Discussions on the history and historiography of Australia's New England

Monday, January 22, 2018

Implications for New England of the latest analysis on the impact of sea level change on Aboriginal Australia - a note

Regular commenter JohnB pointed me to a new paper on historic sea level changes in Australia:  Alan N. Williams, Sean Ulm, Tom Sapienzab, Stephen Lewis and Chris S.M. Turneya, Sea-level change and demography during the last glacial termination and early Holocene across the Australian continent, Quaternary Science Reviews Volume 182, 15 February 2018, Pages 144–154, published on line 12 January 2018.

This third note on understanding the impact of sea level change in Aboriginal New England takes this paper as an entry point for a broader discussion focused on the impact of sea level and associated climatic changes on New England Aboriginal life.

Please note that I have not been able to access the full paper at this point for cost reasons. That will have to wait until I can find a library that will give me access. This is an especial problem for their population modelling.

The Williams et al Paper

Map of Australia by Sean Ulm showing sea-level change and archaeological sites for selected periods between 35,000 and 8,000 years ago. PMSL=Present Mean Sea Level. Note the apparently small shifts in Northern NSW relative to some other areas. . 

The on-line abstract of the paper summarises some of the paper in this way:
  • Investigation of scale, pace and human impacts of post-glacial sea-level change.
  • Presents continental-scale consensus sea-level curve for Sahul between 35-8 ka.
  • Demonstrates some 2.12 million km2 (∼21.6%) of land lost, notably during MWP-1a.
  • Coastlines changed on average by 139 km, and at a rate of up to ∼23.7 m per year.
  • Populations low, but likely severely disrupted, and led to new configurations.
The abstract of the paper reads:
"Future changes in sea-level are projected to have significant environmental and social impacts, but we have limited understanding of comparable rates of change in the past. Using comprehensive palaeoenvironmental and archaeological datasets, we report the first quantitative model of the timing, spatial extent and pace of sea-level change in the Sahul region between 35-8 ka, and explore its effects on hunter-gatherer populations. Results show that the continental landmass (excluding New Guinea) increased to 9.80 million km2 during the Last Glacial Maximum (LGM), before a reduction of 2.12 million km2 (or ∼21.6%) to the early Holocene (8 ka). Almost 90% of this inundation occurs during and immediately following Meltwater Pulse (MWP) 1a between 14.6 and 8 ka. The location of coastlines changed on average by 139 km between the LGM and early Holocene, with some areas >300 km, and at a rate of up to 23.7 m per year (∼0.6 km land lost every 25-year generation). Spatially, inundation was highly variable, with greatest impacts across the northern half of Australia, while large parts of the east, south and west coastal margins were relatively unaffected. Hunter-gatherer populations remained low throughout (<30,000), but following MWP1a, increasing archaeological use of the landscape, comparable to a four-fold increase in populations, and indicative of large-scale migration away from inundated regions (notably the Bass Strait) are evident. Increasing population density resulting from MWP1a (from 1/655 km2 to 1/71 km2) may be implicated in the development of large and complex societies later in the Holocene. Our data support the hypothesis that late Pleistocene coastal populations were low, with use of coastal resources embedded in broad-ranging foraging strategies, and which would have been severely disrupted in some regions and at some time periods by sea-level change outpacing tolerances of mangals and other near-shore ecological communities" .
The authors provided a summary of the  paper in The Conversation, "Australia’s coastal living is at risk from sea level rise, but it’s happened before",  January 16 2018. The map is from this paper.

Some of their key points can be summarised this way:
  • The potential impacts of these past sea-level changes on Aboriginal populations and societies have long been a subject of speculation by archaeologists and historians.
  • Archaeologists have long recognised that Aboriginal people would have occupied the now-drowned continental shelves surrounding Australia, but opinions have been divided about the nature of occupation and the significance of sea-level rise. Most have suggested that the ancient coasts were little-used or underpopulated in the past.
  • Our data show that Aboriginal populations were severely disrupted by sea-level change in many areas. Perhaps surprisingly the initial decrease in sea level prior to the peak of the last ice age resulted in people largely abandoning the coastline, and heading inland, with a number of archaeological sites within the interior becoming established at this time.
  • With the onset of the massive inundation after the end of the last ice age people evacuated the coasts causing markedly increased population densities across Australia (from around 1 person for every 355 square km 20,000 years ago, to 1 person every 147 square km 10,000 years ago).
  • We argue that this squeezing of people into a landmass 22% smaller – into inland areas that were already occupied – required people to adopt new social, settlement and subsistence strategies. This may have been an important element in the development of the complex geographical and religious landscape that European explorers observed in the 18th and 19th centuries.
  • Following the stabilisation of the sea level after 8,000 years ago, we start to see the onset of intensive technological investment and manipulation of the landscape (such as fish traps and landscape burning)
  • We also see the formation of territories (evident by marking of place through rock art) that continues to propagate up until the present time. All signs of more people trying to survive in less space.
But what is the evidence for New England?

Shape of the Continental Shelf

The effect of sea level change varies depending on the size of sea level shifts, shifts in land height as the land adjusts to things such as increased or reduced ice weight and the configuration of the land itself.

In my first note on the impact of sea level changes, I mentioned that I had only just found a 2010 paper: Alan Jordan, Peter Davies, Tim Ingleton, Edwina Foulsham, Joe Neilson and Tim Pritchard,  Seabed habitat mapping of the continental shelf of NSW, Department of Environment, Climate Change and Water NSW, Sydney 2010. This focuses on the current seabed habitat along the NSW, but it contains some of the best local descriptions that I have seen of the shape of the continental shelf itself in Northern NSW.

I am still working my way through this paper, but I would summarise some of the conclusions in this way. In considering them, remember that we are dealing with variations in sea levels over my study period from perhaps 50-60m below current levels to 120-130m below to perhaps 2m above and then a fall back to current levels. In combination with associated climatic changes, this would have had significant habitat changes for local Aboriginal populations as land was revealed and then submerged.

It will be clear from the Sean Ulm map reproduced above, that the effects of sea level changes along the NSW coast were smaller than in other parts of Sahul, but that doesn’t really provide much local information.
To further out thinking here, the overall width of the continental shelf is defined as the distance to the shelf break. From this point, the seabed falls away quickly. The shelf break occurs at varying heights, leading to variations in depth below current sea levels. The continental shelf slopes down from the current shoreline at varying angles. Higher land on the shelf may appear as reefs or islands.

The NSW study shows considerable regional variations in the depth of the shelf break and the certainty with which it can be defined. The shelf break between Cape Howe in the far south of NSW to Forster on the Mid North Coast varied predominantly between depths of 130–170m and gradually decreased in areas further north to depths of approximately 80–110m off the Northern Rivers.

During the LGM, sea levels dropped to 120-130m below current levels. In Southern New England up to Forster, the shelf break with its sharp decline in height would have remained largely below sea level, although it might have been quite close to the shore. Further north, the shelf break would have been marked by a line of cliffs or at least precipitous decline 15-50m above then sea level.

The study did not report on the 130m contour, focusing instead on the 200m contour deep zone. The distance to the outer edge of the 200-metre contour deep zone revealed the smallest zone widths of 13.4 kilometres and 13.9 kilometres were off Hat Head and Sawtell respectively. The broadest areas occurred off the Stockton Bight with a zone width of up to 46.7 kilometres.

The authors also looked at the extent of the intermediate-depth zone, 25–60m. This is significant because 60m below current is around the sea level when the Aborigines spread across Sahul. This ranged from 1.5 kilometres off Botany Bay, 1.6 kilometres off Shellharbour to 17.5 kilometres just north of Yamba.
The majority of wide (greater than 8 kilometres) intermediate-depth areas were located north of Hat Head, reflecting the overall shallower slope of the shelf in that region. The small number of wide intermediate areas south of Port Jackson mostly reflected the presence of shallow embayments and/or offshore reefs.

Some information was also provided on the shallow-depth zone boundary defined as 25m This was universally narrow along the NSW coast, varying between 0.3 kilometres off Sydney Harbour and Botany Bay to 3.5 kilometres off the Shoalhaven Bight south of Nowra. The width of the immediate-depth zones was more variable than the shallow-depth zones: 70 per cent of the widths of shallow-depth zones were between 0.8 and 1.8 kilometres whereas 70 per cent of widths for intermediate-depth zones were between 5 and 15 kilometres.

In considering this information, we need to remember that the coastal land as we know it today did not exist in the past, but has been formed in part by a constant battle between land and sea, the rivers and creeks depositing silt, the sea eroding the shore. Accepting this and very roughly :
·         We have a fall in height of 25m in the first 0.3-3.5 kilometres.
·         Followed by a further fall of up to 35m in the next 3.5-16 kilometres
·         With a further fall of 60-60m in the remaining distance to the LGM sea line.

If we take the period to the LGM, the additional land at then sea levels was most extensive in the Northern Rivers, narrowed south of Nambucca Heads before widening again. The delta-estuarine-lake formations that we know today and which were such rich environment in Aboriginal times did not exist. Lacking the accumulated sand and silt, land levels along much of the existing coastline would have been lower. The path of the rivers and streams would have been different, less meandering. They are likely to have been faster flowing given the slope of the continental shelf. The pre-LGM coastline was established for a considerable period. I do not know to what extent the rivers were able to create estuarine conditions, given the geography that I have described.

The land areas revealed by the LGM drops in sea levels were greatest in the north and south broken by the strip in the middle where the continental shelf is particularly narrow. With steeper slopes and colder temperatures it seems likely if not certain that the land, as suggested by Sandra Bowdler, became relatively inhospitable.[1] I say likely but not certain because of the degree of local variation, the possibility of micro-environments.

With the arrival of the warm Holocene, the seas rushed back, penetrating deep inland in some spots. You can get a feel for this is you remove all the present sand dunes, all the now accumulated silt and sand in the estuaries and flood plains, with a water level 1-2m higher than present sea levels. Now the streams and rivers carrying higher water volumes in the wetter climate began to push back against the seas, progressively creating the coastline we know today.

Archaeological Dates

We do not know when the Aborigines first settled the area that would later be called Northern NSW, the broader New England.

We know from dating at Warren Cave in Tasmania that the Aborigines had reached Tasmania around 35,000 years ago [2]while dates from Willandra Lakes in South West New South Wales suggest occupation as early as 40 to 41,000 years ago, perhaps even later[3]. The dates we have for New England are more recent.

The Cuddie Springs site near Brewarrina suggests occupation as long ago as 35,000 years BP.[4] However, dates here have been subject to considerable dispute and there presently appears to be no agreement on the issue.[5] Excluding Cuddie Springs, we have a date of greater than 20,200 years BP from a hearth at Glennies Creek 35 kilometres north of Branxton in the Hunter, while a site on a former terrace of Wollombi Brook near Singleton suggested a date range of 18,000-30,000 years BP. At Moffats Swamp near Raymond Terrace, a date of 17,000 years BP was obtained. On the Liverpool Plains, Aboriginal occupation has been dated to at least 19,000 years BP[6]. Further north in South-East Queensland, the Wallen Wallen Creek site shows continuous occupation from about 20,000 years ago.[7]

I note that there are some problems with these dates that I have not yet resolved. However, for present purposes it is the broad pattern that I am interested in.

Note, first, that the dates are late Pleistocene dates from the LGM period. However, there are no coastal zone dates from the Hunter to Wallen Wallen Creek. The entire Tablelands and much of the coast is presently an archaeological blank.

Wallen Wallen Creek lies on North Stradbroke Island, then part of the mainland some distance from the coast. The suggestion is that it was a transit camp for people moving to the coast. Something of the same may be true for Moffats Swamp which lies not far from the present coastline but would then have been some distance from the sea.  

The two other Hunter sites are further inland and could have been refuge areas with access to water and game. Richard Wright suggests that the Liverpool Plains was a relatively fertile area even during the harsh climatic conditions of the LGM, although he notes that there were periods when the climate did deteriorate significantly.[8]

Dates then seem to vanish. Then we have a first date of 9,320+/-160 from Stuart’s Point in the Macleay Valley as the coastal zone begins to stabilise.[9]. From around 4,000 years ago a rush of dates begins. Intensification had begun.


Recognising that the continental shelf in New England is far narrower and therefore the scale effects of sea level change are less, my preliminary analysis does seem broadly supportive of the new paper, recognising that I have only read the summary. However, I do have some questions in my mind.:
To begin with, I am reluctant to accept that there were no Aboriginal people on the humid coastal strip between the Hunter and Southern Queensland. I think that we need a lot more micro or local level information on the changing environment including sea levels.

Then we have the language patterns. As best I understand it, the differences between the northern and southern coastal languages could support re-occupation from north and south. But then we have the problem of the length of time required for the languages to differentiate into the pattern that existed at the time of European occupation. I don’t have an answer for this.

[1] Sandra Bowdler, “The empty coast: Conditions for human occupation in southeast Australia during the late Pleistocene”. In O'Connor S, editor, Altered ecologies: fire, climate and human influence on terrestrial landscapes. Vol. Terra Australis; 32. http://epress.anu.edu.au: ANU E Press. 2010. p. 177-186.
[2] John Mulvaney & Johan Kamminga, Prehistory of Australia, Allen & Unwin, Crows Nest, 1999. P186.
[3] Munvaney & Kamminga, op cit, p197. There is debate about the Wilandra Lakes dates, with some arguing for older dates. Josephine Flood, Archaeology of the Dreamtime, J.B. Publishing, Marleston, revised edition, 2004, p1.
[4] Josephine Flood, Archaeology of the Dreamtime, p189.
[5] The Wikipedia article, Cuddie Springs, provides an interesting discussion on this issue. Accessed 15 April 2009.
[6] Dates drawn from Dominic Steele Consulting Archaeology, Preliminary Aboriginal Archaeological Survey & Assessment,  Proposed Upgrade of State Highway SH23, Shortland to Sandgate New South Wales, December 2005, p17
[7] Ian Walters, Antiquity of Marine Fishing in South-East Queensland, QAR, Vol 9, 1992, pp35-39. P35. Accessed on line 4 April 2009.
[8] Richard Wright, The Future of Australia's Prehistoric Past, Arts: The Journal of the Sydney University Arts Association, Vol 12 (1984). pp 22-34
[9] Graham Knuckey, A shell midden at Clybucca, near Kempsey, New South Wales, Journal, Australian Archaeology Volume 48, 1999 - Issue 1, pp1-11


Johnb said...

Purely subjective thoughts Jim but those people whose Country included the coastal fringe should quite naturally move according to wherever the then current coastline was as they well knew how to exploit it as a resource. If this happened then it would be the next inland peoples who would have their range extend or shrink according to the movement of the coastal people. Unless there was a barrier, Wesrward in this case, then a ripple effect would seem logical. One barrier may well have been climate change up on the Tablelands but then adaptation becomes a force and in their settlement of this Australian continent Aboriginal populations have shown significant powers to adapt across all its environments. Maps detailing authoritvely dated archeological sites have not as yet shown any strong bias to earlier settlement in Australia’s Northern half vis a vis its Southern half which would suggest that the original settlement must have been relatively rapid and widespread. If it was possible to plot that rate of settlement it would give some indication as to the rate at which human populations moved into new lands. So much work to be done.

Jim Belshaw said...

Thoughtful comments as always, John. As you say, so much work to be done.

I started to write a fuller response, but I think that I need to know more about the impact of climate change on food and water, recognising local variation. My working hypothesis is that the tablelands were a barrier dividing inland and coastal groups; inland populations concentrated in refuge areas; coast populations retreated north and south; but inland and coastal populations remained separate.