T.G.H. STOCKWELL1 R.J. CLEMENTS2, G.J. CALDER3 and W.H. WINTER4
1Northern Territory Department of Primary Production, P.O. Box 1346, Katherine, N.T. 5780.
2CSIRO, Division of Tropical Crops and Pastures, Cunningham Laboratory, 306 Carmody Road, St. Lucia, Qld 4067.
3Northern Territory Department of Primary Production, P.O. Box 4160, Darwin, N.T. 5794.
4CSIRO, Division of Tropical Crops and Pastures, Darwin Laboratories, P.M.B. 44, Winnellie, N.T. 5789.
Seven bred lines and two parent accessions of Centrosema pascuorum were grown for two years in a cutting trial at six sites in the Northern Territory and at one site near Kununurra, W.A. Stylosanthes hamata cv. Verano and Macroptilium atropurpureum cv. Siratro were included as controls at six sites. Herbage yield was measured towards the end of each wet season, and seed yields were measured during both dry seasons.
In the first year, most C. pascuorum lines yielded more herbage and seed than Verano and Siratro at most sites. Differences between C. pascuorum lines in mean herbage yield were small, but differences in seed yield were substantial. There were no line × site interactions for either herbage or seed yield. In the second year, however, the C. pascuorum lines generally gave poorer yields than Verano at each site, but again outyielded Siratro which performed poorly at all sites. There were significant C. pascuorum line × site interactions for both herbage and seed yield.
One line, subsequently released as cv. Cavalcade, possessed a desirable combination of consistently high herbage yield over sites and years and good seed production, providing 22% more herbage (on average) and 118% more seed than its best parent.