is not valid without accounting for the PEG bound to the residue. For estimation of the adverse effect of tannins on IVND the use of PEG may be more appropriate.