Proteoid root mats stabilise Hawkesbury Sandstone biomantles following fire.
The proteoid roots of Banksia serrata L. f. form a dense root mat which can be thought of as an organic form of geotextile. Previous research on a Hawkesbury Sandstone hillslope revealed that, on average, B. serrata L. f. are spaced at 1 tree per 31.25 [m.sup.2] and demonstrated that proteoid root mats cover 10% of the surface area and a greater percentage of the subsurface area (Gould 1998). Fabric analysis of proteoid root systems has demonstrated that they have a number of binding mechanisms of differing persistence which actively bind biomantle materials. Actively growing roots are densely covered in root hairs which attach to the surface of leaf litter, individual soil particles, and soil aggregates. Rootlets with or without root hairs trap individual particles [is greater than] 250/[micro] m. The most persistent structural components of the root mat continue to bind medium-to-coarse sand even following damage by bushfire or partial breakdown of the root mat (Gould 1998).
Do proteoid root mats provide a stabilising mechanism at the scale of hillslopes? This paper examines the extent and character of proteoid root distribution on the scale of hillslopes by analysing their distribution within repeating landscape elements. The hypothesis investigated is that the proteoid root mats of Banksia serrata L. f. stabilise Hawkesbury Sandstone biomantles following bushfire.
Materials and methods
Two study sites typical of Hawkesbury Sandstone hillslopes with respect to geology, topography, vegetation, and soils were selected. Both study sites have a well-developed biomantle. Soils are discontinuous due to extensive bedrock outcrops and rarely exceed 70 cm in depth. Soil types alternate between uniform coarse-textured soils and duplex soils (in the sense of Northcote 1979) formed by the downslope movement of coarse material over in situ clay material derived from shale or argillaceous sandstone. Vegetation is an open eucalypt woodland with a shrub understorey. One site, at West Pymble in an area of bushland reserve adjacent to Lane Cove National Park, had been burnt 1 year previously in a wildfire. An unburnt control site with comparable topography, soils, and vegetation was selected at Westleigh.
Both study sites consist of a series of sandstone benches. The downslope margin of each bench is defined by outcropping bedrock which forms a low cliff-line. On each bench there is a sandy, low-to-moderate slope (range 3-19 [degrees]) where overland flow either coalesces into shallow tributary ephemeral drainage lines and is channelled through crevices in the cliff-line or remains unconfined and flows over the edge of the cliff-line.
For the purposes of this study, the geomorphology of both hillslopes was characterised by 6 repeating micro-geomorphic elements: (i) litter dam-microterraces; (ii) boulder dams; (iii) joint crevice fans; (iv) mossy ledges; (v) ephemeral drainage lines; (vi) rock overhangs. These micro-geomorphic elements were identified on the basis of characteristic relationships between persistent properties (such as rock exposure) and surface materials.
Litter dam-microterraces form on slopes of [is less than] 10 [degrees] on which overland flow is widespread. Elements of surface roughness trap leaf litter and other floating load carried in overland flow which then accumulate to form a litter dam. An ephemeral pond forms upslope of the litter dam wall and is subsequently infilled by bedload deposits to form a microterrace (Mitchell and Humphreys 1987).
Where the upslope edge of a boulder lies across the direction of water flow, its upslope face acts as a dam wall to both channelled and overland flow. Leaf litter and bedload deposits accumulate upslope of the boulder. Scouring and/or small depositional fans occur around the sides of the boulder.
Joint crevice fans
Vertical joint planes occur in the outcropping bedrock of low cliff-lines at the downslope margins of benches. Overland flow moving downslope toward the lower edge of the bench often coalesces into shallow drainage lines and is channelled through these joint crevices. Water flow initially increases in velocity as it is channelled through the joint crevices but slows as it is released from confinement and base porosity increases. Bedload material is deposited in conspicuous fans at the base of the joint crevices.
Soil water seepage occurs on the top of the low cliff-lines at the edge of each bench. Where the outcropping bedrock is of a low angle, water flow is slow. Moss mats grow along these ledges due to the fairly constant moisture conditions enhanced by seepage through the soil.
Ephemeral drainage lines
Water-repellent soils and extensive rock surfaces cause widespread overland flow to occur on low angle slopes even during low intensity rainfall events. In this sense, much of the slope can be perceived as a line of drainage to some extent. Incision was used to distinguish ephemeral drainage lines from the rest of the slope. Ephemeral drainage lines are incised up to 10 cm below the general slope surface.
Where overland flow occurs, water may run over the top of the low cliff-lines. The erosive potential of the water increases as it drops onto the sandy material below and often forms distinctive impact craters below the overhang. The overhang prevents the accumulation of leaf litter, leaving mineral soil exposed.
Analysing the distribution of micro-geomorphic elements
The distribution of these 6 micro-geomorphic elements was plotted onto a base map and 5 transects were prepared for the West Pymble site to analyse the distribution of micro-geomorphic elements. Using these transects, the base map, and information from excavated sites, a schematic transect of the West Pymble slope was developed characterising the downslope sequence of micro-geomorphic elements and the distribution of soil materials.
Stratigraphic analysis of micro-geomorphic elements
Three replicates of each micro-geomorphic element were excavated at each site. Prior to excavation, microtopography and surface materials were mapped for each replicate. Microtopography was plotted by measuring the distance to ground surface at each intersect from a levelled 1-[m.sup.2] frame strung at 100-mm intervals. Level string lines were used where it was not possible to set up the frame. Representative micro-geomorphic elements were selected from different parts of each site.
Each micro-geomorphic element was excavated (1 [m.sup.2]), matching soil layer characteristics with the mapped surface materials and micro-topography. Each element was vertically sliced at 250-mm intervals to bedrock where possible, so that for each element, 5 consecutive 1-m profiles were dug and examined stratigraphically. The layer sequence of each profile was plotted and a 3-dimensional diagram of each 1-[m.sup.2] plot constructed showing relationships between microtopography, proteoid roots, and soil materials (Fig. 1).
[Figure 1 ILLUSTRATION OMITTED]
Particular attention was paid to the stratigraphic relationships of materials using a modification of the key to soil stratigraphy developed by Mitchell (1985). For each layer of soil material, the following characteristics were recorded: composition, grain size and shape, coherency, packing, presence of stones and roots, and evidence of bioturbation (Table 1). The key for identification of matrix fabric developed by Humphreys (1985) was used to assess coherency and packing in the field. By applying stratigraphic principles, the soil materials provided a record from which sequences of events could be inferred. A charred root mat, for example, indicated the surface at the time of fire. Any material overlying a charred root mat indicated post-fire deposition.
Table 1 Nature of soil materials (after Mitchell 1985) Layer Description 0 Loose, dry leaf litter consisting of entire leaves, twigs, and bark; [is less than or equal to] 5 cm thick. 1 Comminuted leaf litter; [is less than or equal to] 5 mm thick. 2 Charcoal fragments (< 10 mm) 3 Incoherent, clean quartz sand (500-1000 [micro]m), subangular to subrounded; no evidence of bioturbation; clean grains indicate transportation by rainsplash and/or rainwash; occurs as shallow surface lag over Layers 4, 7, and 8 or mixed with Layers 0, 1, and 2; [is less than or equal to] 5 mm thick. 4 Weakly coherent, fine dark crust (62-125 [micro]m); porous plasma support fabric; no evidence of bioturbation; occurs at base of Layer 3; [is less than or equal to] 2 mm thick. 5 Algal mat; may occur as surface layer overlying Layers 3 and 7; also occurs at base of Layer 3 associated with Layer 4; <1 mm thick. 6 Moss mat; sometimes grows on top of Layer 7; often overlain by Layer 3 and/or Layer 2; [is less than or equal to] 20 mm thick. 7 Proteoid root mat; dense mat of tightly curled roots; may be actively growing or in state of decay; often charred on upper surface; often contains quartz sand (500-710 [micro]m) and charcoal fragments (<1 cm); most dense in uppermost 40 mm of the soil profile then intergrades with Layer 8; [is less than or equal to] 150 mm thick. 8 Surface mantle of loamy sands-clayey sands-sandy loams; secondary porous grain support fabric; subangular to subrounded quartz grains (350-710 [micro]m); grains often coated with organic stain; inclusions of charcoal, quartz pebbles, and organics indicate transportational nature of material; commonly contains fragments of sandstone, ironstone, and roots; bioturbation evident in form of organic inclusions, channels, and chambers. 9 Stone layer; stones may be sandstone or ironstone (<100 mm long). 10 Loamy sands-clayey sands-sandy loams-clay loams; secondary porous grain support fabric; boundary between Layers 8 and 10 clear to gradual often separated by Layer 9; differentiated from Layer 8 by absence of charcoal, small stone fragments, and organic inclusions; bioturbation evident in form of maculae; often contains weathered sandstone. 11 Sandstone boulders (>200 mm); often embedded in Layer 8, particularly under rock overhangs. 12 Sandy clay; dark yellow; dense grain support to plasma support fabric; no evidence of bioturbation; indicates presence of weathered shale bedrock
Assessment of soil-water repellency
Undisturbed soil samples were collected in Kubiena tins (90 by 90 by 70 mm) from each of the excavated micro-geomorphic elements. Soil samples were air-dried and then horizontally sectioned at 5-10-mm intervals for fabric analysis. Each layer of soil material was tested for water repellency by placing a single drop of water onto the surface and timing the rate of absorption (Bisdom et al. 1993).
Outcropping bedrock divided the West Pymble hillslope into 4 distinct benches. A schematic cross-section of the slope characterising the bench pattern, soil stratigraphy, and sequence of micro-geomorphic elements is shown in Fig. 2.
[Figure 2 ILLUSTRATION OMITTED]
Distribution of micro-geomorphic elements
Resistant layers of sandstone bedrock defined the structural bench pattern and directly determined the distribution of joint crevices, rock overhangs, and mossy ledges. Rock overhangs and mossy ledges coincided with the outcropping sandstone bedrock which forms low cliff-lines. The direction, dimensions, and position on the slope of ephemeral drainage lines were also strongly influenced by the location of outcropping bedrock and by the presence of persistent features such as joint crevices and boulder dams. Boulder dams occurred in drainage lines and in joint crevice fans and were smaller and further apart with increasing distance from the top of each bench. Litter dam--microterraces mostly occurred on low angle ([is less than] 10 [degrees]), sandy sections of the slope, and generally were oriented parallel to the contours. The distance between consecutive litter dams was 1-3 m.
Distribution of proteoid roots within micro-geomorphic elements
Table 2 shows average values for the surface distribution of materials in all micro-geomorphic elements. At the burnt site, proteoid roots comprised 4.9% of the total surface area compared with 1.8% at the unburnt site. Excavation of micro-geomorphic elements showed that subsurface roots were much more extensive than indicated by surface roots. The average value for the total extent of proteoid roots within micro-geomorphic elements (surface+subsurface) was 31.9% and ranged from 7.1% to 50-6% (Table 3). Overall, the ratio of surface to subsurface roots was 1:8.6.
Table 2. Average extent of surface materials associated with micro-geomorphic elements
Values indicate percentage of 1-[m.sup.2] study plots covered by each material. Acronyms refer to type of micro-geomorphic element: EDL, ephemeral drainage lines; JCF, joint crevice fans; RO, rock overhangs; ML, mossy ledges; BD, boulder dams; LDM, litter dam-microterraces. B, combined data for the burnt site; U, combined data for the unburnt site
Feature Exposed Mineral Leaf litter rocks soil EDL, B 14.6 69.3 13.3 EDL, U 10.3 32.6 52.3 EDL, both sites 12.5 51.0 32.8 JCF. B 39.0 21.0 31.0 JCF. U 48.0 10.3 38.6 JCF. both sites 43.5 15.6 34.8 ROB 13.6 51.6 25.3 RO U 9.3 38.6 47.3 RO both sites 11.5 45.0 36.3 ML B 20.0 50.3 9.3 ML U 13.3 7.0 41.3 ML both sites 16.6 28.6 25.3 BD B 32.0 38.0 25.3 BD. U 37.0 2.3 60.3 BD. both sites 34.5 20.0 42.8 LDM, B 1.3 50.6 42.6 LDM, U 5.0 8.0 83.3 LDM, both sites 3.0 29.3 63.0 All burnt elements 20.1 46.8 24.5 All unburnt elements 20.5 16.5 53.9 All elements 20.3 31.6 39.2 Feature Proteoid Other Moss roots roots EDL, B 0 2.3 0.3 EDL, U 0 1.3 3.3 EDL, both sites 0 1.8 1.8 JCF. B 8.6 0.3 0.0 JCF. U 2.6 0 0.3 JCF. both sites 5.6 0.2 0.2 ROB 9.0 0.3 0 RO U 3.0 1.3 0.3 RO both sites 6.0 0.8 0-2 ML B 3.6 1.0 15.3 ML U 1.0 0.0 37.3 ML both sites 2.3 0.5 26.3 BD B 4.6 0 0 BD. U 0.3 0 0 BD. both sites 2.5 0 0 LDM, B 3.3 1.0 1.0 LDM, U 3.6 0 0 LDM, both sites 3.5 0.5 0.5 All burnt elements 4.9 0.8 2.8 All unburnt elements 1.8 0.4 6.9 All elements 3.3 0.6 4.8
Table 3. Summary data (means) relating to excavated micro-geomorphic elements
Acronyms refer to type of micro-geomorphic element: EDL, ephemeral drainage lines; JCF, joint crevice fans; RO, rock overhangs; ML, mossy ledges; BD, boulder dams; LDM, litter dam-microterraces; prm, proteoid root mat. B, combined data for the burnt site; U, combined data for the unburnt site
Element Max. Max. Max. soil depth depth to thickness top or prm of (mm) prm EDL, B 13.0 EDL, U 0. EDL, both sites 263.0 6.5 8.0 JCF, B 50.0 JCF, U 76.0 JCF, both sites 316.0 63.0 61.0 RO, B 27.0 RO, U 20.0 RO, both sites 333.0 23.5 56.0 ML, B 6.0 ML, U 15.0 ML, both sites 130.0 10.5 33.0 BD, B 30.0 BD, U 16.0 BD, both sites 311.0 23.0 36.0 LDM, B 21.0 LDM, U 33.0 LDM, both sites 313.0 27.0 58.0 All burnt elements 24.5 All unburnt elements 26.6 All elements 277.6 25.6 42.0 Element Extent of Extent of Total exposed buried extent of prm prm prm (% 1-[m.sup.2] study plot covered) EDL, B 0 14.3 EDL, U 0 0.0 EDL, both sites 0 7.2 7.1 JCF, B 8.6 49.3 JCF, U 2.6 17.0 JCF, both sites 5.6 33.2 38.8 RO, B 9.0 46.3 RO, U 3.0 18.3 RO, both sites 6.0 32.3 38.3 ML, B 3.6 17.0 ML, U 1.0 10.0 ML, both sites 2.3 13.5 15.8 BD, B 4.6 57.0 BD, U 0.3 20.0 BD, both sites 2.5 38.5 41.0 LDM, B 3.3 49.6 LDM, U 3.6 44.6 LDM, both sites 3.5 47.1 50.6 All burnt elements 4.9 38.9 All unburnt elements 1.8 18.3 All elements 3.3 28.6 31.9
Excavation of the micro-geomorphic elements revealed the following consistent patterns of proteoid root distribution.
Subsurface proteoid root mats were concentrated in depositional zones: in microterraces associated with litter dams, in microterraces upslope of boulder dams, in fans that occurred to the side of boulder dams, in joint crevice fans, lining the sides of joint crevices, as finger-like projections that tapered towards the front of the bench in mossy ledges, and underneath rock overhangs.
Proteoid root mat exposure occurred in erosion zones: upslope of the microterrace associated with litter dams, in litter dam walls, at the downslope end of joint crevice fans where flow scoured to one side, and in a band coincident with the dripline of rock overhangs.
Maximum thickness of the proteoid root mat generally occurred adjacent to rock surfaces and underneath rock overhangs. The greatest maximum thickness of the proteoid root mat was 150 mm recorded in a litter dam--microterrace. On average, the maximum thickness of the root mat was 42.5 mm. Throughout all features where the proteoid root mat occurred it was densest in the uppermost 20 mm and became sparser with increasing depth.
Proteoid roots were entirely absent from ephemeral drainage lines with the exception of narrow bands overlain by strands of leaf litter parallel to the drainage line in 2 elements at the burnt site.
Proteoid roots in relation to soil stratigraphy
Due to the rapid accumulation of leaf litter at the West Pymble site in the year following the fire in January 1994 fire, there was only a 2-mm difference in average maximum depth of soil material to the top of the root mat between the burnt and the unburnt site. Combining data sets for both sites, the average maximum depth to the top of the root mat was 25.5 mm, consisting mostly of leaf litter.
Generally, proteoid roots were overlain by various combinations of soil Layers 0-6 (Table 1). The most common stratigraphic sequence was dry, entire leaf litter (Layer 0), overlying a combination of comminuted leaf litter (Layer 1), charcoal fragments (Layer 2), and incoherent quartz sand (Layer 3) overlying the proteoid root mat. Proteoid roots were observed actively growing up into the leaf litter forming a tightly bound laminated mat up to 30 mm deep within a litter dam-microterrace and a boulder dam at the unburnt site only. In one litter dam and one boulder dam at the unburnt site only, the proteoid root mat was clearly laminated with light-coloured proteoid root mats overlying older, dark root mats.
Charred proteoid root mats, marking the surface at the time of fire, were overlain by leaf litter, charcoal, and single-grain sands. Charred root mats occurred at both the burnt site and in 2 joint crevice fans at the unburnt site. At the unburnt site, the charred root mat was also overlain by younger, lighter coloured root mat.
The proteoid root mats occurred in both uniform and duplex soils with A horizons of clayey sands--light sandy clay loams and B horizons of clayey sand--sandy clay. All soil fabrics associated with proteoid roots were water-repellent, most of them strongly to severely water-repellent, with water drop penetration time in excess of 60 s.
The surface and subsurface distribution of proteoid roots was not as widespread as initially indicated (Gould 1998); however, these results confirm that subsurface proteoid roots are widespread and persistent following fire. The presence of widespread and highly water-repellent root mats is likely to be a significant contributing factor to the generation of run-off on porous, sandy, low angle slopes. At the same time, the physical structure of the root mat high in the soil profile presents a physical barrier to the processes of erosion. The net effect of the proteoid root mat on soil erosion lies in the distribution of the root mat in relation to zones of net erosion and net deposition and in its temporal distribution in relation to fire cycles.
Position in the soil profile
The most important aspect of proteoid root distribution with respect to soil erosion is their position in the soil profile. Stratigraphic analysis showed that the average depth of material overlying the proteoid root mat was 25.5 mm. The material overlying the leaf litter consisted mostly of organic matter. Consumption of this organic matter by fire would leave a lag layer of incoherent sands [is less than] 5 mm thick. Although this lag material would remain susceptible to erosion by rainwash, it is inferred that the presence of the coherent, persistent root mat establishes a physical limit to the depth of soil over which erosive processes operate. Analysis of soil fabrics associated with proteoid root systems has shown that even where the upper surface of the root mat has been burnt, it is able to retain material [is greater than] 250 [micro]m (Gould 1998). The presence of the root mat protects the sandy soil from the development of rill and gully erosion.
Temporal distribution of proteoid root mats
The other critical aspect of proteoid root distribution with respect to soil stabilisation is their temporal distribution in relation to periods of increased erosion potential. Mitchell and Humphreys (1987) have argued that the chance combination of an intense rainfall event with soils left exposed by bushfire poses the greatest risk of soil erosion. The widespread presence of proteoid root mats at the burnt site clearly demonstrates that proteoid root mats survive bushfire. The presence of coherent root mats high in the soil profile following fire provides protection to soils at a time when they would otherwise be highly susceptible to erosion. The presence of charred root mats overlain by soil materials and younger proteoid roots within 2 elements at the unburnt site dates the proteoid root mat to at least the time of the last fire, which according to fire history maps was 10 years previously.
Persistent proteoid root mats have also been linked to the maintenance of litter dam--microterrace features, which have been shown to be an important control on the timing of sediment yield in relation to fire cycles (Mitchell and Humphreys 1987).
Spatial distribution of proteoid roots in relation to geomorphic processes
On the evidence of bands of micro-geomorphic elements parallel to the slope contours and the presence of sequences of parallel litter dams, the slope is interpreted as consisting of zones of erosion--deposition that lie generally parallel to the contours. On the basis of the distribution of persistent micro-geomorphic elements with associated bedload deposits, the upper section of each bench is interpreted as a zone of net deposition and the lower section as a zone of net erosion. The low cliff-lines formed by resistant beds of sandstone mark the transition from erosion to deposition zones. Superimposed on this broad pattern of lateral bands of erosion--deposition are ephemeral drainage lines. Drainage lines incise through lateral erosion--deposition zones until flow is dissipated by the next set of boulder dams downslope.
Bedrock structure and lithology appear to be the most important factors defining the distribution of geomorphic processes. The concentration of proteoid roots in zones of net deposition and their absence from incised ephemeral drainage lines suggests that at a level of persistent erosion the presence of proteoid roots is determined by the balance between erosion and deposition. The balance between erosion and deposition is dynamic in response to individual weather events but is broadly defined over the long term by the bench structure.
At a lower level of erosive intensity, the distribution of geomorphic processes appears to be defined by surface conditions, including vegetation and soil characteristics. It is at this level that the presence of proteoid roots may influence the distribution of erosive processes through run-off generation and maintenance of overland flow. It is at this level of erosive intensity, too, that proteoid root mats may define the distribution of erosive intensity through the maintenance of ephemeral micro-geomorphic elements such as litter dam-microterraces.
Interaction of proteoid roots with geomorphic processes
The combination of the severe water repellency and the physical structure of the root mat near the soil surface appears to be an important cause of the distribution of erosion and deposition processes. The water repellency of the root mat contributes to the generation of overland flow on sandy, low angle slopes even in response to low intensity rainfall events. At the same time the physical structure of the root mat high in the soil profile assists in the maintenance of run off as overland flow by preventing the development of filling and gullying in the sandy soils (Zierholz and Hairsine 1995).
Exposure of proteoid roots in erosion zones indicates that they transfer erosive intensity downslope. The absence of proteoid roots from zones of persistent or intense erosion such as the ephemeral drainage lines, however, indicates that proteoid roots are limited in their capacity to compete with intense or persistent erosion as determined by slope structure (Thornes 1989).
Absence of proteoid roots from zones of intense erosion suggests that a feedback relationship operates between zones of persistent erosion such that proteoid roots are unable to form due to persistent disturbance. Ongoing incision of the drainage lines in turn may be caused by conversion of overland flow to rill erosion in areas where the root mat is absent.
An alternative explanation for the distribution of proteoid roots in relation to zones of erosion and deposition is that proteoid roots are surface feeding roots that only develop in depositional areas where leaf litter accumulates. In this scenario, the absence of proteoid root mats from drainage lines would be due to the absence of leaf litter. Regardless of the cause of proteoid root distribution in relation to zones of erosion and deposition, the presence of the root mat protects the sandy soil from the development of rill and gully erosion.
From a statistical viewpoint, excavation of a greater number of sites would have been desirable. Sampling was constrained, however, by the considerable amount of time and effort required to excavate and plot each micro-geomorphic element.
Whether or not proteoid root mats are able to provide a stabilising mechanism on the scale of hillslopes depends on their overall extent, and their distribution and coherency in relation to erosive intensity for any particular erosion event. Proteoid roots are clearly patchy in their spatial distribution and interact differently with overlying mobile soil materials over time, depending on whether they are actively growing, coherent, or in decay. However, the present study has shown that they are extensively distributed within micro-geomorphic elements and positioned high in the soil profile at a time when soils might otherwise be susceptible to soil erosion. On the basis of this evidence it is concluded that the proteoid roots of B. serrata L. f. stabilise Hawkesbury Sandstone biomantles following bushfire.
The author would like to acknowledge Peter Mitchell for supervision of the Honours research which forms the basis of this paper, the school of Earth Sciences at Macquarie University for financial assistance, and the staff of Geographics at Macquarie University for assistance with preparation of graphics.
Bisdom, E. B. A., Dekker, L. W., and Schoute, J. F. (1993). Water repellency of sieve fractions from sandy soils and relationships with organic material and soil structure. Geoderma 23, 175-89.
Gould, S. F. (1998). Proteoid root mats bind surface materials in Hawkesbury Sandstone biomantles. Australian Journal of Soil Research 36, 1019-31.
Humphreys, G. S. (1985). Bioturbation, rainwash and texture contrast soils. PhD Thesis, School of Earth Sciences, Macquarie University.
Mitchell, P. B. (1985). Soil bioturbation. PhD Thesis, School of Earth Sciences, Macquarie University.
Mitchell, P. B., and Humphreys, G. S. (1987). Litter dams and microterraces formed on hillslopes subject to rainwash in the Sydney Basin, Australia. Geoderma 39, 331-57.
Northcote, K. H. (1979). `A Factual Key for the Recognition of Australian Soils.' 4th Edn. (Rellim Technical Publications: Glenside, S. Aust.)
Thornes, J. (1989). Solutions to soil erosion. New Scientist 3 June, 27-31.
Zierholz, C., and Hairsine, P. (1995). Runoff and soil erosion in bushland following the Sydney bushfires. Australian Journal of Soil and Water Conservation 8, 28-37.
Manuscript received 12 January 1998, accepted 23 June 1998
Susan F. Gould
PO Box 478, Weipa, Qld 4874, Australia; email: SusanFGould@yahoo.com
|Printer friendly Cite/link Email Feedback|
|Author:||Gould, Susan F.|
|Publication:||Australian Journal of Soil Research|
|Date:||Nov 1, 1998|
|Previous Article:||Enhancement of soil nutrients around nest entrances of the funnel ant Aphaenogaster barbigula (Myrmicinae) in semi-arid eastern Australia.|
|Next Article:||Soil erodibility ([K.sub.m]) values for some Australian soils.|