Agribusiness Perspectives Papers 2002

Paper 52
ISSN 1442-6951

Policy Influences on Genetic Diversity in Australian Wheat Production

David Godden and John P. Brennan*

Department of Agricultural Economics, University of Sydney,
and Wagga Wagga Agricultural Institute, Wagga Wagga NSW.

* This research was funded by the Australian Centre for International Agricultural Research and was conducted in conjunction
with Centro Internacional de Mejoramiento de Maiz y Trigo (CIMMYT).

• 1 Introduction
• 2 Background
• 3 Production in second half of twentieth century
• 4 The policy environment
• 5 Implications for genetic diversity
• 6 Conclusions
• References

1. Introduction

The research reported in this paper has its origins in a wider project (e.g. Brennan et al. 1999a) examining economic dimensions of genetic diversity in the wheat growing industries of Australia and China. Part of this study involved the econometric estimation of the supply of and demand for genetic diversity in the Australian wheat industry. Especially since 1945, government policy has had a major impact on the marketing of Australian wheat, and also on the research and development process. Government policy is therefore a possible shifter of the supply and demand curves, and some “policy” variable would be required to test the significance of this hypothesis. Construction of such a policy variable required the preliminary analysis of the likely forms of government policy that might eventually affect the supply and demand for genetic diversity.

Genetic diversity in the Australian wheat growing industry is interesting for three principal reasons. Firstly, individual farmers face an array of risks and uncertainties including price and production risk. Wheat varieties, because of their different genetic makeup, respond differentially to climatic and other environmental (e.g. pest and disease) conditions. Choice of wheat variety offers some opportunities to manage risk and uncertainty in wheat production. For example, some wheat varieties are optimally sown “early” in a season, whereas others may be sown “late”. The availability of different varietal types allows farmers to exploit different climatic conditions as they emerge. This factor might be denoted “routine” risk and uncertainty.

Secondly, plant breeding is an economic activity in the sense that breeders are continually searching for improved cultivars within the constraints of available financial resources, and within the constraints of what is genetically possible within existing knowledge. Thus breeders are continually making tradeoffs between an array of plant breeding objectives. The choices that they make – e.g. between “more genetically diverse” varieties, and higher yield or quality – govern the array of varieties that farmers have available to manage their production systems.

Thirdly, there is also an issue of the “ecological” sustainability of the wheat production industry. The possibility of major breakdowns of, for example, disease resistance in a crop kind was recognised following 1970 Southern Corn Leaf Blight in the USA. If the response of each genotype to environmental conditions were perfectly known, then there would be no uncertainty about the sustainability of crop production. Without perfect knowledge, however, there will always be some residual uncertainty about the production stability of the existing range of varieties and nearly-available varieties.

The policy environment affects the economic and social conditions in which farmers and plant breeders make decisions about the development and use of varieties. As the policy environment changes, it is possible that these changes affect the kinds of decisions that breeders and farmers make about wheat varieties and, in particular, the genetic diversity of the set of available varieties. The possible impact of policy changes on decisions about the development and use of wheat varieties and associated genetic diversity is explored in this paper.

The post-war Australian wheat industry examined in the wider project was extensively regulated. An evaluation of genetic diversity and production variability in this period requires an indication of what Australian wheat production might have been like without such extensive regulation. Hence the paper commences with a survey of the development of the Australian wheat industry to circa 1950 (section 2). This survey includes (i) the development of the Australian wheat industry at the colony and state level to the mid-twentieth century; (ii) the development of intervention in the industry which was crowned with the first of the peacetime wheat marketing acts in the late 1940s; and (iii) the pattern of production from the mid-twentieth century.

Section 3 broadly outlines the development of the Australian wheat industry 1950-2000. The paper then surveys the development of the Australian agricultural policy environment in the second half of the twentieth century (section 4). This survey is used to explore possible impacts of policy change on the development and use of genetic diversity in the Australian wheat industry. Finally, in section 5, there is a detailed analysis of the possible effects of specifically wheat industry policy on genetic diversity in the Australian wheat industry 1950-2000.

Information about the wheat industry is based on secondary sources such as Dunsdorfs (1956), Rural Reconstruction Commission (1946), and Whitwell and Sydenham (1991) as the intention was not to again account for the development of the industry, but to examine genetic diversity – and its interaction with the policy environment – in the context of the development of the industry.

2. Background

2.1 Nineteenth century

The development of the Australian wheat industry occurred in several phases. In the first half of the nineteenth century, the first two colonies of New South Wales and Tasmania (Van Diemen's Land until 1856) struggled to become self-sufficient, including in wheat production. Wheat area in both colonies increased slowly to the mid-1840s after which Tasmania's plateaued (Figure 1a). After the rapid take-off in wheat growing in the mainland colonies from the mid-nineteenth century, Tasmania became an insignificant wheat producing state despite its geographical suitability for wheat production (cf. Figures 1h and 2c).

In the 1840s, wheat growing accelerated rapidly in South Australia until 1880 after which the area grown plateaued (Figures 1a,b). Growth in wheat area accelerated rapidly in Victoria from 1855; the growth rate in area increased rapidly again from about 1875 and continued at a rapid rate until the First World War (Figures 1a,b). After slow growth throughout the nineteenth century, wheat growing increased rapidly in NSW from the early 1890s (Figure 1b); this development had previously been inhibited by the separation of good wheat growing areas from the population centre of Sydney by the Great Dividing Range, and was overcome by railway expansion.

Figure 1: Australian Wheat Statistics to Mid-Twentieth Century

As most – although not all – Australian soils were nutrient (especially phosphate) deficient or easily nutrient depleted, secular declines in wheat yields occurred from 1850, and were especially severe in Victoria and South Australia (Figures 1h,i and Table 1). Average State yields should, however, be interpreted cautiously since they are influenced not simply by soil fertility, but also changes in cropping practices, short- and longer-run climatic cycles (cf. Godden 1999), and especially changes in the Ricardian extensive margin (e.g. Dunsdorfs 1956, pp.136-7). Brennan and Spohr (1985) showed that changes in State average yields can be misleading because of location shifts.

Table 1: State Mean Yields, Yield Trends and Relative Yield Variability, 1850-1997

+ fewer data points for Queensland (1860-1900) and Tasmania (1851-1900)
** statistically significant
^ statistically insignificant
? low significance
~ borderline significance
Source: computed from data in Dunsdorfs (1956)

2.2 Early twentieth century

By the early twentieth century, the pattern of wheat growing in Australia had been largely established. This pattern involved:

• extensive areas of production in the principal wheat growing states – about 500,000 hectares in each of NSW, Victoria and South Australia; wheat growing expanded rapidly in Western Australia from 1910. Expansion of wheat production in Queensland occurred from the late 1930s (Whitwell and Sydenham 1991).
• wheat growing was a low-input system, dependent on a rainfall pattern which was erratic in annual amount (cf. Godden 1999) and also in seasonal timing and intensity, and interacted with pests and diseases (especially rusts); this variability was reflected in highly erratic wheat yields (Figures 1g,h,i). The variability of yield – estimated as the standard error about the estimated trend line relative to mean yield – is shown in Table 1.
• there had been considerable experimentation and innovation with labour-saving machinery for all aspects of wheat production;[1] and there was subsequently successful experimentation with varietal improvement (see below).
• wheat growing had largely evolved into one (key) enterprise in a mixed-farming system, except in Western Australia (Whitwell and Sydenham 1991, p.32). Unincorporated business structures (sole proprietorship, partnership or family) predominated, and the absence of limited liability increased the financial vulnerability of farms by limiting borrowing options (Whitwell and Sydenham 1991, p.33).
• wheat growing was increasingly oriented to production for export, initially inter-colonial trade (e.g. South Australia to NSW) and subsequently outside Australia (and especially the United Kingdom); wheat exports expanded rapidly after 1900 (cf. Figure 3).

Figure 3: Development of Wheat Exports, Australia, 1910-45

The First World War affected the Australian wheat industry in three principal ways. Firstly, military enlistment from war's outbreak in 1914, which was especially high in country areas, rapidly reduced the farm workforce and led to a substantial reduction in area grown and thus output in the early war years (Figures 1b,d). Secondly, to control wheat marketing, the first Australian Wheat Board was established which, by war's end, provided wheat growers “with the highest prices – and possibly the greatest sense of security – they had enjoyed for 30 years” (Whitwell and Sydenham 1991, p.42).

This marketing experiment was followed by the temporary emergence of “co-operative state-wide pools” in early 1920s (Whitwell and Sydenham 1991, pp.36-37) with more permanent pools only in Queensland (statutory) and Western Australia (voluntary) (Whitwell and Sydenham 1991, pp.42,45-6 respectively). Thirdly, following cessation of hostilities and development of extensive (and generally disastrous) “soldier settlement” schemes, wheat area grew rapidly in the 1920s. Wheat area increased four-fold in Western Australia and 3.5 fold in NSW from the disastrously small area of 1919 to 1930; and area similarly doubled in Victoria and South Australia (Figure 1b). [2]

During the first half of the twentieth century, the declines in average state wheat yields that characterised the nineteenth century were arrested (Figures 1h,i and Table 1). However, except in Queensland and South Australia, there was no statistically significant yield trend in the period 1901-48. In Queensland, average state yield increased approximately 10kg/ha/year, and in South Australia the corresponding estimate was 5 kg/ha/year. Variability – estimated as the ratio of the standard error of a regression of yield against time, relative to mean yield – was low (approximately 0.2) in Western Australia and Tasmania (Table 1). The corresponding variability estimates for the other states were much higher, ranging 0.3-0.4.

2.3 Progress in wheat breeding

The principal uses of wheat were for bread making and for livestock feed (hay and grain for both farm animals and other draught animals). Since the principal costs of growing (e.g. sowing, harvesting) were independent of end-use, the farmer's optimal strategy was to aim for the highest valued use, since this would provide the greatest gross margin to allow transportation to market.[3] Additionally, since crops damaged by weather or disease were still suitable for animal use, there was likely to be sufficient material available for other local uses. Where there was no yield penalty for growing bread wheats, it was optimal to aim for bread wheats. Where there was a bread variety penalty, then the farmer needed to consider the yield/price trade off, and determine which provided the higher gross revenue (and thus gross margin).

Development of commercial wheat growing in Australia provided considerable challenges in the new colonies. Principal among these challenges were to develop early maturing varieties to enable wheat growing to spread to drier areas, to breed for disease resistance (e.g. rust), to obtain higher yields, and to improve quality defined in terms of suitability for bread making. Dunsdorfs (1956, p.193-95) also argued that an important aspect of this early wheat breeding effort was in reducing harvesting costs. For example, Dunsdorfs argued that because of its shorter straw, Farrer's variety Federation stood upright and was better suited to mechanical stripping; further, it held grain after ripening thus extending the harvest period. Extensive introductions and selections were made in the nineteenth century (Dunsdorfs 1956, pp.189-90; Macindoe and Walkden Brown 1968, pp.1-2, and see also the latter's variety listing pp.51ff). Systematic breeding efforts began in the 1880s. While Macindoe and Walkden Brown gave precedence to Farrer in the development of Australian wheat breeding, Dunsdorfs (1968, pp.190-91) argued that there were other breeders of “no less historical significance in blazing the new trail for Australian wheat breeding”.

There is at least the appearance of significant latent genetic diversity in the Australian wheat crop around 1900. This apparent diversity was probably over-stated because of the lack of a systematic documentation of the introduction and origins of wheat varieties brought into Australia, and absence of a systematic description of existing wheat varieties (cf. the large numbers of synonyms recognised in Macindoe and Walkden Brown's (1968) listing of Australian wheat varieties). The latent diversity was not translated into effective diversity since few varieties were suitable for Australian growing conditions and, of those that were, many quickly succumbed to disease.

A very rough concept of genetic diversity in Australian wheats may be gained from Macindoe and Walkden Brown's (1968) listing of named varieties. Discounting synonym names, there were about 130 varieties listed as definite or possible introductions, or where the provenance of the variety was unknown. Most of the dated introductions occurred before 1900. These varieties, and others introduced specifically as parents in breeding, formed the basis of the genetic diversity of subsequent varieties. For introduced varieties, direct genetic observation would be required to investigate their genetic diversity. From about 1900, it would be possible to evaluate the relationships among bred varieties using the coefficient of parentage.

In Macindoe and Walkden Brown's (1968) listing of varieties, 235 varieties were attributed to Farrer as ones he bred, selected or introduced. Pye was attributed with breeding 85 varieties. Macindoe and Walkden Brown attributed 160 varieties to “farmers” of which 1 was introduced by a farmer, 106 were selected by farmers, 53 were bred by farmers, and the status of 3 was unclear. The main period of direct farmer involvement with new varieties was 1880-1940: farmers bred or selected 40 new varieties 1880-1900; 51 new varieties 1900-20; and 47 new varieties 1920-40. Six varieties were attributed to other periods and 16 could not be dated.

The creation of State departments of agriculture at about the time that plant improvement was becoming important was of major significance to Australian wheat breeding. While farmers had made important contributions to selecting, and in some cases breeding, new wheat varieties and would continue to do so for some decades, the application of rapidly advancing sciences relating to plant breeding were beyond the capacity of farmers to integrate into activities which, for them, could only be part-time. The nascent government research stations provided opportunities to exploit economies of size and scope in research, of which plant breeding formed a vital part for a country with a relatively new European agriculture based largely on introduced species. Table 2 summarises the contribution of individual breeders employed in government and university institutions (and institutions where an individual breeder was not identified). While institutionalisation of research provided important economies in the plant breeding process, the divorce of breeding objectives from the activities of farmers provided the opportunity for differences to emerge between the objectives of farmers, grain buyers and wheat breeders.

Table 2: Wheat Breeding in Institutions, 1885-1967

Source: Macindoe and Walkden Brown's (1968)

Note: The varieties attributed to institutions are those varieties where an individual breeder was not identified. There may be some double-counting of varieties where more than one breeder cooperated in the development of a variety, for example where one breeder made a cross and another breeder undertook selection and/or fixing of the variety. Locations without an institutional type are state government experiment farms/stations.

2.4 Evolution of regulatory intervention

The rapid growth in wheat production in the 1920s led to increasing tensions between growers and “parasitical” merchants (Whitwell and Sydenham 1991, p.38). The Great Depression, and the catastrophic attempt by the Commonwealth Government to use wheat industry policy as a tool of macroeconomic management – the highly successful but disastrous “grow more wheat” campaign of 1930-31 (Whitwell and Sydenham 1991, pp.35-6, 50-3) – resulted in rapid reductions in wheat area in Victoria, South Australia and Western Australia in the early 1930s and in NSW from the mid-1930s. Despite these difficulties, wheat growers in the principal wheat growing states continued to reject state intervention as a policy response to income difficulties and perceived market imperfections (e.g. Whitwell and Sydenham 1991, pp.43-5).

A proposal to introduce a home consumption price support scheme for wheat similar to those in dairying and dried fruits foundered in the mid-1930s when the latter was declared unconstitutional (Whitwell and Sydenham 1991, pp.54-5). The Commonwealth's Wheat Industry Assistance Act of 1938, together with complementary state legislation, provided for a flour tax which was used to create a home consumption price for flour, an export tax on wheat when export prices exceeded domestic prices and a stabilisation fund to hold and disburse these taxes (Whitwell and Sydenham 1991, pp.55-6).

With the outbreak of World War Two, the Commonwealth Government immediately established another Australian Wheat Board under national security regulations. The Board was responsible for marketing, storage and shipping arrangements, compulsory pooling, and introduced an advance payment on deliveries (Whitwell and Sydenham 1991, p.59). In late 1940, stabilisation arrangements were implemented, starting with the 1941-42 harvest, whose key features were a guaranteed price f.o.b. less charges for a specified maximum crop, a stabilisation fund based on wheat production taxes, and a growing licence with basic acreage. In 1942, stabilisation arrangements were modified via a quota scheme with a differential first advance, coupled with a stockfeed wheat subsidy (Whitwell and Sydenham 1991, p.61). “By 1945 it was taken for granted by government and opposition parties alike and by the states and grower organisations that there would and should be a postwar wheat stabilisation scheme” (Whitwell and Sydenham 1991, p.62). In a 1946 referendum, the Commonwealth failed to gain powers to enable it to enact national commodity stabilisation schemes (Whitwell and Sydenham 1991, p.63).

Finally, in 1948, national wheat marketing arrangements involving the essential aspects of the arrangements of WW2 were enacted under peacetime powers. Two major concessions by the Commonwealth Government, which had significant impacts for the following two decades, were the abandonment of demands for production controls and acceptance of cost of production as the basis for farm-gate pricing (Whitwell and Sydenham 1991, pp.62-3).

3. Production in second half of twentieth century

The development of the Australian wheat industry in the second half of the twentieth century is illustrated in Figure 2. In the early 1950s, wheat area was either static (Western Australia and Queensland) or declining (Victoria and South Australia, and especially NSW where wheat area fell 65 per cent from 1947 to 1956) (Figure 2a). From the mid-1950s, by contrast, wheat area grew rapidly in NSW (nearly sixfold) and Queensland (fivefold increase) from 1956-68, with area more than doubling in the Victoria, South Australia and Western Australia.

The imposition of quotas temporarily reduced wheat area from 1969, but rapid growth recommenced from the early 1970s until the early 1980s. Area doubled in WA and Queensland, and increased 50-80 per cent in the other states. Under the combined pressure of falling international prices, especially resulting from fierce international competition from highly-protected wheat industries in developed countries, high wool prices until the late 1980s, and severe drought in eastern Australia in the first half of the 1990s, wheat areas tumbled in NSW, Victoria and SA in the period 1980-91 (falling 55-65 per cent), fell 25 per cent in Western Australia, but increased nearly 80 per cent in Queensland. Following the end of the severe drought in the northern part of the eastern wheat belt in the 1990s, and the lift in relative wheat price, wheat area grew rapidly in NSW.

Figure 2: Wheat Area, Production and Yield, Australian States, 1939-95

As shown in Table 1, average wheat yields in eastern states (NSW, Queensland) and Western Australia, increased approximately 15-18 kg/ha/year over the period 1950-97 (cf. Figures 2c,d). In Tasmania, average state yields trended at 37 kg/ha/year over the same period. In South Australia and Victoria, there was little evidence of trend (estimated trend was low and of low statistical significance). The variability of wheat yields relative to state mean yield was similar to the 1901-48 period for NSW, South Australia and Western Australia, and declined substantially in Victoria and Queensland (Table 1).

The combination of area changes and average yield changes created different patterns of state level output. In Western Australia, state wheat output increased consistently over the period with relatively little inter-year variability (Figure 2b). By contrast, in the second major wheat producing state, New South Wales, wheat output increased commensurately with that of WA to the early 1980s, tumbled 75 per cent to the early 1990s, and recovered spectacularly in the mid-1990s; the variability about the trend was large compared to WA. The smaller wheat producing states (Victoria, SA and Queensland) increased wheat output to the early 1980s and, like in NSW, wheat output fell substantially to the mid-1990s, especially in SA.

3.1 Variability in Australian agriculture

3.1.1 Sources of variability

As shown in the Green Paper (Harris et al. 1974, chapter 4), Australian agriculture experiences substantial output and price variability. As suggested by the Green Paper's analysis, the major components of income variability can be decomposed into output and price variability. These components may also be further subdivided.

The elements of price variability include:

• domestic prices and export prices, and (changes to) the share weights on these two; and

• classes of wheat and their relative price variabilities, and (changes to) the share weights of these classes;

The elements of output variability include area and yield variability:

• the principal determinant of area variability is the relative profitability of wheat relative to key alternative enterprises (wool, beef, sheepmeats), and thus a key component of area variability is relative price variability;

• the principal determinant of yield variability is weather operating directly through the level, seasonal distribution and intensity of rainfall, and indirectly through the influence of rainfall on pest and disease incidence and damage, production-related attributes such as sowing time, flowering time (particularly relative to the last frost), and harvest;
• yield variability may also be influenced by area variability as, for example, increased wheat area prompts expansion into relatively less-favoured areas, induces farmers to extend wheat sequences (increasing pest and disease problems), or brings area out of other crops, pasture or fallow more quickly;
• yield variability may also be affected by output (or input) price variability if the optimal application rate of key inputs (chemicals and fertilisers) is responsive to relative output:input prices, and farmers respond to changes in optimal application rates rather than being guided by past practice; and
• yield variability may also be affected by technology although generally adoption of new technology is slow; however, emergence of diseases could, if farmers were unable to respond quickly enough, be observed as one-off yield decreases that would observed statistically as increased variability.

Because Australian wheat production is spread over a large and climatically-diverse geographical area, seasonal conditions are not uniform across wheat production areas. This diversity is increasingly important since the major expansion of wheat growing in Western Australia with its relatively less variable climate. Thus observed variability at the national level is likely to be lower than at the farm level (cf. Harris et al. 1974, para. 4.7).

3.1.2 Estimates

At the national level, output variability was substantially greater for broadacre cropping (including wheat) than for all other industries except cotton. This variability was greater 1960-61 to 1972-73 (0.3-0.5 for broadacre cropping) than over the longer period 1949-50 to 1972-73 (0.3-0.4) (Harris et al. 1974, Figures 4.1-4.2). Price variability was low in wheat (0.05) compared to other broadacre cropping (0.1-0.25); surprisingly, the variability of average export prices was also low (Harris et al. 1974, Appendix Tables A4.1-A4.2). Since the analysis was conducted over the period of the wheat price stabilisation scheme it is not surprising that prices were more stable for wheat than other broadacre cropping. The price variability of other broadacre cropping industries was similar to or greater than most other industries with the exception of wool and potatoes (Harris et al. 1974, Figures 4.1-4.2).

Corresponding to the Green Paper analysis, variability in the post-WW2 period for the wheat industry at an aggregate level is reported in Table 3. Because of the operation of a highly regulated wheat market through national wheat marketing arrangements, with an objective of “stabilisation”, for most of the post-WW2 period, observed variability should be treated cautiously. For most states in most of the sub-periods reported, relative area and yield variability exceed 0.2, and exceed 0.3 in NSW and Queensland. Western Australia consistently exhibits the lowest relative variability of area and yield (below 0.2).

The relative variability of state wheat area tends to be lowest in the period of rapidly increasing yields; conversely, the relative variability of wheat yield tends to be highest in the period of rapidly increasing yields. Thus there appears to be a negative correlation between the relative variability of both area and the rate of yield increase at the state and national level (Figure 4). However, since the higher rates of yield increase occur in the latter part of the period, this apparent negative correlation may be an artefact of some other relationship. These higher rates of yield increase are associated with the adoption of semi-dwarf, high-yielding wheat varieties which generally require higher levels of inputs (e.g. fertilisers and chemical weed control); use of these inputs may also reduce variability. Additionally, these periods are also associated with rapid increases in machinery size which may also affect variability, although the a priori effect is ambiguous.

Table 3: Sources of Relative Variability in Australian Wheat Industry

Variability estimated as standard error of regression residuals relative to mean of series

Sources: Harris (1974), and computed from ABARE data.

Figure 4: Variablity vs. Yield, Australian States

3.2 Farm structure in late twentieth century

In a 1947 enquiry, the Simpson Committee on the cost of wheat production found only 1 out of 635 farmers solely producing wheat. The Committee concluded that “We are satisfied that the pure wheat farmer has ceased to play any part in the production of wheat in Australia” (per Whitwell and Sydenham 1991, p.139). However, structure is not immutable. In 1996-97, 15,358 Australian specialist cropping farms produced an estimated 16.233 million tonnes of wheat. In the same year, 14,014 mixed livestock-cropping farms produced an estimated 4.541 million tonnes of wheat. The enterprise nature of these two farming types is shown in Figure 5.

Figure 5: Categorisation of Farms by Size and Livestock Enterprises, by State

The “mixed livestock-cropping farms” produced, as their class name suggests, a mix of crops and livestock (Figure 5b). The area of wheat grown on these farms in each state averaged 100-200 hectares in 1996-97. In NSW, Victoria and South Australia, farms of this type averaged 1500-2000 sheep, with an average of 3,500 sheep in Western Australia. Queensland farms of this type had small numbers of sheep, but averaged 400 head of beef cattle. In NSW, mixed livestock-cropping farms averaged over 150 head of beef cattle, while average beef numbers were low in the other states. About 100 hectares of grain legumes on average were grown on this farm type in South Australia and Western Australia, with 150 hectares of other crops in Queensland.

On specialist cropping farms, the area of wheat grown on these farms in each state in 1996-97 averaged 250-450 hectares in the eastern states, and approximately 1100 hectares in Western Australia (Figure 5a). Wheat area represented approximately 20 per cent of total farm area in the eastern states, and 34 per cent in Western Australia. Despite this greater cropping specialisation, the average number of sheep on these specialist cropping farms was 2600 in Western Australia (average farm area 3029 hectares) and approximately 1600 sheep and 90 head of cattle in NSW (average farm area 1870 hectares).

In states where average farm area was smaller, wheat area and livestock numbers were smaller: Victoria averaged 258 hectares of wheat and 755 sheep; Queensland averaged 268 hectares of wheat and 61 cattle; and South Australia averaged 286 hectares of wheat and 808 sheep. NSW farms also averaged 220 hectares of crops other than wheat; comparable figures for the other states were Victoria (410 hectares), Queensland (255 hectares), Western Australia (540 hectares) and South Australia (280 hectares)

In Table 4 is reported State-level breakdowns of specialist cropping farms by size of farm (measured as gross receipts) and size of sheep flock (in some cases, numbers of respondents are too small to report details).

• in NSW, 22.5% of specialist cropping farms reported no sheep although, in the largest turnover category of farms, 234 sheep were sold in the year. More importantly, in both cases, there were significant numbers of cattle. On other farms, there were significant numbers of sheep and/or cattle.
• in Victoria, 31.3% of specialist cropping farms reported no sheep although, in the smallest and largest turnover categories of farms, 170 and 225 sheep were sold in the year respectively. Only in the middle turnover category where no sheep were reported was there no obvious grazing enterprise. On average, all other categories of specialist farms reported sheep flocks averaging 270-910 head and small beef herds.
• n Queensland, most of the 1,973 specialist cropping farms reported no sheep and, on average, small to modest beef herds (30-235 head).
• in South Australia, 40.2% of specialist cropping farms reported no sheep; in the middle turnover category of these farms there were small numbers of beef and sheep sold and in the largest turnover category, modest numbers of sheep. Amongst other categories of farm there were sheep flocks averaging 660-2100 head and wool clips averaging 3600-11900 kg.
• in Western Australia, 14.7% of specialist cropping farms reported no sheep and where detailed data was available (only for the middle turnover category) there were on average modest beef herds. Amongst other categories of farm there were sheep flocks averaging 1600-4100 head and wool clips averaging 6800-23100 kg.

Table 4: Categorisation of Specialist Cropping Farms by Size and Livestock Enterprise, by State, 1997-98