| Abstract: | A series of subtropical grasses and temperate grasses, herbs and legumes were analysed for the presence of extractable and bound condensed tannin (CT) using colorimetric analysis by the butanol–HCl method. Condensed tannins are routinely purified using affinity chromatography with Sephadex LH-20 as a matrix. Therefore, Sephadex LH-20 extracts were further analysed for the presence of CT by 13C nuclear magnetic resonance, for anthocyanidin formation after butanol–HCl treatment and for their ability to precipitate ribulose-1,5-bisphosphate carboxylase (Rubisco) protein from lucerne, at pH 7·0. Criteria for the presence or absence of CT were defined. Trace amounts of CT (0·2–2·5 g kg−1 dry matter; DM) were identified and confirmed in summer grass (Digiteria sanguinalis), perennial ryegrass (Lolium perenne) and red clover (Trifolium pretense), with chicory (Chicorium intybus), lucerne (Medicago sativa) and plantain (Plantago lanceolata) identified as probably containing CT. It was concluded that the subtropical grasses kikuyu (Pennisetum clandestinum), paspalum (Paspalum diatatum), smooth witchgrass (Panicum dichotomiflorum) and crowfoot (Eleusine indica) and the temperate grass, Yorkshire fog (Holcus lanatus) probably did not contain CT. Analysis of the extractable fractions by vanillin–HCl gave higher values for CT than analysis by butanol–HCl and wrongly identified some forages as containing trace levels of CT. It was concluded that vanillin–HCl was not specific enough for the detection of trace levels of CT in forages. These results raise the possibility of plant selection programmes to increase the level of CT in grazed forages to approximately 5 g kg−1 DM, the suggested minimum level required to prevent bloat in cattle and to increase wool growth in sheep. It is suggested that this be considered for perennial ryegrass, chicory, red clover and lucerne. |
