The extractable and bound condensed tannin content of leaves from tropical tree, shrub and forage legumes

The extractable, protein-bound and fibre-bound condensed tannin (CT) concentrations in the leaves of tropical legumes grown in both Colombia and Northern Australia were determined by the butanol-HCl method, whilst extractable CT was also determined by the vanillin-HCl method. With the exception of Senna siamea all species contained CT. The very high CT concentration found in many plants growing in Colombia may have been partly due to soil fertility being much lower at the Colombian than the Northern Australian site. Acacia boliviana, Arachis pintoi, Centrosema latidens, Senna velutina and Gliricidia sepium contained <55 g total CT kg(-1) DM, which suggests that they could comprise a reasonable proportion of ruminant diets. All other species grown in South America contained 100-240 g CT kg(-1) DM, which suggests that they should only be fed in small amounts as supplements to dilute the CT concentration. Leucaena species and Calliandra calothyrsus grown in Northern Australia contained intermediate concentrations of total CT (60-90 g kg(-1) DM). Most species contained 70-95% of total CT as extractable CT, with the exception of Flemingia macrophylla, where 60% was extractable and 40% bound, and Gliricidia sepium, where almost all the CT was bound to protein. Values for Flemingia macrophylla differed between accessions. Extractable CT determined with vanillin-HCl was generally higher than extractable CT determined with butanol-HCl Three accessions showed negligible(<1 g kg(-1) DM) extractable CT with butanol-HCl but 10-12 g extractable CT kg(-1) DM with vanillin-HCl. Two accessions showed undetectable levels of extractable CT but substantial levels of protein-bound CT, illustrating the importance of using a bound CT method for identifying forages containing CT. Relative to freeze drying, oven drying of Leucaena species reduced the concentration of extractable CT and increased concentrations of bound CT. The significance of the results for the nutrition of ruminant livestock are discussed, including the possible roles of protein-bound and fibre-bound CT.