Effect of stocking rate and autumn saving on the carrying capacity of a Jemalong barrel medic pasture.

Abstract

EFFECT OF STOCKING RATE AND AUTUMN SAVING ON THE CARRYING CAPACITY OF A JEMALONG BARREL MEDIC PASTURE H. BROWNLEE* and G. E. ROBARDS Summary Over a two year period a Jemalong barrel medic pasture was not damaged by set stocking at three or four dry sheep per hectare. Sheep set stocked at six per hectare damaged the pasture during a drought year, 1965-66, but were successfully supported on similar pasture in a second experiment commenced a year later. Closing to livestock two-thirds of the pasture area in autumn at the highest stocking rate did not result in increased animal production compared with set stocking, but after the drought year the pasture on the autumn saved area was able to re-establish successfully whereas that under continuous grazing did not do so. I. INTRODUCTION In the low rainfall areas of the Australian wheat belt, medic pastures in the cropping rotation are valuable in restoring soil fertility and improving subsequent crop yields (Crawford 1962; Amor 1965). Although the medics can be established readily under cover crops, and produce large quantities of good quality feed during the pasture phase of the rotation, little is known of their potential stock carrying capacity. At Condobolin, in an area of central-western New South Wales which has a mean annual rainfall of 40 cm, Jemalong barrel medic (Medicago trunca.tula Gaertn) produces more dry matter than Commercial barrel medic and is more productive and persistent than Dwalganup subclover (Trifolium subterraneum L.) (unpublished data). This paper presents results from two experiments to determine the carrying capacity of a Jemalong barrel medic pasture on the Agricultural Research Station at Condobolin. II. MATERIALS AND METHODS The pasture was established in 1962 under a wheat cover crop sown at the rate of 28 kg seed/ha. The pasture seed was sown at rates per hectare of 2.2 kg Jemalong barrel medic, 1 .l kg Dwalganup subclover, 0.6 kg Lucerne (Medicago saliva L.) and 1 .l kg Wimmera ryegrass (Lolium rigidum Gaudn). An annual top-dressing of 126 kg/ha of superphosphate was applied from 1962 onwards. *Department of Agriculture, Agricultural Research Station, Condobolin, New South Wales. IDepartment of Agriculture, Agricultural Research Station, Trangie, New South Wales. . 114 (a) Experiment 1 After two years of stocking at between 1.3-1.9 sheep/ha, the *pasture consisted predominantly of Jemalong barrel medic with some subclover, ryegrass, barley grass (Hordeum leporinum L.) and saffron thistle (Ca,rthamus lanatus L.) Four grazing treatments each replicated twice were commenced in 1965. Three of these treatments were set `stocked with groups of eight Merino wethers at 3.1, 4.1 and 6.2 sheep/ha throughout the year. The remaining treatment was also stocked with eight Merino wethers at 6.2 sheep/ha, but two-thirds of the area was closed to livestock during the autumn (autumn saving) and the other one-third during the winter. After September 1, the whole area was available for grazing. (b) Experiment 2 0, A similar experiment was commenced in autumn 1966 on what was originally part of the same pasture area. By then, this area had been stocked at approxi\\ mately 1.9 sheep/ha for the previous two years. Clean wool production at shearing in autumn each year and seasonal bodyweight changes were recorded. The weight of barrel medic pods on the soil surface just before autumn germination was determined each year. Oats and hay were fed to the sheep in measured amounts during periods of pasture shortage. Split plot analyses of variance where measurements were repeated on the same sheep, were used to examine differences in clean wool production between treatments. III. RESULTS When Experiment 1 commenced in autumn 1965, there was 696 kg/ha of barrel medic pod on the area. By autumn' 1966, on the areas set stocked at 6.2 sheep/ha, there was only 21 kg/ha of pod which was not sufficient to germinate and form a pasture and the treatment was discontinued; at that time there was 453 kg/ha of pod on the lowest stocking rate areas. The high stocking rate treatment with autumn saving continued for a second year but only 45 kg/ha of pod remained at the end of that year. When Experiment 2 commenced in autumn `1966, there was 613 kg/ha of pod on the area; after one year of set stocking at 6.2 ahd 3.1 sheep/ha, the amount of pod on the ground was 550 kg and 1,078 kg/ha, respectively. Clean wool production and amounts of oat grain and hay given are shown in Table 1. During 1965166, sheep at 4.1 and 6.2/ha, both set stocked and autumn saved, were hand fed. Despite better seasonal conditions, these treatments were again hand fed during 1966-67 due to a carry-over effect on the pasture of the previous drought year. None of the treatments in Experiment 2 required hand feeding during their first year (1966-67). Wool production per head varied significantly (P < 0.05) between stocking rates during both years of Experiment 1 and the one year of Experiment 2. At no time was there a significant effect of autumn saving on wool production. Comparison between corresponding treatments in the first and second year of Experiment 1 showed that in each case, clean wool production was significantly higher in the second year. These differences appear to be directly due to season, because 115 (c) Records TABLE 1 Wo61 production on Jemalong barrel medic pastures during 1965-67, and amounts of oat grain and hjay given to the sheep IV. DISCUSSION In both experiments, increases in stocking rate resulted in decreased production per head but increased production per hectare. However, in Experiment 1, the increased production per hectare at the higher stocking rates was partly due to hand feeding during the drought year. To assess the economic implications of the treatments, simple gross margins analyses were made in which running costs and depreciation of stock were subtracted from net returns from sale of wool. Such analyses are subject to seasonal and district fluctuations in wool prices, running costs and availability of fodder, but within the likely limits of variation in prices and costs, the lowest stocking rate of Experiment 1 would have been economically the most sound during the drought year 1965-66. Again in 1966-67, due to the need for hand feeding at 4.1 and 6.2 (autumn saved) sheep/ha, the lowest stocking rate of 3.1 sheep/ha would have given the greatest monetary returns. In Experiment 2 on the other hand, net monetary returns per hectare increased with increases in stocking rate. This contrasting result occurred because the second experiment was started in a year more favourable for pasture growth, and on a pasture which had not previously been damaged by severe grazing. Although autumn saving at the highest stocking rate aided the survival of the pasture by allowing the saved area to germinate and re-establish while free of grazing animals, the need to hand-feed the stock while they were confined to the unsaved area would have reduced the profitability of the treatment. Since the pasture was able to re-establish while set stocked at 3 .l and 4.1 sheep/ha, the practice of autumn saving appears to be unnecessary and uneconomic under the conditions examined. 116 These experiments have indicated that a Jemalong barrel medic pasture at Condobolin can support three to four sheep/ha under set stocking. They suggest that the initial stocking rate' of less than two sheep/ha was not fully utilizing the pasture once it had established following sowing. These experiments have also shown that a topdressed barrel medic pasture is more productive than unimproved native pastures in the area which support 1.4 sheep/ha (unpublished data), and that greater returns per hectare can be expected. As the main use of arable land in the area is for wheat production, the effect of intensive grazing during the pasture phase on subsequent crop yields is an important aspect of these experiments which remains to be examined. V. ACKNOWLEDGMENTS The assistance given by Messrs. M. E. Bourke and J. C. England in the planning of this work is gratefully acknowledged. The authors would also like to thank the Manager and staff of the Agricultural Research Station, Condobolin, for their co-operation and assistance. VI. REFERENCES AMOR, R. L. ( 1965). J. Aust. Inst. a&c. Sci. 31: 25. C RAWFORD , E. J. (1962). J. Agric. S. Aust. 65: 214. I17