Age estimates for Willandra Lakes human bones

by Richard Gillespie

Lake Mungo excavation 1973

Human bones from more than 100 burials have been found in many locations around the margins of now dry lakes in the World Heritage Willandra Lakes region of southwestern NSW, Australia. The regional geomorphology, environmental reconstruction and archaeological overprint is summarised in Bowler (1998) and the dating in Gillespie (1998). In this discussion, the WLH numbering system for skeletons employed by Webb (1989) will be used, except for 2 skeletons which have become better known under alternative names: Mungo 1 (WLH-1) and Mungo 3 or LM3 (WLH-3).

Bone has been notoriously difficult for all dating methods, particularly in open semi-arid locations like the Willandra Lakes, because the material is fragmented and often badly degraded. This usually means a loss of the organic components (mainly the protein collagen), alterations to the mineral components and contamination with organics and inorganics from groundwater and sediments.

Radiocarbon dating (14C)

In well-preserved bone the best fraction for 14C dating is collagen or it's constituent amino acids. The carbon in these molecules was originally part of the living body and can be isolated in high purity. A good overview of bone dating progress is in Stafford et al (1991). Of 56 Willandra human bone samples analysed, only 2 contained more than 0.2% nitrogen and are known to contain well-preserved collagen: WLH-15 & 55 are probably Late Holocene (Gillespie, 1998).

The next best choice for 14C dating (if collagen is absent) comes from the carbon preserved by burning or cremation of the skeleton. Again this carbon was originally part of the living body, but it is difficult to identify a particular molecular structure to isolate. Early workers used simple acid washing to remove carbonates, preferably an alkaline wash is also used to remove humic acids, with final measurements made on the insoluble residue.

Bones from burials Mungo 1, WLH-9, 23, 24, 28, 44 & 122 were burnt and contained sufficient carbon to perform acid/base/acid pretreatment on small fragments for AMS 14C analysis. In all samples more than 90% of the total carbon present was contained in the base-soluble humic acids fraction, WLH-44 was completely base-soluble (Gillespie, 1997).

This acid/base/acid chemistry was originally designed for 14C dating of charcoal and assumes that strongly acidic and basic solutions will remove all contamination. For charcoal, dates on base-soluble humic acids fractions are generally younger than the residual charcoal, so we can say that younger contamination has been removed and the insoluble residue date is closer to the real age.

When measurements on both soluble and insoluble fractions for Mungo 1 and WLH- 122 were made, the dominant base-soluble fractions were both significantly older than their insoluble residues. The original date of ~25,000 BP was based on the acid insoluble residue (Bowler et al, 1972) and contained mostly humic acids of that age. As it turns out, many of the so-called charcoal samples from the Willandra also exhibit this contrary behaviour of humic acids older than insoluble residue dates. These hearth or fireplace samples contain mostly base-soluble material with little or no real charcoal and the insoluble residues contain unburnt fine organic particulates.

It is also possible to get 14C dates from the carbonate fraction of bone apatite, although there is evidence that this is unreliable because of ion exchange with groundwater carbonates. The apatite date (Bowler et al, 1972) on Mungo 1 of 19,000 BP is between the insoluble residue date (17,000 BP) and the humic acids at 25,000 BP.

Should we believe the older dates (which always seem curiously more desirable to archaeologists) measured on base-soluble humic acids fractions, or the younger dates on insoluble residues? The unpalatable choice for both black sediment and burnt bone dates seems to be between soluble or insoluble organic humic substances, both of which are of unknown composition and dubious origin.

Webb (1989) argued for the older soluble humic acids dates on these burnt bones, on my advice at the time. I subsequently changed my view (Gillespie, 1997; 1998) and supported the younger insoluble dates, on the grounds that humic acids of whatever age are not likely to represent the burnt carbon we seek in charcoal or burnt bone. Not much of a choice, and neither may be the real age.

Other dating methods

Have the results from other direct dating measurements on skeletal remains helped to resolve the Willandra chronology? An added confusion here is that calibration of radiocarbon dates is necessary for comparison with other methods supposedly operating on calendar time, an approximate calibration curve used here is described in Gillespie (1998) with units of ka (thousands of years ago).

Electron spin resonance (ESR) and uranium/thorium series (U/Th) dating methods are secondary in the sense that they rely on measurement of isotopes not originally present in the living body and the modelling of processes taking place after burial. Thermoluminescence (TL) and optically stimulated luminescence (OSL) dating methods are applied to quartz grains in sediments containing skeletal remains but not directly on the bones. These methods have advantages in a longer useful time range than 14C, sometimes with a signal increasing with time, and can sometimes be non-destructive.

On WLH-50, an ESR estimate of 29 ka on bone by Caddie et al (1987) is contradicted by a very consistent set of dates centred around 14 ka from TIMS & Gamma spec U/Th series measurements (Simpson & Grün, 1998). This would place the 'robust' WLH-50 burial in a similar age bracket with 'gracile' burials in the Garnpung/Leaghur area such as WLH-23, 24 & 122.

More controversial are the results from LM3, which was originally estimated from regional stratigraphic evidence at about 30-35 ka (Bowler & Thorne, 1977). A combined age estimate of 62±6 ka based on TIMS & Gamma spec U/Th on bone, ESR on tooth enamel and OSL on quartz in underlying sediments was proposed by Thorne et al (1999). These results have been strongly challenged on sampling, methodology and modelling grounds by Bowler & Magee (2000) and Gillespie & Roberts (2000), who favour the fairly robust 14C, TL and OSL regional chronology suggesting that LM3 is less than 43±3 ka. On the available evidence, with LM3 still of undecided age (and sex, according to Brown, 2000), 9 skeletons from the Willandra Lakes with direct 14C or U/Th dates (including Mungo 1 & WLH-50) are younger than 20 ka.

Commercial dating laboratory

BETA Analytic: AMS Dating Bones, Antler, and Teeth


Bowler, J M (1998) Willandra Lakes revisited: environmental framework for human occupation. Archaeology in Oceania, 33, 120-155.

Bowler, J M & Magee, J W (2000) Redating Australia's oldest human remains: A sceptics' view. Journal of Human Evolution 38, 719-726.

Bowler, J M, Thorne, A G & Polach, H A (1972) Pleistocene man in Australia: age and significance of the Mungo skeleton: Nature 240, 48-50.

Brown, P (2000) Pleistocene variation and the sex of Lake Mungo 3. Journal of Human Evolution 38, 743-750.

Gillespie, R (1997) Burnt and unburnt carbon: dating charcoal and burnt bone from the Willandra Lakes, Australia: Radiocarbon 39, 225-236.

Gillespie, R (1998) Alternative timescales: a critical review of Willandra Lakes dating. Archaeology in Oceania, 33, 169-182.

Gillespie, R & Roberts, R G (2000) On the reliability of age estimates for human remains at Lake Mungo. Journal of Human Evolution 38, 727-732.

Simpson, J.J & Grün, R (1998) Non-destructive gamma spectrometric U-series dating. Quaternary Geochronology 17, 1009-1022.

Stafford, T W Jr, Hare, P E, Currie, L Jull, A J T & Donahue, D J (1991). Accelerator radiocarbon dating at the molecular level. Journal of Archaeological Science 18, 35-72.

Thorne, A, Grün, R, Mortimer, G, Simpson, J J, McCulloch, M, Taylor, L and Curnoe, D (1999) Australia's oldest human remains: age of the Lake Mungo 3 skeleton. Journal of Human Evolution, 36, 591-692.

Webb, S G (1989) The Willandra Lakes Hominids, Monograph, Dept. of Prehistory, RSPAS, Australian National University, Canberra.


Willandra Lakes bone dates

Burial Lab. No Fraction dated 14C Age BP

Cal Age (ka)

WLH-1 ANU-618A Apatite (carbonate) 19,030 ± 1300

22.2 ± 2.6

(Mungo 1) ANU-618B Acid insoluble organics 24,710 ± 1200

28.5 ± 2.4
NZA-246 Humic acids 24,750 ± 2400

28.5 ± 4.8

NZA-230 Humic acids 25,120 ± 1380

28.9 ± 2.8
NZA-231 Humic-free residue 16,940 ± 635

19.8 ± 1.3
WLH-9 NZA-160 Humic-free residue 15,780 ± 430

18.4 ± 0.9
WLH-23 NZA-165 Humic-free residue 11,690 ± 580

13.5 ± 1.2
WLH-24 NZA-163 Humic-free residue 11,910 ± 340

13.7 ± 0.7
WLH-122 NZA-194 Humic-free residue 11,095 ± 370

12.7 ± 0.7
NZA-164 Humic acids 16,540 ± 540

19.3 ± 1.1
WLH-44 NZA-159 Humic acids 18,640 ± 530

21.8 ± 1.1
WLH-50 Postcranial fragments

13.5 ± 1.3

3 (a)

13.3 ± 0.9

14.3 ± 3.0

13.5 ± 0.6
Partial cranium

14.0 ± 6.0


14.0 ± 6.0

14.8 ± 2.0

15.8 ± 2.0

11.1 ± 1.3


14.2 ± 0.7
Postcranial fragments

29.0 ± 5.0

WLH-3 Long bone shavings

69.8 ± 2.1

5 (a)
(Mungo 3,

58.3 ± 1.2

50.7 ± 0.9

54.5 ± 0.7
Partial cranium

69.5 ± 2.9


64.1 ± 3.7

74.0 ± 7.0


60.0 ± 5.0
Tooth enamel fragments

63.0 ± 6.0


78.0 ± 5.0

References and dating methods used:
1. Bowler, Thorne & Polach (1972); radiocarbon.
2. Web (1989), Gillespie (1997); AMS radiocarbon.
3. Simpson & Grun (1998)
(a) TIMS 230Th/234U
(b) Gamma spec 230Th/234U
(c) Gamma spec, 231Pa/235U
4. Caddie et al (1987); ESR
5. Thorne et al (1999)
(a) TIMS 230Th/234U
(b) Gamma spec 230Th/234U
(c) Gamma spec, 231Pa/235U
(d) ESR
This page last updated 21 May, 2000.