R.J. JONES¹ and B. PALMER²

¹CSIRO Sustainable Ecosystems, Townsville, Queensland
²formerly CSIRO Division of Tropical Crops and Pastures, Townsville, Queensland

Abstract

Condensed tannin (CT) in 26 Leucaena accessions (14 species and 3 hybrids) grown near Townsville, north Queensland was measured using ¹⁴C-labelled polyethylene glycol. Results showed a wide range in the amount of PEG bound (PEG-b) (0–167 mg/g DM) both between and within species, confirming other results indicating that selection and breeding for low CT levels is possible.
Lowest PEG-b values (< 25 mg/g) were measured in L. collinsii, L. lempirana, L. salvadorensis, L. magnifica, L. trichodes and L. trichandra OFI 4/91. Highest values (> 120 mg/g) were measured in L. pulverulenta, L. involucrata, L. diversifolia K156 and OFI 82/92, and L. trichandra CPI 46568. L. pallida, L. diversifolia CPI 33820, L. trichandra OFI 53/88, L. macrophylla istmensis, the KX2 F1 hybrid between L. pallida and L. leucocephala and L. leucocephala had intermediate values. The cultivars Cunningham and Tarramba and the L. pallida × L. leucocephala F1 hybrid had similar values (81–87 mg/kg).
The results were compared with reported values from the same species grown in south-east Queensland and in Honduras where the CT was measured by different techniques (Dalzell et al. 1998; Stewart and Dunsdon 1998). For the Honduran data, with 12 common accessions, there was an excellent linear relationship between PEG-b and total tannin estimated by radial diffusion protein-precipitation assay (r² = 0.881), but a poorer relationship with CT estimated by a butanol/HCl assay (r² = 0.649). For the south-east Queensland data, the linear relationship between CT estimated by a butanol/HCl method and PEG-b was significant (r² = 0.60), but there were notable departures from the relationship. Reasons for these are discussed.