Source DocumentPrevious PageTable Of ContentsNext Page

Thirty years of change in South Australian broadacre agriculture

Ian Black and Chris Dyson

South Australian Research and Development Institute, GPO Box 397, Adelaide SA 5001.


We interrogated ABARE and other agronomic statistics from 1977 to 2005. Both wheat and barley production grew by over 100 percent and “other grains” by over 700 percent, the latter essentially from increases in grain legumes up until the early nineties and from canola since. Annual rates of increase were: wheat yield 1.8 %; barley yield 2.3%; stocking rates per grazed hectare 1.1% and net turnoff 1.8% percent in the high rainfall zone; but only 0.1% and 0.3% in the wheat/sheep zone. In 1977-80 livestock products contributed around 60 percent of revenue, with 40 percent from grains. By 1996-2005 the positions were reversed. In the context of an overall 30 percent reduction in the number of properties, the number of crop specialist and beef specialist properties increased while the remainder declined. Fertilizer per cropped ha increased three and a half-fold and volumes of chemicals sixfold. Fertilizer amounts applied to pastures have fluctuated widely over time. A significant driver is livestock revenue per ha – a 1 percent increase in revenue being associated with a 1.3 percent increase in fertilizer applied. Labour per ha in the settled areas has declined since the mid eighties and is now roughly 75 percent of what it was then. An index of machinery in use per cropped ha and per DSE dropped by 20-25 percent. Paid advisory services per farm increased ninefold between 1989 and 2005 in real terms. These and other changes are discussed in the context of broadacre productivity improvements.

Key Words

Grains, livestock, fertiliser, chemicals, labour, productivity


Our primary objective in interrogating historical ABARE and other sources of data was to derive policy implications for R&D resource allocations. However, the work has revealed historical trends that will interest a wider range of audience. In this paper we draw attention to SA broadacre farming trends of interest to agronomists.


ABARE data come from two electronic sources: ASPIRE: 1977-1999 and AGSurf: 1990-2005. In summary, these databases contain statistics on physical production and its value (e.g. tonnes of wheat produced and wheat receipts), areas of crops and farm costs (e.g. payments for repairs and maintenance). These two databases were joined to form a coherent series from 1977 to 2005 inclusive – 29 years of data. ABARE provide data categorised in specialised property groupings (sheep, beef, sheep-beef, mixed crop-livestock and crop) as well as in zones (pastoral, sheep-wheat, high rainfall). Data analysis results from both types of categories are discussed. The ABARE categories discussed are self-explanatory. A “machinery in use” category was also derived, comprising the ABARE “depreciation”, “fuel, oil and grease” and “repairs and maintenance” categories, with due partitioning in proportion to fixed capital improvements and mobile machinery.

Results and Discussion

Land use and revenue

An average of 1.5% of properties per year were lost from SA broadacre agriculture, from 11,400 in 1977-79 to 8,000 in 2003-05. The losses were in the sheep, sheep-beef and mixed crop-livestock property categories; beef specialist (11 percent of the total in 2003-05) and crop specialist (42 percent) property numbers actually rose. The fact that there was no significant increase in property size in the crop specialist category indicates that productivity improvement was the most important way that cropping properties remained viable over a period of generally deteriorating terms of trade. Very recent market developments may substantially alter trends.

Figure 1 shows that in 1977-80 livestock products contributed around 60 percent of revenue; between 1996 and 2005 they contributed around 40 percent. In particular, wool revenue has markedly declined as a proportion of the total since 1990. After a relative decline in the middle years, cattle and sheep sales have undergone a minor revival in terms of their proportional contribution. Wheat and barley revenues contributed an increasing proportion of total revenue until around 1997-1998, as have the “other grains” category from a low base.

Figure 1

Production and yield increases

Wheat and barley production increased by over 100 percent and “other grains” by over 700 percent, the latter consisting of increases in grain legume production up until the early nineties and canola production increases since then. Wheat yields have increased by 1.8 percent p.a. and barley by 2.2 percent on average. However both increases appear to have started to diminish (Figure 2).

Figure 2

The large differences in stocking rate and turnoff improvements between the high rainfall and wheat-sheep zones (Table 1) reflects the fact that the high rainfall zone is still largely comprised of livestock specialist properties, and there is adequate rainfall to allow intensification of grazing industries. By contrast in the wheat-sheep zone there has been a large increase in cropping intensity, forcing grazing enterprises onto less productive land. Table 1 also shows that farmers have become progressively more efficient in turning livestock off for market.

Drivers of farm inputs

Owner/operator, family and hired labour per ha in the settled areas has progressively declined since the mid eighties and is now roughly 75 percent of what it was in the 1977-85 period. Machinery in use per cropped ha and per DSE has dropped by 20-25 percent over the 1977-2005 period. We infer from these results that the proportionate gain from increasingly expensive machinery purchases and running costs is more than outweighed by the consequent improvements in labour productivity. Paid advisory services per farm increased linearly by nearly nine-fold between 1989 and 2005, from $65 to $560, in real terms. Clearly, farmers increasingly perceive that personalised technical advice is a good investment.

The quantity of fertilizer per ha applied to cropping land increased approximately three and half times and that of chemicals increased approximately sixfold (Figure 3). The ABARE data reveal that increased fertiliser nitrogen use was the cause. Since the late 1990s Figure 3 shows that the increases in both fertilisers and chemicals have diminished, and this may be a contributing factor to the plateauing of wheat and barley yields.

Figure 3

The amount of fertiliser applied to pastures fluctuates widely over time (Figure 4). It is strongly related to livestock revenue per ha – the statistically significant analysis shows that a 1 percent increase in revenue produces (assuming causality) a 1.3 percent increase in the index of fertilizer applied. Given that the legume component of pastures, and therefore their nutritional value and productivity, are highly dependent on regular applications of P, the behaviour revealed by these data show that pasture productivity is not the highest priority for managers of SA livestock production systems.

Figure 4

Ordinary least squares analyses revealed that there was a considerable degree of commonality in the drivers of cash costs both for zones and property types, except in the case of the crop specialist properties. These included, variously, livestock purchases, hired labour, machinery in use, cropping contracts, fertiliser, purchased fodder and livestock materials. The best model for predominantly cropping properties included only machinery in use, chemicals and cropping contracts, although fertiliser usage was a near perfect substitute for chemicals usage. This near perfect substitution suggests that, in any given year, crop farmers cannot afford to increase both at the same time to the desired level.

Productivity increases

Estimated total factor productivity growth was not significantly different from zero on sheep properties, 0.8 percent per year on beef-sheep properties (P<0.1), and was only 1.1 percent per year on beef properties. These results are disappointing, particularly in relation to those achieved by cropping properties (3.2 percent) and, to a lesser extent, mixed crop-livestock properties (1.9 percent). There is no lack of research in SA devoted to livestock and pasture systems, as well as interstate and overseas, resulting in new technology. This new technology, if attractive to farmers, should be adopted and eventually be measured as TFP gains. Unfortunately, these results show little evidence for substantial productivity gains in the livestock industries. Crop yield increases are the major cause of the superior TFP performance on crop and crop-livestock properties compared to livestock specialist properties (inputs per cropped ha have remained relatively constant). However, yields of wheat and barley, the predominant crops in SA, have plateaued in the last few years, causing a plateauing of productivity in these property categories.


Broadacre farming in SA has shown itself to be dynamic and adaptable over the last 30 years. Major drivers of this adaptability include the relative profitability of livestock and cropping enterprises, together with the relative productivity improvements in these industries over time. Productivity improvements are in turn driven by technical innovation, through R&D, and the willingness of farmers to adopt such technical innovation. With the onset of significant climate change, and a rapidly changing market, broadacre agriculture will increasingly need to show further flexibility and adaptability in terms of farming systems in use.


ABARE ASPIRE-AAGIS. Electronic data base. Australian Bureau of Agricultural and Resource Economics, Canberra (This data base is no longer serviced by ABARE; it is available from the senior author on request).

ABARE AGSURF. Electronic data base. Australian Bureau of Agricultural and Resource Economics, Canberra.

Previous PageTop Of PageNext Page