Recent research on condensed tannin content of Lotus uliginosus L.

Walter Kelman¹ and Kevin Smith²

¹ CSIRO Plant Industry, GPO Box 1600, Canberra, ACT 2601, Australia.

² Agriculture Victoria, Pastoral and Veterinary Research Institute, Private Bag 105, Hamilton, Victoria 3300, Australia.

Introduction

The protein binding properties of condensed tannins (CT) have important influences on the prevention of bloat and the utilisation of dietary nitrogen in ruminants. CT occur commonly in the tribe Loteae and are present in varying concentrations in Lotus. Lotus uliginosus (syn L. pedunculatus) has generally high levels of CT that are predicted to negatively affect palatability and digestibility of forage. One of objectives of a plant breeding program based in Canberra, Australia, is thus to select for lower concentrations of CT.

Measurement of condensed tannins.

The effectiveness of selection is dependent on accurate and rapid measurement of CT. Near infrared reflectance spectroscopy (NIRS) is routinely used for forage quality analysis and we have provided equations to predict CT concentrations in L. uliginosus (1). Calibrations were obtained using the butan-1-ol/HCl hydrolysis technique. A partial least squares technique was found to be more satisfactory for L. uliginosus than the step-wise multiple linear regression approach. The standard errors of calibration and prediction were small enough to be satisfactory for the purpose of genotype screening in a breeding program and it was unlikely that genotypes with low levels of CT would be overlooked using NIRS.

Selection for lower CT content

A sample of L. uliginosus accessions was grown in three environments in south-eastern Australia to assess their general adaptability to these environments (2). CT were measured as part of this assessment. The variance components for CT revealed significant variance among accessions and relatively lower accession x locality interaction effects. Estimates for repeatability for CT were high (76-88%) in cuts made in each of summer, autumn and spring growth periods. Strong positive associations were found between CT concentration, herbage yield and plant height in autumn but significant variance for CT within high-yielding groups of accessions indicated that selection within these groups for lower CT would be effective.

Population crosses between Portuguese (high CT)and New Zealand (low CT) germplasm sources were advanced to produce F2 and back-cross populations for generation mean analysis (3). In this study the distribution of CT concentration in the F2 population was markedly skewed towards the high CT parent and dominance effects were significant in the generation analysis. However there was no indication from the F2 distribution of the action of major genes for CT and additive effects were also significant in the quantitative analysis. Thus several genes for CT appeared to be segregating in these crosses

In the development of the breeding program two cycles of recurrent selection for lower CT content in the Portuguese x New Zealand material were completed. Half sib families derived from polycrossing the second cycle selections were grown at one site and vegetative clones of the polycross parents were transplanted to two sites. Variance component analysis was used to provide estimates of broad-sense (BS) heritability from the clonal material and narrow-sense (NS) heritability from the half-sib family means and the regression of offspring on parent means (4). In the clonal analysis genotype x locality effects were significant but were not attributable to large changes in genotype rankings across sites, thus suggesting that low CT selections would show this attribute in the target environment. The heritability estimates were significantly greater than zero and ranged from 0.46 in the parent/offspring analysis to 0.57 for the clones. In this study it was also predicted that selection both within and between families would be more effective than selection on a family mean basis alone.

References

1. Smith, K.F. and Kelman, W.M. 1997. Predicting condensed tannin concentrations in Lotus uliginosus Schkuhr using near-infrared reflectance spectroscopy. Journal of the Science of Food and Agriculture 75: 263-267.

2. Kelman W.M., Blumenthal, M.J. and Harris, C.A. 1997. Genetic variation for seasonal herbage yield, growth habit, and condensed tannins in Lotus pedunculatus Cav. and Lotus corniculatus L. Australian Journal of Agricultural Research 48: 959-968.

3. Kelman, W.M. 1995. Inheritance of condensed tannin concentration and rhizome number in Lotus pedunculatus Cav. 1995. SABRAO Journal 27: 23-29.

4. Kelman, W.M. 1997. Estimates of heritability and genetic gain from selection for lower condensed tannins in greater lotus (Lotus uliginosus Schkuhr). Journal of Genetics and Breeding (in press).

Back to the Table of contents