Landuse Change, Productivity & Development - Recent Land Use Changes (1983-1997)

Final Report of Theme 5.1 to the National Land & Water Resources Audit
August 2001

3. Recent land Use Changes (1983-1997)

Summary

Land use has changed during the 1982-83 to 1996-97 timeframe to different extents across the area of intensive agricultural land uses. Least change was noted in areas furthest inland where total rainfall and its reliability were lowest. Greatest change was noted in areas that had access to significant irrigation or close to major centres of population.

A large number of factors have likely influenced those changes. Some of note include deregulation of the financial system and of commodity marketing boards, two widespread droughts, major shifts in international markets, shifts in public perceptions and attitudes, and technology innovations to productivity. Shifts in the price of products and continuing declines in terms of trade have often been the most immediate evidence of the influence of these factors.

Continuing trends included slight declines in the total area of agricultural land, with farm sizes increasing on average, with a decrease particularly noticed in the smallest farms (less than 100 hectares). The area irrigated has increased, particularly in New South Wales and Queensland.

Diversity, of agricultural plant species and farm activities, was highest in the grain cropping belt with individual farm diversity generally reflected at the shire level. Most of the southern cropping belt increased diversity by at least 5 percent by increasing proportion of non-cereal grain crops.

The high proportion of the total area of agricultural holdings that is not designated in the Agricultural Census - the 'residual' - means that many natural resource ecosystems have very sparse information about land use for agriculture. This residual area needs to be given more attention to determine actual land use and potential for diversification more accurately.

Although there has been considerable changes in land use in some parts of Australia, there is little to indicate the effect of land degradation issues.

The previous section described the major challenges and responses in agricultural history. What have been the important changes during the last 20 years? Has the pioneering development phase finished?

The best sources of information to provide a profile of change over the last 20 years was the annual Australian Bureau of Statistics (ABS) Agricultural Census (AgStats) and the Australian Agricultural and Grazing Industry Survey (AAGIS) undertaken by the Australian Bureau of Agricultural and Resource Economics (ABARE).

The AgStats data was available in electronic form from 1982-83 until 1996-97. It provides details on area and production collected from individual agricultural holdings but aggregated to Statistical Local Areas (SLA - usually the same as local government areas such as shires) and larger groups such as Statistical Division, States and national. Here we use the convention of the ABS to use the year of recording as nominated year, ie 1982-83 year was recorded at March 31 1983 and is referred to as 1983. As the Audit commenced in 1997 the last census year (1996-97) is used as the base year. (Since 1997 ABS have conducted surveys rather than census and only report results at Statistical Division Level.)

During the time frame of the AgStats available there have been changes to the questions asked, in the value of goods due to inflation and in the boundaries of Statistical Local Areas (SLAs). In response to these changes a minimum data set of 475 items that was common to the majority of years was selected, adjusted to 1996-97 dollar values with an cutoff for inclusion of farm Estimated Value of Agricultural Operation (EVAO) of $22,500. This data set was further concorded to account for changes in boundaries of the SLAs. Further details are given in the Methods in Appendix 2.

A limitation in using the SLA unit for showing land use change is that their boundaries tend to follow convenient features such as rivers and roads rather than catchments and landscapes. However, considered as part of a mosaic with adjoining SLAs, they provide a broad contextual picture. Furthermore they cover an important jurisdiction influencing land use, namely local government.

The seasonal weather varied greatly during the 15 years under study. As an example, Figure 1 in Appendix 1 shows the total rainfall experienced across Australia for a major growing season (June to October inclusive) for each of the years as a percentile of that recorded from 1900. The severe droughts of 1982 (ABS year 1982-83) and 1994 (ABS year 1994-95) over most of the areas of intensive agriculture show out strongly. In 1996 (ABS year 1996-97 - the base year for the Audit) rainfall appears to be mostly above average. By contrast the remaining years were quite variable, with some regions experiencing above average whilst at the same time others experienced very dry conditions. Other combinations of historical rainfall in comparison to the long term record can be viewed from the web-sites of the Bureau of Meteorology (www.bom.gov.au) or the Queensland Department of Natural Resources and Mines (www.nrm.qld.gov.au/longpdk/index.html).

3.1 Pattern of Land Use in Australia (1996-97)

The use of agricultural land as presented in ABS AgStats is summarised in Table 3-1. The largest land area for agricultural use is in Queensland followed by Western Australia. Agroforestry, which consists of seedlings and seeds sown on agricultural holdings, was very small in comparison to other land uses. Sown pastures are usually composed of fertilised, introduced pasture species for the intent of more intensively grazing dairy cattle, beef cattle and sheep and appears most prominently in New South Wales and Western Australia. Native or naturalised pastures are not sown, although likely to be modified in composition by sustained grazing, rarely fertilised, and are extensively grazed. Crop areas are highly modified to either annual or perennial species, usually fertilised and more intensively managed for productivity. The residual area - that is the area not identified as having a use in the Census - is likely to be low productivity woodlands or tussock grasslands and to be used for extensive grazing. It represents by far the largest proportion of agricultural land in New South Wales, Queensland, South Australia and Western Australia; only in Victoria and Tasmania does other uses such as sown pastures approach the same level as the residual. Included in uses for sown pastures and crops is irrigation, however, in aggregate this amounts to a small fraction of the total area of each State.

Table 3-1. Area (million hectares) of agricultural land and its major uses for each State in 1996-97. Source ABS AgStats

	New South Wales	Victoria	Queensland	South Australia	Western Australia	Tasmania
Total agricultural area	60.901	12.745	149.560	56.219	112.482	1.920
Agroforestry	0.022	0.018	0.005	0.011	0.040	0.001
Sown pastures	4.436	3.945	3.219	2.214	4.542	0.639
Native or naturalised pasture	10.314	0.158	28.770	7.996	12.778	0.379
Crops	5.589	2.552	2.685	3.278	6.950	0.073
Residual area	40.540	4.654	114.882	42.720	88.171	0.834
Irrigation	0.907	0.546	0.404	0.115	0.030	0.051

A comprehensive map of land use, derived from the AgStats 1996-97, was produced by another Audit project. This map is summarised in Figure 3-1 to illustrate some of the main features of the more detailed full map available at http://environment.gov.au/atlas. The land uses in constructing this map from ABS AgStats 1996-97 are based on the Australian Land Use and Management Classification (available at http://www.daff.gov.au/content/output.cfm?ObjectID=D2C48F86-BA1A-11A1-A2200060B0A05976). The products from this present project are designed to complement and to elaborate on that work. For instance, the map was used to concord previous AgStats census years to the common 1996 SLA boundaries.

Within the crop category a wide variety of crop types are grown around Australia (Table 3-2). The most widely grown are the broadacre crops particularly cereals of which the largest single crop is wheat. Oilseeds, such as canola, and pulses, such as lupins and field peas, have been increasing in popularity and adding diversity to cropping systems.

Table 3-2 Area (thousands of hectares) of major crops types grown in Australia during 1996-97. Source ABS AgStats.

	New South Wales	Victoria	Queensland	South Australia	Western Australia	Tasmania
Broadacre crops	5079	2453	2097	320	6921	41
Hay	271	490	62	202	202	54
Cereals for grain	4676	1826	1676	2781	5519	27
Wheat	3192	962	980	1535	4264	2
Oilseeds	247	115	112	42	107
Pulses	138	429	61	257	1172
Semi-intensive crops	424	12	5059	9	4	7
Rice	151	0.8
Sugar cane	18		371		1
Cotton	249		129
Potatoes	6	11	47	9	3	7
Horticultural crops	85	86	83	73	24	24
Vegetables	21	30	35	13	11	19
Fruit	60	50	44	57	11	4
Grapes	20	25	1	38	4
Plantation fruit	4	0.8	13	0.1	0.6
Orchard trees	35	24	29	18	6	3

Figure 3-1 Summary map of land use patterns in Australia during 1996-97. (source National Land and Water Resources Audit, 2001 - Land Use Map of Australia, version 2 1:1 million scale)

3.2 Farm size

The total number of agricultural holdings decreased from 178,025 in 1982 to 145,086 in 1997. This appears to have come about by a reduction in the number of smaller area farms (particularly those less than 1,000 hectares) as shown in Figure 3-2. In relative terms the greatest reduction occurred in those farms less than 100 hectares in area.

Figure 3-2 The distribution of number of farms by farm size in 1982 and 1997. (source ABS)

Figure 3-3 Changes in the distribution of farm sizes in different land uses between 1982 and 1997. (source ABS)

The reduced number of farms appears to have occurred in all of the major groups of land use with area of operations less than 100 hectares (Figure 3-3). Sheep and beef cattle producing farms with area between 100 and 5,000 hectares have increased in the time. Grain cropping was another group that showed a substantial increase in numbers, in this case in the categories between 2,500 hectares and 25,000 hectares.

3.3 Changes in land use between 1982-83 and 1996-97.

Data from the annual Agricultural Census (AgStats) provides information on many different agricultural products. To provide some meaningful indications of land use over the 20 years and across the areas of intensive agriculture it was necessary to aggregate the data in different ways. The Audit has identified 21 such groups that represent intensifications of land use similar to that used by Australian Land Use and Management Classification.

For easier discussion this project has further reduced that to 5 larger categories that have similar levels of intensity, natural resource management issues and sometimes location. They are: extensive grazing; sown pastures; broadacre crops; semi-intensive crops; and horticulture.

The changes, shown at the State level over the timeframe, are aggregates of many individual changes that occur at local levels. To be concise, we present maps that show changes as the difference between the beginning and the end of the period. For the beginning we used averages of the 3 years 1983,1984 and 1985 and for the end 1994, 1996 and 1997. The year 1995 was omitted because of an incomplete number of categories collected that year, which was unfortunate because drought effects would have occurred in both sets ie 1982-83 and 1994-95. The maps show the percentage changes in the proportion of the total agricultural land reported for each year.

More detailed changes will be illustrated through the case studies, which are presented in Chapter 5. Within the SLAs there are regions with different rates and direction of change, that reflect different resource conditions, soil, aspect and access to water.

Extensive grazing land use
The area of extensive grazing land was estimated by the sum of native (and naturalised) pasture and the ‘residual’ (total agricultural area less all nominated land use areas) as discussed in section 3.1 above. The areas of this category show up in the Land Use Map (figure 3-1) as category 2.1 ‘livestock grazing’. The changes in the area that can be designated as extensive grazing in the areas of intensive agriculture of the states are shown in Figure 3-4. This indicates gradual declines in this extensive grazing area of the areas of intensive agriculture in South Australia and Victoria, possibly in Queensland and considerable variation within New South Wales over this time. However, the method of defining this land use suggests a cautious interpretation of this figure. Some of the variations in the chart are due to changes in the categories collected in the different census years. For instance reduced categories of land uses collected for most states in 1995, for Victoria in 1991 and for Western Australia in 1992.

Figure 3-4 Changes in the area of extensive grazing lands (native pasture plus residual) for the areas of intensive land use of each State during 1983 to 1997. (Source ABS.)

The map of changes in Figure 3-5 indicates that in much of the areas of intensive agriculture there have been only small changes (of 5% or less). Larger increases have occurred in the Great Southern and Swan Coastal Plain of Western Australia, in eastern parts of the Eyre peninsular, the Lower Southeast and the Murray-Mallee of South Australia, the Western Districts of Victoria and the western Darling Downs of Queensland. Larger decreases have occurred in the Mallee of Victoria, the Riverina, Southern Tablelands, Northern Tablelands and South Coast of New South Wales and the Burnett of Queensland.

Figure 3-5 Changes in the area of extensive grazing within the areas of intensive land use in Australia, comparing the 3 years 1983-85 with the 3 years 1994, 1996 and 1997. (source ABS)

Sown pasture area
Sown pasture, wherein land has been deliberately sown to pasture species to improve productivity, represents another state of intensification. Improved pastures add value through the extra livestock they feed – dairy cows, beef cattle, wool sheep and fat lambs. The location of this land use is shown in the National Land Use map of 1996/97 (Figure 3-1) as category 3.3 ‘grazing modified pastures’.

This category includes lucerne as well as other pasture legumes and grasses. However, no data was collected in 1995 except in the Northern Territory. There were also changes in definition of some components for 1996 and 1997 (Table 3-1 ).

1997 & 1996 categories and description		1983 to 1994 categories and descrption
1000401	Lucerne and other legumes - area at 31 March	1000501	Pure lucerne at 31 March - area
1001201	Pasture legumes (exluding lucerne) - area at 31 March	1001201	Pasture legumes excluding pure lucerne at 31 March - area
1001301	Pasture grasses - area at 31 March	1001301	Sown grasses at 31 March - area
1002101	Mixture of perennial grasses and legumes - area at 31 March	1001401	Mixture of grasses and legumes at 31 March - area
1002201	Mixture of annual grasses and legumes - area at 31 March
1002511	Sown pastures at 31 March 1997 - total area	1002751	Sown pastures and grasses at 31 March - total area

The area of sown pastures increased in all States until about 1970 when the areas reached plateaux except in Queensland which increased until 1994 (Figure 3.6). The sharp decline in 1996 and following years reflects the change in definition of the items collected (see Table 3-1). It is unlikely that the drought of 1994 would have had an impact that lasted over the succeeding four years. However, it is not obvious why the changes that seem rather minor should have had such a large impact on the area recorded, particularly when the area of crops did not change very dramatically at the same time (see following section). Given the relative stability in area of sown pastures in most of the States up until then the areas could be assumed to be similar for the following years. In the case of Queensland if the areas of sown pasture that continued to increase up to 1993-94 were sown on cleared woodlands, then it may be best to also assume that the current areas are likely to be slightly larger than indicated in Figure 3-6.

Figure 3-6 Area of sown pasture in the States of Australia since 1950. (source ABS)

There are, of course, many different changes in the area of sown pasture across Australia (Figure 3-7). This map indicates that there are areas to show increases (greater than 5% of total area) in sown pastures during 1983 to 1997 and they are widespread in eastern New South Wales and Queensland. There were also increases in northern Victoria along the New South Wales border, and western Tasmania. Reductions (greater than 5% of total area) in area of sown pasture appeared in most of south-west Western Australia, the Murray Mallee and the Lower Southeast of South Australia, and adjoining areas in Victoria.

The number of sheep supported in the areas of intensive agriculture (Figure 3.8) generally increased in all states from 1983 until 1990, and in Queensland until 1991. Numbers generally declined from then with each succeeding year until 1994 in Victoria and Queensland, until 1995 in New South Wales and South Australia and 1997 in Western Australia. Most of this should be ascribed to the decline in wool prices since 1990 when the wool support price scheme was abolished. However other factors such as a wide-spread drought in New South Wales and Queensland during the mid-1990s would also have had an influence.

Beef cattle are the other major grazer of pastures and changes in their number within the areas of intensive agriculture in the states are shown in Figure 3.9. In some contrast to sheep numbers, beef cattle have generally increased in number over much of the timeframe. There is a decline evident during 1987 and 1988 in New South Wales, Victoria and Queensland, which reversed in the succeeding years, that is likely a response to allow sheep numbers to increase as a result of high wool prices at that time (Figure 2-11).

Figure 3-7 Changes in the area of sown pasture in the areas of intensive land use of Australia comparing the 3 years 1983-85 and the 3 years 1994, 1996, 1997. (source ABS)

Figure 3-8 Changes in sheep numbers within the areas of intensive land use of each State during 1982-83 to 1996-97. (source ABS)

Figure 3-9 Changes in the number of cattle within the areas of intensive land use of each State during 1982-83 to 1996-97. (source ABS)

Broadacre crop area
The broadacre crops category includes all the annual crops such as wheat and other cereals, oilseeds, and pulses grown in both winter and summer, but excludes cotton, rice and sugar cane. Its extent is shown in the Land Use map of 1996/97 (Figure 3-1) as category 3.4 ‘cropping’ which in the map also includes 3.5 ‘perennial horticulture’ and 3.6 ‘seasonal horticulture’.

There are considerable differences between the States in the changes in this land use over the timeframe (Figure 3-10). In Western Australia the area sown increased each year up to 1985, with a decline in area sown until 1989 whereafter there has been a steady increase in area to achieve a record area sown in 1997. Victoria, New South Wales and Queensland had broadly similar patterns although recovery during the 1990s was less pronounced in Queensland. In New South Wales and Queensland there is a dramatic impact of the 1994/95 drought, and by 1997 had still not returned to the areas sown during 1984-86 nor in Queensland to areas of crop areas sown during the drought of 1983. In South Australia the pattern is much more stable, with few fluctuations and relatively small increases occurring over the time frame. Tasmania has such a relatively small area of this category that changes do not show on the scale of this figure.

Figure 3-10 Changes in the area sown to broadacre grain crops within each State during 1982-83 to 1996-97. (source ABS)

The map in Figure 3-12 shows where the changes occurred between the beginning (3 years of 1983 to 1985) and the end (3 years of 1994, 1996 and 1997) of the sequence. About half of the areas of intensive agriculture do not show any net change in the area of cropping undertaken in most cases because there is minimal amount of cropping. Changes of less than 5% of total area were also common across the cropping belt. Significant increases (greater than 5%) appeared to occur in the wider mallee lands from south-east Western Australia to north-west Victoria. These include the western Great Southern, Eyre Peninsular, Flinders Ranges, into the Murray Mallee, the Mallee and Wimmera of Victoria. By contrast in some areas, such as in the Darling Downs and Fitzroy Basin in Queensland, the Central West and eastern Riverina of New South Wales, and the western part of the Central Wheat Belt of Western Australia there appears a general reduction in the area under crop over the 15 years.

The profile of farm sizes of Australian grain-growing farms (which includes sheep-wheat groups) has changed from 1983 to 1997 (Figure 3-11). The total numbers have decreased from 38,365 in 1983 to 36,868 in 1997. This has occurred in the smaller farm area categories of less than 1,000 hectares while there have been increases, although fewer, in the categories greater than 2,500 hectares. The figure also shows that most farms producing grain are between 100 and 2,500 hectares in area.

Figure 3-11 Profile the sizes of farms producing mainly grain in 1982 and 1997. (source ABS)

Figure 3-12 Changes in area sown to broadacre crops in the area of intensive land use of Australia comparing the 3 years 1983-85 and the 3 years 1994, 1996 and 1997. (source ABS)

Semi-intensive crop areas
The category of semi-intensive crops includes cotton, sugar cane, rice and potatoes. This category is included in the National Land Use map of 1996-97 (Figure 3-1) as 4.0.0 ‘production from irrigated agriculture and plantations (undifferentiated)’. In general the area sown to the semi-intensive crops has increased over the time, with the trend particularly noticeable in Queensland and New South Wales over the period, perhaps with more variability in NSW (Figure 3-13). There appears very little area sown to these semi-intensive crops in the other states.

Figure 3-13 Changes during 1982-83 to 1996-97 in the area sown to semi-intensive crops in the States of Australia. (source ABS)

Many of the areas of intensive agriculture do not grow any of these semi-intensive crops (Figure 3-1) and so do not show a change. The areas that do show a change in area sown to this group are illustrated in Figure 3-14.

Cotton areas, located on the Moree Plains down to Warren in New South Wales, on the Darling Downs and the Central Highlands of Queensland, have all increased by at least 5% over the 15 years.

Sugar cane production is located along the eastern coast of Australia from northern New South Wales to just north of Cairns in Queensland. Most of the areas have increased, sometimes by more than 10%. However, within these regions small areas near Innisfail, Townsville and Bundaberg have declined due to conversion to horticultural activities and urban development.

Rice production is located almost solely in the MIA of New South Wales in the SLAs around Deniliquin where increases have generally been about 5%.

Potato production has increased by about 5% in the Albany coast of Western Australia, in the Murray Mallee, the Adelaide Hills and Lower Southeast of South Australia (Mount Gambier), the north west of Tasmania around Devonport, in southern Victoria around Melbourne, around the west of Sydney, the Dorrigo in northern New South Wales, and around Brisbane in Queensland. Larger increases of 10% have occurred near Devonport, southern Victoria, and southeast Queensland.

Figure 3-14 Changes in the area sown to semi-intensive crops in the ares of intensive land use of Australia comparing the 3 years 1983-85 and the 3 years 1994, 1996, 1997. (source ABS)

Horticultural crop area
Horticultural crops here include all vegetables (except potatoes), fruit, nuts, vines, nurseries and turf. The modified National Land Use Map for 1996-97 (Figure 3-1) does not show these areas because they are too small (there is a maximum of 80,000 hectares in any one State) to show up at the scale used.

There is a trend for the area planted to these crops to increase over the time frame in all states (Figure 3-15). It is possible that the large variations noted for South Australia in 1994 and Victoria in 1995 are anomalies in the collection of the agricultural statistics. It is also highly likely that the areas of vegetables are significantly under-reported in AgStats.

Figure 3-15 Areas sown to horticultural crops between 1982-83 and 1996-97 in the States of Australia. (source ABS)

The total number of horticultural farms declined from 20,231 in 1982 to 17,891 in 1997. This was effected mostly through a reduction in the number of farms in the most numerous and smallest area category of less than 50 hectares (Figure 3-16).

Figure 3-16 Profile of horticultural farm sizes for Australia in 1982 and 1997. (source ABS)

Location of the SLAs that had changes in the area under horticultural crops is shown in Figure 3-17. This map further highlights that only a relatively small area of the areas of intensive agriculture is used to cultivate horticultural crops. Increases appear to occur where irrigation water was abundant and around major population centres. Modest increases of 5% occurred in the larger regions: such as Margaret River, Albany coast (wines at Frankland and Mount Barker)in Western Australia; the Adelaide hills, Riverland and Coonawarra/Padthaway of South Australia; the Huon valley and north coast (Devonport) of Tasmania; central Gippsland, Goulburn Valley (to Shepparton) and around Melbourne former tobacco areas of the Ovens and Kiewa valleys in Victoria; around Batlow, Griffith, Orange, and Coffs Harbour in New South Wales; and in Queensland around Stanthorpe, the rim of the Great Dividing Range up to Toowoomba, around Brisbane, Bundaberg, Townsville and Innisfail. Some areas that experienced increases of 10% were near Perth, around Adelaide, in the Riverland, the Sunraysia and Swan Hill areas, around Devonport, Melbourne, Sydney, Lismore, and Brisbane.

There were very small areas close to Perth, Melbourne, Sydney and Brisbane that declined in the area of horticultural crops most likely due to urban encroachment.

Figure 3-17 Changes in the area under horticulture in the areas of intensive land use of Australia comparing the 3 years 1983 - 85 and the 3 years 1994, 1996, 1997. (source ABS).

3.4 Changes in irrigation

Since much of Australian agriculture is limited by rainfall, supplying extra water through irrigation increases potential productivity enormously. Moreover, the cost of obtaining such water usually means that higher value activities are undertaken, resulting in intensification of land use. However, the capital costs involved also require that supplies of water be reliable. Consequently most irrigation is limited to dedicated regions where the supply is controlled by dams and weirs. The Murray-Darling Basin Commission, in partnership with the National Land and Water Resources Audit, initiated Phase 1 of the Irrigation Management Information Reporting System. As part of this activity summary statistics and information has been compiled for Australia and the MDB.

Between 1955 and 1967 an increase in storage capacity of reservoirs resulted in an increase of irrigated land from 500,000 hectares to 1,280,000 hectares, mostly in New South Wales and Victoria with the Murray and Murrumbidgee Rivers the major source of water. By 1983-84 this had steadily increased to 1,625,000 hectares, with over 70% of all irrigation water in Australia were withdrawn from the Murray-Darling system of which over half was used to irrigate pasture rather than crops (Sewell 1985). At the end of the period of interest, in 1996-97 the area had reached 2,056,000 hectares under irrigation. At this time a ‘cap’ was imposed by the Murray-Darling Basin Commission on the diversions of surface water for irrigation set at the level of development of the 1993-94 financial year. The use of water for irrigation purposes has changed considerably since the turn of the nineteenth century with intensive irrigated agriculture such as wine grapes, horticulture, sugar and cotton becoming increasingly more profitable than other types of crops.

ABS has collected information about irrigation somewhat irregularly over the timeframe. Also, the questions that were asked in the agricultural census about irrigation activities have changed with time. Changes in the areas of irrigation are calculated using the difference in the mean of 1984, 1987 and 1990 with the mean of 1991, 1992, 1993, 1994 and 1997. Because of incomplete irrigation datasets, boundary and coding changes, the statistics were standardised to the total area, pastures, cereals (includes rice), vegetables, fruits (includes nuts and vines), and other crops (includes cotton, sugar cane, soybean etc).

Total irrigation area
As shown in Figure 3-18 about 45% of the total area irrigated is in New South Wales, 27% in Victoria, nearly 20% in Queensland , 5% in South Australia, and less than 3% in Western Australia and Tasmania. Most of the changes in areas irrigated appear to have occurred in New South Wales and Queensland.

Changes in the total area of irrigation are shown in Figure 3-19. Approximately 1,300 Statistical Local Areas (SLAs) throughout Australia are recorded as receiving some type of irrigation during the years 1984-1997. However, only about 600 show a pattern of regular, significant irrigation throughout those years. It is clear that some irrigation only occurs during dry years in some parts of Australia. There was a general increase in the irrigation of crops and pastures in 1994, particularly in the eastern states (Figure 3-19) where rainfall was below average during the years 1991-95.

Figure 3-18 Areas irrigated in the States of Australia during 1983-84 to 1996-97. (source ABS)

The regions where irrigation increased in the timeframe by more than 10,000 hectares include Burdekin, (Qld), Warren, Carrathool, Wakool and Hay in the Murrumbidgee Irrigation Area (NSW). Narrabri, Moree and Waggamba in northern NSW also had increases of 7000 or more hectares. The larger increases were usually associated with major rivers such as the Murrumbidgee, Lachlan, McIntyre, Barwon, Macquarie, Burdekin and the Daly. The most significant decreases occurred in the regions of Gippsland, Mildura and Gannawarra in Victoria, of Singleton/Hawkesbury, Richmond, Copmanhurst and Lismore in NSW, and of Paroo in Queensland.

The total area of major commodity groups that were irrigated at the beginning (1983-84) and end (1996-97)of the Audit period is given in Table 3-4. The category of other crops has doubled in area during the time frame due mostly to cotton (314,956 ha in 1996-97) and sugar cane (173,2223 ha in 1996-97), which were not recorded separately for the whole of Australia in 1983-84.

Table 3-4 Total area (in thousands of hectares) of commodity groups in Australian that were irrigated in 1983-84 and 1996-97. (source ABS)

Commodity group	1983-84	1996-1997	Increase (ha)	Increase (%)
Pastures	871	935	64	7.3
Cereals	315	337	22	6.9
Vegetables	76	87	11	14.5
Fruit	97	151	54	55.7
Total	1625	2056	431	26.5

Figure 3-19 Changes in the areas irrigated between the 1980s (1984, 1987, 1990) and the 1990s (1991, 1992, 1993, 1994, 1997). (source ABS)

Irrigated pastures
Pastures still constitute the largest area that is irrigated (Table 3-4). It is concentrated in southern New South Wales and Victoria (Figure 3-20), where it is important for the dairy industry. In fact, in the Loddon and Moira West most of the irrigation area is used for pastures, and little for crops. The largest increases in the area of irrigated pastures have occurred in the south-east of the continent and in northern Tasmania.

Irrigated cereals
The largest area, around 280 000 hectares or 80%, of irrigated cereals occurs in New South Wales, with a large proportion (between a third to a half) in rice. The largest increases of more than 1 000 hectares have occurred in this area of the Murrumbidgee Irrigation Area (MIA). Some other increases of 500 to 1 000 hectares have occurred in the Eyre Peninsula, the Central West and Moree Plains of New South Wales, and the Central Highlands of Queensland. Most other areas have experienced little (ie less than 500 hectares) change. Significant declines of more than 500 hectares have occurred in the Murray-Mallee of South Australia and the adjacent Mallee of Victoria, in the Atherton Tablelands of Queensland and in the Central West, the northern and central tablelands of New South Wales. Cereals other than rice have decreased in area irrigated since 1984 so that the total area of cereals irrigated has undergone only a modest increase since 1982-83 (Table 3-4).

Irrigated fruit
Irrigation of fruit has increased in all states although the major region in Australia (Figure 3-22) occurs within the Murray-Darling Basin. South Australia, Victoria and New South Wales continue to be the largest producers of fruit, although changes are occurring within irrigation districts. South Australia has experienced growth in the area of irrigated fruit within the SLA’s of Penola, Tatiara, Waikerie and Loxton where increases of between 500 to 1,000 hectares have occurred. Griffith, New South Wales, has also experienced significant growth as have the Huon Valley in Tasmania and the Margaret River region in Western Australia. The largest decrease in area of irrigation of fruit has occurred in Mildura, Victoria (over 5,000 hectare decrease) while the Lismore/Richmond area in northern NSW also shows a large decrease. The Burdekin region is a major producer of tropical fruits and the SLA’s of Mareeba, Johnstone and Cardwell have increased by more than 1,000 hectares.

Irrigated vegetables
There is a trend for a general increase in the area of irrigated vegetables (Table 3-4). Although the area for irrigated vegetables is comparatively small, it is a high value crop and can provide more than one harvest per season. Queensland has the highest production of irrigated vegetables where large increases (of up to 1,700 hectares - Figure 3-23) have occurred in the Burdekin/Bowen region. In Western Australia, increased irrigation for vegetable production has occurred in the south-west in several SLAs close to the Margaret River. While vegetable production in New South Wales has increased around the border with Queensland, and in the Griffith region, other areas , such as Narromine, Forbes, Dubbo, Singleton and Hawkesbury, have decreased. Victoria shows little change. Tasmania has experienced some decreases in the north-west part of the state, while there have been increase in the north-east. However, some of this variation in individual SLA’s may reflect boundary changes rather than variation in production.

Figure 3-20 Change in the area of irrigated pasture between the 1980s (1984,1987, 1990) and the 1990s (1991, 1992, 1993, 1994, 1997). (Source ABS)

Figure 3-21 Changes in the area of irrigated cereals between the 1980s (1984, 1987, 1990) and the 1990s (1991, 1992, 1993, 1994 and 1997). (source ABS)

Figure 3-22 Changes in the area of irrigated fruit between the 1980s (1984, 1987, 1990) and the 1990s (1991, 1992, 1993, 1994 and 1997). (source ABS)

Figure 3-23 Changes in the area of irrigated vegetables between the 1980s (1984, 1987, 1990) and the 1990s (1991, 1992,1993,1994, and 1997). (source ABS)

Other crops irrigated
New South Wales has the largest irrigated area of ‘Other crops’ (Table 3-4), which consists of mostly of cotton (314,956 ha in 1996-97) but also sugar cane (173,2223 ha in 1996-97), soybeans and other commodities not classified under the major crop categories. The increases in this category have been sufficiently large to now constitute the second largest irrigation land use. In New South Wales increases have occurred (Figure 3-24) in the Macquarie-Barwon region and in the northern border district. Queensland has the next highest area, the majority in sugar cane and cotton, and has significantly increased the area from approximately 9,000 hectares in 1984 to 106,000 hectares in 1997, especially Emerald and Balonne and areas on the New South Wales border. All other states are only small producers of these crops.

In concluding it may be possible that there is much improvement in productivity still possible for many crops and vegetables (except for rice only half to two-thirds of potential yields) in the irrigation areas of northern Victoria and southern New South Wales (Lamble 1999).

Figure 3-24 Changes in the area of 'Other Crops' irrigated between the 1980s (1984, 1987, 1990) and the 1990s (1991, 1992, 1993, 1994 and 1997). (source ABS)

3.5 Intensification of land use

One of the responses to the challenges of the cost-price squeeze and increasing population is likely to be an intensification of land use - a consequence of attempting to secure more economic yield from each hectare. Such intensification implies greater use of inputs, including land resources. Intensification can also be considered as the speeding up some of the processes in agricultural and natural systems. While this leads to larger risks from residues of unused inputs, or a wider range of environmental costs or reduced resources, it also provides more income and financial capacity for addressing environmental issues.

To identify and summarise any such changes during the period from 1983-84 to 1996-97 an index was developed that compared changes between categories of land use as a proportion of the total agricultural area. The data from the 1995 agricultural census was omitted from calculations because a reduced range of items was collected in AgStats compared to other years. The categories were based on the Australian Land Use and Management Classification used in the National Land Use map. The index was calculated (for details see Methods in Appendix ) as
_L_i x F_I

where Li are the proportions of the different land use categories in each region, and Fi are the corresponding intensity factors. The intensity factor was based upon the average cost of production, as a surrogate for the level of intensity of land use, for 1991-1994 taken from the ABS Farm Financial Survey.

Table 3-5 Grouping of land uses into categories and the intensity factor applied to derive the agricultural land use Intensity Index.

Land use category	Major components	Intensity factor
Extensive grazing	Native pasture, residual	0.5
Sown pasture	Lucerne, grasses, legumes	7.5
Broadacre crop	Cereals, oilseeds, pulses	7.5
Semi-intensive crop	Cotton, rice, sugar cane, potatoes	125
Horticulture	Fruit, nuts, vegetables	275

The greatest changes in agricultural land use Intensity Index occurs in a broad crescent that curves around inside the east coast, around the south coast to the southern part of the west coast of Australia and including Tasmania (see the map in Figure 3-25). Further inland, most areas appear less changeable and likely reflects having fewer viable land use options. The areas of greatest change surround large population centres and often occur near irrigation and thus most likely reflect the changes in semi-intensive cropping and horticulture shown above. Within the crescent however, there are also pockets such as in south-east Queensland extending into the northern tablelands of New South Wales, in eastern Victoria, the Eyre Peninsula of South Australia and eastern Western Australia that seem to have changed little in agricultural land use intensity over this time.

Figure 3-25 The range (maximum less minimum) in values of agricultural land use Intensity Index that occurred during 1982-83 to 1996-97. (source ABS)

Changes in intensity index of agricultural land use with time are shown in some examples for individual SLAs (Figure 3-26). The changes occurring at often different times often indicate variable responses to the different challenges experienced, rather than any general trend,. However, the Burdekin SLA is an example of high and increasing intensification - a result of the increasing availability of irrigation water from the Burdekin Dam and its use for sugar production in particular. A comparison of the different lines in the graph indicates a wide range of land use intensity between areas around Australia.

In Figure 3-30 the major components that contribute to the intensity index are shown, over the period 1982-83 to 1996-97 for four SLAs. All four have low levels of semi-intensive and horticultural land uses. There are considerable differences between the shires in the proportions of broadacre cropping from around 40-50 percent in Mingenew to around 10 percent in Boyup Brook and Corangomite shires. Likewise the proportions of sown pasture vary between shires. Within the individual shires there are changes between years but not often in any underlying trend.

Figure 3-26 Comparison of changes in Intensity Index - Value for different Statistical Local Areas (SLAs) between 1982-83 and 1996-97. (source ABS)

Figure 3-27 Comparison of different land use categories over 1982-83 to 1996-97 in four Statistical Local Areas (SLAs) that contribute to the Intensity Index. (source ABS)

3.6 Changes in the diversity of land use

Diversification is an important mechanism for managing risks in production, markets and income. Within the Ecologically Sustainable Development (ESD) process diversity is an important indicator of sustainability, particularly for the facilitation of multiple and sequential land uses. Diversification of plant species used in agricultural production contributes some biological resilience into agricultural systems, through the ability to withstand adverse climatic conditions and minimises the build up of diseases, pests and weed populations, and thus to the maintenance of the resource base (Swanson 1997, quoted in SCARM 1998).

To identify changes of these responses two indices were developed that group land uses and compared them to the total agricultural area. Data from the 1995 collection year was omitted from calculations involving intensiveness and diversification measures because AgStats collected a reduced range of items compared to other years.

Agricultural plant species diversity
Agricultural plant species diversity is defined as the diversity of plant specie used for agricultural production in a region. The major application is to monitor trends within a region. An Agricultural Plant Species Diversity Index, derived from the Shannon Index of Diversity using AgStats data was developed by the National Collaborative Project on Indicators of Sustainable Agriculture (NCPISA ) to quantify this attribute. Although a recognised weakness in the method is that pastures, though they usually contain a range of species, were counted as a single species this project updated the data for the full time-frame of the Audit. The method of calculation and results up to 1995 are presented in SCARM (1998).

Sh2 = - _i sigma ^k Pi log(Pi /Ni )
where k is the number of agricultural species in the Statistical Local Area (SLA), Pi is the proportion of total area in the ith species class, and Ni is the number of holdings on which species i occurs. Total area is the sum of areas of the k species across the SLA.

The present distribution (Figure 3-28) reflects the pattern of winter rainfall. The low values for diversity index in the inland regions reflect high proportions of native pasture, or ‘residual’ areas not attributed to a land use, and should not be interpreted as low biological resilience as these pastures often comprise a well-adapted range of species. Sown pastures, because they constitute grasses, legumes and lucerne, will show a higher diversity than native pasture in this index.

There were some positive changes that occurred in the southern parts of mainland Australia (Figure 3-29), probably associated with diversification in cropping rotations. There were also some, generally smaller, negative changes in Tasmania and in south eastern Queensland that may be a consequence of less sown pasture and reduced cropping diversity. It would also seem that the change in diversity index is mostly reflecting area changes that have flowed from changes to questions in the Agricultural Census during the mid 1990s.

Figure 3-28 Distribution of agricultural diversity (as measured by modified Shannon Index) across the areas of intensive land use of Australia in 1996-97. (source ABS)

Figure 3-29 Changes in diversity (as measured by the Shannon Index) across areas of intensive land use of Australia between the 3 years 1983-85 and the 3 years, 1994, 1996, 1997. (source data ABS)

Broadacre cropping systems diversity
Diversity within the land use category of ‘broadacre crops’ is illustrated by the relative contribution to total area of broadacre crops by crops from different biological families such as cereals, pulses (grain legumes) and oilseeds. As an example, the ratio of non-cereal (includes pulses and oilseeds) to total crop gives some indication of biological diversity that has implications for sustainable cropping. Including non-cereals in cropping rotations provides breaks many disease and weed life-cycles, helps develop soil structure through different rooting patterns, and pulse crops may also improve nitrogen status of soils. However, in general such crops are less adaptable than cereals, require more husbandry to be productive and are more sensitive to land degradation factors such as salinity and acidification.

In 1997, as shown in Figure 3-30, the most diverse areas of winter cropping appear in: the northern part of the Western Australian wheat belt, and reflect intensive use of lupins in crop rotations; in a the Lower Southeastern region of South Australia and the Wimmera of Victoria with peas and canola; and in the Central Highlands of Queensland where sunflowers have increased considerably but where winter cereal areas are not large.

Figure 3-30 An estimate of diversity in broadacre crops using the proportion (%) of total crop that is non-cereal in areas on intensive land use during 1996-97. (Source ABS)

Diversity on farms
Diversification within farms was explored using some of the detailed data collected in the Australian Agricultural and Grazing Industry Survey (AAGIS) by the Australian Bureau of Agricultural and Resource Economics (ABARE). The set covered the full 21 year period 1978-79 - 1998-99 with a 5 category classification system. Details are provided in the Methodology.

The degree of farm diversity, as measured by this diversity ratio, ranges from 1 to 5 - the number of enterprise activities included in the measure. A value of 1 for a farm would indicate that a farm was a specialist in a specified broadacre industry and a higher diversity ratio value for a farm would indicate that the farm undertook more activities and was more diverse. Broadacre farms only were included in the target population As a caution, unlike the ABS categories that were concorded to a constant estimated value of agricultural operations (EVAO) of $22,500, the ABARE survey data was not concorded but ranged from $5,000 to $22,500.

Figures 3-31 and 3-32 show that regions of highest diversity within broadacre farms consistently lie within the traditional sheep-wheat belt of southern NSW and northern Victoria. The map for 1997 (Figure 3-32) indicates moderately high diversity of broadacre farms within the cropping belt, with higher diversity areas within southern New South Wales, northern Victoria and southern Western Australia.

Figure 3-31 Changes in financial diversity ratios calculated from ABARE Farm surveys for 1982-83 interpolated to give a smoothed distribution.

Figure 3-32 Changes in financial diversity ratios calculated from ABARE Farm surveys for 1996-97 interpolated to give a smoothed distribution.

Some areas show a high level of diversity consistent with that shown for plant diversity at a Statistical Local area in Figure 3-28. These included the south coast of Western Australia, the Wimmera and Goulburn of Victoria, and the south eastern Riverina of NSW. This would support the idea that individual farms in these areas undertook a range of industries and so showed in both analyses. Areas that did not show up in the SLA map include the northern cropping region of Western Australia, western Riverina and central northern plains of NSW and south-east and Flinders Ranges of South Australia. This would suggest that farms in these areas were more specialist in their industry and that diversity was contributed by different farms having different specialisations.

Interpolated values of physical diversity for each of the 15 years mapped for the whole of Australia are given in Figure 7 in Appendix 1. It was also possible to calculate physical diversity by converting the different activities to Dry Sheep Equivalents (DSE). The conversions were: each hectare of crop was 12 DSE, cattle were counted as 8 DSE, and sheep as 1 DSE. In this form diversity increases from a low of 1 to a maximum of 5 - the number of categories included. The results of interpolated values of physical diversity for each of the 15 years mapped for the whole of Australia are given in Figure 8 in Appendix 1.

In general there is great similarity in the patterns of both physical and financial diversity. There is considerable variation in both attributes from year to year, in each year the cropping belt stands out for its higher diversity in this method. Areas that are consistently high for diversity appear to be the Great Southern region, the lower South East of South Australia, the Wimmera and Goulburn Valley of Victoria, the Riverina and Central West of New South Wales. More variable for diversity included the central wheat belt of Western Australia, the Mallee and Western District of Victoria, the Moree Plains of New South Wales, and the Darling Downs and Central Highlands of Queensland. Areas that appeared low in diversity include the Eyre Peninsula, Esperance, Gippsland, Southern Tablelands and Far West of New South Wales and much of Queensland.

3.7 Conclusions

The Steering Committee on Natural Resource Management Policy Statement (1999) discussion paper, and subsequently supported by public responses (2000), forecast that fundamental changes in land use and management practices may be required at the farm level across regions in order to sustain an area’s capacity for long-term sustainable production or to protect areas that are critical for other uses- such as water supply and biodiversity.

The changes in land use during the 15 year timeframe indicate that there is capacity to change given incentives to do so. While a large number of factors have likely influenced those changes, shifts in the price of products and continuing declines in terms of trade have often been the most immediate evidence of the influence of these factors.

There is a high proportion of the total area of agricultural holdings that is not designated in the Agricultural Census - called ‘residual’ and allocated to ‘extensive grazing’ land use in this report - which even in Victoria it amounts to 37 per cent. They are likely to be areas of low rainfall and/or poor soils and thus to have ecosystems that are the most fragile and with poorest resilience to stress. While it is assumed that these areas are most likely to be used for extensive grazing and have a relatively low productivity for agriculture, the degree of change of biodiversity and soil degradation is unknown but unlikely to be negligible. It may also include areas of private conservation that would contribute to diversity. This residual area needs to be given more attention to determine actual land use and potential for diversification more accurately.

Although there has been considerable changes in land use in some parts of Australia, there is little to indicate the effect of land degradation issues.

Australian Natural Resources Atlas

Natural Resource Topics

This web site is no longer being updated