Geological Impacts on Early Settlement on the South Coast of NSW; Lessons to be Learned.

BRIAN LEES¹, ANTHONY GREENHALGH², DAVID MOORE³, BRONWEN WICKS⁴ and DAVID PRICE^{5 1} Australian National University ²Dubbo City Council ³Geo Mapping Technologies ⁴University of Wollongong ⁵Environment Australia

Abstract

The planning of settlement location on the NSW South Coast appears to have been underpinned by a simple model. An association of igneous rock with an adjacent potential port site is common to all of Florance’s settlements. The subsequent development history of two of these, Milton and Kioloa, is so different that an examination of the Kioloa site was undertaken. The soils around Kioloa have little relationship to the local Termeil Essexite and are, instead, derived from Pleistocene estuarine infill. In review, it is clear that the simple model, like all such models, was only capable of predicting tendencies and thus some predictions were wrong. The spread of geographic technologies is currently enabling simple models of this type in land planning, and the long debate between environmental determinists and possibilists, which led to the moderation of both positions, needs to be understood by those involved. The use and application of geographic technology, without an understanding of geographic theory and the history of geographic thought, can easily lead, through economic impacts, to quite adverse and unintended social impacts of the type described in the paper.

Introduction

Looking at the NSW South Coast today, one is struck by the architectural legacies of the early settlers (Fig., 1). At Kiama and Milton, attractive and substantial Victorian public buildings dominate the centres of what were clearly prosperous small towns. Large, well built hotels and emporia surround them and the elegant private houses of that time still grace their streetscapes.

However, not all settlement of that era survives so well. The scattered settlements of the cedar getters have mostly vanished and, of those towns laid out by the early surveyors, not all have prospered.

Fig., 1: The Victorian-era courthouse at Milton. One of many substantial buildings in the town.

Murramarang-Kioloa (Fig., 2 & 3), one of those locations proposed for early settlement, is now growing as a holiday and retirement settlement, but conspicuously lacks the evidence of past prosperity so apparent in nearby Milton . Indeed, the contrast between Murramarang-Kioloa, and Milton and Kiama is so great as to make one wonder why this should be so, given their similar early histories.

Fig 2: The Avenue; the ‘main street’ of the original Kioloa village in 1976.

Figure 3: The location of Kioloa on the South Coast of NSW. Murramarang is the area between Kioloa and Bawley point.

Thomas Florance surveyed the South Coast of NSW in the late 1820’s. Surveyors

of that time were tasked with land evaluation, the identification of possible town sites and good harbours. Florance surveyed the current Milton-Ulladulla region in late 1827, and reached the Murramarang-Kioloa area in early 1828. Another surveyor, Robert Hoddle, also surveyed the Milton-Ulladulla region in 1828.

These were certainly not the first Europeans in either area. Thomas Davidson, Hamilton Hume and Alexander Berry are recorded as having ascended Pigeon House Mountain in January 1821 (Berry, 1825; in Cambage, 1916) and various parties are recorded as having passed through the area, or sailed along the coast, during the previous fifty years. However, in the expansion outwards from Sydney it had been decided that the area to the North Sydney offered easier access and it was not until the better land to the north was taken that attention turned to the south coast.

The early experiences of the colony had made people very sensitive to the importance of soil type for agricultural success. Most of the soils in the Sydney Basin are very poor and the discovery of better soils associated with an exposure of igneous rocks, at Prospect Hill, by Phillip in 1788 was an important part of the history of the colony. Much of the exploration carried out by the military during the late 18^th-century was concerned with land evaluation, the search for good agricultural soils. This emphasis on land evaluation continued as military exploration gave way to civilian surveys.

Florance and the other surveyors of that time were good at their jobs. They were well aware of the relationship between igneous rocks and fertile soil. They identified the basalts at Kiama, the Monzanite at Milton, and the Essexite at Murramarang as indicating fertile soils. They also identified associated port locations. The harbours at Ulladulla and Kiama are now fine, and seem sensible choices but, in their undeveloped states, were initially as dangerous as Kioloa is today. The undeveloped harbour at Kioloa, dangerously situated in the lee of Belowla Island, only makes sense as part of a ‘package’ where good soil, good water, and a potentially good harbour were the preconditions for the establishment of a settlement. The state map of the time (1862) makes it clear that high hopes were held for settlement at Kioloa with it being shown as the only port site reserved on the South Coast south of Ulladulla.

Settlers were already moving in to these areas in search of cedar and these surveyors’ reports seem to have largely focussed later settlement. Florance himself was one of the first settlers in the Milton area and his father-in-law, Thomas Kendall, settled his family in the area. As at Kiama and Braidwood, the forest overlying the igneous rock was rapidly cleared and for more than 150 years afterwards the limit of clearing almost exactly coincided with the boundary between the igneous and sedimentary rocks. Only in recent decades is this boundary breaking down. A topographic map of the South Coast from 1942 (A.H.Q.) shows cleared and uncleared areas quite explicitly, and a brief comparison of this map to the geology map still shows a very high level of spatial correlation between the two (Fig., 4).

A similar pattern of clearing took place in the Murramarang-Kioloa region. The north-south extent of the blocks selected by Stephen, Morris and Carr from 1828 almost exactly matches the north-south extent of the Essexite exposure. Although the east-west extent of their selections extends much further inland than the Essexite, their subsequent clearing of forest does not.

The soils around Milton were derived from the weathering of the Monzanite, which is very rich in phosphorous and potassium. These highly productive soils meant that, by the end of the 19^th century, Milton had become an important, and prosperous, dairy centre. Local produce was exported through the local port which was established at Ulladulla. Exports at the time included wheat, maize, potatoes, onions, butter, bacon, cheese, honey and poultry (Milton-Ulladulla and District Historical Society, 1988).

Figure 4: A comparison with cleared areas mapped in 1942 and the distribution of igneous rocks at Kiama, Milton and Kioloa. The limits to clearing at Kiama appear to be topographic, but the limits to clearing at Milton and Kioloa are related to the occurrence of igneous rocks or alluvium.

The mid-19^th-century boom and bust of Kiama was well-documented by Jervis (1942) in the journal of the Royal Australian Historical society. Its subsequent recovery to prosperity was, like Milton, based on highly productive agriculture on the rich basalt soils.

Whilst settlement at Murramarang-Kioloa was taking place at the same time as Milton was becoming established, Hamon (1994) notes that there were contemporary references to considerable difficulties in marketing the output of Carr’s Murramarang property. He ascribes this to the obvious problem that the harbour at Kioloa identified by Florance was quite dangerous and exposed. Whilst this is true, it is only part of the problem.

A surveyor (Larmer) had been contracted by Carr to survey and lay out a village at Kioloa. Larmer’s report (1843, in Hamon, 1994) described the proposed village site at Kioloa as scrubby, barren and unproductive. He correctly warned that the harbour site was unprotected and shallow. However, given its use to export timber over the subsequent ninety years, it was not unusable. If there had been the potential for a significant economic surplus from farming in the area, the Kioloa harbour, like many up and down the coast, would have been improved. That it was not, suggests that the potential of the local economy was not promising. The clue lies in his description of the land itself.

There were huts on the beach at Kioloa during Larmer’s survey which were used as small warehouses to stockpile bark and other produce from Murramarang for shipping. Wattle bark was a very low value commodity and they ceased shipping it out when a large consignment brought them only two pounds (Jones, unpub; in Hamon, 1994). The problem seems to be the return on the export rather than an inability to do so. Such a conclusion is supported by comments published in the Town and Country Journal (1870; in Hamon, 1994) that the grass on the Murramarang property was not first class, either in quality and quantity. There was an early realisation amongst the landholders in the area that it was fairly mediocre farming land (Hamon, 1994). In 1910, the owners of the Murramarang property were happy to sell part of it to McKenzie’s for timber getting.

Timber getting was the only industry in the Batemans Bay region at this time. The industry was run as an extractive, rather than sustainable, industry. Mills were highly mobile and as soon as the good timber was extracted from an area, the mills closed and moved to another location. It was therefore not in the interests of the timber companies to establish expensive, substantial, infrastructure.

When McKenzie’s took Kioloa over, William Walker, the manager, built a row of modest wooden houses to house his senior employees and a house for himself on the hill at the end of the row, dominating the settlement. The building style was typical of forestry settlements of the time.

Given the expected life of the mill in the area, this was naturally a modest investment. Nevertheless, it stands out in stark contrast to the untidy collection of shacks which housed the workers employed in the earlier Goodlet and Smith mill at Kioloa. The McKenzie mill operated for only 18 years. When it closed, the mill equipment was dismantled and sold and the Kioloa property was put up for sale.

During the period of operation of the mill the local population of four or five farm families and their employees grew considerably. The mill at Kioloa employed 72 men (Milton-Ulladulla and District Historical Society, 1988) in various roles, and the mill at Bawley Point employed 30. There were enough families with children to justify the establishment of a school at Murramarang. When the mill closed, the population dropped again and the school closed. There was no demand for new housing, and the local economy did not support the upgrading of the existing housing stock.

During the same period, the settlers at Milton had experimented with several forms of agriculture, finally settling on dairy farming. By 1892 a butter factory was established in Milton and in 1896 the Ulladulla dairy factory had opened. Pasture improvement, herd improvement and mechanisation increased profitability and Milton-Ulladulla entered a very prosperous era of consolidation and growth.

The substantial investment in infrastructure and building at Milton-Ulladulla that we can see today reflected the confidence that the community had in the long-term future of the prosperity of the region. This period of prosperity lasted, with a few ups and downs, until the 1970s.

The 150 years of settlement led to the quite striking differences between Milton-Ulladulla and Murramarang-Kioloa that we see today. In the former, a strong and sustainable agriculture based on the rich Monzanite soils underpinned prosperity, growth and a substantial investment in urban infrastructure, in the latter, poor agricultural prospects led to the use of the land for unsustainable extractive industry and a minimal investment in infrastructure.

Florance’s initial survey had recommended both Milton-Ulladulla and Murramarang-Kioloa for settlement, for much the same reasons, but 150 years later their paths had diverged significantly. It’s clear that Florance’s inference that igneous rock was going to be associated with fertile soils had not held true at Murramarang-Kioloa. What was there about Murramarang-Kioloa that misled Florance?

Murramarang-Kioloa

Looking at the cleared areas to the north and south of the east-west ridge on which the original Kioloa village was built, the topography is remarkably suggestive of river terraces. Broad, flat paddocks lying at about 4-6 m above present sea level, and again at about 1-2 m above present sea level, are typical of the Murramarang-Kioloa area. Higher paddocks are formed on rolling country.

Figure 5: The catchment of Butlers Creek, Kioloa, showing sediment types and the location of the cross section, A₁ – A, shown in figure 6.

Murramarang and Kioloa are located on an outcrop of Termeil Essexite, an igneous rock not too dissimilar to the Monzanite of Milton. Essexite is a dark coloured basic intrusive rock primarily composed of plagioclase, hornblende, biotite, and titanaugite with magnetite as a major accessory mineral. Initial weathering releases silts, clays and a magnetite sand. Magnetite has an iron content of 65% (Fe,Mg)Fe₂O_4, often contains variable amounts of titanium oxide, and is chemically weathered very readily in this acid coastal environment. Most of the weathering products of magnetite will be progressively removed in solution in a freely drained environment.

We examined a core taken from the sedimentary terraces at Kioloa to see why, despite the Essexite bedrock, the soils are so poor. The study area consists of a broad shallow embayment adjacent to narrow coastal plan, and rises approximately hundred and 20 m to steeply dissected Permian sandstones in the west (Fig., 5). The catchment is approximately 160 hectares in area and supports a diversity of forest types including Wet and Dry Sclerophyll forest communities. The catchment is drained by Butler’s Creek, a small perennial stream that falls rapidly from the forest catchment to the cleared coastal plain where it meanders through river flats into a coastal lagoon.

The coastal plain consists of a number of sedimentary units. The beach (Avenue Beach) is narrow and rises steeply to the foredune. The dune ridge is a single line of low, narrow dunes covered with a low littoral heath. The dune and beach units comprise the marine barrier. A back-barrier depression runs parallel to the coast behind the barrier. Butlers Lagoon occurs at the southern end of this depression and a shallow, muddy, swamp supporting closed Casuarina glauca forest extends northwards. To the west of the back-barrier depression are two sedimentary terraces. The lower terrace (one to two metres above sea level), through which Butlers Creek flows, covers an area of approximately four hectares. The upper terrace (four to six metres above sea level) covers approximately six hectares, is flat topped and slopes into the lower terrace.

Figure 6: The drilled cross section showing the location of drill and auger holes.

Figure 7: The core from which the samples for TL dating were taken.

The thin coastal sequence overlies an erosion scarp where, at some stage, the paddock sequence has been truncated by wave action. Trenches cut across this contact show that slumping of the erosional scarp underlies a beach deposit of black sand This sits at 1-2m above the modern beach. The black sand, which is composed of the iron mineral magnetite, is derived from the Essexite. At the water table, the black sand has almost completely disappeared due to post depositional weathering leaving an iron stained quartz sand. In the slightly acid environment of this coastal deposit, the magnetite seems to have been almost as easily removed by post depositional weathering as carbonate.

Sitting on the seaward face of the black sand beach, and under the overlying foredune deposit, is a cobble beach. This sits slightly lower in elevation, about 1-1.5m above the modern beach. Seawards of this is a beach deposit of pale yellow quartz sand indistinguishable from the modern beach.

Where the groundwater drains into the lagoon at the rear of this beach-foredune sequence, iron concretions have been found around the roots of some of the vegetation. In form, these are similar to the root calcarenite often found in the profiles of late Holocene dunes. It seems that the iron concretions are similar ephemeral features formed as highly mobile components are weathered from the coastal deposit. Given that the plants around whose roots these concretions have been found are living acacia sp., of fairly recent age, the amount of iron being removed in solution must be relatively high.

The cobble beach consists of well rounded pebbles and cobbles of a mixture of rock types. Eliminating the Essexite, the material closely resembles the well rounded fluvial material which forms a conglomerate in the local Permian sandstones. Most of the catchments of both Butler’s Creek and Prosser’s Gully, which are the adjacent watercourses, lie in Permian sandstone. It seems probable that the cobbles formed the stream bed of the watercourses, having been eroded from the Permian sandstone, and were reworked during sea level rise into this beach deposit. Given the stability of sandstone to weathering in this environment, these cobbles may have been released over some considerable time.

The Core

A north-south cross section of the Kioloa property (Fig., 6) was surveyed. Twenty holes were drilled along the cross-section. This allowed a cross-sectional diagram of sediments to be drawn. The sequence is up to 10 m deep and consists of a shallow coarse basal gravel deposit overlain by several meters of undifferentiated clayey sand, capped by a shallow sandy clay.

A 10 meter core from the upper sedimentary terrace (Fig., 7) was subjected to detailed analysis and dating. The first 2.6 m of the core were extracted intact and stored in poly pipe to maintain its integrity. Below 2.6 m, which correspond to the water table, the consistency of the sediment changed becoming more fluid. As a consequence, intact cores could not be extracted below this point and samples were taken at one meter intervals from the drill tip. This sampling method resulted in the loss of stratigraphic information and uniformly coloured samples were obtained.

In the laboratory the core was dissected lengthwise and divided into stratigraphic sample units corresponding to changes in colour and texture. A detailed description of the core is shown in table 1. Visually, the core was variable with depth. It ranged from a very dark brown A horizon (Munsell 2/2:10YR) through a mixture of yellow-brown (4/4:YR) and reddish-brown (3/4:5YR), to a uniform yellowish-brown (4/4 to 5/6: 10YR) below 2.6 meters. The colour changes suggest an increasingly reducing environment with depth. A pH of 4.5 was consistent throughout the core, indicating a uniformly acid environment. Banding of yellow and red-brown sediments was observed between 0.75 and 1.00 meters. This pattern probably indicates fluctuating water tables. Below one meter, mottling of the red-brown and yellow sediments, interspersed with grey sandy lenses, occurred down to 2.6 m. In samples below 2.6 m no stratigraphy could be identified due to the method of sampling.

Sediment Analysis

Portions of each sample unit were used for sediment analysis, the exact amount being determined to ensure that a sufficient sample of the sand fraction was obtained (30 to 50 g). Samples were soaked, disaggregated and the sand fraction was separated from the silt and clay by decantation (Folk, 1968). Samples were also obtained from the beach fronting the property, the bank of Butlers Creek, and the upper reaches of the Creek and surrounding catchment. These additional samples were processed using the same procedure.

The sand fraction samples were analysed using a settling tube (A Waikato ‘Rapid Sediment Analyser’) and descriptive parameters of the size distribution were calculated (Table 1 and Figs 8 & 9). The results were calibrated using the technique proposed by Bryant et al, (1987) and the difference between the initial results and the calibration was negligible.

A significant feature of the results is a difference between the mean particle size for the core sediments, ranging from 2.08 to 2.72 phi with a mean value of 2.5 phi, and that of the river sediment (1.13 phi) and the beach sand (1.7 phi). The maximum particle size of the core samples is generally greater, or at least equal to those occurring in the source samples. This suggests that although the lower mean particle size of the terrace sediments suggests a low energy regime, erosional and depositional events still occur which were capable of transporting and depositing large particles.

As expected, the beach material is the most well sorted sample. The sand fractions in the core were found to be less well sorted and samples from all of the potential sources, the beach, the catchment, and the upper reaches of the creek.

Interesting trends in the size distributions of particles are apparent from the 2.6 m sample through the bottom of the core. Mean particle size increases, the sediments become progressively more leptokurtic and well sorted with depth through to 8 m. Through to 8 m, the particle size distribution becomes bought more bimodal. From 8 to 10 m, there is an abrupt change. The mean particle size increases greatly and the size distribution becomes less bimodal, whilst remaining well sorted.

The sand grains from each of the sample units were compared with the samples from the beach, the catchment, and the upper reaches of the creek. The shape of the grains was examined under a stereo microscope and the predominant shape noted and the relative proportions of differentially rounded grains assessed subjectively. The observations are displayed in table 1.

The catchment grains were found to be angular to sub-angular, and the beach grains were predominantly sub-rounded, with some sub-angular grains. The core sand fraction was predominantly angular to sub-angular in the first 2.6 m, with a general tendency towards greater proportion of sub-rounded grains with increasing depth.

The sediment analysis is interpreted as suggesting that the lower part of the core, 8-10 m, is a marine environment. As depth decreases, the amount of marine sand also decreases. The upper part of the core is consistent with the alluvial material in the upper catchment. The change from sandy clay to clayey sand at 2.6 m is quite marked.

The wide range of particle sizes in the upper part of the core, with up to 70 percent silt and clay and a relatively poorly sorted sand fraction in comparison to fluvial, aeolian, and beach deposits from the local area, suggest that the material was introduced from a high-energy suspension to an environment of low energy. This is consistent with a fluvial source depositing in an estuarine environment.

Figure 8: Percentage of Sand, Silt and Clay in Core.

Figure 9: Mean sand particle size with depth.

Sample Depth	Mean Grain Size	Max Grain Size	Skew	Kurtosis	Sorting	Shape
0.15	2.46	-2.00	-0.12	1.10	0.86	Angular
0.76	2.52	-0.75	0.01	1.05	0.68	Angular
1.20	2.72	0.50	0.08	0.99	0.61	Angular
1.40	2.66	-1.00	0.01	1.01	0.66	Angular
1.50	2.60	-2.00	0.02	0.97	0.68	Angular
1.96	2.53	0.00	-0.06	1.01	0.69	Angular
2.40	2.51	-1.25	-0.09	0.94	0.76	Angular
2.60	2.60	-2.00	0.01	0.95	0.72	angular- subangular
3.80	2.62	-2.00	0.02	1.03	0.72	angular- subangular
6.00	2.51	-1.75	-0.08	1.14	0.82	angular- subrounded
8.00	2.48	-2.00	-0.24	1.28	0.94	angular- subrounded
9.00	2.39	-2.00	-0.05	0.83	0.87	angular- subrounded
10.00	2.08	-2.00	-0.03	0.87	0.93	angular- subrounded
Beach	1.65	-0.75	0.12	0.87	0.33	subrounded
Creek Flats	1.70	-1.25	0.16	1.02	0.56	angular
Catchment	1.20	-1.25	0.13	1.13	0.59	angular

Table 1: Summary of Core Analysis

Sample	Reference	Sediment Type	TL Age	Depth
W1725	Kioloa A1	Alluvium	92.3+/-11.8ka	1.5m
W1726	Kioloa B1	Alluvium	>122+/-46ka	2.14m
W1727	Kioloa B2	Alluvium	>127+/-23ka	3.5m

Table 2: TL Ages

Diatom Analysis

Samples for diatom analysis were taken on either side of distinct stratigraphic boundaries, and at every meter down the core. These were prepared using the procedure outlined in Batterbee (1986). After the carbonate and organic material had been removed, the samples were centrifuged at 2000 to 200 rpm for 30 seconds to separate the clay sized particles from the rest of the sample. Slides of the sample material were scanned for diatom frustules under a light microscope at 1400 magnification.

No diatoms were observed in the slides. The preparation procedure was repeated on several samples, but diatoms were not found. The slides were then examined for any other biogenic indicators, although the diatom specific preparation could be expected to have destroyed some of the organic material that may have been present. Unidentified phytoliths were present in the 10.0,9.0, 8.0 and 2.6 m examples suggesting terrestrial sentiment input. Pollen identified as Chenopodacea was observed in the 10 m sample. This family is notable for its salt tolerance and is presently associated with coastal environments in the area. Its presence suggests a nearby coastal swampy environment at the time of deposition, and an anoxic deposition environment conducive to the preservation of pollen subsequently.

The lack of diatoms probably reflects post deposition weathering in a moderately acid environment. The brackish environments of estuaries are not considered suitable for the preservation of diatom skeletal material and, though other interpretations are possible, this seems the most probable explanation. No marine or estuarine fossils were found in the core nor, given the pH of the sediments, would one expect them to have survived.

Dating

Samples from the core were collected for TL dating. These were processed at Wollongong using the procedure outlined in detail in Lees et al, (1990; 1992). The ages of the three samples are shown in table 2. The sample taken from the upper sandy clay unit (W1725) gave an age of 92.3+/-11.8ka (at 1.5 m depth), the two samples taken from the lower clayey sand unit gave ages of 122+/-46ka (W1726 at 2.14 m depth) and 127+/-23ka (W1727 at 3.5 m depth). The lower samples were close to TL saturation giving them large error ranges. No surface reference sample was taken.

Discussion

The flat surface of the deposits, and the lack of gravel within the sediments, indicate deposition in aqueous environment. The mixing of marine and terrestrial sands, with a decreasing proportion of marine material through time, and the wide range of particle sizes, suggests that this is estuarine infill. The dates of the lower two samples indicate that this infill is of last interglacial age, or older. Sea level was estimated to be about 3 to 5 m above present sea level between 122 and 128 ka B.P., dropped and then rose briefly to between 0 and 5 m above present sea level between 118 and 120 ka B.P. (Esat et al, 1999). The upper terrace, in all aspects, is consistent with a marine-estuarine-fluvial, fining-upwards sequence of this age.

The lower terrace, whose sediments are of fluvial origin, appears to be a Holocene infilling of a valley cut through the Pleistocene deposits during lower sea levels. Although it was not investigated in detail during this study, the upper 1-2 m of the lower terrace are visible along the stream channel. The sand particles are highly angular and show close similarity to the catchment slope deposits. These dark brown silty sands clearly indicate a fluvial origin for this deposit, at least in its middle to upper reaches. Closer to the coast, the fluvial sands grade into back barrier and modern estuary sands and probably reflect sediment input during a late Holocene high sea level of 1-2 m above present sea level (Baker and Haworth, 1997)

The Pleistocene barrier behind which the upper terrace formed as an estuarine infill appears to have been completely removed. The presence of a surviving Magnetite black sand beach on the erosional contact suggests recession at a fairly recent date. All the evidence suggests that the Magnetite sand weathers almost as rapidly as carbonate sand in this environment.

Conclusion

The deep sandy terraces, with their origins in the mixing of sands from the upper catchment sandstones and the sea, have clay contents varying from less than 2 percent up to as much as 40 percent in some layers. Part of this clay is derived from weathering of shale layers in the Permian deposits and only part from the weathering of Essexite. The minor contribution to nutrient store by the Essexite has been removed in solution fairly rapidly, in this acid environment, leaving fairly poor soils. The capping of Essexite-derived alluvial material, which Florance may have seen, is quite untypical of the material which lies beneath.

The soils of the Milton-Ulladulla area, where they overlie Monzanite, have retained their high nutrient status and continue to be productive. Given the collapse of the dairy industry, both the Milton-Ulladulla and Murramarang-Kioloa regions are now growing as recreational and retirement centres and there is little difference in their relative prosperity. However, the styles of infrastructure developed over the first hundred years of European settlement forever emphasise the difference in their geological history.

If this has echoes of environmental determinism it is because it reflects a failure of environmental determinism in practice. The simple reductionist model implicit in the selection of sites for settlement by Florance is environmental determinism inverted and used as a predictive tool. Like all abstractions of reality it was so generalised that there must, inevitably, be exceptions to it. It was unfortunate for the first European settlers of the Murramarang-Kioloa area that this was an exception to the rule, but the rule/model itself tended to be very powerful and enabled the successful planning of many settlements up and down the coast with only limited resources. This tension between the efficacy of simple models, and the costs of making mistakes as a result of their overgeneralisation, is still present in many of our land use planning discussions today.

It is difficult to fault Florance however, as he was very much a man of his time with a broad interest in natural philosophy. This is demonstrated by the fact that the difficulties which arose during his surveys of the Port Stephens area (Dowd, 1972) did not prevent him returning to Sydney with a substantial collection of geological samples in addition to his plans and surveys. His education at Chichester, where his father was a barrister, was obviously good as it allowed him to join the army as a ‘surveyor-engineer’ and serve in Canada during the American War. The education of many of the other younger officers in that force was almost certainly heavily influenced by the rationalist thinking of the Scottish Enlightenment. Thus his education and experience were such that the environmental determinism of the time was a natural, and very modern, mode of thought. Florance and his colleagues were using environmental determinism as a useful and practical tool some forty years before Marsh’s (1864) seminal volume on the subject, and some hundred years before Griffith-Taylor’s work on the settlement of Australia and Canada (1911, 1917, 1940). It is worth making the point that, although Griffith-Taylor himself claims to be a determinist, and was roundly criticised for it, his argument ‘What Australia teaches the Geographer’ (in chapter XIX of Taylor, 1966) reads more like ‘probabilism’, a more defensible position.

Today’s land use modellers use these same techniques. The basis of many projects, if looked at critically, is fundamentally (environmental) probabilism and in some cases, pure environmental determinism. Whilst it was in part the extension of the environmental determinism paradigm by Griffith-Taylor and others to race that caused such a negative reaction to it in the middle of the last century, many of those using its weaker derivative, environmental probabilism, in its current revival, are unaware of that debate, or its detail. Both environmental determinism and its possibilist opposite, also undergoing a revival currently, are flawed theories and their application leads to error. Many young geographers, now in the workforce, are being equipped with advanced geographic tools and given responsibilities to plan like Florance. They need to be aware that the use and application of geographic technology, without an understanding of geographic theory and the history of geographic thought, can easily lead, through economic impacts, to quite adverse and unintended social impacts of the type described above. There is a need to heed the lessons implicit in the striking contrast between the inherited infrastructures of Milton and Kioloa and the planned intent. Similarly, those who argue that environmental outcomes can be chosen and negotiated, need to revisit the arguments against the possibilist theory, but that must remain a topic for another paper.

References

A.H.Q. Cartographic Company, 1942: Ulladulla, N.S.W. 1:253,000 sheet. Revised Second addition. I 56/13 Zone 8.

A.H.Q. Cartographic Company, 1942: Wollongong, N.S.W. 1:253,000 sheet. Revised Second addition. I 56/9 Zone 8.

Baker, R.G.V. and Haworth, R.J., 1997: Further evidence from relict shellcrust sequences for a late Holocene higher sea-level for eastern Australia. Marine Geology 141: 1-9.

Battarbee, R.W., 1986: Diatom analysis, in Berglund, B.E. (ed) Handbook of Holocene Palaeoecology and Palaeohydrology, John Wiley & Sons, New York, 527-570.

Cambage, R.H., 1916: Captain Cook’s Pigeon House. Samuel Lee, Sydney.

Dowd, B.T., 1972: ‘Thomas Florance’, 1783-1867. Journal of the Royal Australian Historical Society 58(2) : 89-100.

Esat TM, McCulloch MT, Chappell J, Pillans B, Omura A, 1999: Rapid fluctuations in sea level recorded at Huon Peninsula during the penultimate deglaciation. SCIENCE, 283 (5399): 197-201.

Folk, R.L., 1968: The petrology of sedimentary rocks. Hemphill’s, Austin, Texas, p170.

Griffith-Taylor, T. 1911: Australia in its physiographic and economic aspects. Oxford, Clarendon Press, 256p.

Griffith-Taylor, T. 1917: Canada : a study of cool continental environments and their effect on British and French settlement. London, Methuen, 524p.

Griffith-Taylor, T. 1940 (1966): Australia : a study of warm environments and their effect on British settlement. London : Methuen, 489p.

Hamon, B. V. 1994: They came to Murramarang. Centre for Resource and Environmental Studies and the Edith and Joy London Foundation, Australian National University, Canberra, p137.

Jervis, J., 1942: Illawarra: A century of change. Journal of the Royal Australian Historical Society, 18(2).

Lees, B.G., Head, J. & Lu Yanchow, 1990: Reconnaissance Thermoluminescence dating of northern Australian coastal dune systems. Quaternary Research 34, 169-185.

Lees, B.G., Price, D. & Lu Yanchow. 1992: Thermoluminescence dating of dunes at Cape St. Lambert, East Kimberleys, northwestern Australia. Marine Geology 106,131-139.

Marsh, G.P. 1864: Man and Nature or Physical Geography as Modified by Human Action. New York: Charles Scribner.

Milton-Ulladulla and District Historical Society, 1988. Nulladulla. Milton-Ulladulla and District Historical Society, Milton, N.S.W.