Volume 37, 65-83

Tropical Grasslands (2003) Volume 37, 65–83

Effect of stocking rates on animal gain, pasture yield and composition, and soil properties from setaria-nitrogen and setaria-legume pastures in coastal south-east Queensland

R.M. JONES¹ and R.J. JONES²

¹ CSIRO Livestock Industries, Brisbane
² CSIRO Sustainable Ecosystems, Townsville,Queensland, Australia

Abstract

Animal production, pasture production and pasture composition were measured from 3 fertilised sown pasture systems in the subtropics of south-east Queensland. The pastures were Setaria sphacelata cv. Nandi (setaria) + Desmodium intortum cv. Greenleaf (desmodium), setaria + Macroptilium atropurpureum cv. Siratro (siratro) and setaria + 333 kg N/ha/yr as urea (N-fert). The legume treatments were initially stocked continuously at 4 stocking rates (SR) of 1.1, 1.7, 2.3 and 3.0 head/ha, and the N-fert at 5 rates of 3.0, 3.6, 4.2, 4.8 and 5.4 head/ha in a completely randomised design with no replications.
For the first 4 years, all treatments were stocked as planned. For the next 8 years, the highest legume SR was reduced to 2 head/ha and only the lowest and highest SRs for desmodium were maintained. For the remaining 12 years, measurements were made on only 3 SRs on siratro and 2 SRs on the N-fert treatment. Each November, new yearling heifers (in later years steers also) replaced those on the trial. The 2-yr-old heifers were then mated and breeding success measured.
For periods in the experiment where 3 or more SRs were operating, annual animal gain from all 3 systems declined linearly with increasing SR. Rates of decline/unit of SR with the legume treatments were 2–3 times those for the N-fert treatment. Annual animal gain (Y) was more closely related to pasture presentation yield than to mineral composition of the pasture or faeces. The best fit for all pastures at all SRs over a 12-year period was with the log of DM presentation yield/head in autumn (X): Y = 89.8 + 72.7X (r² = 0.87; P<0.0001).
Pattern of gain through the year was similar for all pastures from November–May, with only a slightly reduced gain with increasing stocking rate. Thereafter, the effects of SR became more apparent and gains from May–October declined at a rate related to SR. Annual animal gains also declined over time, especially on the legume systems. This resulted in a decrease in the estimated SR for maximum gain/ha from 2.2 and 5.2 head/ha in the siratro and N-fert treatments in 1969–1973 to 1.3 and 3.7 head/ha for the same pastures in 1977–1981. Breeding performance was good for heifers from all treatments except for animals grazing at the highest stocking rate on siratro-based pastures for 1 year when heifer weights were below 250 kg at mating.
The contribution to pasture yield of both legumes declined with increasing SR and over time. Siratro was far more persistent than desmodium but declined due to inadequate recruitment from seeds or stolons. A subsidiary experiment showed that spelling in autumn enabled partial recovery of siratro. Associated with the yield decline at the higher SRs was an increase in grazing-tolerant grasses and native legumes. However, these did not support good liveweight gains as pasture yields were low.
Compared with the legume-based siratro pasture, nitrogen application reduced topsoil pH, cation exchange capacity and exchangeable Ca, Mg and K. In contrast, topsoil of the N-fertilised pasture had higher levels of organic carbon, NO³–N, exchangeable acidity, exchangeable aluminium and extractable manganese.
In the light of these results, management options for Siratro to persist better and be productive under grazing in this environment are discussed.