The Kyeamba Landcare Group represents landowners in an area of some 70,000 ha which occupies most of the Kyeamba Creek catchment and a small area of the Tarcutta Creek catchment in the north-east portion. It extends from the Murrumbidgee River (of which Kyeamba Creek is a tributary) in the north to Kyeamba Gap on the Hume Highway in the south (Figures 1 and 2).

Figure 1. Kyeamba Landcare Locality Plan

The area was cleared for agriculture in the latter part of the last century, and has been the site of continuous, but changing, farming practices since then. Wheat was an important crop during the 1960s and 1970s but the emphasis is now on cattle and sheep grazing, with some lucerne and oat cropping. An important feature is the recent establishment of a feedlot operation.

During the past few years, small localised patches of waterlogged land have developed and in many cases they have progressed to a state of land salinisation. The increased rate of new occurrences, and rapidity of expansion of existing occurrences, prompted action by a group of farmers which led to the formation of the Kyeamba Landcare Group. The Group now includes some 150 landholders, who comprise the majority of farmers in the Landcare area.

The Group has, together with several State Government agencies, worked to provide a framework of action within which farmers can take remedial measures to halt or reverse the spread of land degradation. The aim of the project is to prepare and publish, for use by all concerned with land management in the area, a management strategy document which will provide a basis for positive action on individual farms. Various programs in the area have been supported by funds provided by the State Government under its Salt Action Program, and by the Commonwealth Government through the Murray-Darling Basin Commission's Natural Resource Management Strategy Program. Contributions have been made by NSW Departments of Water Resources, and Conservation and Land Management.

A considerable body of knowledge on the causes and development of dryland salinity has been built up during the past decade, as various States come to grips with the problems it causes. It is now recognised that groundwater processes are a central part of the wider problem, and that the impact of agricultural practices can result in an observable response in groundwater movement patterns. Accordingly, the Department of Water Resources was asked to contribute an analysis of groundwater conditions as its part of the study. The following summarises that analysis.

The pattern of groundwater occurrence and movement is strongly influenced by the geology of an area, and this is certainly the case in the Kyeamba catchment. The distribution of geological formations is shown in Figure 2 (based on Adamson and Laudon, 1966).

Granite underlies a large proportion of the south-western third of the area. It is essentially impermeable - i.e. it will not transmit water - and there are no productive bores in the granite areas. Nor are there any identified waterlogging or salinisation problems.

Figure 2. Kyeamba Landcare Geology and Bore Locations

Most of the area is underlain by slate, which is highly fractured and a moderately good water carrier. It is generally weathered to a moderate to strong degree near the land surface, and this weathered zone may be relatively impermeable. Salinity of groundwater in the State aquifers ranges from a few hundred to several thousand mg/L.

Much of the slate and granite is covered by younger deposits, to a relatively shallow depth. In the main valleys of Kyeamba Creek and its tributaries, and along the northern margin of the Landcare area in the valleys of Tarcutta Creek and the Murrumbidgee River, there is an accumulation of alluvial deposits with a maximum thickness of some 50 m and whose upper surface is the alluvial plains of these valleys. These deposits are dominated by clay and silt, but do contain some sand and gravel lenses particularly in the thicker central areas. The sand and gravel lenses are useful aquifers and commonly provide water sufficient for stock watering. In a limited area, supplies sufficient for irrigation can be obtained.

Flanking the alluvial plains, and grading into the higher hilly areas of slate and granite, there is a zone of colluvium, essentially a disorganised mix of weathered rock remnants which are either in situ or have been transported over very short distances. The material is likely to be essentially impermeable.

Water enters the groundwater system in the Ordovician rocks by infiltration of rainfall, possibly concentrated in the higher ridge areas. Once in the groundwater system the water drains towards the local and regional drainage. The Kyeamba and O'Brien's Creek valleys form the main drain for the Kyeamba Catchment, while Coreinbob and Mates Creeks form small drains in the north-east part of the Landcare area. The Tarcutta and Murrumbidgee Valleys form the regional drainage system and it is clear the groundwater from the Landcare area is draining into the alluvial deposits along these valleys.

The response of the groundwater system in the fractured rock aquifers to clearing and development of agriculture in the area was examined by comparing past and present water levels in a set of bores. Past levels were obtained from DWR records which, in most cases, include a water level measured at the time of drilling. A field inspection was made to measure present water level in those bores for which an initial value was available. Bore locations are shown in Figure 2, together with a table showing initial and present water levels. Very little intermediate data could be obtained, and the history of water level rise is therefore uncertain. Such data as are available are plotted in Figure 3.

Figure 3. Bore hydrographs, bedrock aquifers

Water levels have risen by as much as 40 m since 1921 in some places and it seems likely that much of the rise has occurred since the early 1960s. There are now large areas in which the water pressure head is above ground level (Figure 4). As a consequence, water is discharging at ground surface in many places. This causes waterlogging of the soil and subsequent salinisation, as the initial salt concentration is increased by evaporation.

Where such areas coincide with the occurrence of alluvial deposits, water will drain upwards into the alluvium. Once in the alluvium, the alluvial aquifer system will dominate and the water will tend to move laterally. The general direction will be to the north, towards the regional drain (i.e. the Murrumbidgee Valley). At some stage the capacity of the alluvial deposits to transport the extra water will be exceeded and water levels will rise in this aquifer system too. As the water table rises to within 1-2 m of ground level, concentration of salinity will commence, leading to salinisation as well as waterlogging.

Figure 4. Kyeamba Landcare - Depth to standing water level, Dec. 1990

Figure 5. Bore hydrographs, alluvial aquifers

Figure 6. Bore hydrographs, alluvial aquifers

The response of the alluvial aquifer system is shown by the hydrographs of DWR bores in Figures 5 and 6. Bores 30385 and 30386 are in the Tarcutta Creek alluvium, and 30383 is within the Murrumbidgee River alluvium. Water levels in these bores show a peak during the 1974 floods, with a subsequent gradual decline interrupted by a further major recharge event in 1983. Since then, water level in 25383 has declined to the pre-1983 level, but water levels at 30385 and 30386 in the Tarcutta alluvium have remained approximately at the later 1983 level, with minor fluctuations.

At bore 30351, in the Kyeamba alluvium just south of the O'Brien's Creek confluence, the 1973/74 peak is clear, but the decline to 1983 levels is less pronounced. There was a rapid rise in 1983, but this was followed by a general rising trend with 1990 levels almost as high as the 1974 peak.

At bore 30355, which is in the Gregadoo Creek/O'Brien's Creek alluvium, the rapid rise of water level in 1973/74 is shown, but since then there has been a rising trend rather than a falling one. At this location water level declined at a low rate until early 1978 when a recharge event returned water level to the 1974 peak level. A further slow decline was interrupted by recharge in 1981 and 1983, and since 1983 there has been a clear upward trend with only small fluctuations. The December 1990 level is the highest yet recorded (approximately 1 m higher than the late 1989 level shown in the hydrograph).

It is apparent that in the alluvium with the lowest transmissivity (i.e. capacity to transport water laterally), represented by bore 30355, the inflow of water from the fractured rock aquifers has already exceeded its transmitting capacity. Development of further areas of waterlogging and salinisation can be expected, therefore, within alluvial areas where aquifers are thin and where the pressure head of the rock aquifers is at or near ground level. At 30351, in the main Kyeamba alluvium, it is possible that the capacity of the alluvium to handle the additional water entering from fractured rocks has nearly been reached, and there is reason for some concern even in the Tarcutta Creek alluvium. The Murrumbidgee alluvium does not seem to have been affected.

An excellent demonstration of development of salinity in the low permeability parts of the alluvium was observed in the Kyeamba Creek alluvium, some 10 km south of Ladysmith, where a line of poplars has been planted along a fenceline normal to the valley margin. They extend from a point several hundred metres from the alluvial boundary to the valley margin and up the valley sideslope. There are indications of waterlogging across the alluvium. All trees on the alluvium are dead (except one, which is sick) while all those on the valley sideslope are flourishing. Since poplars will survive in waterlogged conditions, the evidence clearly suggests saline conditions occur here.

All areas of alluvium are vulnerable to this effect if the pressure head of groundwater is above the base of the alluvial deposits. The effect is independent of scale - it has been observed in the main Kyeamba alluvium; it is presumed to be the cause of observed salinisation in some very small tributary gullies; and it is bound to be occurring within the Tarcutta Creek/Murrumbidgee River alluvium, although in this case the effects may not be observable for some time.

Where there is no alluvial cover and the pressure head is above ground level, water will travel upwards to ground surface if there is a path. A path can be provided by fractures or joints, perhaps by weathering zones, or any other discontinuity in the rock mass. Where such paths reach the surface and provide a passage for water, waterlogging will develop and salinisation will shortly follow.

Comparison of the history of water level rise, as known, with factors which may have influenced it, leads to the conclusion that the rise is at least partially, and probably mainly, attributed to land clearing and subsequent agricultural activity. Thus, the rainfall pattern shows a very wet period from 1968 to 1978, and it is possible that groundwater levels rose substantially during this period. But water levels continued to rise during the following decade when rainfall was below average. Wheat grazing was also an important activity during the late 1970s and early 1980s, but the subsequent reduction has not caused a reversal of the rising water level trend.

A small area near Yass with similar geological features, underlain by the weathered Ordovician slate, has recently been examined in detail by Scott (1991). Figure 7, reproduced from his work, neatly encapsulates the general conditions in the Kyeamba Valley area. In the higher topographic areas, the aquifer system in the Ordovician rocks is unconfined, and recharge occurs by direct infiltration through thin permeable soil into the open rock fractures. Colluvial material, which is essentially a residual clay from the weathering of the rock formations, forms a low permeability barrier (aquitard) above the fractured rock, and acts as a confining layer. This enables the development of artesian conditions in the fractured rock, as increased recharge causes rising water levels in the areas where the aquifer system is unconfined. Upward leakage through this layer, in the lower topographic areas, has two effects:

1. a direct effect, i.e. tendency to saturate the overlying soil (or alluvial) material;

2. an indirect effect, i.e. to prevent the downward drainage of any excess water from the soil (excess water which may itself result from increased infiltration rates in cultivated, as opposed to native, vegetation).

Figure 7. Representative Catchment Profile (after Scott, 1991).

The combination of these two effects, which may have a varying degree of relative importance across the catchment, is the root cause of the waterlogging and salinity problem. The prime driving force, however, is the artesian head condition of the rock aquifer. That this is so is demonstrated by the time scale over which the problem has developed - waterlogging and salinisation did not develop until artesian heads had risen close to, or above, ground level. If waterlogging of the soil by increased infiltration on the lower lying areas had been the main cause, problems would have developed much sooner.

Acceptance of this conceptual model leads to some important conclusions, as follows:

1. Priorities

The first priority for changed land management practice should be to reduce the pressure head on the water in the rock aquifer, in the lower part of the catchment. This can be done by reducing recharge in the areas where the aquifer is unconfined.

The second priority is supplementary action to reduce infiltration to the soil/alluvium in the lower lying parts of the catchment where the colluvial layer is present. While reduction of pressure head (the first priority) has the potential to rehabilitate the area by itself, the process would be faster if supplemented in this way.

2. Recharge Areas

Recharge to the rock aquifer will not occur where the potentiometric surface is above ground level, although there may be local recharge of shallow alluvial aquifers from infiltration of rainfall. Recharge to the rock aquifers will occur mainly, if not entirely, in areas where there is no development of impermeable colluvial material. Such areas are most likely to coincide with the higher topographic areas.

It is therefore concluded that recharge to rock aquifers can best be reduced by the use of high water-use vegetation concentrated in the higher parts of the area. Sufficient reduction of recharge in these areas should eventually result in a lowering of the potentiometric surface below ground level in the lower catchment areas and hence a cessation of saline seepage.

3. Required Rate of Recharge Reduction

It is not possible to provide precise estimates, because of the complexity of the natural system and lack of sufficient data. Nevertheless, it is possible to provide estimates which may be good enough to provide useful targets for development of land management strategies.

The rise of water level in the higher topographic areas where the aquifer system is unconfined, can be used as an indicator of the rate of recharge, if a value for the storativity (i.e. capacity to store water or, effectively, the porosity) of the rocks is known. As a first approximation, it could be assumed on general principles that the storativity would be between 0.01 and 0.1. In his analysis, Scott (op cit) has used bore hydrograph data to derive a value of 0.005. Given the similarity of the rock types, and lack of any means of deriving a better estimate for the Kyeamba area, it is reasonable to adopt a value of 0.01 for this area.

The rate of rise of water levels in the Kyeamba area is uncertain. While rises of up to 42 m have been recorded, the amount of rise is uneven over the area, and it is not clear that it has been a steady rise. As discussed in a previous section, it seems likely that most of the rise took place during the mid-1960s to late 1970s. For estimation purposes, a value of 1 m/year has been adopted, i.e. a 20 m rise over a period of 20 years.

For a water level rise of 1 m, in rocks with a storativity of 0.01, the recharge required is 10 mm. That is, the amount of extra recharge required to produce the observed water level rises is about 10 mm per year. Given the approximations used, the estimate should be regarded as 10-20 mm.

Catchment management strategies should therefore aim at a reduction of infiltration over the main recharge areas of 10-20 mm to stabilise it at its current status. Doubling of this reduction could be expected to see a significant improvement over a period of 10-20 years. These estimates are of a similar order of magnitude to those derived by Schofield (1990) for a small salinised catchment in Western Australia.

4. Reduction of Infiltration in Lower Areas

As indicated previously, lowering the pressure head on the rock aquifer system should be sufficient, in time, to alleviate the salinisation problems. The process will be quicker if there is a concurrent reduction of infiltration to the shallow soil/alluvium in the lower parts of the catchment.

From the groundwater analysis conducted in this study, it is not possible to estimate an optimum reduction, but the speed of recovery will be proportional to the reduction achieved.

Means of achieving the necessary reduction of infiltration and recharge rates are currently being addressed by the Soil Conservation Service. Use of trees as part of the management strategy is clearly an essential aspect. It is not feasible, however, and neither is it necessary, to plant trees on the whole catchment. But the high water use capability of eucalypts especially suggests that they must have a place.

The difficulty is in providing a strategy for farm scale action based on an analysis made on a regional scale with very limited data, and to ensure that maximum benefit is derived from the trees which are planted. Considerable research effort is being expended on this problem at present, and solutions at individual farm scale have been achieved by CSIRO using the TOPOG computer simulation techniques (O'Loughlin et al., 1989). Such techniques, combined with detailed soil and terrain mapping, such as has been conducted by the Soil Conservation Service in the Kyeamba area, offer the possibility of providing useful answers to the questions of landholders who are now saying, "I have been convinced there is a problem - please tell me how I can conduct my farm in such a way that there is the best chance for remediation of that problem".

References

1. Adamson, C.L. and Loudon, A.G. (1966). Wagga Wagga 1:250 000 Geological Map. Geological Survey of NSW.

2. Schofield, N.J. (1990). Determining reforestation area and distribution for salinity control. J.Hydrological Sciences, 35:1.

3. Scott, D. (1991). Hydrogeological, Geophysical and Water Balance Study of Dryland Salinity at Dicks Creek, Yass Valley, NSW. M.App.Sc. Thesis, UNSW (unpublished).

4. O'Loughlin, E.M., Short, D.L. and Dawes, W.R. (1989). Hydrology and Water Resources Symposium, University of Canterbury, Christchurch, NZ, 23-30 November 1989.