SELECTION OF CHAMAECRISTA-ROTUNDIFOLIA BY CATTLE

Hereford cattle fistulated at the oesophagus were used to study the pr oportion of roundleaf cassia (Chamaecrista rotundifolia) cv. Wynn in t he diet when they were grazing a cassia-green panic (Panicum maximum v ar. trichoglume) cv. Petrie pasture during 2 years. The percentage of cassia in the diet was estimated from analysis of carbon isotope ratio s of extrusa in both experiments and of faeces in Experiment 2 only. I n the first year, 1986-87, animals rotationally grazed the pasture for 7 months (early summer-early winter) and there was ample opportunity for selection. The cattle ate very little cassia in summer. Although t he percentage of cassia in the extrusa increased in autumn to 22%, it failed to reach the percentage of cassia in the top ''grazed layer'' o f the pasture (50-60%). In the second year, 1987-88, grazing was conti nuous from early autumn to early winter, and the grazing pressure was much higher. The percentage of cassia in the extrusa was again much lo wer than the percentage in the grazed layer until mid-autumn, but then increased to approximately equal the percentage in the pasture (10-20 %). However, the estimate of percentage legume in the diet based on an alysis of carbon isotope ratios of extrusa was consistently lower than the same estimate based on faeces. Possible reasons for this are disc ussed. Cattle discriminated against cassia in summer and early autumn, especially when there was ample opportunity for selection. Cassia was eaten more readily in late autumn, particularly when grazing pressure was high. The role of plant P and S concentrations in affecting cassi a acceptability warrants further study.