AbstractsPhysics

This thesis emphasizes the need to use a systems approach when examining the effects of management on the efficiency of animal production from grazed pastures. The three experiments that are described were designed to increase the understanding and precision of some of the relationships that are used in the simulation of grazing systems. Conventional tests of farm management systems are usually of small scale and involve considerable handling of stock. Extrapolation of results to normal farm practice is therefore uncertain. An experiment carried out over 12 months with sheep in flocks of various sizes and managed under various grazing systems was analysed by estimation of the variance components of a non-orthogonal linear model. No effect on liveweight gains or wool production was detected in weaner sheep whether they were in flocks of 3, 9, 27 or 135 or whether they were set-stocked, rotated around three subdivisions of a pasture which was continuously grazed. Theses results indicate that within these limits neither flock size nor movement of animals has any important effect on animal production and thus can be ignored when establishing relationships between variables of basic biological importance to grazing systems. Increased precision in the relationships describing the interaction between the grazing animal and the pasture is limited in part by the technical difficulties in measuring food intake and in interpreting liveweight changes in terms of tissue energy retention. A review of the literature suggests that estimates of intake obtained by current techniques may be subject to biases that are at present undefined. Although these techniques may be useful in comparative studies of animal production, they are unlikely to be sufficiently precise for estimating the growth of pastures under grazing. A substantial improvement to the technique for predicting chemical composition in the live sheep was achieved in the second experiment. New and published data from sheep of widely different ages were used to derive regressions of body components on parameters measured in the live animal. Allometric models including body weight, tritiated water space and maturity as predictors substantially reduced the residual standard deviation for all components and largely eliminiated bias from the predicted values, yielding equations of general rather than local application. These regressions and the best available methods for measuring intake were used in the third experiment to examine the effect of body condition (energy status) on intake by mature sheep. Energy status within the range 220-289MJ had no effect on the intake of herbage by grazing animals or of a standard feed offered to animals in pens. Sheep which were moved from a sparse pasture (low energy status) to an abundant pasture gained at least as much liveweight, but only 20% as much energy as sheep in better condition (high energy status) because more of the gain was water. The results indicated that feed intake is likely to be unaffected by body condition under a…