RESEARCH Survey of Land Degradation in New South Wales, Australia OWEN P. GRAHAM Soil Conservation Service of NSW PO Box 1416 Parramatta, NSW 2124, Australia
ABSTRACT/A survey of land degradation was undertaken in New South Wales, Australia during 1987-1988. The aims of the survey were to assess the location, extent, and severity of ten forms of degradation and to present the data in map and statistical form. Sample points were located on a regular grid. The method was designed so that data could be acquired from aerial photographs, expert local knowledge, and limited field checking. Individual statewide maps were prepared for each form of degradation. Map data were shown in pixel
In 1988, the first systematic survey of land degradation that encompassed a wide range of degradation forms was carried out in New South Wales (NSW). T h e aims of the survey were to assess the location, extent, and severity of land degradation, to publicize and raise community awareness of the problems, to assist management in resource allocation, and to present data in map and statistical form. This article outlines the methods, data collection, and summarizes the results. Land degradation is defined in this survey as a decline in the condition or quality of the land as a consequence of misuse. T h e assumption of misuse is qualified for individual forms of degradation. T h e forms of land degradation included in the survey were sheet and rill erosion by water, gully erosion, mass movement, wind erosion, soil salinity (both dryland and irrigation), decline of soil structure, induced soil acidification, scalding, and infestation of woody shrubs. In addition, data were collected on land use, tree regeneration, and loss of perennial saltbush and bluebush cover in semiarid lands. Data were obtained from remote sensing, specialist local knowledge, and existing data rather than from intensive field observations. T h e method was not designed for precise monitoring of trends in the status of degradation.
KEY WORDS: Erosion; Land degradation; Survey; Methodology; Map; Soil
Environmental Management Vol. 16, No. 2, pp. 205-223
form. Sheet and rill erosion and soil structure decline were confined mostly to lands used for cropping. Gully erosion was commonly found across the state, while mass movement was confined to steeper lands. There were three severe areas of dryland salinity; irrigation salinity was mapped in parts of the southern irrigation lands. Induced soil acidity was severe in some cropping and pasture lands. Absence of tree regrowth was a noticeable feature of lands used for cropping. The survey enabled community awareness of the problems of land degradation to be increased, in addition to assisting regional land managers in resource allocation. The survey also provided the basis for the future location of sites that could be used to monitor the trends in the status of land degradation.
Previous Surveys
There has been concern in Australia about land degradation since the early to mid-1800s. Most of the concern related to soil erosion in the older settled areas, principally widespread gullying in the uplands, sheet and rill erosion on cropping lands, sedimentation in streams and waterbodies, and wind erosion of sand and soil during periods of extended drought. T h e NSW Soil Conservation Act was passed in 1938 in response to these issues and contained powerful legislation for environmental management. Part of its charter was to undertake surveys of soil erosion. Surveys in the 1940s and 1950s were concerned specifically with water and wind erosion and were aimed to promote interest in, and political commitment to, soil conservation and in establishing soil conservation priorities. T h e earliest comprehensive erosion survey in Australia was by the NSW Soil Conservation Service (SCS) in 1941-1943 (Kaleski 1945). Although the survey was updated in 1967 (Stewart 1968), comparisons between the surveys are impossible owing to a lack of consistent criteria for the assessment of type and severity of erosion. This lack of standardisation is apparent across the entire nation. There is no standard method for erosion assessment between the states. Between 1975 and 1977 the Australian Soil Conservation Standing Committee prepared a National Statement on Land Degradation in Australia (Department of National Development 1979). This study used a treat-
9 1992 Springer-Verlag New York Inc.
206
o.P. Graham
ment measures approach in its assessment of land degradation condition. This approach allowed the integration of preventive and reclamation measures (Woods 1984). T h e survey, however, contained inconsistencies and problems of interpretation, particularly between one state and another. This was attributed to differences in approach. Since the 1975-1977 study, it has now been recognized by soil conservation agencies that there are more forms of land degradation affecting agriculture than those included in previous surveys. In particular, soil structure decline and induced soil acidity have been cited as causing significant productivity losses in the Murray-Darling Basin (Murray-Darling Basin Ministerial Council 1987). Some forms of degradation, such as salinity in irrigation areas, have only become apparent in the years since the early surveys were completed. In the mid-1980s, a method for the assessment and mapping of land degradation was developed under a program funded by the Australian National Soil Conservation Program (NSCP) (Graham 1987). A three-tier approach was proposed for mapping and monitoring-each tier being appropriate for specific purposes. The first tier was designed to cover a whole state or country at small scale. This was used as the basis of the 19871988 survey of the State of New South Wales (Graham 1989).
.oo... ~
1
l ~
f qs~J
,~,
9 BROKEN HILL
i
+ "~
,....,,-(,-
#[
C
~.9v
-
l/
,__-.) @
9
r"-a~,
"J
..,"
AR M-IOALE 9
/.o,.~
,.,k
-/''"
~
If"
/
n
I
I iI
•
~" / 7
) ~. /BATHURST.
,c,", .,,., WAfiGA WAGGAe % -."
_~.
[,, I
/g JI'NEWCAST LE
.~ | f
"/(O
I ,~ oCANBERRA
Figure 1. Boundaries of the Eastern, Central, and Western
Divisions in NSW, Australia, with principal rainfall isohyets superimposed.
Location of Survey Area NSW covers an area o f approximately 800,000 sq km
(Figure 1). Summer rainfall is dominant in the northeastern corner of the state and evenly distributed to slightly autumn-winter dominant in the southern parts. Sydney has an average annual rainfall of 1209 mm while Broken Hill, approximately 950 km west of Sydney, has an annual rainfall of only 260 mm (Harriman and Clifford 1987). The Eastern Division of the state is composed of a narrow coastal strip and the mountains of the Great Dividing Range, which extend up to approximately 150 km inland (Figure 1). T h e r e are extensive areas of forested lands in national parks, reserves, and state forests. While these may occur on poorer quality, steeply sloping land, in their undisturbed condition, with dense tree and shrub cover, they generally yield extremely low levels of sediment. After wildfires, however, sedimentation levels can be extremely high (Atkinson 1984). On many of the lower sloping lands extensive clearing of the hafive forests has occurred to establish grasslands of either native or improved pastures for grazing. T h e recurring land degradation problem in these areas is gully erosion with mass movement on the steeper slopes.
The Central Division is composed of gently to moderately sloping land that is used predominantly for the cultivation of wheat. Many of the more obvious degradation problems, such as water and wind erosion and salinity, have been recognized for a long time. T h e more recently identified problems of soil structure decline and induced soil acidification are mainly concentrated in this Division. The Western Division occupies the remaining half of the state. This Division is mainly semiarid to arid grassland (Young 1979) with occasional cropping of small areas in the southwestern corner. The landscape is flat to gently undulating with occasional hills. Most degradation arises from the loss of native vegetation by sheep, cattle, and rabbit grazing. Wind and water erosion can be common, particularly where stock concentrate around watering points. Wind erosion in the cropping areas can be extreme.
Methods The method was designed so that data could be collected without visiting each sample point. Data were ob-
Land Degradation Survey, NSW
Table 1. Land-use classes identified in the survey National park/nature reserve Timber/scrub Logged native forest Hardwood plantation Softwood plantation Native/voluntary pasture Improved pasture Cropping--frequent Cropping--infrequent Orchards/vineyards Vegetables/flowers Urban Industrial Quarry/mining Spray irrigation Flood irrigation Water
rained directly from aerial photographs, personnel with local specialist knowledge, or from existing data, while other data were derived. Data directly recorded included land use, soil group, gully density, depth to saline watertables in irrigation areas, scalding, woody shrub infestation, and occurrence of perennial bush. Derived assessments were made for induced soil acidity, wind erosion hazard, soil structure decline, and some assessments of tree regrowth. Particular land characteristics were determined to help assess the severity of some land degradation types. Surface soil textures were identified to provide data that were necessary for the assessment of sheet and rill erosion, wind erosion, induced soil acidity, and soil structure decline. Soils were grouped into clays, loams, or sands from aerial photographs, published maps, existing knowledge, or field inspection. From aerial photographs, clays could usually be determined by their position in the landscape---such as floodplains and areas of known heavily textured soils. Similarly, sands could be readily determined by their morphology, as in dunes or sand plains. Soils not fitting these general criteria were usually loams. Local specialist knowledge played an important role in verifying assessments from aerial photographs. Land use also was recorded for each sample point. These data were used to derive the sheet and rill erosion, wind erosion, induced soil acidity, and soil structure decline assessments. Techniques used for the identification of land use by interpretation of aerial photographs were based on Emery and others (1986). Seventeen land uses classes were identified in this survey (Table 1). Most of these land uses could be readily determined from aerial photographs with the exception of the two cropping classes: frequent or in-
207
frequent. In NSW, this terminology is important in describing land-use patterns and making degradation predictions. Frequent cropping refers to the practice o f consecutive annual cropping for 2-3 yr or more followed by 1-2 yr under pasture. Infrequent cropping occurs when an area is cropped only once every 3-4 yr. Degradation rates are higher under repeated cultivations because of the greater amounts of vehicular traffic causing compaction at depth, increased tillage causing disaggregation and loss of organic matter, and the repeated absence of protective vegetative cover during fallow, all of which increase the water and wind erosion hazard. Cropping land was classified as cropping--frequent unless it was known that the cropping problems in a particular paddock were part of a regional trend. Such a situation occurs in marginal cropping areas where cropping may only be practiced 1 yr in every 3-4 yr because of unreliable rainfall.
Assessment Criteria Status assessments were undertaken for the current condition of gully erosion, mass movement, dryland and irrigation salinity, induced soil acidity, soil structure decline, and woody shrub infestation. T h e data recorded the presence or morphology of the attribute; in effect, it was an instantaneous view of the form and severity of degradation at the time of the survey. For gully erosion, mass movement, salinity, and scalding, the assessment was based on measurement of the form or observation of its occurrence from aerial photographs or existing data. For induced soil acidity and soil structure decline, the status assessment was inferred and based on soil test results, land management, recorded soils data, and other environmental factors. Assessments for sheet and rill erosion and wind erosion were made on the basis of hazard, rather than status. Since the effects of sheet and rill erosion and wind erosion were often masked by cultivation and crop or pasture establishment, status assessments for these attributes could not be reliably determined from aerial photographs. Status assessments could only have been established for these forms of degradation with field checking at each point, a task outside the scope of this project. Degradation hazard was used as a measure of the level of degradation that was predicted to occur over a time period with a given set of field parameters. An example is the universal soil loss equation (USLE) (Wischmeier and Snfith 1978), which predicts the rate of soil loss from a point given certain variables including land management and cropping practice. It does not consider how degraded the site is now or what has hap-
208
O.P. Graham
pened in the past; rather, the amount of soil that can be lost with a nominated land use. Hazard assessments are particularly useful when land management recommendations to minimize soil loss are undertaken.
Sampling Strategy The data were collected from sample points at regular intersections of the Australian Map Grid (AMG). Sample point refers to the area within a circle of up to 100 ha surrounding the map grid point when translated to aerial photographs or the ground. Sampling was at twice the minimum acceptable rate in the Eastern and Central Divisions because of the availability of resources. These sample points were located at 5-kin spacings on the easting AMG and 10-kin spacings on the northing AMG. T h e grid was expanded to 10 x 10 km in the Western Division. There were approximately 13,600 sample points across the state,
Establishment of Sample Points The sample base was the 1:100,000 topographic sheet, This was the largest uniform map scale. The sample points were identified and marked on 1:100,000 topographic map sheets. The points then were transferred to the most recent aerial photographs, which varied in scale between 1:40,000 and 1:80,000. It was generally possible to locate the points with an accuracy better than 100 m. Sheet and rill erosion and wind erosion were assessed at the exact sample point, while for other forms of land degradation a larger area around the sample point was used. This variation in the area was necessary because of the areal distribution of the particular forms of land degradation at each sample point. The assessments for gully erosion required a measure of gully density, and it was most practical to measure this within a 100-ha circle centered on the sample point. Since the assessments for mass movement, salinity, scalding, and tree regrowth identified an areal feature, these were undertaken within the same 100-ha cirde. A sample point size of 4 ha was used for the assessment of induced soil acidity and soil structure decline. This size was the smallest nominal area that could reasonably be used for a representative sample of land for assessment. A 4-ha quadrat also was used for assessing woody shrub infestation and the occurrence of perennial bush. In these latter cases, the assessment was by road traverse rather than by aerial photograph interpretation, and 4 ha was considered to be a reasonable area for visual estimation from the roadside.
Table 2. Categories of soil loss rates Class 1 (negligible) 2 (minor) 3 (moderate) 4 (severe) 5 (very severe)
Rate (t/ha/yr) <1 1-<5 5-< 10 10-<25 >25
Data Recording Data were obtained from each sample point using the latest aerial photographs, specialist field knowledge, and existing data. In all cases assessments were made either by, or in liaison with, specialists working in the field. Time constraints prevented field visits to each sample point, although a minimum of 10% of the sample points were field checked either on site or visually from some distance. As this was not a monitoring survey, no detailed soil descriptions or soil samples were taken. While data were collected by several different field staff, one person acted as project coordinator to ensure that uniform standards were maintained. Data were recorded by hand on a standard data card. One card was used for all the sample points on each 1:100,000 topographic sheet. The card was printed with fixed format fields to facilitate computer data entry at a later stage. Data cards were used in preference to direct computer entry to facilitate note taking, general editing, and corrections after field checking. This systematic sampling procedure facilitated entry of the data into a geographic information system (GIS) with corresponding computerized handling and display of the data.
Assessment Techniques Sheet and Rill Erosion Sheet and rill erosion were assessed on tile basis of hazard using a modified form of the universal soil loss equation (USLE) (Wischmeier and Smith 1978). While acknowledging the shortcomings in the USLE (Edwards and Charman 1980), it was considered a satisfactory method for deriving relative erosion hazard figures. T h e U S L E i s A = R x K x L x S x C x P , where A = soil loss in t/ha/yr, R = rainfall erosivity, K = soil erodibility, L = slope length factor, S --- slope gradient factor, C = cover and management practice, and P = support practice. The estimated rate of soil loss was calculated for each sample point and categorized into one of five groups (Table 2). Owing to the lack of calibrated data for large portions of the state, it was necessary to use average values
Land Degradation Survey, NSW
for several factors in the USLE where measured values were not available. Interpretation of aerial photographs was used to identify soil group, slope length, slope gradient, land use, and management practices at the date of photography. Local knowledge and field checking were used to upgrade land-use information. Techniques for estimating slope gradient, slope length, and land use are discussed in Emery and others (1986). A standard value for rainfall erosivity (R) was used for each 1:100,000 topographic sheet from data prepared by Rosewell and Edwards (1988). T h e K value was obtained either from laboratory data and applied to the nomograph of Foster and others (1981) or from average default values for the three broad groupings of soils, identified on the basis of soil texture. T h e L factor was measured from aerial photographs and categorized into one of five groups. T h e slope length was the slope distance within the paddock from the point where overland flow originated to the point where soil material could be deposited. Where soil conservation banks had been installed to control water erosion problems, the slope length was taken as the disLance between the banks. T h e S factor was measured at each sample point. The gradient was determined from aerial photographs and field checked where necessary. Use was made of slope range categories rather than absolute values in order to simplify estimation. Slope gradients were measured from large-scale (1:25,000 to 1:50,000) topographic maps or from aerial photographs using stereo slope dip comparators (Emery and others 1986). T h e C factor was based on the land use at each sample point (Table 1), and values were determined using the SOILOSS program of Rosewell and Edwards (1988). I f land management data were not available for the sample points being assessed, average default values were specified for each land use. T h e average values were allowed to be increased or decreased if it was known that land management practices at that point were better or worse than for the district as a whole. T h e P factor determined in accordance with Wischmeier and Smith (1978).
Gully Erosion Gully erosion was assessed from aerial photographs. Only active gullies were included in the assessment because stable, fully grassed gullies did not represent an erosion problem during the time of the survey. Streambank erosion was included in this degradation category because the processes that produce it are similar to those occurring along gully sidewalls. Gully density was recorded within a circular sample
209
area of 100 ha, with the center of the circle being the sample point on the AMG intersection. A transparent plastic overlay with 100-ha circles conforming to the various scales of aerial photographs was constructed. T h e appropriate scaled circle was overlaid on the aerial photograph, and tile total length of active gully channel was measured within the circle. T h e unit of measurement was the total length of active gully channel per 100 ha. Gully density was recorded in meters per 100 ha. Active gullies could be identified in aerial photographs by the tone of the sidewalls and their general shape. A whitish tone near the headwall indicated bare active sidewalls, whereas a duller tone generally indicated vegetated sidewalls. Sidewalls and headwalls that were near vertical indicated recent active erosion. Rounded forms of sidewalls near the gully mouth indicated stability. T h e length of gully channel was recorded on the data card, and this was divided into one of seven categories for severity ratings.
Mass Movement Mass movement included all categories of slope failure such as slumps, creep, earthflow, and debris slides (Hadley and others 1985). Only those forms of mass movement that were considered to be induced by land clearing and land-use practices were included. Consequently, rockslide avalanches and individual rock falls were excluded, whereas soil creep (including terracetting), earthflow, debris slide, and slumps were included. Mass movement was determined from aerial photographs. T h e same 100-ha circle as used for the gully erosion classification was used for mass movement. Mass movement of any form within the circle was recorded simply as present or not present.
Wind Erosion Wind erosion hazard is a function of land use, soil group, and ground cover. It was estimated on a fourclass scale that ranked the relative susceptibility of the surface soil to erosion when the soil was in its most vulnerable condition during any 12-month period. T h e relative differences in wind erosion hazard between the various soil groups and land uses were established from wind-tunnel studies (Davidson 1989). T h e hazard rating was modified according to cultivation conditions or presence of windbreaks. T h e assessment was related exclusively to surface soil type and vegetation cover. It was assumed that climatic conditions were the same over the whole country. Aspect and prevailing winds were ignored.
210
o.P. Graham
Dryland Seepage Salinity This form of salinity is associated with nonirrigated land. It is generally accentuated or caused by the clearing of hillslopes, which allows more water to infiltrate to the water table. When the groundwaters rise to the surface in low points in the landscape, capillary action draws water to the surface. Dissolved salts in the water are concentrated by evaporation. Salinity was assessed within the same 100-ha circle as used for some other forms of land degradation. Only presence or absence was recorded. T h e three recorded classes were based on the occurrence and form of the salt outbreaks. No attempt was made to rate its severity. Naturally occurring salt lakes were not included in this assessment.
Irrigation Salinity This type of salting is associated with irrigated land. T h e assessment was based on existing data from boreholes and classified the depth below the surface to the saline water table. T h r e e categories were recorded. T h e most severe rating was applied if the saline water table was closer to the surface than 2 m depth. This was the maximum depth of the water table below the surface at which evaporation causes significant salt accumulation on the surface in irrigation areas of southwestern NSW (Gutteridge, Haskins and Davey 1985).
Scalding A scald is a bare area produced by the removal of the surface soil by wind and/or water. T h e result is exposure of the more clayey subsoil which is, or becomes, relatively impermeable to water. Most scalds are formed by erosion of the duplex soils in the arid or semiarid regions of NSW. They vary in size from a few square meters to hundreds of hectares and are very difficult to revegetate due to the lack of topsoil, low permeability, and often saline surface (Houghton and Charman 1986). Scalding was assessed from aerial photographs using a 100-ha circle centered on the sample point. T h r e e classes of severity were recorded based on the proportion of scalded land within the circle.
Induced Soil Acidity This was an assessment of the degree of induced acidification of the land. Absolute levels of acidity were not recorded. T h e assessments related to the status of the soil in relation to its degraded condition when it could be regarded as problem acid. This assessment was productivity-based, since problem acid conditions restrict the growth of some of the major pasture and cereal species.
T h e first class was of soils with no induced acidity problem; the second class was for soils that were not problem acid, although they were likely to proceed to a problem status with current management techniques; and the third class was for soils which were classified as problem acid. The assessment was a function of local specialist knowledge of soil p H in farmed paddocks compared with undisturbed areas, soil buffer capacity (soils with high buffering capacities being more resistant to p H decline than those with lower buffering capacity), and toxic element levels, particularly exchangeable aluminum and manganese. While field samples were not taken, the assessment was made with knowledge of regional trends for soil types based upon soil tests and the current land use. Assessments were made in consultation with research personnel and with local specialists from agricultural organizations and fertilizer companies who provided information on the soil conditions and specific practices likely to cause a change in soil acidity conditions. For example, areas with timber or native pastures were rated in the lowest class by definition (no inputs such as fertilizers that could have affected the acidity status). In winter rainfall areas, soils with improved pastures were likely to have developed acid problems owing to applications of fertilizers and the leaching of nitrate nitrogen from legume-dominated pastures over many years.
Soil Structure Decline Soil structure decline is a change in the inherent structure of the soil as the result of a series of land-use practices. T h e soil develops either a more massive condition, which is demonstrated by a plough pan at the base of the cultivation layer; a denser overall structure as a result of stock or vehicular pressure; or a partial to complete shattering or disaggregation of the structure to a finer or single-grained fabric. Soil structure decline was determined empirically from characteristics of the sample site. These were soil type, land use, land-use rotations, cultivation history and intensity (if applicable), evidence of hardpans or surface crusts, research studies, and local knowledge. Where local knowledge of conditions was poor, a twolevel matrix relating the land use to soil group was used to place soils into one of three classes for soil structure decline (Table 3). For cropping lands, this used the general assumption that the longer an area had been under crop, and hence the greater the n u m b e r of cultivation cycles, the greater would be the degree of soil structure decline. T h e structure decline rating could be altered if it was known that management practices had resulted in soil
Land Degradation Survey, NSW
Table 3. Criteria for assessing soil structure decline a Land use National park/nature reserve Timber/scrub Logged native forest Hardwood plantation Softwood plantation Native/voluntary pasture Improved pasture Cropping--frequent Cropping--infrequent Orchards/vineyards Vegetables/flowers Urban Industrial Quarry/mining Spray irrigation Flood irrigation
Sands
Loams
Clays
S S S S S S T T T T T T T T T n/a
S S S S S S T U T T U U U U U U
S S S S S S S U S S U U U U U U
aS = nil to minor structure decline. T = moderate structure decline, U = severe structure decline, n/a = not applicable.
structure properties that differed from the rules listed above.
Woody Shrub Infestation W o o d y shrubs are inedible shrub species that have proliferated or are rapidly spreading in parts o f the semiarid and arid zones o f NSW. T h e shrubs are native species that proliferate to greater densities than would occur naturally as a result o f changed environmental conditions such as the incidence o f fire. T h e y restrict the growth o f understorey vegetation and increase the hazard o f sheet, rill, and gully erosion. Six species o f shrubs were listed for inclusion in this survey. Owing to the inability to detect infestations o f woody shrubs accurately f r o m small-scale aerial p h o t o g r a p h y or satellite imagery, g r o u n d sampling techniques using road traverses were adopted. At regular intervals and consistently on one side o f the road, a 4-ha quadrat was sampled away f r o m the immediate influence o f the road and associated drainage and services. T h e assessment at each road traverse point was based on the reduction in grazing capacity caused by the shrub presence. W o o d y shrubs were g r o u p e d into one o f four severity classes and between 25 and 50 points were recorded on each 1:100,000 sheet. T h e inherent bias associated with a road traverse was reduced as far as possible by selecting a route that covered a representative proportion o f the various land systems within the particular m a p sheet [land systems are areas or groups of areas t h r o u g h o u t which there is a recurring pattern o f topography, soils, and vegetation (Houghton and Charman 1986)].
211
Since data points for woody shrubs were not located at AMG intersections, a weighted average value for shrub infestation was made for each 1:100,000 topographic sheet. That value was applied to all A M G sampie points on that sheet. If there was a road traverse data point within 5 km of an A M G sample point, then that severity rating was transferred to the sample point. This assessment was restricted to the semiarid and arid areas of the state. In the wetter areas, problems o f shrub definition precluded any reliable assessment o f what constituted a problem.
Occurrence of Perennial Bush T h e death o f edible peremfial saltbush (Atriplex vesicarla) and bluebush (Maireana spp.) is a significant f o r m of land degradation in the semiarid and arid parts o f NSW. Good perennial bush cover provides protection o f the soil surface from wind erosion, creates more favorable g r o u n d surface conditions for pasture plants, and provides an important source of stock fodder during droughts when annual and other perennial forage species have died out. Occurrence o f perennial bush was not regarded as a form o f land degradation. It was implied that the denser the occurrence o f perennial bush, the better the quality o f the land. T h e presence o f perennial bush was assessed at the same road traverse points as for woody shrubs since small-scale aerial p h o t o g r a p h y was not adequate for showing perennial bush. T h e occurrence rating was based on the increase in grazing capacity offered by its presence (Young 1985). Four classes were recorded.
Tree Regrowth T r e e regrowth is defined as the emergence o f saplings underneath or in close proximity to a mature tree of the same species. It is an assessment o f the regeneration o f existing species and ecotypes. Tree decline is a feature of land degradation in NSW, although soils may be unaffected. In fact, the lack o f tree regrowth in areas that were heavily timbered prior to E u r o p e a n settlement is an indication o f the intensity of land use. L a n d intensively used was prone to higher levels o f land degradation. Tree regrowth was assessed for each sample point from aerial photographs within a 100-ha circle m a r k e d on the aerial photographs. A single instance o f regrowth was sufficient to code the sample area as having regrowth. Regrowth was inferred from the land use when access to check the sample point was not possible. For example, timbered areas were assumed to have regrowth and single isolated trees in intensively grazed paddocks were assumed to have no regrowth because o f
212
o.P. Graham
Figure 2. Land use. grazing by animals. Naturally treeless areas were mapped as having no regrowth. This needs to be considered when interpreting the survey data.
Results The results were compiled as a series of maps, one for each form of degradation (Figures 2-14). T h e maps were composed o f pixels and produced by a color inkjet printer. One pixel contained the data from each sample point. Each pixel covered the equivalent scaled down area of 50 sq km for the Eastern and Central Divisions, or 100 sq km for the Western Division. Since the maps were composed of pixels, the final map scale was dependent on printer legibility and paper size. Statistics on the areas and relative proportions o f each form o f land degradation were derived from the data. T h e land use map (Figure 2) shows the distinct belt of cropping land in the Central Division, the predominance of national parks and timbered land in the Eastern Division, and the extensive pasture lands in the Western Division.
Sheet and rill erosion were largely confined to lands used for cropping in the Central Division and some isolated pastures on steeper slopes and forested areas that recently had been burned (Figure 3). T h e most severe sheet and rill erosion hazards were on the steeper sloping lands used for cropping. T h e generally good seasonal conditions prevailing in the semiarid grasslands of the Western Division at the time of the survey meant that its erosion hazard was recorded as negligible to minor. Gully erosion was most extensive on the tablelands o f NSW, the sloping country west of the Great Dividing Range, and parts of the Western Division with hilly terrain (Figure 4). The high4ntensity summer rainstorms in the northeast of the state probably contributed to the generally high density of gullies in this area. T h e mixed grazing and cropping lands in the southeastern part of the Central Division also had a high density o f gullies. Although rainfall was generally of lower intensity in the southern parts of the state than in the north and more evenly distributed (Edwards 1979), the higher incidence of gully erosion is probably related to the more erodible soils and a longer history o f intensive agriculture.
Figure 3. Sheet and rill erosion.
Figure 4. Gully erosion.
214
o.P. Graham
Most incidences of mass movement occurred in the northeastern corner and central coastal areas of the state, generally in the headwaters of catchment areas where the land is steep (Figure 5). Many of these slope failures appear to have occurred at varying intervals of time after the original timber cover had been cleared. General observations suggested that this could be up to 70 yr after clearing, depending upon the rates of decay of the root material. Slope failures have been noted commonly during or immediately after long periods of heavy rain when the soil has been wetter than normal. T h e most severe wind erosion problems occurred in the central and southern parts of the Central Division, and the southern and northwestern parts of the Western Division (Figure 6). In the Central Division, the northern cropping lands generally had a lower wind erosion hazard than the central and southern areas, probably because of the higher clay contents of the soils. T h e cropping lands in the south of the Western Division generally had a very severe wind erosion hazard, probably because of the sandy textured soils and long fallowing practices. T h r e e distinct areas of dryland seepage salinity were evident in the state (Figure 7). T h e most severe area occurred in the southwest of the Eastern Division. This area has a long history of European settlement with extensive tree clearing in valleys and surrounding hillsides. Irrigation-induced salinity occurred mainly in the southern irrigation areas of the state in large established irrigation schemes (Figure 8). In parts of some irrigation areas there had been an estimated 50% increase in the area with high saline water tables over a 15-yr period (Gutteridge, Haskins and Davey 1985). T h e saline water table had risen above the surface in some lowlying depressions. T h e most extensive areas of scalding occurred in the vicinity of the floodplains of major river systems or prior streams (Figure 9). T h e r e were also widespread occurrences throughout the sandplains and outwash areas of ephemeral streams in the west of the Western Division. Induced soil acidity was most severe in the cropping and improved pasture lands in the southeastern section of the Central Division (Figure 10). T h e problems were associated with a long history of superphosphate applications on legume-dominated pastures experiencing winter rainfall. Isolated pasture lands along the Great Dividing Range with lighter-textured soils also had a potential soil acidity problem while current land use was maintained owing to low soil buffering capacity and the acidifying effects of agriculture. T h e greatest soil structure decline problems were on
lands used regularly for cropping (Figure 11). T h e central and southern areas in the Central Division had the most serious and extensive soil structure decline problems, and these were associated with the generally loamy soils found throughout this area. Soil structure decline problems were not as severe in the northern parts of the Central Division because, it is thought, the soils have higher clay contents. Woody shrub infestation is shown in Figure 12. Almost 70% of the Western Division was affected by shrub infestation. Shrubs generally were found on the lightertextured soils rather than on heavier-textured clay soils. Severe infestations occurred on the sandplains and dunefields in the northwest of the state. In some instances, shrub infestation had caused complete abandonment of the land--the cost of clearing the shrubs being more than the value of the land, Moderate infestations occurred over large areas of the central parts of the Western Division. Perennial bush occurrence is shown in Figure 13. T h e r e was a moderate to dense cover of perennial bush in several parts of the Western Division. Perennial bush was generally absent from the lighter-textured sandy soil areas in the north and northwest of the Western Division. T h e general absence of tree regrowth was a noticeable feature of areas under cropping, particularly on the central and southwestern parts of the Central Division (Figure 14). T h e r e were extensive areas of grassland in the Western Division with large numbers of mature trees but no regrowth, probably resulting from grazing animals.
Statistics Statistical results are presented as percentage and total areas affected. T h e sampling procedure allowed areas to be determined by applying these percentages to the total area of NSW to obtain the absolute areas of each severity class of each type of land degradation. T h e Australian Capital Territory (within southeastern NSW) was not included in these figures. Not all features are given an area basis because of sampling procedure: those recorded as presence or absence within 100-ha circles are shown as not applicable (n/a) because of statistical unreliability. T h e results are given in Table 4.
Discussion Results These data have, for the first time, provided an overview of the extent and severity of the various forms of
Figure 5. Mass m o v e m e n t .
Figure 6. W i n d erosion.
Figure 7. Dryland seepage salinity.
Figure 8. Irrigation-induced salinity.
Figure 9. Scalding.
Figure 10. Induced soil acidity.
Figure 11. Soil structure decline,
Figure 12. Woody shrub infestation.
Figure 13. Perennial bush occurrence.
Figure 14. Tree regrowth.
220
Table 4.
o.P. Graham
Percentages and areas affected for land degradation attributes Eastern and Central Divisions
Attribute Sheet and rill erosion Negligible Minor Moderate Severe Very severe Gully erosion No appreciable Minor Minor to moderate Moderate Severe Very severe Extreme Mass movement Not present Present Wind erosion Nil to minor Moderate Severe Very severe Dryland salinity Nil to minor Moderate Severe Irrigation salinity Nil to minor Moderate Severe Scalding Nil to minor Moderate Severe Induced soil acidity None Potential to become acid Severe Soil structure decline Nil to minor Moderate Severe Woody shrub infestation Nil Minor Moderate Severe Tree regrowth Regrowth No regrowth Occurrence of perennial bush Nil Scattered Frequent Dense
Western Division
Whole state
%
ha x 1000
%
ha x 1000
%
ha x 1000
79.8 15.6 3.1 1.0 0.5
39,457 7,738 1,517 482 269
99.7 0.2 0.1 0.0 0.0
30,608 50 20 0 0
87.5 9.7 1.9 0.6 0.3
70,065 7,788 1,537 482 269
68.0 3.6 12.7 7.1 7.3 1.2 0.1
33,638 1,780 6,281 3,531 3,601 572 60
90.9 0.2 3.4 4.0 1.5 <0.1 0.0
27,874 50 1,054 1,243 447 10 0
76.8 2.3 9.2 6.0 5.1 0.7 0.1
61,512 1,830 7,335 4,774 4,048 582 6O
95.3 4.7
n/a n/a
I00.0 0.0
n/a n/a
97.1 2.9
n/a n/a
74.7 10.9 12.1 2.3
36,956 5,396 5,967 1,144
75.4 18.7 5.5 0.4
23,140 5,728 1,681 129
75.0 13.9 9.5 1.6
60,096 11,124 7,648 1,273
97.9 0.9 1.2
n/a n/a n/a
99.9 <0.1 0.0
n/a nla n/a
98.7 0.6 0.7
n/a n/a n/a
98.7 0.9 0.4
rda rda n/a
100.0 0.0 0.0
n/a n/a n/a
99.1 0.6 0.3
n/a n/a n/a
94.9 4.4 0.7
46,932 2,198 333
81.4 17.4 1.2
24,970 5,340 368
89.7 9.4 0.9
71,902 7,538 701
82.9 11.3 5.8
41,043 5,570 2,850
100.0 0.0 0.0
30,678 0 0
89.4 7.0 3.6
71,721 5,570 2,850
71.1 10.4 18.5
35,175 5,162 9,126
98.6 1.0 0.4
30,271 298 109
81.7 6.8 11.5
65,446 5,460 9,235
95.4 2.5 1.8 0.3
47,186 1,248 900 129
33.2 29.2 27.3 10.3
10,183 8,960 8,373 3,162
71.6 12.7 11.6 4.1
57,369 10,208 9,273 3,291
62.9 37.1
31,097 18,366
31.3 68.7
9,606 21,072
50.8 49.2
40,703 39,438
98.2 0.6 1.0 0.2
48,588 288 473 114
72.4 17.0 8.7 1.9
22,225 5,201 2,675 577
88.4 6.8 3.9 0.9
70,813 5,489 3,148 691
Land Degradation Survey, NSW
221
Table 4. Continued Eastern and Central Divisions Attribute Land use National park/nat, reserve Timber/scrub Logged native forest Hardwood plantation Softwood plantation Native pasture Improved pasture Cropping--frequent Cropping--infrequent Orchards/vineyards Vegetables/flowers Irrigation--spray Irrigation--flood Urban Industrial Quarry/mining Water Swamp---wet Swamp---dry
Western Division
Whole state
%
ha x 1000
%
ha x 1000
%
ha x 1000
6.2 19.2 2.1 0.1 0.5 35.2 9.5 19.0 3.8 0.2 0.1 0.2 2.4 0.4 <0.1 0.2 0.7 0.1 0. I
3,073 9,508 1,093 40 229 17,406 4,685 9,399 1,870 119 30 85 1,199 229 15 94 338 65 40
1.9 8.9 <0.1 0.0 0.0 87.6 0.0 0.4 0.7 0.0 0.0 0.0 0.1 0.0 0.0 0.1 0.3 0.0 0.0
587 2,735 l0 0 0 26,879 0 109 219 0 0 0 20 0 0 20 99 0 0
4.6 15.3 1.3 <0.1 0.3 55.3 5.8 11.9 2.6 0.2 <0.1 0.1 1.5 0.3 <0.1 0.1 0.6 0.1 <0.1
3,660 12,243 1,049 40 229 44,285 4,685 9,508 2,089 119 30 85 1,219 229 15 114 437 65 40
land degradation across the state. T h e data offer a more objective rationale for resource allocation. This is a significant i m p r o v e m e n t over the old situation where such data were lacking. Extensive publicity associated with publication o f results in the f o r m o f a popular booklet b r o u g h t to the attention o f the urban and rural community the existing and developing problems o f our rural lands. Similarly, the presentation o f separate maps for each f o r m o f land degradation has enabled all of the forms o f land degradation to be treated with equal emphasis, a situation particularly useful for raising public awareness o f the various land degradation issues. O f all the forms o f land degradation, the MurrayDarling Basin Ministerial Council (1987) has identified induced soil acidification a n d soil structure decline as causing the greatest productuvity losses in the basin. In the past, these two forms o f land degradation received little attention when c o m p a r e d with water and wind erosion, because they are not immediately obvious. Changes in land quality occur gradually and only recently have become apparent. This survey has demonstrated how extensive and severe the two forms o f land degradation are. T h e need for a greater commitment by land-management agencies, governments, and individuals to overcome these and the other forms o f land degradation is evident f r o m this survey.
Since each f o r m o f land degradation was considered separately, there was no consistency between the n u m ber o f severity classes assessed for each attribute. T h e range o f classes varied f r o m two (such as mass movement or tree regrowth) to seven (gully erosion). Likewise, several different forms o f land degradation were often recorded at the one sample point. Hence individual percentages or areas o f any one form o f degradation should not be added to other forms o f degradation to attempt to derive a total picture for the state. T h e data and maps show the general locations where problems exist. T h e y are not suitable, however, for monitoring trends in the status o f land degradation, except at an extremely broad scale. For precise monitoring purposes, relocatable field sites need to be established in areas identified from the survey maps as having significant problems. Recognized field tests should then be undertaken to provide data for trend analysis. T h e methodology used in ~his survey is not suitable for such a task because detailed field tests were not undertaken.
Methods Sample points were stratified in such a way that the points in the east o f the state were at twice the density o f those in the west. Land resource data that would have allowed more sophisticated stratification were lacking. T h e degradation m a p p i n g might have been based on
222
o.P. Graham
polygonal units (Eilers and others 1989), but such a technique would have been impractical for this survey because of the lack of statewide large-scale soil data from which polygons could be constructed. Similarly, a mathematically more correct grid sampling approach would have been to use a hexagonal pattern to select grid points, in which there would have been an even distribution of points that were equidistant from each other, instead of a rectangular pattern, where neighboring points are at varying distances from each other. Given the need to collect data with the least delay and the need to issue standard instructions to the data collectors, it was most practical to use a rectangular grid pattern where regular intervals could be specified. T h e r e were various limitations in the method for the assessment of particular forms of land degradation: One of the more obvious results of the survey is the concentration of many forms of land degradation in the principal cropping areas of the Central Division. T h e data should be viewed with caution, however. T h e circularity of reasoning when several forms of land degradation are compared must be acknowledged. For example, the default value for soil structure decline on loamy soil cropping lands was a severe rating. Despite this, the relationship between soil structure decline and cropping is real and significant. T h e wind erosion severity criteria assumed that winds were of equal intensity across the country. Similarly, aspect and wind direction were not considered. T h e r e is scope for the method to be improved by incorporation of these factors. It is inevitable that local variations in soil structure due to management techniques may have been overlooked in this method. This has to be accepted in a survey where it is not feasible to field-examine each sample point. Since the tree regrowth assessment was based simply on presence or absence data, an objective statement, naturally treeless areas such as the Riverine Plain in the southern part of the state were m a p p e d as having no regrowth. This limitation has to be considered when reading the maps and statistics. T h e assessments for many forms of land degradation were based on relationships with other factors rather than direct observations, for example, soil structure decline. Because of this, these data could not be used for monitoring trends in the status of degradation. It is only with a methodology developed around recognized field and statistical methods that reliable data can be obtained for such monitoring. Because the various attributes in the survey were not assessed in the same way, any comparisons between them should be made with caution. Any comparison
among the various forms of degradation could only be made if such comparisons were on the basis of their costs to landholders or the general community. Such economic analyses were beyond the scope of this survey.
Conclusion This survey of land degradation has provided the basis for the future establishment of monitoring sites to assess trends in the status of land degradation. Placement of monitoring sites would logically be in areas that have been identified as having problems or sites on the margins of problem areas where trends are likely to become apparent. Furthermore, the survey has provided basic data that will allow regional land managers to make better decisions about tackling specific landdegradation problems.
Acknowledgments Keith Emery, Vicki Pattemore, and Oeoff Cunningham provided valuable editorial assistance. Individuals from many Australian state, federal and overseas agencies assisted with the preparation of the methodology. Data for the survey were collected by 16 individuals from the Soil Conservation Service of NSW. Personnel from NSW Agriculture and Fisheries assisted with interpretation of field data for the induced soil acidity and soil structure decline sections. Watertable height data for the irrigation salinity section were supplied by the NSW Department of Water Resources.
Literature Cited Atkinson, G. 1984. Erosion damage following bushfires.Jourhal of Soil Conservation New South Wales 40:4--9. Davidson, S. 1989. Wind erosion: Dust, dunes, and tell-tale sediments. Rural Research 144:13-23. Department of National Development. 1979. Commonwealth-states collaborative soil conservation study 1975-77. Report No. 1. Australian Government Publishing Service, Canberra, 193 pp. Edwards, K. 1979. Rainfall in New South Wales--with special reference to soil conservation. Technical Handbook 3. Soil Conservation Service of NSW, 116 pp. Edwards, K., and P. E. V. Charman. 1980. The future of the soil loss prediction in Australia. Journal of Soil Conservation New South Wales 36:211-218. Eilers, R. G., M. N. Langman, and D. R. Coote. 1989. Water erosion risk. Canada--Manitoba soil inventory. Land Resource Research Centre, Research Branch, Contribution 87-12, Agriculture Canada, 11 pp. Emery, K. A., R.J. Morse, and P. D. Houghton. 1986. Aerial photograph interpretation for land resource mapping.
Land Degradation Survey, NSW
223
Technical Handbook 8. Soil Conservation Service of NSW, 68 pp.
Kaleski, L. G. 1945. The erosion survey of NSW.Jour~al of Soil Conservation New South Wales 1: 12-20.
Foster, G. R., D. K. McCool, K. G. Renard, and W. C. Moldenhauer. 1981. Conversion of the universal soil loss equation to SI metric units.Journal of Soil and Water Conservation 36:355-359.
Murray-Darling Basin Ministerial Council. 1987. Murray-Darling Basin environmental resources study. Murray-Darling Basin Ministerial Council, Canberra, 426 pp.
Graham, O. P. 1987. Land Degradation Assessment: Methodology. Pages 192-204 in Proceedings of the annual conference, 7-8 July 1987. Soil Conservation Service of NSW. Graham, O. P. 1989. Land degradation survey of NSW 198788: Methodology. Technical Report 7. Soil Conservation Service of NSW, 45 pp. Gutteridge, Haskins, and Davey. 1985. Waterlogging and land salinisation in irrigation areas of NSW. Vol. 1, The problems. Water Resources Commission, New South Wales, 150 PP. Hadley, R. F., R. Lal, C. A. Onstad, D. E. Walling, and A. Yair. 1985. Recent developments in erosion and sediment yield studies. Technical Documents in Hydrology. UNESCO, Paris, 127 pp. Harriman, R.J., and E. S. Clifford (eds.). 1987. Atlas of New South Wales. Central Mapping Authority, Department of Lands, New South Wales, 135 pp. Houghton, P.D., and P. E. V. Charman. 1986. Glossary of terms used in soil conservation. Soil Conservation Service of NSW, 147 pp.
Rosewell, C.J., and K. Edwards. 1988. SOILOSS--a program to assist in the selection of management practices to reduce erosion. Technical Handbook 11. Soil Conservation Service of NSW, 71 pp. Stewart, J. 1968. Erosion survey of N.S.W. eastern and central divisions re-assessment--1967. Journal of Soil Conservation New South Wales 24:139--154. Wischmeier, W. H., and D. D. Smith. 1978. Predicting rainfall erosion losses--a guide to conservation planning. Agricultural Handbook 537. US Department of Agriculture, Washington, DC, 58 pp. Woods, L. E. 1984. Land degradation in Australia. Department of Home Affairs and Environment. Australian Government Publishing Service, Canberra, 105 pp. Young, M. D. 1979. Influencingland use in pastoral Australia. Journal of Arid Environments 2:279-288. Young, M. D. 1985. The influence of farm size on vegetation condition in an arid area. Journal of Environmental Management 21:191-203.