Email j.aiken@uws.edu.au

Abstract

Key Words

UWS Hawkesbury, soil series, soil classification, soil type, soil map, HAC soil types.

Introduction

The University of Western Sydney (UWS) Hawkesbury campus site at Richmond New South Wales has been an educational institution since the inauguration of the Hawkesbury Agricultural College (HAC) in 1892. Farm areas covered 1412 ha (3,490 acre). Soils are alluvial depositions. They represent regional and site-specific geomorphological history; associated with a change of course by the Nepean Hawkesbury River between Penrith and Windsor, subsidence of the Sydney Basin at Richmond (Branagan & Packham, 2000; Hall, 1926; Jensen, 1912), a lake formation (Hall, 1926) and Recent depositions of the river’s current course (Walker, 1960). First described as red, yellow and white sand and clay soils (Musson, 1907) with areas defined on an undated map, HAC soils are located within the Clarendon Formation, one of six alluvial formations occurring between Penrith and Windsor (Walker & Hawkins, 1957). Walker (1960) mapped these regional soils as associations and series in his survey of the County of Cumberland Sydney Basin, a survey that locates the entire site of UWS Hawkesbury campus within soils of the Elderslie association. Walker mapped this site as the HAC farm (unpublished, 1954) from eleven types; consisting of five loams, five sands, and a clay. No record of an accompanying report has to date been found in support of the detail contained in the soils survey of 1954. Seven of the eleven HAC mapped types have the same name as published series profiles Londonderry loam, Clarendon Sand, Yarramundi loam and Rickamundi loam, Glenlee loam, Austral loam, and Moore Sand (Walker, 1960). The remaining four, Blackendon Sand, Rickaby Sand, Yarraby Sand, and Glebe Clay are specific to the UWS Hawkesbury campus (Table 1) and thus have no documented profile description.

To utilize the published series profiles for current soils research comparison, the first step was to redevelop key descriptors relevant to mapped types. The second step for current soils research comparison was to validate site soils against their predicted mapped type, by cross-referencing to profile characteristics from the generic series descriptions. Combining the two strategies, redeveloping key descriptors for soil type identification and validating map locations, relies on several assumptions. (1) That a descriptive key for classing the eleven mapped soil types can be identified, (2) that soil types were accurately mapped and (3), methodological differences do not influence the level of accuracy necessary to cross reference between historic and current data.

Walker’s County of Cumberland survey (Walker, 1960) provided details of analytical laboratory methods but not field techniques and classification methods. Knowledge of how the soil classifications were determined means that differences between final classifications can be better understood. The clue to Walker’s survey methods was the note that series were classed using the techniques of Butler (1953). The book “Soil Classification for Soil Survey” (Butler, 1980) details the theory of categorizing soil descriptors for preparing a soil map and is the method most likely undertaken by Walker for his 1954 survey of the HAC farm. This methodology of observational soil grouping, statistical confirmation of preliminary groups, developing a classification system based on key descriptors, and then objectively identifying field profiles against the descriptive key and mapping these locations, provides the theoretical basis for utilizing the HAC soils map to inform current investigations.

The primary identifiers for key descriptors of the seven soil series profiles recorded by Walker (1960) (Appendix A) are textural class, colour, and pH. Texture as the descriptor for A-horizon is included in the soil name. Soil colour groupings are the red, yellow podsolics and solodic soils. Difference within these groups is the changing soil pH reaction with horizon depth. These characteristics provide a basis for re-developing a surrogate soil-type key for cross-referencing current field samples to mapped HAC soil types.

Since the 1970s many areas of UWS Hawkesbury campus have a history of effluent irrigation. Theoretically, soil pH will increase with secondary treated municipal sewage irrigation (Falkiner & Smith, 1997). Thus validation of mapped soil delineations (as predicted by Walker’s 1954 map), using horizonal change of pH as a key descriptor, would be best assessed from soils never irrigated with effluent.

The aim of this paper is to establish a process of validating mapped soil types; cross referenced to generic series profile information, so that the extensive studies already undertaken for the region and UWS Hawkesbury can be used to support current investigations.

Method

Soil series from Walker (1960) were tabulated (Appendix A) to summarise the soil properties pH, horizon and depth, texture and colour, and presence of concretions. A soil character key was re-developed based on descriptors A-horizon topsoil, soil name, horizon texture profile, colour, profile pH, and B-horizon texture. Soil colours were grouped as reds and reddish brown; grey browns and browns; yellow, yellow brown and brown. An increase, decrease, or no change with soil depth was noted for profile pH reaction.

Predicted mapped soil types were compared to soils sampled from non-irrigated areas of the UWS Hawkesbury campus from an area to the west, unambiguously located centrally within the mapped area of Clarendon Sand (sites S1 and 10A) and an area to the east, potentially comprising five different loam soil types (a 17 ha field known as Clarendon paddock number 5). In the year 2000 soils were collected for a spatial variation study from both these areas. The sample sites within paddock areas were randomly located. Soil cores were taken at intervals of 0, 0.05, 0.15, 0.45, 1.35, 4.05 metres from transects configured as a triangle orientated north south. Core locations and field topography were surveyed using electronic distance measuring equipment and data were overlaid with the soils map as a digital image. Profile information was obtained using 5 cm diameter soil cores drilled to 1m. Samples from 0 – 10 cm and 60 – 70 cm depths were removed from each profile, air-dried in the laboratory, then ground using a mortar and pestle to pass through a 2 mm square mesh sieve (Endecotts Ltd, London). Soil texture was determined mechanically with the hydrometer method (Standards Australia, 1994) from 25 g soil. Percentages of clay and sand were derived from readings at 5 min and 90 min against a temperature corrected 0.002 % Calgon blank solution. Texture classes of the USDA textural triangle (Soil Survey Staff, 1975) were determined for each sample calculated using a computer-based algorithm as programmed by (Gerakis & Baer, 1999) accessed via the World Wide Web at http://nowlin.css.msu.edu/software/triangle_form.html. Values of pH were determined for 1:5 soil:0.01 M CaCl₂ extracts (Method 4A2 (Rayment & Higginson, 1992), from analytical duplicates of 5.000 g + 0.006 g air-dry soil and averaged for each soil plot. Soil colour was determined under good light using the Munsell^® colour chart system 1994 edition (Macbeth division of Kollmorgen Instruments Corporation, New Windsor, USA) from a small portion of air-dried sieved soil moistened with distilled water to below ‘plastic limit’ and compared to the colour chart. Hue, chroma and value were converted to a soil colour description.

Results

Soil profile characteristics for HAC soil types were collated as a descriptive HAC Soil Types 2004 Key (Table 1).

Table 1 HAC Soil Types 2004 Key UWS Hawkesbury Campus Soils

A horizon Topsoil	Soil Name	Publication Source	Key Descriptors
			Horizon Texture Profile	Red, Reddish Browns	Grey Browns, Brown	Yellow, Yellow Browns, Brown	Profile pH with depth	Light Textured B Horizon
Sand	Moore Sand	Walker (1960)	Sand		Moore		No change
	Clarendon Sand	Walker (1960)1	Sand over clayey sand			Clarendon	Decrease1
	Rickaby Sand	This paper1	Sandy clay	Rickaby			Increase1
	Blackendon Sand	Not documented	-
	Yarraby Sand	Not documented	-				-	-
Loam	Austral Loam	Walker (1960)	Medium over heavy over fine sandy clay			Austral	Increase
	Glenlee Loam	Walker (1960)	Medium over light clay			Glenlee	No change
	Londonderry Loam	Walker (1960)	Medium over heavy over sandy clay			Londonderry	Decrease
	Rickamundi Loam	Walker (1960)1	Medium over heavy clay		Rickamundi		Decrease1
	Yarramundi Loam	Walker (1960)	Light medium clay becoming loam	Yarramundi			Increase	Yarramundi
Clay	Glebe Clay	Not documented	-				-	-

¹ Data obtained from results presented in this paper

Clarendon Sand

The published profile information of Clarendon Sand was that the soils were coarse sand brownish grey to a depth of 75 cm overlying light grey and yellow brown sandy clay (Appendix A). The field investigation found that topsoil depth was 20 cm with a decrease of organic colour at 45 cm. Soils were coarse loamy sand brownish to light greys including dark grey brown, brown, dark brown, covering a Munsell^® colour range between 10YR: 4/2, 4/3, 3/2, 3/3, 5/2, 5/3 and 7.5YR: 4/2, 5/3. A saturated zone was evident above 60 – 70 cm. At 60 – 70cm soils were sand or loamy sand. Below the sand at 80 – 90 cm was sand clay, brown, yellowish brown, (10YR 4/3, 5/4 and 7.5YR 5/4) with some grey mottle. In some profiles at 75 cm, layers of ironstone nodules, gravel, or manganese were evident; saturated zones were at 70 – 80 cm; the sand horizon was a deeper profile than 1m. Soil pH increased slightly at depth. Reaction range in 0 - 10cm soils at site S1 was pH 4.24 – 4.55 and in the 60 – 70 cm zone 4.5 – 4.91; at site 10A soil reaction was pH 4.6 – 4.71 in 0 - 10cm soils and 4.87 – 5.17 at 60 – 70 cm. In these soils pH reaction decreased with depth. Profile comparison between the published series and actual field data with respect to mapped soil location confirms that published soil series profiles were consistent with mapped soil types.

Map Profile Cross Referencing

Sample locations within the Clarendon 5 paddock corresponded to mapped soil types Londonderry, Yarramundi, Rickamundi loams, and Rickaby Sand (Fig. 1). Cross-comparisons between series profile characteristics as summarised with the HAC Soil Types 2004 Key and field data revealed a difference was associated with the mapped location and key descriptors for site 4 mapped as Yarramundi loam (Table 2). It is therefore necessary to validate map predicted soil types because differences between actual and the predicted types can occur.

Key properties of the Austral loam (Table 1) are an increasing pH reaction with depth and yellow 2.5Y subsoil colour. Yarramundi loam is a red podsolic soil. Austral loam as a yellow podsolic comprises colours from olive to olive to yellow brown. As both Yarramundi loam and Austral loam have profile pH reactions that increase with horizon depth their difference is determined with colour. However, B-horizon texture differences are also evident (Appendix A). The Austral loam texture profile is a medium becoming heavy clay compared to the Yarramundi loam as a light and medium clay becoming loam. For the current data presented, the level of description provided by the USDA textural classification did not allow for a sub-classification to light, medium or heavy clay between these two loams. This issue highlights the importance of using the detail of a clay class for soils classification and that data obtained from the USDA texture triangle for this purpose was not adequate.

For all soils surveys regional and site topography is one of the principles of soil distribution (P.H. Walker personal comm. 17 July, 2004). Within the Clarendon 5 paddock there was a topographic relationship for soil types. Rickaby Sand occurred on the higher elevated contours and Rickamundi loam was present along the lowest of the across slope contour (Fig. 1). Therefore the cross-referencing of profile types has as its basis one of the fundamental principles of soil survey for soil classification, that of regional topographic character.

Rickaby Sand

Rickaby Sand was not a series profile described by Walker (1960), but a local UWS Hawkesbury soil type. Rickaby Sand is characterised by its light textured 0 – 10 cm sandy loam, loam or clay loam surface soil of brown and dark brown 7.5YR 3/2, 4/2 colour (Table 2). Soil pH was slightly lower in surface soils than at 60 – 70 cm by approximately one pH unit. Subsoil B horizons were sandy clay, reddish brown, brown, yellowish red, strong brown with Munsell^® codes that range across 7.5YR 4/3, 4/4, 4/6/ 5/4, 5/6 to 5YR, 4/6, 4/8, 5/6, 5/8 (Table 2). These colour ranges of red and red browns are consistent with a red podsolic group.

Colour Confirmation

Field soil colour variations of subsoils by Munsell^® colour descriptions indicate that the validated soil types have dominant colour ranges. Colours for Londonderry loam subsoil were dark yellowish brown to strong brown to yellowish red (10 YR 4/6 to 7.5YR 5/6 to 5YR 4/6). Austral loam range was dark greyish brown to light olive brown (2.5Y 4/2 to 2.5Y 5/4). Rickamundi loam was olive brown to dark greyish brown to grey (2.5Y 4/4 to 10YR 4/2 to 5YR 5/1). Rickaby Sandy was yellow red to strong brown (5YR 4/6 to 7.5YR 5/6) (Table 2).

Table 2 Soil Profile Characteristics Clarendon 5 Paddock

Clarendon 5	Transect	Depth	Soil Texture	Munsell Soil Colour Present Along 5 m Transect	pH CaCl	Predicted Type
Site 1	x Sept	0 - 10	Clay loam, clay	Dark greyish brown 10YR 4/2, Brown 7.5YR 4/2	4.24	Rickamundi
	y				4.35
	x	60 - 70	Clay	Olive brown 2.5Y 4/4, Dark greyish brown 2.5Y 4/2, 10YR 4/2	3.88
	y				3.82
Site 7	x Dec	0 - 10	Clay loam, clay	Brown 7.5YR 4/2	4.21	Rickamundi
	y				4.42
	x	60 - 70	Clay	Grey 5YR 5/1	3.95
	y				3.79
Site 2	x Sept	0 - 10	Sandy clay loam, clay loam	Brown 7.5YR 4/2	4.56	Londonderry
	y				4.41
	x	60 - 70	Clay	Yellowish red 5YR 4/6	5.02
	y				4.18
Site 8	x Dec	0 - 10	Loam, sandy clay loam, loam	Brown 7.5YR 4/2	4.35	Londonderry
	y				4.46
	x	60 - 70	Sandy clay, clay	Strong brown 7.5YR 5/6, Brown 7.5YR 5/4, Yellowish red 5YR 4/6, 5YR 4/8, 5YR 5/8, 5YR 5/6	4.89
	y				4.29
Site 3	x Sept	0 - 10	Sandy loam, sandy clay loam,	Brown 7.5YR 4/2, Dark brown 7.5YR 3/2	4.39	Rickaby Sand
	y				4.85
	x	60 - 70	Sandy clay	Reddish brown 7.5YR 5/4, 5YR 4/4, Yellowish red 7.5YR 5/6, 5YR 4/6, 5YR 4/8, Strong Brown 7.5YR 4/6	4.51
	y				5.83
Site 9	x Dec	0 - 10	Sandy loam	Brown 7.5YR 4/2	4.46	Rickaby Sand
	y				4.80
	x	60 - 70	Sandy Clay, sandy clay loam, clay	Yellowish red 5YR 4/8, 5YR 4/6, Reddish brown 5YR 4/4, 5YR 4/6, 5YR 5/6, 5YR 4/3	5.04
	y				5.48
Site 4	x Sept	0 - 10	Clay loam, clay	Dark greyish brown 2.5YR 4/2	4.30	Austral loam1
	y				4.42
	x	60 - 70	Clay, sandy clay loam	Light olive brown 2.5Y 5/4, Olive brown 2.5Y 4/4, Dark greyish brown 2.5Y 4/2	6.07
	y				6.24
Site 5	x Sept	0 - 10	Clay loam, sandy clay loam, clay	Dark greyish brown 10YR 4/2, Dark brown 10YR 3/3.	4.56	Londonderry
	y				4.51
	x	60 - 70	Clay, clay loam	Strong Brown 7.5YR 5/6, 7.5YR 4/6	4.38
	y			Brownish yellow 10YR 6/6, Dark yellowish brown 10YR 4/6, Yellowish brown 10YR 5/6	4.44
Site 6	x Sept	0 - 10	Clay loam, clay	Dark Brown 10YR 3/3, Dark greyish brown 10YR 4/2, Brown 10YR 5/3	4.39	Londonderry
	y				4.53
	x	60 - 70	Clay	Yellowish brown 10YR 5/8, 10YR 5/6, 10YR 5/8	4.48
	y				4.46

Figure 1 Soil Type and Topography Part Clarendon 5 Paddock.

Conclusion

For UWS Hawkesbury soils, although there were profile descriptions, samples taken and a detailed scheme of relationships between soil types prepared (P.H. Walker, personal com. 17 July, 2004), the only available data that remains is the soil map. In order to utilize the detail provided by this map and the published series profiles of the same name, it was first necessary to understand the soils survey process, redevelop a soil-type key interpreted from the data available and then cross-reference field data for soil type identification. This process of validation highlighted that differences between field types and map predicted types do occur. It also confirmed a high level of similarity between the historic published series profile and the UWS Hawkesbury soil type Clarendon Sand. For soils with undocumented profiles such as the Rickaby Sand, current field data added to the HAC Soil Types 2004 Key (Table 1) now provides a basis to future map validations.

Although there were methodological differences (an issue in part with relevance to the use of hand methods for field texture and soil colour classifications as opposed to an analytical process of comparative laboratory testing), the level of accuracy necessary for most cross-comparisons with the analysis methods of the spatial variation investigation was adequate. For soil colour, a 2 mm ground sample can only ever be an average method that loses the detail of aggregate structure and mottle character. With respect to grouping reds, yellows and brown greys, the soil colour descriptions converted from Munsell^® codes were definitive and consistent with the visual descriptions by Walker (Appendix A). At the regional scale Munsell^® colour hues were consistent with data relevant to the UWS Hawkesbury site for soil landscape units Upper Castlereagh (Up) (2.5Y - 10YR, 10YR - 7.5YR) and Berkshire Park (Bp) (10YR - 7.5YR - 5YR) (Bannerman & Hazelton, 1990). For soil pH, a profile typical character for Clarendon Sand and Rickamundi loam was unknown (Appendix A). This current investigation identified that their profile pH reactions decreased with increasing depth (Table 2). Measured in a calcium chloride solution for this current investigation, the pH reactions for Walker’s series profiles were in water. This methodological difference did not limit classification because it was horizonal change with depth and not the actual pH value that was the key descriptor. It was however, a lack of information about clay class (heavy, medium or light) with the USDA textural classifications that posed the greatest limitation to cross-comparison between the hand textured descriptions by Walker (1960). Whilst it is possible that the validation of soil criteria between Austral and Yarramundi loams is colour based, it is evident for loam soils that texture values should indicate a clay class.

Current theory of soil classification requires that Australian soils surveys be reported with an Australian Soil Classification (ASC) according to (Isbell, 2002). To maximize the knowledge contained within Walker’s HAC soils map, future investigations not only need to include as a component of soil type classification at minimum the primary descriptors texture, colour and pH; the cross-referencing of soil properties between an historic dataset and current data needs also to include the scope for updating to a current classification method.

References

Bannerman S, Hazelton PA (1990) Soil Landscapes of the Penrith 1:100 000 Sheet. (Soil Conservation Service of New South Wales: Sydney, Australia)

Branagan DK, Packham GH (2000) Field Geology of New South Wales, 3^rd edition (New South Wales Department of Mineral Resources: Sydney)

Butler BE (1953) Structure and consistence. (Australian Conference in Soil Science: Adelaide, Australia)

Butler BE (1980) Soil Classification for Soil Survey (Clarendon Press: Oxford)

Falkiner RA, Smith CJ (1997) Changes in soil chemistry in effluent-irrigated Pinus radiata and Eucalyptus grandis plantations. Australian Journal Soil Research 35,131-147.

Gerakis A, Baer B (1999) A computer program for soil textural classification. Journal of Soil Science Society of America 63,807-808.

Hall LD (1926) The physiography and geography of the Hawkesbury River between Windsor and Wiseman's Ferry. Proceedings Linnean Society of New South Wales 60.

Isbell RF (2002) The Australian Soil Classification. (Revised Edition) (CSIRO Publishing: Melbourne, Australia)

Jensen HI (1912) The river gravels between Penrith and Windsor. Royal Society of New South Wales - Journal and Proceedings 45,249-257.

Musson CT (1907) Our College And Its History. II - Site and Surroundings. The Hawkesbury Agricultural College Journal pp 220-223.

Rayment GE, Higginson FR (1992) Australian Land Survey Handbook - Australian Laboratory Handbook of Soil and Water Chemical Methods. (Commonwealth of Australia, Inkata Press, Reed Books International, Australia, Pty. Ltd.: Sydney, Australia)

Soil Survey Staff (1975) Soil taxonomy: A basic system of soil classification for making and interpreting soil surveys. (U.S. Government Printing Office: Washington, DC)

Standards Australia (1994) Determination of the particle size distribution of a soil-Standard method of fine analysis using a hydrometer, AS 1289.3.6.3 – 1994, Sydney, Australia.

Walker PH, Hawkins CA (1957) A study of river terraces and soil development of the Nepean River NSW. Journal and Proceedings Royal Society of New South Wales 91,67-84.

Walker PH (1960) A Soil Survey of the County of Cumberland, Sydney Region New South Wales. Soil Survey Unit Bulletin No 2 (New South Wales Department of Agriculture Chemist’s Branch: Sydney Australia)

Appendix A Soil Series Summary Profile Character (Walker, 1960) UWS Hawkesbury Campus

pH	Depth cm	Texture Description	Colour Description	Horizon	Concretions	Amount
1. Austral Loam: Yellow Podsolic Soil- Parent material colluvium of depressions between shale hills (Wianamatta shale)
5.3	0 – 5	Loam	Grey brown	A1	Ironstone	low
5.3	5 – 17.5	Loam	Light grey	A2	Ironstone	low
5.5	17.5 – 32.5	Medium clay	Dark grey	B1	Ironstone	low
6.4	32.5 - 1050	Heavy clay	Olive to yellow brown	B2	Ironstone	low to moderate
6.8	1050 – 1300	Fine sandy clay	Red brown and yellow brown	B2	Ironstone	moderate
2. Clarendon Sand: Parent material is coarse sandy alluvium of the Nepean River (Pleistocene)
-	0 – 22.5	Coarse sand	Brownish grey	A1	No	-
-	22.5 - 55	Coarse Sand	Light grey	A2	No	-
-	55 - 75	Coarse sandy loam	Light grey and pink mottled	B1	No	-
-	75 - 210	Clayey sand	Grey and yellow brown	B2	No	-
-	210 - > 240	Clayey sand	Yellow brown, slight grey mottling	B2	No	-
3. Glenlee Loam - Parent material is Upper Wianamatta Group, the soil formed in depressions calcareous sandstone
5.3	0 – 6.25	Fine sandy loam	Drab brown, dark grey mottling	A1	No	-
5.6	6.25 – 15	Fine sandy loam	Light brown, light grey mottling	A2	No	-
5.7	15 – 20	Medium to heavy clay	Brown, yellowish brown and grey mottling	B1	No	-
5.9	20 – 97.5	Light to medium clay	Yellow grey,; red brown and yellow brown mottling	B2	Lime	trace
5.2	97.5 – 140	Light clay	Yellow brown, yellow grey and red brown mottle, black inclusions	trace
4. Londonderry loam: Parent material is fine flood plain alluvium (Pleistocene)
5.5	0 – 10	Fine sandy loam	Dark grey brown	A1	Ironstone	low
5.4	10 - 22.5	Find sandy loam	Very light yellow grey	A2	Ironstone	heavy
5.1	22.5 - 25	Medium clay	Brownish yellow, brownish red mottled	B1	Ironstone	moderate
5	25 - 60	Heavy clay	Brownish yellow, brownish red mottled	B2	Ironstone	trace
4.4	60 - 75	Heavy clay	Light grey , yellow brown mottled	B2	-
4.2	75 - 105	Fine sandy clay	Light grey, yellow brown mottled	C	Cementation
5. Moore Sand: Parent material is beach sand - windblown dune material
4.5	0 – 30	Sand	Grey	A1	No	-
4.8	30 - 60	Sand	Light grey (ashy)	A2	No	-
4.6	60 – 67.5	Sand	Dark brown, yellowish brown mottling (coffee coloured)	-
4.9	67 - 80	Sand	Yellow brown, light yellow mottling	B2	Cementation	pockets
4.8	80 - 205	Sand	Very light yellow, light grey mottling	C	No	-
6. Rickamundi loam - Solodic soil - Parent material river alluvium (Pleistocene)
-	0 - 5	Fine sandy loam	Dark grey brown	A	No	-
-	5 - 30	Medium clay	Dark greyish brown	A	No	-
-	30 - 75	Heavy clay	Dark grey brown	B	No	-
-	75 - 120	Medium to heavy clay	Yellow brown, dark grey brown mottling	BC	No	-
7. Yarramundi Loam (also sand): Red Podsolic - Parent material is Nepean River alluvium (Pleistocene)
5.2	0 - 12.5	Fine sandy loam	Grey brown	A1	No	-
5.5	12.5 - 47.5	Clay loam to light clay	Light yellowish brown	A2	Ironstone	medium - heavy
6.3	47.5 - 80	Light clay	Light brown, mottled and light red brown	B1	Ironstone	slight
6.6	80 - 97.5	Medium clay	Light brown, red brown mottling	B2	No	-
6.8	97.5 - 180	Fine sandy clay, becoming clay loam and loam	Light reddish brown	B2C	No	-