A comparison of twenty-seven introduced grasses in two dry-tropical environments in northern Queensland

In a grazed sward trial to compare dry matter production from 25 grass accessions fertilized with 250 kg ha-1 urea annually over a period of four years at 'Lansdown', the highest yielding grasses were Cenchrus ciliaris (buffel grass) cvv. Nunbank and Biloela. Chloris gayana (Rhodes grass) accession 16059, Panicum maximum (green panic) cv. Petrie and C. gayana cv. Callide in that order. Several species failed to persist and some accessions of Chloris, Paspalum and Melinis were virtually non-existent by the end of the second season. In another experiment comparing dry matter production under grazing of 23 accessions sown with a standard legume mixture at two sites ('Lansdown' and 'Fanning River') over periods of five and four years respectively, the highest yields were obtained from Urochloa mosambicensis 6559, C. ciliaris cv. Tarewinnabar and accession 18019 at both sites. Urochloa was top yielder in the higher rainfall site at 'Lansdown' and the two buffel grasses were best at the drier site 'Fanning River'. Of the 23 accessions sown, only 12 at 'Lansdown' and only 7 at 'Fanning River' were persisting strongly when the experiment was completed. Grasses common to both experiments at 'Lansdown' were higher in nitrogen when grown with associated legumes than when grown alone and fertilized with urea. Grass nitrogen levels were low enough to reduce intake by cattle of nearly all accessions throughout the growing season under the urea fertilizer system and phosphorus levels in the grass were generally below that considered adequate for animal nutrition in the legume plots; however, the legumes would probably supply sufficient to meet the animal's requirements for at least part of the year. At 'Fanning River' the accessions were generally lower in nitrogen and phosphorus than at 'Lansdown'.