This article was downloaded by: [University of ] On: 12 September 2008 Access details: Access Details: [subscription number 778575344] Publisher Taylor & Francis Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Australian Journal of Earth Sciences Publication details, including instructions for authors and subscription information: http://www.informaworld.com/smpp/title~content=t716100753

Geochronology of the Australian Cenozoic: a history of tectonic and igneous activity, weathering, erosion, and sedimentation* P. M. Vasconcelos a; K. M. Knesel a; B. E. Cohen a; J. A. Heim a a Earth Sciences, University of Queensland, Australia

Online Publication Date: 01 January 2008

To cite this Article Vasconcelos, P. M., Knesel, K. M., Cohen, B. E. and Heim, J. A.(2008)'Geochronology of the Australian Cenozoic: a history of tectonic and igneous activity, weathering, erosion, and sedimentation*',Australian Journal of Earth Sciences,55:6,865 — 914 To link to this Article: DOI: 10.1080/08120090802120120 URL: http://dx.doi.org/10.1080/08120090802120120

PLEASE SCROLL DOWN FOR ARTICLE

Full terms and conditions of use: http://www.informaworld.com/terms-and-conditions-of-access.pdf

This article may be used for research, teaching and private study purposes. Any substantial or systematic reproduction, re-distribution, re-selling, loan or sub-licensing, systematic supply or distribution in any form to anyone is expressly forbidden.

The publisher does not give any warranty express or implied or make any representation that the contents will be complete or accurate or up to date. The accuracy of any instructions, formulae and drug doses should be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims, proceedings, demand or costs or damages whatsoever or howsoever caused arising directly or indirectly in connection with or arising out of the use of this material. Australian Journal of Earth Sciences (2008) 55, (865 – 914)

Geochronology of the Australian Cenozoic: a history of tectonic and igneous activity, weathering, erosion, and sedimentation*

P. M. VASCONCELOS{, K. M. KNESEL, B. E. COHEN AND J. A. HEIM

Earth Sciences, University of Queensland, Qld 4072, Australia.

The development and application of geochronological tools suitable for dating Cenozoic rocks and processes have been instrumental to our understanding of the modern history of Australia. Geochronology reveals a dynamic continent that traced a long and rapid trajectory from a position adjacent to Antarctica in the early Cenozoic to its present position near the tropics. The average travel velocity along this path is revealed by the age of hotspot volcanoes, derived by the K–Ar method, and is complemented by measured geomagnetic pole positions on dated igneous rocks and sedimentary deposits. K–Ar dating of volcanic rocks also provided constraints on rates of landscape evolution before and after volcanism and the timing and pattern of dispersion of life—including human inhabitation. K– Ar geochronological results reveal a history of faunal and floral evolution suggestive of a continent undergoing progressive cooling and dehydration with a few brief warm and humid excursions. In contrast, 40Ar/39Ar, SHRIMP U–Pb, fission-track thermochronology, luminescence techniques, and cosmogenic-isotope methods have played relatively minor roles in reconstructing the chronology of Cenozoic volcanism in Australia. Integrated application of these techniques will be critical to providing more precise constraints on the volcanic history of the continent and its climatic and biological evolution. While Cenozoic volcanism, uplift, and denudation were active along the eastern and southeastern margins, a significant part of Australia west of the Tasman Line remained relatively quiescent. The history of this part of the continent is marked by slow and subdued uplift and subsidence, with subtle displacements along major continental structures, and occasional invasion by shallow seas. Despite the general absence of Cenozoic igneous rocks west of the Tasman Line, Australia (east and west) is blanketed by Cenozoic sedimentary covers and weathering profiles. If we consider weathering as a fourth rock-forming process (in addition to igneous, metamorphic and sedimentary), Australia has one of the most complete Cenozoic rock covers of any continent. Retrieving information recorded in these weathering profiles is essential for unravelling its Cenozoic history. Paleomagnetic studies, calibrated d18O curves, and weathering geochronology by K–Ar, 40Ar/39Ar, and (U–Th)/He provide insights into the imprint of climatic events and tectonic processes and illustrate the importance of erosion and weathering to the formation or enrichment of ore and mineral deposits. Except for diamondiferous lamproites of Western Australia and sapphire-rich volcanic rocks in eastern Australia, all other Cenozoic ore and mineral deposits in Australia are related to weathering and erosion. The widespread weathered blanket in Australia suggests low Cenozoic erosion rates. Numerical constraints on chronology and erosion rates are derived from the cooling and denudation histories retrieved from apatite and zircon fission-track and, more recently, (U–Th–Sm)/He thermochronology and cosmogenic Downloaded By: [University of Queensland] At: 05:12 12 September 2008 isotope studies. Geochronological studies of veneers of sediments, lake and cave deposits, marine carbonates, organic matter and groundwaters provide information on sediment provenance, subtle tectonic movements, and the Australian Cenozoic climatic history. These studies reveal a continent sensitive to global climatic cycles and subject to active, but subtle, tectonism and erosion. This record shows that Australia suffered periods of extreme aridity during cyclical glaciation at high latitudes and precise dating of carbonate sediments and speleothems reveals the exact timing and duration of these glacial and interglacial periods. Cosmogenic isotopes also provide constraints on the age and migration paths of Australia’s limited and finite groundwater resources. Lastly, age information extracted from surficial deposits reveals a protracted history of human occupation. KEY WORDS: Australia, geochronology, magmatism, weathering, erosion, sedimentation, climate.

*Appendices 1 and 2 [indicated by an asterisk (*) in the text and listed at the end of the paper] are Supplementary Papers; copies may be obtained from the Geological Society of Australia’s website (5http://www.gsa.org.au4) or from the National Library of Australia’s Pandora archive (5http://nla.gov.au/nla.arc-251944). {Corresponding author: [email protected]

ISSN 0812-0099 print/ISSN 1440-0952 online Ó 2008 Geological Society of Australia DOI: 10.1080/08120090802120120 866 P. M. Vasconcelos et al.

INTRODUCTION trajectory from high-latitude zones (70–358S) in the Cretaceous to its present position at mid to low latitudes The interplay between endogenous tectonic and igneous (43–108S) (Figure 1). Geochronological techniques rele- activity, and exogenous weathering, erosion, deposi- vant to the study of the Australian Cenozoic are, tional, and diagenetic processes has shaped the Austra- therefore, those techniques suitable for dating igneous lian landscape. In turn, the superimposed effects of activity; uplift and fault movement; and weathering, these processes through time provide the record erosion, depositional, and diagenetic processes. Some of that permits tracing Australia’s geological, tectonic, these techniques (e.g. K–Ar, Rb–Sr, U–Pb) are based on geomorphological and climatic histories during its absolute clocks (i.e. the radioactive decay of a parent Downloaded By: [University of Queensland] At: 05:12 12 September 2008

Figure 1 Cenozoic plate recon- structions for Australia and Ant- arctica (after Duncan & McDougall 1989a), showing the position of Australia at high lati- tudes for much of the Early Cen- ozoic, before fast seafloor spreading between the two conti- nents, commencing around 55 Ma, led to rapid migration of Australia into tropical and subtropical latitudes. Geochronology of the Australian Cenozoic 867

isotope into a daughter product), while other techniques namic processes and constraints on climatic events that are based on relative methods that require external shaped the Australian continent in the more recent past calibration [e.g. paleomagnetic dating, thermolumines- (Wellman & McDougall 1974a; Johnson 1989). Perhaps cence (TL), optical emission spectroscopy (OES)]. for these reasons, volcanic rocks are the most intensely Australian geoscientists have played a leading role in dated Cenozoic features in Australia, with nearly 1300 the development, refinement and application of geochron- geochronological analyses undertaken over the last 40 ological tools suitable for unravelling Australian and years (Appendix 1*). Most of the geochronological global Cenozoic histories. These include early improve- record is derived from K–Ar and, to a lesser extent, ments in K–Ar dating (McDougall 1961, 1966); the seminal 40Ar/39Ar and fission-track dating; Rb–Sr, U–Pb, cosmo- development of sensitive high-resolution ion microprobe genic isotopes, 14C and TL have been only sporadically (SHRIMP) U–Pb methodologies (Compston et al. 1984; employed in dating Australian Cenozoic volcanic units. Ireland et al. 2008); methodological advances in 40Ar/39Ar dating (McDougall 1974; McDougall & Harrison 1999); K–Ar developments in theoretical and practical aspects of fission-track thermochronology (Gleadow et al. 2002a, b; The most important contribution of geochronology to Kohn et al. 2002, 2005); demonstration of the utility of understanding the Australian Cenozoic has arisen from using d18O of pedogenic minerals to indirectly date the careful, precise, accurate and comprehensive dating of weathering profiles (Bird & Chivas 1989); development extrusive magmatic units by the K–Ar method (Wellman of the (U–Th)/He thermochronological method (Zeitler & McDougall 1974a, b). K–Ar geochronology of igneous et al. 1987); refinement of 40Ar/39Ar (Vasconcelos 1999a, b; suites started in 1963 on an analcite basalt in northeast Vasconcelos & Conroy 2003) and (U–Th)/He (Shuster et al. New South Wales (Cooper et al. 1963) and continues to the 2005; Heim et al. 2006) weathering geochronology; high- present (Sutherland 2003; Sutherland et al. 2005). Notable resolution U-series dating of carbonates (Stirling et al. laboratories, researchers in this field and some of their 1995, 1998); laser-ablation U-series disequilibria dating of major contributions are listed in Appendix 1*. phosphates (Eggins et al. 2003; Gru¨n et al. 2005) and Fe- In the K–Ar method, time is obtained by the oxyhydroxides (Bernal et al. 2006); and OSL dating of relationship: single crystals of detrital minerals (Roberts et al. 1994). 1 40 40 The geochronological tools developed, perfected or widely t ¼ðl Þln½ Ar = KÞðl=leÞþ1 applied in Australia underpin the application of isotope geochemistry to unravel the Cenozoic history of this where l represents the total decay constant of K 710 71 unique continent. (5.543 + 0.010 6 10 a ) (the sum of lb and le); lb In order to fully evaluate the contribution of geochro- represents the 40Kto40Ca decay constant 710 71 40 40 nology to our present understanding of Australian (4.962 + 0.009 6 10 a ); le represents the Kto Ar geology, we must first review the contribution of decay constant (5.508 + 0.004 6 10711 a71); 40Ar* is the geochronology (K–Ar, 40Ar/39Ar, Rb–Sr, U–Pb, etc.) to total amount of radiogenic 40Ar (obtained by subtracting the study of volcanism, the most extensively dated the atmospheric 40Ar from the total amount of 40Ar Cenozoic process in Australia. To unravel the tectonic measured from an aliquot of the sample); and 40K is the and climatic evolution of the continent, we must also total amount of 40K in the sample (obtained by measur- assess the history of weathering and erosion. For that, ing the total amount of K for an aliquot of the sample weathering geochronology [paleomagnetism, calibrated and multiplying by 0.01167, the proportion of 40KinK). d18O curves, K–Ar, 40Ar/39Ar, (U–Th)/He and U-series Reliable determination of t (the age of a sample) disequilibria] and low- to intermediate-temperature depends on: (i) the precise and accurate determination Downloaded By: [University of Queensland] At: 05:12 12 September 2008 40 40 40 40 thermochronology [AFTT, ZFTT, (U–Th)/He)] and in situ of the Kto Ca (lb) and Kto Ar* (le) decay cosmogenic isotopes are the most direct and relevant constants; (ii) the assumption that the 40K decay

approaches. Finally, to understand the effects of tectonic constants (lb and le) have remained constant through process and climatic changes shaping the Australian time; (iii) the assumption that the isotopic composition landscape and environment, we examine dating techni- of K in the sample is known and it has changed only due ques (magnetostratigraphy, 14C, TL, OSL, ESR, U-series to radioactive decay; (iv) the correction for atmospheric disequilibria, and cosmogenic isotopes) suitable for the 40Ar adsorbed onto or included into the mineral study of soils, sedimentary deposits, speleothems and structure at the time of precipitation; (v) the assumption waters. We have attempted to evaluate and include all that the sample became closed to gains or losses of the geochronological contributions to the study of the radiogenic 40Ar* and K soon after its formation; (vi) the Australian Cenozoic available in the published litera- assumption that excess 40Ar* was not incorporated into ture, but as it is common in a review of such a broad the sample at the time of formation; and (vii) on the topic, we may have inadvertently omitted important homogenisation and splitting of the sample into two contributions, and we apologise for those omissions. exactly equivalent aliquots for the independent mea- surement of K and Ar. A major limitation of dating Australian volcanism by CHRONOLOGY OF CENOZOIC VOLCANISM the K–Ar method derives from the fact that Cenozoic volcanic rocks in Australia display varying degrees The Cenozoic igneous history is instrumental in of weathering, leading to potential Ar losses and understanding Australia’s trajectory from high- to underestimation of ages (Wellman & McDougall 1974b). low-latitude zones, and it provides evidence for geody- To assess the likely accuracy of K–Ar ages obtained 868 P. M. Vasconcelos et al.

from volcanic rocks, Wellman & McDougall (1974b p. study on the Nebo volcano in northern Queensland, 249) proposed a set of stringent petrographic criteria (A– Sutherland et al. (1977) were able to analyse only 31 out D) for assessing the rocks to be dated: ‘A, where of 4150 samples collected in the field because most potassium-bearing phases are fresh and the K–Ar age samples did not pass the suitability criteria. As a is likely to be correct; B, where potassium-bearing consequence, the problems associated with dating vari- phases are slightly altered and the K–Ar age could be ably weathered volcanic rocks by the K–Ar method raises low; C, where potassium-bearing phases are consider- the possibility that much of the record so far obtained for ably altered, and the age is likely to be low; and D, where the volcanic history of eastern Australia is fragmentary the potassium-bearing phases are extensively altered and incomplete. and the measured K–Ar age is expected to be much too In some cases, researchers eager to date an important low.’ This careful selection of suitably pristine samples unit may be tempted to compromise and analyse has been instrumental in developing a picture of the samples in the C and D categories, which are likely or time–space–composition patterns of Cenozoic volcanism very likely to yield dates younger than the true in Australia and in extending the application of K–Ar extrusion ages. On publishing such results, care must dating of mafic volcanic sequences well into the Late be taken to identify these dates as minimum ages. Pleistocene (Griffin & McDougall 1975). However, when these results are later transcribed, cited, Using the screening process above, Wellman & or used in general compilations, the concept of mini- McDougall (1974a, b and references therein) dated a wide mum age may be forgotten, and an unreliable date may spectrum of volcanic rocks in eastern Australia. The most be promoted to an extrusion age. Utilisation of such significant result of those studies was the identification of results may then lead to interpretations involving three classes of Cenozoic volcanism: central volcanoes, complicated geological histories, which may be a lava fields, and leucitite magmas, each associated with consequence of incorrect geochronology and not com- distinct age distributions (Figure 2). The central volca- plex geology. An approach suitable for overcoming some noes form variably preserved mafic shield volcanoes, of these difficulties is the 40Ar/39Ar technique. with minor silicic flows and plugs. Their ages become systematically younger towards the south, with an age vs 40Ar/39Ar latitude trend of 65 + 3 mm/y from 32 Ma to the present (Wellman & McDougall 1974a; Duncan & McDougall In the 40Ar/39Ar method, K is indirectly measured by the 1989b). These results were interpreted as evidence for a generation of the 39Ar isotope by the reaction 39K (n,p) plume-related origin for the Australian central volcanoes 39Ar (Merrihue & Turner 1966). A sample (single crystal, and are consistent with the northward velocity of the cluster of crystals, rock fragment, glass, etc.) is irra- Australian Plate. The leucitite occurrences in central diated by a neutron flux, together with an appropriate New South Wales and Victoria also appear to show a neutron fluence standard, in a nuclear reactor. Time is north-to-south age progression (Figure 2), with a trend obtained by the relationship: similar, albeit less well defined, to that of central volcanoes (Wellman et al. 1970; Duncan & McDougall 1 40 39 t ¼ðl Þ ln ½Jð Ar = ArkÞþ1 1989b). The lava-field volcanoes, on the other hand, form mainly mafic-only provinces, and their ages do not show any systematic correlation with latitude (Figure 2b) where l is as defined for K–Ar (above); 40Ar* is the total (Wellman & McDougall 1974a; Duncan & McDougall amount of radiogenic 40Ar (obtained by subtracting the 40 40 1989b). Importantly, K–Ar dating of lava fields in Queens- atmospheric Ar and ArK from the total amount of 40 40 land (Griffin & McDougall 1975; Robertson et al. 1985; Ar measured from the sample, where ArK is the Downloaded By: [University of Queensland] At: 05:12 12 September 2008 Robertson 1985; Coventry et al. 1985; Whitehead et al. 2007) amount of 40Ar generated by the reaction 40K (n,p) 40Ar); 39 39 39 and Victoria (McDougall et al. 1966; Aziz-Ur-Rahman & Ark is the amount of Ar generated by the reaction K 39 39 McDougall 1972; McDougall & Gill 1975) reveals that (n,p) Ar [obtained by subtracting the ArCa from the volcanic activity in these areas continued well into the total amount of 39Ar measured from an aliquot of the 39 39 Pleistocene, suggesting that the tectonic conditions con- sample, where ArCa is the amount of Ar generated by ducive to magmatism in eastern Australia span the entire the reaction 42Ca (n,a) 39Ar]. Cenozoic. J is a dimensionless irradiation parameter defined as The chronological history extracted from Cenozoic (Mitchell 1968): volcanism has been instrumental in understanding the Z geological and climatic evolution of the continent. This 39K l J ¼ 40 D fðEÞsðEÞ dE wide range of applications is only possible because of K le þ lb Wellman & McDougall’s (1974b) selective criteria for screening suitable samples for K–Ar geochronology. where 39K is the amount of 39K present in the sample; 40K 40 Unfortunately, however, the application of this important is the amount of K present in the sample; l and le are screening criterion results in the elimination of many as defined for K–Ar (above); D is the duration of the field samples from further consideration. For example, irradiation; f(E) is the neutron flux at energy E; s(E)is only 17% of the samples analysed by Wellman & the neutron capture cross-section at energy E for the McDougall (1974b) probably represent dates likely to be reaction 39K (n,p) 39Ar. The irradiation parameter J is true eruption ages (category A); nearly 50% of the samples determined by irradiating a neutron fluence standard of probably yield dates younger than the eruption ages known age placed in the same irradiation container as, [categories C (41.5%) and D (7.8%)]. Similarly, in a major and in close proximity to, the unknown. Measurements Geochronology of the Australian Cenozoic 869 Downloaded By: [University of Queensland] At: 05:12 12 September 2008

Figure 2 Distribution and volume of Cenozoic volcanism in (a) space and (b, c) time showing relationships to topographic and tectonic features of eastern Australia and the Tasman and Coral Seas (after Johnson 1989 p.8, p.15, p.23, and Wellman & McDougall 1974a). Lava fields, shown in grey, were mainly emplaced along the crest of the Great Divide (blue dotted line) in the east Australian highlands, and began to erupt at about 70 Ma, following initial rifting and opening of the Tasman Sea, although the peak of activity was between 55 and 35 Ma. As lava-field activity began to wane, Australia began its rapid northward migration away from Antarctica and the earliest bimodal central-volcano, shown in black, erupted along the coast in at Cape Hillsborough (CH) at about 34 Ma. Central-volcano activity then migrated southward at about 65 km/Ma, arriving at Mt Macedon (M) in central Victoria at about 6 Ma. These volcanoes track the northward motion of the Australian plate over a fixed melting anomaly (or anomalies) in the underlying upper mantle. Similar age-progressive volcanism is recorded by the parallel seamount chains of the Tasman Sea and leucite-bearing lavas and plugs in the central New South Wales and Victoria. Geochronological studies of these three classes of volcanism, which mainly occurred to the east of the Tasman Line (black broken line), have played an important role in unravelling the tectonic, weathering, erosional, sedimentary, climatic and biological evolution of the Australian Cenozoic. Bathymetry is from Petkovic & Buchanan (2002); age data and references are compiled in Appendix 1*; accumulative volume data are from Wellman & McDougall (1974a). 870 P. M. Vasconcelos et al.

40 39 of the amounts of K and Ark produced in the standard geochronology (McDougall & Roksandic 1974). McDou- provides the information necessary for calculating J: gall & Duncan (1988) followed this work in a study of the Tasmantid Seamounts by total-fusion 40Ar/39Ar and lt 40 39 J ¼ðe 1Þ=ð Ar = ArkÞ whole-rock K–Ar to resolve a linear decrease in age with latitude at a rate of 67 + 5 mm/y, which is where t is the known age of the standard, determined compatible with the record of south-younging of the independently; l is defined as above; 40Ar* is the total continental central volcanoes. With this result, one can amount of radiogenic 40Ar measured from the standard; project the present location of the hotspot responsible 39 39 39 Ark is the amount of Ar generated by the reaction K for the Tasmantid Seamounts volcanism to 40.48S and (n,p) 39Ar in the standard. Measurement of 40Ar* and 155–1568E, a site coincident with the epicentre of a large 39 Ark in the sample and standard requires correction for recent earthquake (McDougall & Duncan 1988). interfering isotopes created by the reactions 40Ca (n,na) More recently, incremental-heating 40Ar/39Ar ana- 36Ar, 40Ca (n,a) 37Ar, 42Ca (n, a) 39Ar, and 40K (n,p) 40Ar. lyses have been applied to whole-rock samples from the Correction procedures are discussed by McDougall & Ebor volcano (Ashley et al. 1995), to a plagioclase Harrison (1999). separate from southern Victoria (Hare et al. 2005), and Successfully determining t in a sample by the leucite separates from the El Capitain leucitite in New 40Ar/39Ar method relies on assumptions (i)–(iii) listed South Wales (McQueen et al. 2007). However, the only for the K–Ar method, and it may also require meeting comprehensive 40Ar/39Ar study of Cenozoic volcanism assumptions (v) and (vi). However, in the 40Ar/39Ar in Australia is that by Cohen (2007) and Cohen et al. method, (v) and (vi) are not assumed; they are tested. (2007, 2008). These studies were inspired and facilitated Samples that do not pass the test are discarded. by the seminal K–Ar investigations of McDougall, Webb Detecting and quantifying 40Ar* and K losses and excess and Wellman, and are only beginning to tap the full 40Ar* greatly increases the geological accuracy and potential of Ar geochronology to study the history of reliability of 40Ar/39Ar results. In addition, in the volcanism in eastern Australia. For instance, in a 40Ar/39Ar method, only isotopic ratios, and not the detailed study of central-volcano activity in southeast absolute amounts of Ar isotopes, need to be determined Queensland, Cohen et al. (2007) showed that over a for a sample; and all isotope ratios are measured from a restricted range of latitude (*320 km from north to single sample aliquot in one instrument (noble-gas mass south), the 40Ar/39Ar incremental-heating analysis of 30 spectrometer), diminishing analytical uncertainties. groundmass and 31 single-crystal aliquots from 21 The large time-span potentially datable by the method samples, and the total-fusion analysis of 99 single (to ca 2 ka in ideal circumstances, with no upper limit) crystals from 10 samples, reveal a linear north-to-south permits reliable comparisons of self-consistent geo- age progression not apparent from K–Ar dates from the chronological results determined by a single technique. same region (Figure 3). Similarly, 40Ar/39Ar dating of Importantly, the fine-scale resolution of the method seven leucite-bearing volcanic centres, spanning over allows dating of very small grains (typically in the 0.2– 600 km from New South Wales to Victoria, more 2 mm range), and therefore permits dating of single precisely constrain the age and duration of leucitite fresh phenocrysts or visibly unaltered or unweathered volcanism (Cohen et al. 2008). In doing so, these results groundmass fragments from partially altered or weath- confirm and refine the picture from the K–Ar record that ered volcanic rocks. More importantly, incremental the leucite-bearing volcanoes progressively young to the heating permits determining the Ar-retention history south at a rate similar to the central-volcanoes to for a mineral or rock and identifying the presence of the east. Importantly, many of the K–Ar dates for the excess/inherited Ar, further obviating problems asso- leucitites span age ranges of up to several million years Downloaded By: [University of Queensland] At: 05:12 12 September 2008 ciated with partial deuteric alteration, weathering, or for individual lavas or even individual samples, and incorporation of excess Ar during extrusion. Multiple- they are often younger than the 40Ar/39Ar ages for the step analysis also enables selective extraction and same localities (Appendix 1*). The 40Ar/39Ar results thus analysis of Ar isotopes hosted in different reservoirs appear to indicate that many of the leucitite samples in the sample, facilitating the separation of atmospheric analysed by K–Ar suffered from varying degrees of Ar from radiogenic Ar and increasing the precision of the loss (Wellman et al. 1970; McQueen et al. 2007; Cohen analysis. Finally, the full automation of the method et al. 2008). One of the most significant outcomes of this permits dating a large number of samples, providing ongoing margin-wide 40Ar/39Ar campaign is the demon- improved statistics and the acquisition of a comprehen- stration that the increased resolution possible with the sive geochronological database in a fraction of the time fully automated incremental-heating method permits required by conventional analysis. The improved ana- resolving volcanic propagation rates over time windows lytical precision of the 40Ar/39Ar technique (50.1% is as brief as a few million years and thus provides a tool routinely achievable with the incremental heating for tracking variation in plate velocity during Austra- method), as compared with the K–Ar method, makes it lia’s northward migration (Knesel et al. in press). We ideal for addressing Cenozoic geological problems. revisit this exciting prospect in our discussion in the Despite these capabilities and the fact that many second half of the paper. notable advances in 40Ar/39Ar geochronology took place Before continuing the geochronological review, it is in Australia (McDougall & Harrison 1999), the 40Ar/39Ar worth commenting on an additional important contri- method has seen limited use in the study of Australian bution of Australian K–Ar and 40Ar/39Ar geochronology Cenozoic volcanism. The first application was a pilot laboratories to the global geochronological community. study on the use of the HIFAR reactor for 40Ar/39Ar In 40Ar/39Ar geochronology, primary standards must be Geochronology of the Australian Cenozoic 871

Figure 3 Comparison of (a) individual K–Ar ages and (b) 40Ar/39Ar ages vs latitude for southeast Queensland central volcanoes (after Cohen et al. 2007), showing the improved resolution provided by the laser incremental-heating 40Ar/39Ar method. The 40Ar/39Ar results resolve a southerly age progression of these hotspot-derived central volcanoes (R2 ¼ 0.7) not apparent from the K–Ar data (R2 ¼ 0.02). FI, Fraser Island; N, Noosa; M, Maleny; GH, Glasshouse Mountains; FP, Flinders Peak.

calibrated by the K/Ar method (Renne et al. 1998), and mond deposits on Earth (Wellman 1973; Jaques et al. K–Ar dating at the Australian National University 1984). The Rb–Sr system was used by Dasch & Millar laboratory has provided crucial data for the calibration (1977) to date the soda-trachytes in central Victoria of primary (GA1550) 40Ar/39Ar standards (McDougall & (Hanging Rock, Camel’s Hump, Brock’s Monument), Roksandic 1974; Renne et al. 1998; Spell & McDougall yielding an isochron result (6.2 + 0.6 Ma) within error 2003) and for assessing the viability of leucite as a of feldspar K–Ar dates (6.48 + 0.15 to 6.17 + 0.11 Ma) Quaternary 40Ar/39Ar standard (Karner et al. 2001). obtained by Wellman (1974). However, both results are Further calibration of 40Ar/39Ar standards will depend older than some of the whole-rock K–Ar dates from the on the maintenance of a high-quality K–Ar laboratory in same rocks (6.13 + 0.10 to 5.43 + 0.09 Ma) (Wellman 1974; Australia. Ewart et al. 1985). This comparison indicates that some whole-rock samples had apparently lost varying Rb–Sr amounts of radiogenic argon since eruption, and high- lights once again some of the problems associated with The Rb–Sr dating method is based on the b7 decay of establishing emplacement ages for variably altered or naturally occurring 87Rb to stable 87Sr, with a half-life of weathered igneous rocks. þ0:06 48:80:10 Ga (Steiger & Ja¨ger 1977). Time (t) is determined by the equation: U–Pb 1 86Sr 87Sr 86Sr U–Pb geochronology is based on the natural radioactive t ¼ ln 1 þ 238 235 232 206 207 208 l 87Rb 86Sr M 87Rb I decay of U, U and Th to Pb, Pb and Pb, respectively. These three decay schemes have decay 710 71 710 where M indicates a measured ratio, the subscript I constants of l1 ¼ 1.55125 6 10 a , l2 ¼ 9.8485 6 10 Downloaded By: [University of Queensland] At: 05:12 12 September 2008 71 711 71 indicates the initial ratio (i.e. at the time of rock a and l3 ¼ 4.9475 6 10 a , respectively (Steiger & formation), and l is the 87Rb decay constant. The initial Ja¨ger 1977). Time is determined for each decay scheme 87Sr/86Sr ratio can be either assumed (e.g. a value of by the following equations: 0.704 + 0.001 for oceanic basalts: Faure 1991) or deter- mined by isochron analysis of cogenetic minerals and/ 206 204 206 204 t206 ¼ð1=l1Þ ln ðf½ Pb= PbÞm ð Pb= PbÞ0= or whole-rocks with different initial Rb/Sr ratios. Like ð238U=204PbÞ gþ1Þ other radiometric methods, it is assumed that the m mineral/rock systems were closed to either Rb or Sr 207 204 207 204 gains or losses since their formation, apart from t207 ¼ð1=l2Þ ln ðf½ Pb= PbÞm ð Pb= PbÞ0= 235 204 changes brought about by the decay of Rb to Sr, that ð U= PbÞmgþ1Þ the 87Rb decay constant is accurately known, and that the decay constant has remained the same through time. 208 204 208 204 The Rb–Sr method has seen only limited use in the t208 ¼ð1=l3Þ ln ðf½ Pb= PbÞm ð Pb= PbÞ0= 232 204 Australian Cenozoic, largely because the long half-life of ð U= PbÞmgþ1Þ 87Rb limits the method to Rb-rich materials when dating

young rocks. However, the method was instrumental, where t206, t207 and t208 are three independent dates, along with K–Ar, in establishing an Early Miocene age subscript m identifies measured isotope ratios, sub-

for the diamond-rich Ellendale group of lamproite script 0 identifies the initial isotope ratios, and l1, l2 and diatremes in Western Australia [as opposed to a Jurassic l3 are the decay constants above. If the three indepen- age proposed by Prider (1960) and Kaplan et al. (1967)], dent dates above are concordant, they determine the age making these pipes the youngest known primary dia- of the sample. When they are discordant due to partial 872 P. M. Vasconcelos et al.

238 gains or losses of radioactive parent isotope(s) or where la is the total decay constant for U; rs is the 238 206 radiogenic daughter isotope(s), the U ! Pb and spontaneous fission-track density; ri is the induced 235 207 U ! Pb isotopic development can be treated fission-track density; lf is the rate of spontaneous fission simultaneously, and details of the geologic history of of 238U; 235U and 238U are the present number of atoms in the samples can be recovered from concordia diagrams a cubic centimetre of sample; j is the thermal neutron (Dickin 1995). dose used to induce fission of 235U; and s is the reaction U–Pb geochronology has seen limited use in Cenozoic cross-section for induced fission of 235U. volcanism in eastern Australia. A notable exception is Hurford & Green (1983) simplified this equation to: the application of SHRIMP-based techniques (Ireland et al. 2008) to date magmatic zircons and zircon inclu- 1 rs t ¼ ln 1 þ Zla sions in sapphire crystals in New South Wales. For la ri example, Coenraads et al. (1990) dated, by SHRIMP U–Pb geochronology, two zircon inclusions in sapphires from by introducing a calibration factor, the zeta factor (Z),

the New England/Central Province, New South Wales. Z ¼ jsI/lf, which is measured through the irradiation of The results, 35.9 + 1.9 and 33.7 + 2.1 Ma, are compatible an apatite or zircon grain of known age together with with K–Ar ages (38–19 Ma) and ZFT ages (49–2 Ma) the unknown samples. obtained for basaltic rocks in the region, leading the Despite its inherent lower precision, as compared authors to suggest a genetic link between sapphire with K–Ar and 40Ar/39Ar dating, the fission-track method mineralisation and alkali basalt magmatism in eastern has been used to date samples too weathered for K–Ar Australia. The bulk of the K–Ar ages for this region are analysis (Graham et al. 2003), or where an independent from Wellman et al. (1969), Wellman & McDougall (1974b) dating method was sought to cross-check other techni- and Young & McDougall (1993). Similarly, zircons from ques (Cundari et al. 1978; Gleadow & Ollier 1987). The alluvial deposits associated with the Barrington shield fission-track method has also been used to date popula- volcano in New South Wales yield U–Pb SHRIMP ages tions of detrital zircons derived from volcanic rocks that (60–45 Ma) (Sutherland & Fanning 2001) that match might have blanketed an area. Fission-track dating of basalt K–Ar ages (59–44 Ma) (Wellman et al. 1969; alluvial zircons associated with alluvial sapphires and/ Wellman & McDougall 1974b; Sutherland & Fanning or diamonds, particularly the New England gemfields in 2001). Zircon inclusions in sapphire crystals from the northern New South Wales (Yim et al. 1985; Seward & Tumburumba region of southern New South Wales yield Korsch 1989; Sutherland 1993; Sutherland et al. 1993), more complex age distributions (Sutherland et al. 2002). provides maximum age constrains on the formation of Out of a total of 12 spot analyses from eight inclusions, the placer deposits. In addition, if sapphires and zircons four analyses yield ages from 27.6 + 0.3 to 22.6 + 0.4 Ma are derived from the same magmas (Coenraads et al. (Sutherland et al. 2002), some of which overlap with the 1990), the zircon fission-track results provide constraints period of basalt activity determined by K–Ar (24–19 Ma) on the timing of magma extrusion. (Wellman & McDougall 1974b; Young & McDougall 1993; Unfortunately, the low annealing temperature of Sutherland et al. 2002). The remaining eight spot fission tracks that is so useful for thermochronological analyses yield ages ranging from 902.7 + 10.2 to studies renders the technique susceptible to resetting by 67.2 + 0.9 Ma, and are interpreted as inherited zircons post-extrusion heating events and may complicate the from the Paleozoic country rocks (Sutherland et al. 2002). use of this technique in dating volcanic eruptions. For instance, Sutherland et al. (2005) showed that while some Fission-track chronology zircon crystals from the Belmore volcano in northern New South Wales yield fission-track dates between Downloaded By: [University of Queensland] At: 05:12 12 September 2008 The fission-track method is based on the accumulation 20.5 + 2.4 and 22.8 + 6.1 Ma, which are concordant with of crystallographic damage tracks generated by the zircon SHRIMP and whole-rock K–Ar ages, some zircons spontaneous fission of 238U after a sample has cooled yield much younger results (down to ca 0 Ma). These below its annealing temperature (*110 + 108C for F- were interpreted as geologically insignificant results and apatite, 125–1508C for Cl-apatite and *2408C for zircon: attributed to a relatively recent heating event (Suther- Gleadow et al. 2002b). In the case of rapidly cooled land et al. 2005). In other cases, such as the Barrington extrusive or hypabyssal igneous rocks that have not shield volcano, a wide range of zircon fission-track dates been subsequently reheated, apatite and zircon fission- have been interpreted to reflect episodic gem-bearing track thermochronology (AFTT and ZFTT) provide eruptive events at about 57, 43, 38, 28 and 5 Ma (Suther- reliable cooling ages equivalent to the time of emplace- land & Fanning 2001). Although only the three older ment (Tagami & O’Sullivan 2005). These ages are fission-track results overlap with the age range defined obtained by measuring the number and length of tracks by K–Ar dating of mafic lavas and U–Pb SHRIMP dating in zircon or apatite crystals etched with appropriate of alluvial zircons in the region (see above), the authors acids, and comparing the number of tracks with the maintained that the younger ages indicate that eruptive number of artificially induced tracks in U-bearing activity continued until 5 Ma. standards irradiated together with the sample (Gleadow et al. 2002a, b). In fission-track methods, time (t)is 14C obtained according to: Direct dating of volcanic eruptions by the 14C technique 235 1 rs la U is not possible. However, during eruptions, lavas or t ¼ ln 1 þ 238 js la ri lf U tephra deposits incorporate organic matter, or bury Geochronology of the Australian Cenozoic 873

vegetated soils and weathering profiles, killing the the potential of contamination, but very few studies organisms and immediately setting the 14C clock, which on dating volcanic rocks in Australia have utilised permits indirect dating of extrusion events. Minimum the AMS technique (Leaney et al. 1995; Sherwood et al. ages of volcanism can also be obtained by dating peat 2004). swamps and maar-lake sediments overlying volcanic As a result of the limitations and caveats above, the units. Thus, carefully selected materials, with clear field 14C method has been used with varying success to date relationships to the volcanic event, can be used to young volcanic events in South Australia, Victoria and bracket the age of eruption. northern Queensland. For example, in the Newer The 14C technique is based on the continuous produc- Volcanics Province in Victoria, minimum ages for Mt tion of 14C via interaction by cosmic rays in the atmo- Eccles (6325 + 120 a BP) and Mt Napier (7240 + 140 a BP) 14 14 sphere. The C is rapidly oxidised to CO2, which then were obtained through dating of basal peat deposits forms part of the carbon cycle. During their lifetime, from Condah and Buckley Swamps, which formed by

organisms will exchange CO2 with the atmosphere, and damming of the pre-volcanic drainage by the basaltic remain in equilibrium with it, but upon the death of the flows (Gill & Gibbons 1969; Gill & Elmore 1973). In organism, the uptake of new 14C stops, and the amount of contrast, much greater minimum 14C dates have been radioactive 14C decays with time by b7 emission back to reported for swamp samples overlying the Mt Eccles and 14 14 N. If the initial amount of C activity at death (Ao) can Mt Napier lavas, including: 19 550 (no error reported: be predicted (van der Plicht et al. 2004), and if the sample Ollier 1981), 26 240 + 480 and 27 510 + 240 a BP (Head has subsequently remained a closed system, then its age et al. 1991). Importantly, the age of Mt Napier was (t) can be determined by measuring its present level of independently estimated at 31.9 + 2.4 ka by cosmogenic activity (A), according to the equation: isotope exposure dating (Stone et al. 1997), lending support for the older 14C results. lt 14 A ¼ A0e Conflicting C results have also been reported for Mt Gambier, South Australia. Plant material buried be- where l is the decay constant for 14C [equivalent to neath the Mt Gambier tuff yielded 14C ages in the range 14 0.693/t1/2, where t1/2 is the half-life of C, 5568 + 30 a of 8–0 ka, with a strong peak at 5–4 ka (Blackburn et al. (Libby 1955) or the presently accepted value of 5730 + 40 a 1982). In contrast, Leaney et al. (1995) obtained a (Godwin 1962)]. The present level of activity (A) can be minimum age of 28 ka BP for acid-washed AMS 14C measured by direct-counting of b7 decays from 14C, or by from sediments inside the volcanic crater-lake. In this direct measurement of the remaining amount of 14Cin case, independent age constraints derived from thermo- the sample by Accelerator Mass Spectrometry (AMS) luminescent-methods, discussed below, support the (Trumbore 2000). The direct counting technique requires younger 14C result. Murray-Wallace et al. (1998, p. 263) a minimum of 10–100 g of material, depending on the offer an explanation for this discrepancy: ‘Despite their sample’s age, but AMS can be performed on as little as careful work, the possibility remains that Leaney et al. *1 mg (Trumbore 2000). The upper limit for 14C dating is (1995) underestimated the amount of contamination by *50 000–60 000 years BP, beyond which there is detrital and authigenic calcium carbonate, depleted in insufficient 14C remaining in the sample for accurate 14C, from external sources such as the Miocene Gambier measurement (Trumbore 2000). Limestone, which thus yielded an apparently older There are two major difficulties with 14C geochronol- radiocarbon age.’ Similar discrepancies between old ogy. First, the 14C content of the atmosphere has varied 14C and young thermoluminescence dates have also been through time because of changes in the production rate reported from the Mt Schank volcano (see below) and of 14C (due to variations in the flux of cosmic rays and highlight the difficulties and uncertainties plaguing Downloaded By: [University of Queensland] At: 05:12 12 September 2008 the strength of Earth’s magnetic field) and because of attempts to date Quaternary volcanic activity. changes in the carbon distribution among terrestrial reservoirs (biosphere, atmosphere and ocean). Thus, a 14 Thermoluminescence (TL), optically stimulated C age must be transferred to a calendar age based on luminescence (OSL) and electron spin empirically produced calibration curves (van der Plicht resonance (ESR) dating et al. 2004) before it can be directly compared with other dating techniques. Most 14C studies on Australian Thermoluminescence (TL), optically stimulated lumi- volcanism were undertaken prior to 1990 and have not nescence (OSL) and electron spin resonance (ESR) utilised these calibrations (Gill & Gibbons 1969; Barton dating are promising techniques that work on the same & McElhinny 1980). principle: natural radiation from U, Th and K from The second difficulty inherent to 14C dating is sample the surrounding environment causes electrons to be- contamination by ‘young’ (recently released from organ- come trapped in structural defects in a mineral; the isms) or ‘dead’ (released to groundwater from weath- longer the exposure to radiation, the more electrons are ering carbonate rocks) carbon (Trumbore 2000). Various trapped. However, the three methods differ in the washing techniques have been developed to attempt to approach employed for measuring the amount of remove these contaminants (Bird et al. 1999; Trumbore trapped electrons: TL uses heat, OSL uses light and 2000), but many of the earlier 14C dates in Australia were ESR uses a magnetic field to release (TL and OSL) or performed prior to the introduction of these washing excite (ESR) and detect the electrons (Geyh & Schleicher procedures; therefore, the reliability of many of these 1990). results is difficult to assess. The much smaller amount In TL, OSL, and ESR dating, an age is calculated of material required for AMS 14C significantly decreases from the equation: 874 P. M. Vasconcelos et al.

De sediments in a crater of the volcano, and conventional t ¼ annual dose 14C dating of plant remains in paleosol buried by volcanic ash, which establish confidence in an eruption

where the De, the total radiation dose, is measured from date of 35 + 3 ka BP. This study underscores the the total TL, OSL, or ESR intensity and the annual dose importance and potential power of an integrated is obtained from the measured concentrations of U, Th approach to Quaternary geochronology. and K (Faure & Mensing 2005). The luminescent clock is set to zero by exposure to Cosmogenic isotopes sunlight or heat; thus for a volcanic rock, minerals formed (e.g. phenocrysts) or heated (e.g. xenocrysts) Cosmogenic isotopes represent another potentially during eruption can be dated, as can sediments exposed fruitful avenue for dating young volcanism in Aus- to sunlight before burial by pyroclastic material or tralia. Lavas extruded at the earth’s surface ascend lava flows (Geyh & Schleicher 1990). Luminescence from depths beyond the penetration length of most methods are particularly useful for Quaternary dating, cosmic rays (520 m for muons, and mostly less than as they can be applied to materials older than the upper 2 m for neutrons and protons). Once exposed at the limit of 14C(450 000 a) but younger than commonly surface, these lavas will accumulate cosmogenic iso- dated by K–Ar and 40Ar/39Ar (*100 000 a) (Geyh & topes (3He, 10Be, 14C, 21Ne, 26Al, 36Cl, etc.), whose Schleicher 1990). abundances can be used to date the exposure (extru- As with all dating techniques, some caution must be sion) by measuring the accumulation of the cosmic- used when interpreting results from luminescence ray-generated nuclide: studies. The luminescent clock may not have been fully reset if the heating event was not strong enough, or if C ¼½P=ðl þ re=LÞ=ð1 eð1þre=LÞtÞ some of the sediments had not been fully exposed to sunlight (i.e. were undergoing active reworking at the where C is the number of product nuclides per gram; P time of burial). In this respect, OSL and ESR have a is the production rate of nuclide at a flat horizontal distinct advantage over TL, as electrons in defects surface in atoms g71 a71; l is the decay constant of the suitable for OSL and ESR measurements are much cosmogenic nuclide in a71; t is the duration of irradia- more easily reset by the heating and/or sunlight tion in years (a); e is the erosion rate of the target area in exposure event than the total defect-occupying electron cm a71; r is the density (g/cm3) of the target mineral; population measured by TL. OSL and ESR have addi- and L is the absorption mean free path of neutrons in tional advantages over TL, as described in Gru¨n & the target rock (*165 g cm72) (Lal 1991). This approach Roberts (2007). Finally, it is important to note that post- can be used to calculate exposure ages (by assuming the depositional weathering can significantly perturb all erosion rate, e, to be zero) or erosion rates, by assuming three systems by changing the radionuclide concentra- steady-state erosion when t (l þ me)71. In dating tion in the surrounding environment through leaching volcanic rocks, it is generally assumed, based on flow or deposition of U, Th or K; therefore, horizons visibly textures, that e ¼ 0 and that the target surface has been affected by weathering should be avoided in lumines- continuously exposed since extrusion. A major diffi- cence studies. culty in applying this approach in volcanic units is, In Australia, thermoluminescence has been used to however, the uncertainty in ascertaining that the sur- place constraints on the age of a small number of erup- face exposed today is the original uneroded lava flow, tions in the Newer Volcanics Province of western and that sediments or soils have not sporadically Victoria and southeastern South Australia. At Mt blanketed the surface since the timing of extrusion, Downloaded By: [University of Queensland] At: 05:12 12 September 2008 Schank, South Australia, TL of baked quartz beneath which would partially shield the surface from cosmic basaltic lava yielded a date of 4930 + 540 a (Smith & rays (Stone et al. 1997). Prescott 1987), which is considerably younger than Application of 36Cl dating to two basalt flows in the originally inferred from the conventional radiocarbon Newer Volcanics Province, western Victoria, yielded result of 18 100 + 350 a BP for a sample of soft charcoal ages, corrected for geomagnetic variations, of 31.9 + 2.4 below the tuff from Mt Schank (Polach et al. 1978). At the and 58.5 + 5 ka for Mt Napier and Mt Porndon, nearby Mt Gambier volcanic complex, Robertson et al. respectively (Stone et al. 1997). The exposure ages (1996) reported a similarly young TL result of 4.2 + 0.5 obtained are consistent with stratigraphic constraints, ka on baked tuff underlying a lava flow at Valley Lake. since the cosmogenic isotope results are older than the In this case, however, the TL dates from Mt Gambier greatest 14C age of stratigraphically overlying lake support the conventional 14C dates of Blackburn et al. sediments. The results also show that the volcanic units (1982), rather than the AMS 14C on lake material (Leaney are considerably older than initially estimated based on et al. 1995). 14C dating of the sediments stratigraphically overlying Discrepancies between TL (Mortlock 1977) and 14C the flows (Stone et al. 1997). (Gill 1972; Coutts 1981; Edney et al. 1985) dates have also fuelled controversy over the age of Tower Hill volcano Summary in Victoria and highlight the difficulty in establishing reliable ages for Quaternary volcanism. Importantly, This review highlights that the chronology of volcanism however, Sherwood et al. (2004) recently reported in Australia, while the most intensely studied part of the excellent agreement between TL dating of quartz sand Cenozoic record, still requires substantial refinement. buried by Tower Hill ash, AMS 14C dating of lake core Improved temporal resolution of the volcanic record Geochronology of the Australian Cenozoic 875

will in turn provide important information about the the underlying and overlying lithologies. K–Ar dating of geological, geomorphic and climatic processes that Cenozoic volcanic rocks associated with weathered shaped the continent, as expanded in the discussion sequences in eastern Australia has been extensively below. One area where chronology of volcanism has used to indirectly date weathering profiles (Langford- played, and continues to play, an important and original Smith et al. 1966; Exon et al. 1967, 1970; Dury et al. 1969; role is in providing constrains on the evolution of Wyatt & Webb 1970; Wellman & McDougall 1974b; weathering profiles. Sutherland et al. 1977; Coventry 1979; Stephenson et al. 1980; Coventry et al. 1985; Pillans 1997). McQueen et al. (2007) have recently expanded such indirect studies to 40 39 CHRONOLOGY OF CENOZOIC WEATHERING include the high-resolution Ar/ Ar method. The age of the volcanic units in such studies provides broad The mineralogical complexity, physical competence, constraints on the age of the underlying weathering and recognisable and universal characteristics of ferri- profiles and, in most cases, shows that Australian cretes, silcretes, calcretes, gypcretes, manganocretes, weathering profiles are older than Early to Middle mottled zones and other rocks formed by intense Miocene, and many profiles are older than Eocene. weathering of a variety of protoliths reveal that weath- The approach is best applied in thick basaltic ering in Australia is not simply a process promoting the sequences, where a well-defined weathering history physical and chemical degradation of rocks. It is in fact can be extracted from the depth and chemical composi- a rock-forming process. The vast spatial distribution of tions of soils and weathered rocks preserved between weathered rocks and profiles, suggestions that these dated basalt flows. For example, in a comprehensive rocks and profiles may be correlated at great distances investigation of a 215.8 m-thick volcanic sequence (Idnurm & Senior 1978; Grimes 1979; Twidale & Bourne through the Monaro basalt, Brown et al. (1992) identified 1998) and the economic significance of weathering- 22 flow boundaries and showed that 55% of the volcanic related ore deposits (iron, manganese, nickel, alumi- sequence is weathered, that a significant number of the nium, gold, etc.) drive interest in unravelling the weathering profiles are up to 12.5 m-thick, and that some weathering history in Australia. The timing of mineral of these weathering profiles contain bauxite zones up to precipitation in weathering profiles, if accurately and 3.5 m thick. Despite the fact that individual basaltic precisely determined, provides information on Austra- flows were not dated and the fact that the number of lia’s paleoclimatic history, complementing information flows has been questioned (Sharp 1994), the range in K– obtained from the sedimentary and paleontological Ar ages (54–36 Ma: Wellman & McDougall 1974b) for the records. Paleoclimatic information derives from the fact basalts of the Monaro region and the occurrence of that weathering results from water–rock interaction weathering profiles throughout the entire sequence and is enhanced by abundant vegetation and high suggest that climatic conditions were conducive for ambient temperatures (Vasconcelos 1999a). The tempor- deep weathering throughout the Eocene, a time when al distribution of supergene mineral ages in weathering this part of Australia was at very high latitudes. profiles should, therefore, reveal favourable climatic Weathering profiles underlying the basalt flows also conditions in the past (Vasconcelos 1999a; Feng & indicate that weathering-prone conditions prevailed Vasconcelos 2007). The spatial distribution of supergene before 54 Ma, when Australia was adjacent to Antarctica mineral ages in weathering profiles should, on the other (Figure 1) (Taylor et al. 1990). hand, identify the mode of propagation of weathering fronts through time, and the duration of the weathering Direct-dating methods processes that formed the profile. Downloaded By: [University of Queensland] At: 05:12 12 September 2008 Dated weathering profiles also provide tectonic Direct-dating methods extract geochronological data information. When profiles are faulted or tilted, or when from the target weathering profile, as opposed to using coeval profiles assumed to have formed on a same dated stratigraphic constraints. Direct-dating methods surface are vertically offset by tectonism (e.g. at Tick can be either relative or absolute. In relative direct- Hill: Lilley 1997; Vasconcelos 1998), dated weathering dating methods, the property measured from the profile profiles provide useful stratigraphic markers delimiting (e.g. chemical remanent magnetisation) has to be the time of the deformation event. Dated weathering translated into an age through a calibration step [e.g. profiles provide additional information on tectonic comparison with well-dated apparent polar wonder path evolution when they can be used to constrain the ages (APWP) for methods using paleomagnetism]. In absolute of topographically distinct land surfaces (Vasconcelos & direct-dating methods, based on radiogenic-isotope tech- Conroy 2003) interpreted to result from differential niques, the age of precipitation of a mineral present in isostatic rebound through time. the profile is measured directly. Both methods are currently used to date weathering profiles in Australia. Indirect-dating methods RELATIVE DIRECT-DATING METHODS Early attempts to date weathering profiles relied on indirect methods. Ages of these profiles were bracketed The two relative direct-dating methods used in the study by the ages of the underlying bedrock and the ages of of weathering profiles are based on the measurement of datable volcanic or sedimentary units overlying the chemical remanent magnetisation and oxygen-isotope profiles. The size of the bracketing window, and the composition of pedogenic minerals. The long northward usefulness of the results, depends largely on the ages of trajectory traced by Australia during the Cenozoic 876 P. M. Vasconcelos et al.

(Figure 1) permits, in principle, resolving ages of with the present pole, irrespective of the sediment weathering if a precise pole position can be measured deposition ages. These authors interpreted the results for the profile of interest and compared with a time- to indicate a strong episode of chemical remanent calibrated APWP (Figure 4). Australia’s Cenozoic tra- magnetisation of the samples acquired during lateritisa- jectory also resulted in variations in the isotopic tion in the Late Oligocene–Early Miocene. Extension of signatures for pedogenic minerals, providing the means this approach to a sub-basaltic weathering profile in for using d18O to indirectly date weathering profiles Victoria yielded Mid-Cenozoic lateritisation ages (Figure 5). (Schmidt et al. 1976). A subsequent study by Idnurm and Senior (1978) substantiated the usefulness of the approach, yielding distinct ages for two regionally Paleomagnetic dating recognised and distinct weathering profiles in the The first extensive and successful use of paleomagnet- Eromanga Basin, Queensland: a Maastrichtian to Early ism to directly date weathering profiles in Australia was Eocene (60 + 10 Ma) age for the Morney Profile; and serendipitous (Schmidt & Embleton 1976). In attempting a Late Oligocene (30 + 15 Ma) age for the Canaway to extend paleomagnetic investigation of Late Paleozoic Profile. and Mesozoic units in Western Australia, Schmidt & Following these initial studies, the approach has Embleton (1976) analysed 222 oriented samples from been intermittently used to infer paleoclimatic and sedimentary deposits in the Perth Basin. The samples paleoenvironmental conditions and the duration of the showed remanent magnetism with poles closely aligned processes controlling weathering-profile evolution in Australia. Paleomagnetic dating of paleo-lacustrine sequences cropping out between Cooma and Bredbo, in the southern New South Wales Tablelands, yields broad age estimates of mid-late Cenozoic for weath- ering and ferruginisation of these sediments (Schmidt þ20 et al. 1982). Schmidt & Ollier (1988) provided a 6030 Ma age for weathering profiles in the New England area of New South Wales, which they correlated with the Morney profile dated by Idnurm & Senior (1978). In the absence of more precise dating methods, these studies suggest that paleomagnetism may provide useful albeit imprecise constraints on the ages of weathering profiles. However, Embleton (1981 pp. 85–86) warned against the use of paleomagnetic dating of weathering profiles: ‘During the past five years much interest has developed in the potential use of Australian Cenozoic APW as an aid to provide absolute chronological control to time the origin of laterites and weathered profiles . . . . Since the absolute chronology provided by the track, i.e., the location of the pole positions in the relation to their K– Ar ages, has now been brought into question we are no longer justified in using this approach.’ Idnurm & Downloaded By: [University of Queensland] At: 05:12 12 September 2008 Schmidt (1986) also identified two main problems limit- ing the use of paleomagnetism to date weathering profiles. The first reflects the limited accuracy and precision of the Cenozoic segment of the Australian APWP. As explained below, there are currently three competing APWPs for Australia and difficulties in determining a most reliable one (Figure 4). As a consequence, estimated ages and associated uncertain- ties are strongly dependent on the APWP used for calibrating the paleomagnetic poles measured from weathering profiles (Acton & Kettles 1996). Paleomag- netic ages obtained in earlier studies have been recently revised using currently accepted versions of the Cen- Figure 4 Cenozoic Australian apparent polar wander paths, ozoic APWP. For example, Pillans (2005) adjusted the 25– after Acton & Kettles (1996), showing discordance among 20 Ma ages of regional lateritisation for Western those derived from dated volcanic rocks with apparent pole Australia obtained by Schmidt & Embleton (1976) in positions calculated at 25 and 50 Ma (Embleton 1981), from the Perth Basin to 10 5Ma(in Anand & Paine 2002), sedimentary sequences and weathering profiles with pole + positions at 20 and 60 Ma (Idnurm 1986) and statistical and to 4 + 2 Ma (Pillans 2005). Such revisions not only treatment of paleomagnetic data from dated volcanic rocks raise questions about the accuracy and precision of and weathering profiles with pole positions at 20, 40, 60 and other results in the literature, but also raise uncertain- 100 Ma (Musgrave 1989). ties about the intrinsic reliability of paleomagnetic Geochronology of the Australian Cenozoic 877

Figure 5 (a) The isotopic signature of kaolinite defines a trend in d18O vs dD space parallel to the meteo- ric-water line, reflecting the equi- libration of the pedogenic phases with rainwater. The isotopic com- position of kaolinite from Austra- lia indicates a progressive enrichment in heavier isotopes from the Permian to the present day. This trend can be used to date weathering profiles by comparing the isotope composition of kaoli- nite with the ranges in the cali- brated curve illustrated in (b) (modified from Bird & Chivas 1989, 1993). Downloaded By: [University of Queensland] At: 05:12 12 September 2008

dates. Thus, it is always critical to publish the paleo- certain stage of weathering is reached (Idnurm 1986; magnetic data on which the ages are based, including Idnurm & Schmidt 1986; Acton & Kettles 1996). Paleo- the paleomagneitc measurements, calculated pole posi- magnetic dates are often considered to represent the tion, and details of the polar wonder path, and ages of terminal phase of lateritisation because stable rema- reference poles (B. Pillans pers. comm. 2007). nence appears to require aging of iron oxides and a The second problem arises from the limited under- minimum degree of induration (Idnurm & Schmidt 1986; standing of the origin of chemical remanent magnetisa- Acton & Kettles 1996), conditions likely to prevail during tion (CRM) in weathering profiles and the physical dehydration reactions. Consequently, paleomagnetic significance of paleomagnetic dates (Idnurm & Schmidt ages may record when a weathering profile undergoes 1986). Paleomagnetic dating records the time when dehydration, but not the onset of weathering or the weathering minerals acquire a stable chemical rema- main period of mineral precipitation. Moreover, recent nent magnetisation. However, the exact stage at which a geochronological studies (Shuster et al. 2005; Heim et al. stable chemical remanent magnetisation is acquired by 2006; Bernal et al. 2006) reveal a prolonged and complex a weathering mineral is poorly understood. Acquisition history of precipitation and recrystallisation of iron- of CRM is generally assumed to involve ferruginisation bearing phases in weathering profiles. The high-spatial under relatively oxidising conditions (Idnurm & resolution techniques used in these studies show that Schmidt 1986). Acquisition of chemical remanence may several generations of iron oxides and hydroxides, occur continuously throughout the history of a weath- precipitated at very different times in the past, may ering profile evolution, or it may commence after a coexist at the millimetre and even micrometre scale. 878 P. M. Vasconcelos et al.

Since most paleomagnetic dates on weathering profiles should take into account the effects of grainsize or rely on the analyses of centimetre-scale samples (gen- crystallite size on magnetic remanence stability. It erally 2 6 2 6 2 cm), it is likely that the CRM measured should also be applied in conjunction with independent was acquired at different times, possibly over a pro- dating techniques suitable for providing reliable ages longed period of weathering. These measurements, for the phases recording the ChRM. therefore, may not record a single pole position that can be related to a time-calibrated APWP. Oxygen-isotope dating Paleomagnetic dating of weathering profiles may also be complicated by the presence of primary minerals. The lengthy trajectory traced by Australia during the Petrographic observations, confirmed by (U–Th)/He Cenozoic (Figure 1) also permits indirect dating of results (H. Monteiro unpubl. data), reveal that even in weathering profiles through the analysis of the d18O extremely advanced stages of weathering, lateritic signatures of authigenic kaolinite (Bird & Chivas 1989). crusts overlying banded iron-formations (cangas) or The rationale for this approach is that, since approxi- Cu–Au deposits may contain unweathered grains of mately 95 million years ago, Australia has moved primary hematite and magnetite. Thus, it is important progressively northward across a latitudinal tempera- to apply a series of tests to ascertain that the measured ture gradient (Figure 1). Air temperature at the site of remanent magnetisation is characteristic remanent precipitation is strongly correlated with the isotopic magnetisation (ChRM) acquired during weathering, as composition of meteoric water. Therefore, meteoric outlined in Butler (1992). Some standard checks for water in Australia is interpreted to have become ChRM are: lack of response to alternate field demagne- progressively enriched in d18O since the mid-Cretac- tisation, which is typical of weathered materials domi- eous. The isotopic composition of weathering minerals nated by hematite; the unblocking temperature, which formed in equilibrium with these meteoric waters is different for distinct phases; time dependence of would have also become progressively enriched in d18O magnetic susceptibility; petrographic and mineralogical (Figure 5a). Bird & Chivas (1989) calibrated the change (e.g. XRD) investigations to determine the main miner- in isotopic composition of authigenic clays over time by als present in the profile (B. Pillans pers. comm. 2007). analysing samples from independently dated weath- Standard checks are also necessary to detect inherited ering profiles (Figure 5b). Four broad d18O groups of magnetisation, such as: testing whether multiple speci- kaolinitic clays can be distinguished from Early Per- mens from one site give similar results; testing whether mian to post-mid-Tertiary (Figure 5b). Matching the normal and reverse polarities in specimens from the d18O signature of kaolinite of unknown ages with these same site give the same pole position; testing whether four broad groups permits indirect dating of weathering results lie on the APWP; in folded strata, the paleomag- profiles. It is prudent to note, however, that the d18O netic fold test can be applied to determine if remanence calibration is dependent on the interpretation that the was acquired before or after folding (B. Pillans pers. ages of the weathering profiles containing the clay comm. 2007). These tests should contribute to ascertain- minerals used to derive the curve are well known, and ing that pole positions measured for weathering se- that these clay minerals have not undergone recrystal- quences are indeed recording ChRM acquired during lisation during more recent weathering events. The weathering. However, inherited magnetisation from method also relies on the assumption that the isotope primary minerals may still preclude extracting correct fractionation recorded by the present meteoric water pole directions from weathering profiles, even in cases line accurately depicts the variation in the isotopic where these procedures are used, since some of the composition of precipitation as far back as the Permian, primary phases carrying the inherited magnetisation and that an altitude-effect correction based on present Downloaded By: [University of Queensland] At: 05:12 12 September 2008 are the same minerals (e.g. hematite) believed to carry altitude applies to the study sites at the time of weath- the chemical remanent magnetisation acquired during ering (Bird & Chivas 1988). Lastly, it is implicit in the weathering. Distinguishing supergene from hypogene method that the isotopic fractionation of meteoric water hematite petrographically or by standard mineralogical was entirely latitude- and temperature-dependent, and tests is not often possible, particularly in the case of fine- did not reflect other mechanisms of fractionation, such grained hematite intergrown with other phases. In as continentality or orographic effects. addition, if the primary mineral is evenly distributed The application of this approach reveals that weath- throughout the bedrock, reproducible pole positions do ering profiles in Australia span a range of ages, dating not exclude the presence of inherited primary minerals. as far back as the Permian. Bird & Chivas (1988) Thus, pole positions measured for a weathering profile estimate, from the d18O and dD values of kaolinite in falling outside accepted values for the various Cenozoic Permian deposits, that low-temperature weathering or Mesozoic Australian APWP may reflect, rather than processes, associated with meteoric water indicating ancient weathering (Pillans et al. 1999; O’Sullivan et al. polar to sub-polar temperatures of condensation, were 2000), inherited magnetisation. responsible for the genesis of clay deposits when Not withstanding issues related to refining an appro- Australia was at high latitudes. Bird et al. (1990b) and priate APWP, future advances in paleomagnetic dating Bird & Chivas (1993) further suggest that deeply require the application of suitable experiments to better weathered profiles developed at extra-tropical latitudes determine at what stage stable magnetic remanence is formed in cool to cold and presumedly moist climatic acquired by weathering minerals (Idnurm & Schmidt conditions. To account for the lack of formation of such 1986). Ideally this characterisation should be carried out weathering profiles today, they suggest that atmo-

on single crystals or monominerallic clusters, and it spheric CO2 concentrations were higher in Mesozoic Geochronology of the Australian Cenozoic 879

and Paleogene times, when some of these profiles were periods of intense weathering related to warm and

forming. The aggressive H2CO3-bearing solutions pro- humid climatic conditions: a pre-Late Eocene period moted more active weathering, even at the presumedly (443.7 + 1.2 Ma); an Oligocene episode (ca 30 Ma); and a lower temperatures at higher latitudes. strong episode at the Miocene (ca 18–6 Ma). A similar approach by Dammer et al. (1999) showed that supergene manganese deposits in the interior of Western Australia Summary yield roughly similar results. Collectively, these studies The studies outlined above illustrate how relative suggest that episodic weathering controlled the dissolu- dating methods can provide constraints on the approx- tion and reprecipitation of Mn oxides at a continental imate ages of weathering profiles. However, in order to scale. extract reliable information on continental paleocli- Similarly, through K–Ar dating of supergene alunite, matic history from the weathering record, comparable Bird et al. (1990a) showed that weathering profiles in with the oceanic record, the resolution of weathering Australia span a broad range of ages [from ca 91.8 + 1.0 geochronology must be significantly improved. There- Ma (corrected for the presence of detrital micas to 60.9 fore, our goal must be to numerically calibrate the Ma) for Wonyulgunna Hill, Western Australia; and continental weathering record with sufficient precision 63.6 + 0.7 Ma (similarly corrected to 62.0 Ma) at Spring- and accuracy to permit comparisons with the detailed sure, Queensland; to zero age at some salt lakes in record obtained from ocean sediments. Such precision southwestern Australia]. These results provide further and accuracy can only be obtained through radiogenic- evidence for the antiquity of Australian weathering isotope techniques. profiles, as postulated from d18O measurements of kaolinite (Bird & Chivas 1989). Importantly, Bird et al. (1990a) argued that alunite most likely precipitated late ABSOLUTE DIRECT-DATING METHODS in the weathering history and that alunite ages provide The widespread application of radiogenic isotopes to only a minimum estimate for the age of formation of the study of weathering reactions (weathering geochro- weathering profiles. nology) is a relatively new development, despite the Despite the successes above, a major limitation to the long-established recognition that supergene minerals application of the bulk K–Ar method in weathering are suitable for such geochronology (Strutt 1910; geochronology arrives from the possible contamination Chukhrov et al. 1966, 1969). The most widely used of a supergene mineral sample by hypogene contami- weathering geochronology methods are K–Ar, nants, particularly micas; the presence of intimately 40Ar/39Ar, and U-series disequilibria. More recently, intergrown but distinct generations of supergene miner- (U–Th)/He, 4He/3He, and laser-ablation U-series dis- als; and the fact that some supergene minerals occur equilibria have also been applied to a variety of only as minor phases in weathering profiles. To obviate supergene minerals (sulfates, Fe- and Mn-oxyhydrox- the contamination problem, Bird et al. (1990a) and ides, carbonates, etc.). The great advantage of these Dammer et al. (1996, 1999) carefully applied a two-stage methods, particularly those that afford high spatial analytical process proposed by Chukhrov et al. (1966) to resolution, is that the phases dated can be positively identify and subtract the contribution of hypogene identified, and petrographic relationships among the contaminants to the date obtained for the weathered dated phases and surrounding minerals permit deter- assemblage. Although this approach is critical for mining the chemical process whose product is being calculation of K–Ar ages for weathering, it lengthens dated (Vasconcelos 1999b; Bernal et al. 2006; Heim et al. the dating process, preventing the analysis of a large 2006). Despite its recent history of application, a number of samples, and it also increases the uncertainty Downloaded By: [University of Queensland] At: 05:12 12 September 2008 substantial weathering geochronology database is of the final result. More importantly, it is not suitable emerging for Australia (Appendix 2*). for dating distinct generations of supergene minerals or minerals that occur as minor phases in weathering profiles. However, the 40Ar/39Ar method overcomes K7Ar and 40Ar/39Ar these limitations. Varentsov & Golovin (1987) pioneered weathering The laser-heating 40Ar/39Ar dating of single crystals geochronology in Australia through K–Ar dating of or clusters of alunite, jarosite and K–Mn-oxides (Vas- supergene cryptomelane from the Groote Eylandt depos- concelos et al. 1992; 1994a, b, c) obviates contamination it, Northern Territory. Pracejus (1989) substantiated problems and permits high-resolution analysis of a very these results and demonstrated that some of the manga- large number of samples, providing a more detailed nese minerals present in the Groote Eylandt deposit insight into the evolution of weathering profiles. This underwent partial recrystallisation during Cenozoic single-crystal or single-grain approach, possible due to weathering and thus were not solely derived from the high spatial resolution and low blanks inherent to precipitation in a shallow Cretaceous marine environ- laser-extraction systems, allows the study of weathering ment, as previously proposed (Frakes & Bolton 1984). profiles where K-bearing supergene minerals occur as Dammer et al. (1996) further extended this approach by minor phases. The incremental-heating 40Ar/39Ar meth- combining K–Ar dating with age corrections for od also provides information about the Ar and K detrital contamination and total-fusion 40Ar/39Ar dating retention histories of the mineral analysed and permits of manganese oxides. This important step forward the presence of excess/inherited Ar as well as possible revealed that supergene enrichment at Groote Eylandt contaminants to be detected. In addition, the full progressed episodically during three broadly defined automation in modern 40Ar/39Ar systems enables rapid 880 P. M. Vasconcelos et al.

analysis of multiple samples from each site, providing a strated the internal consistency of the 40Ar/39Ar geo- comprehensive database. chronology results, where different grains from the The utility of 40Ar/39Ar weathering geochronology same sample, different samples from the same site, was illustrated in a regional study of the distribution of different sites from the same weathering profile (or weathering profiles in the Mt Isa region (Vasconcelos landsurface) and different minerals from the same site 1998) (Figure 6). 40Ar/39Ar laser incremental-heating and weathering profile yield internally consistent ages, analysis of 97 individual grains of cryptomelane and conclusively demonstrating the soundness of weath- jarosite from 26 samples demonstrates that deeply ering geochronology. This internal consistency was ferruginised weathering profiles in the Mt Isa region further tested and confirmed by Li & Vasconcelos may be as old as 66.9 + 0.1 Ma. These results also show (2002), who used a large database of weathering ages that sites at higher elevations (4400 m) contain older from Mt Tabor, central Queensland (45 samples, 125 (67–30 Ma) supergene minerals; weathering profiles grains analysed by laser incremental-heating) to demon- located at intermediate elevations (350–270 m) yield ages strate that distinct weathering events can be unequi- in the 21–12 Ma range; and the samples collected from vocally identified and quantified in space and time. With lower elevation sites (5200 m) yield results younger application of appropriate analytical approaches, than 10 Ma (Figure 6). This topographic control on the 40Ar/39Ar weathering geochronology can be extended distribution of supergene minerals was further tested at into the Quaternary and permits dating samples as Dugald River, Queensland (Vasconcelos & Conroy 2003). young as ca 200 ka (Feng & Vasconcelos 2001, 2007). The analysis of 125 grains of cryptomelane, alunite and These studies show that K–Ar and 40Ar/39Ar geo- jarosite extracted from 35 samples collected at 11 chronology of supergene sulfates and Mn-oxides provide distinct sites confirms a strong topographic control on reliable constraints on the ages of weathering profiles weathering ages (Figure 7). This study also demon- and land surfaces; in turn, this information permits Downloaded By: [University of Queensland] At: 05:12 12 September 2008

Figure 6 40Ar/39Ar geochronology for widely distributed sites, sitting at different elevations, reveals a protracted history of weathering and erosion for the Mt Isa region, Queensland, spanning as far back as ca 65 Ma. The results also show a consistent pattern, where older weathering profiles occur at higher elevations, while younger weathering results are obtained for lower elevation sites (after Vasconcelos 1998). Geochronology of the Australian Cenozoic 881

Figure 7 40Ar/39Ar geochronology results on 125 grains from 35 samples collected from several sites located at three distinct elevations in the Dugald River catchment, Mt Isa region, Queensland, illustrate that the higher elevation sites host the older supergene minerals, intermediate elevation sites host minerals displaying intermediate ages, and the lowest elevation sites contain the most recently precipitated minerals. This age vs elevation relationship suggests that weathering geochronology may be successfully used to date and differentiate weathered erosion surfaces, providing much needed time constraints on landscape-evolution models. The age vs elevation relationship also permits inferring rates of erosion and weathering for the Dugald River sites, as discussed in Vasconcelos & Conroy (2003).

inferring climatic conditions in the past and may also l238, l235 and l232 are accurately known and have contribute to the resolution of longstanding landscape remained constant through time; and that the sample evolution problems. However, these methods are solely did not contain any significant initial 4He at the time applicable to K-bearing supergene minerals. Weath- when the mineral became a closed system. ering profiles host a variety of supergene phases, each In (U–Th)/He dating, the long stopping distances potentially retaining information about slightly differ- (*15–20 mm on average) of recoiled alpha particles (or ent processes. This mineralogical diversity requires the helium nuclei), compared with the size of crystals search for radiogenic-isotope techniques suitable for commonly dated (typically 100–200 mm), requires the dating minerals devoid of K. One such technique that application of corrections to account for the fraction of has much to offer in the study of the weathering history a particles ejected from a crystal. Farley et al. (1996) of Australia is (U–Th)/He dating of supergene iron defined the Ft parameter to account for a ejection, oxides and hydroxides. where Ft is the factor by which a measured age must be divided to obtain the a-ejection-corrected age. The computed Ft parameter is a function of crystal (U7Th)/He and 4He/3He geometry and dimensions, and it works well in most The (U–Th)/He isotopic system was recognised as a cases, but complications arise when U and Th are Downloaded By: [University of Queensland] At: 05:12 12 September 2008 potential geochronometer more than a century ago heterogeneously distributed (Farley 2002; Farley et al. (Rutherford 1905). (U–Th)/He geochronology of super- 1996; Reiners et al. 2004; Hourigan et al. 2005), as in gene Fe oxides is not new either (Strutt 1910), but it was zoned crystals, and when crystals have undergone only recently revisited (Lippolt et al. 1998). In the (U– extremely slow cooling (Meesters & Dunai 2002a, b). Th)/He dating method, helium is generated by radio- Selecting interior aliquots from large grains or crys- active decay of U and Th to Pb and, to a lesser extent, by tals, where helium loss by ejection of recoiled alpha Sm to Nd. Samarium contributes a negligible proportion particles is presumably balanced by implantation, may of the helium budget in most minerals and can often be overcome these effects. ignored. At secular equilibrium, U and Th release Lippolt et al. (1998) revived the application of (U–Th)/ multiple alpha particles through distinct decay series, He dating to supergene iron hydroxides by demonstrat- so the fundamental 4He ingrowth equation is ing that goethite contains significant concentrations of U, Th and He, and by showing that the calculated He 4 238 l t 235 l t 232 l t He ¼ 8 Uðe 238 1Þþ7 Uðe 235 1Þþ6 Uðe 232 1Þ results qualitatively suggest He retention. Pidgeon et al. (2004) applied similar methods to analyse four ferrugi- where 4He, U and Th refer to the measured present-day nous pisoliths and nodules from four hand specimens of amounts, t is the time of accumulation or He age, and ls a lateritic duricrust sampled within *1km2 in the 710 71 are the decay constants (l238 ¼ 1.55125 6 10 a , Darling Range, southwestern Australia. The (U–Th)/He 710 71 711 71 l235 ¼ 9.8458 6 10 a and l232 ¼ 4.9475 6 10 a ). dates ranged from 10.0 + 0.7 to 7.5 + 0.4 Ma and To calculate a sample’s age, the equation above is solved were interpreted to represent the timing of formation by iteration, assuming that the sample was closed to U, of the ferruginous pisoliths and nodules (Pidgeon et al. Th and 4He gains or losses since their formation; that 2004). 882 P. M. Vasconcelos et al.

Despite these encouraging results, the lack of quanti- Shuster et al. (2005) showed that despite the spatial tative data on He and U retentivity and information on variability of the U and Th distribution and U–Th ratios, the crystallographic sites occupied by U raises uncer- (U–Th)/He dating of goethite yields internally consis- tainties regarding the reliability of (U–Th)/He ages for tent and reliable apparent helium ages. To assess supergene oxides. In addition, the open-system beha- possible helium loss of goethite, stepped-heating diffu- viour of weathering profiles, evidence for superimposed sion experiments were performed on proton-irradiated weathering episodes through time, the suspected long- samples using the 4He/3He methodology (Shuster et al. evity of weathering profiles in Australia and the 2005). The results suggest that goethite quantitatively influence of environmental factors (solar heating, bush retains 486% of natural radiogenic helium. The results fires, lightning strikes, etc.) that might affect 4He also demonstrate that 4He diffusion losses can be retention in an iron oxide or hydroxide require careful rigorously quantified, and (U–Th)/He dating of goethite assessment before (U–Th)/He results can be interpreted yields reliable precipitation ages if corrections for as ages. The incorporation of detrital or inherited diffusive helium losses are applied. The reliability of phases in supergene Fe-oxyhydroxides (Bernal et al. the (U–Th)/He goethite precipitation ages was verified 2006) adds further complexity in interpreting (U–Th)/He by agreement with 40Ar/39Ar ages obtained from coex- results as ages. Analytical approaches suitable for isting Mn-oxides in samples from , Mt overcoming the difficulties above include ascertaining Isa (Shuster et al. 2005). that only pure minerals are dated; multiple analysis of Heim et al. (2006) also used 4He/3He diffusion loss grains from the same sample; analysis of samples corrections to date late-stage authigenic goethite in the displaying clear paragenetic sequences to test whether Yandicoogina paleochannel iron deposit, Western Aus- the relative timing relationships are honoured by tralia. Corrected (U–Th)/He ages (21 grains analysed absolute dating results; dating distinct minerals in the from eight distinct field samples) range from ca 18 Ma same profile, through independent methods, to test the near the surface to 5 Ma at the bottom of the profile, compatibility between independent results; and devel- indicating that ferruginisation of the channel sediments oping suitable analytical approaches to test 4He diffu- becomes progressively younger with depth in the sivity and retentivity history in specific mineral phases. profile. The results strongly suggest that ferruginisation Recent experimental investigation on uranium co- of channel sediments was driven by drawdown of the precipitation with Fe oxyhydroxides provides compel- water-table during the Neogene, with goethite cementa- ling evidence that U is incorporated into oxide struc- tion occurring at the paleo-groundwater interface. The tures and is not simply adsorbed onto mineral surfaces geochronological results, when extrapolated to the sur- (Duff et al. 2002). A crucial step in ascertaining the face of the channel, also provide an estimated minimum reliability of (U–Th)/He ages, however, is testing the age for aggradation of the paleochannel at approxi- quantitative retention of 4He during a sample’s geologi- mately 36 Ma (Heim et al. 2006). cal history. The 4He/3He methodology (Shuster et al. Despite the great utility of the (U–Th)/He and 2004; Shuster & Farley 2004) provides the basis for 4He/3He methods for dating ferruginisation in weath- quantitatively assessing the He retention properties of ering profiles, the method is not easily applied in minerals. The approach is based on measuring the measuring ages of young samples not in secular spatial concentration of radiogenic 4He and 3He gener- equilibrium. Therefore, to date the precipitation of iron ated by proton-irradiation in the same grain. The spatial oxides and hydroxides younger than ca 500 ka, an concentration of radiogenic helium in rocks and miner- alternative approach is necessary. This approach is als reflects a delicate balance between the production of provided by the U-series disequilibria method. radiogenic 4He and removal of this helium by thermally Downloaded By: [University of Queensland] At: 05:12 12 September 2008 activated volume diffusion and alpha ejection. The U-series disequilibria spatial distribution of helium in a single U- and Th- bearing crystal is thus an evolving function of the The radioactive nature of intermediate daughter pro- sample’s time–temperature history. This relationship ducts in the decay series of 238U, 235U and 232Th to 206Pb, can be summarised by the following function: 207Pb and 208Pb, respectively, can be used to date samples Z precipitated by geological processes that cause the today nuclides in a decay series to be fractionated (weath- ½Production ðx; y; z; tÞRemoval ðx; y; z; T; tÞ dt to ering, sedimentation, melting, etc.), resulting in secular ¼ Distribution ðx; y; z; todayÞ disequilibrium. Samples in secular disequilibrium can be dated by the quantification of the excess or deficiency where Production represents the time-dependent accu- of a daughter product. Modern U-series disequilibria mulation of radiogenic 4He from U and Th (and Sm); dating using TIMS and LA-MC-ICPMS (laser ablation Removal is the time (t)- and temperature (T)-dependent multi-collector ICPMS) obviates problems associated function of diffusive 4He loss and a-ejection; Distribu- with large samples and dissolution of complex mineral tion is the spatial concentration distribution of 4He assemblages encountered by techniques that rely on a-

measured in a sample today; and the subscript to is the counting, and it provides reliable information on a time helium initially accumulates. 4He/3He thermo- variety of weathering-related (carbonates, oxides, opal, chronometry provides the analytical tools to constrain etc.) and sedimentary minerals. the Distribution function in a sample today and the Pioneering the use of U-series disequilibria to date Removal function (via the 4He diffusion kinetics) during supergene minerals in Australia, Short et al. (1989) a sample’s history (Shuster & Farley 2005). demonstrated that supergene Fe and Mn oxides from the Geochronology of the Australian Cenozoic 883

71 Alligators Rivers Uranium Province, Northern Terri- the product isotope in y (equal to ln2 / t1/2, where t1/2 tory, contained irreversibly adsorbed U and that these for 10Be is 1.5 6 106 y and for 26Al, 7.01 6 105 y) (Figure 8) oxides quantitatively retained authigenic 234U and 230Th, (Heimsath et al. 2000). For steady-state cases, when the but not 226Ra. They also demonstrated that 234U/238Uin erosion rate (e) and the soil depth remain constant and t groundwaters was consistent with the 230Th/234U activ- 44 (l þ me)71, the cosmogenic nuclide concentration is ity in supergene oxides. Importantly, the oxyhydroxides given by: 14 provided Th/U ages compatible with TL and C ages of the hosting sediments, indicating that supergene oxides lt 1 C ¼ C0e þ PðH; yÞ may be suitable for dating weathering reactions. l þ mes Further application of this method to coastal dune fields at Groote Eylandt confirmed its validity in dating where P(H,y) is the nuclide production rate (atom g71 71 pedogenic ferruginous pisoliths (Shulmeister et al. y ) at the soil–bedrock interface on a slope, y , and C0 is 1993). However, the difficulties and uncertainties asso- the initial concentration of the nuclide (Lal 1991; ciated with sample dissolution and the a-counting Heimsath et al. 2000) (Figure 8).

technique resulted in limited application of the method. Assuming C0 ¼ 0 and steady state, it is possible to use A significant advance in U-series dating of pedogenic 10Be and 26Al concentrations to measure soil formation material is the recent development and application rates (Heimsath et al. 2000). Using a study site in the of in situ LA-MC-ICPMS analysis of supergene Bega Basin, New South Wales, Heimsath et al. (2000) Fe-oxyhydroxides (Bernal et al. 2005, 2006). The app- calculated soil-production and erosion rates on the roach provides all the advantages of modern analysis: coastal plains east of the Great Escarpment (Figure 8). in situ analysis provides and preserves spatial informa- From 26Al and 10Be concentration measurements in 14 tion; laser-ablation permits high spatial resolution; MC- bedrock or saprolite samples from the soil–bedrock ICPMS affords excellent detection limits and analytical interface, Heimsath et al. (2000) showed that soil speed; and the full-automation yields statistically sig- production decreases exponentially with soil depth nificant databases. Furthermore, 230Th/238U measure- according to the function: e(H) ¼ (53 + 3)e7(0.020 + 0.001)H, ments in the same ablation pits used for calculating where soil-production rate is given in m/Ma and soil 230Th ages provide additional controls on the open- or depth in centimetres. The results reveal maximum soil closed-system behaviour of the oxides analysed. The 115 production and erosion when there is no soil mantle, 230Th-ages measured by Bernal et al. (2006) provide a while production and erosion rates decrease to a statistically significant database that reveals an episodic minimum of 7 m/Ma under a local soil maximum depth distribution of ages. These ages are consistent with of 100 cm (Figure 8). A similar study at a site in the warm periods defined by the global d18O composition of highland region of southeastern Australia (Frogs Hol- oceanic sediments (SPECMAP: Imbrie et al. 1984) and low, near Canberra, at 930 m elevation and 12 km west were interpreted as evidence for global control on of the Great Escarpment) yields a soil-production continental chemical weathering (Bernal et al. 2006). function of e(H) ¼ (143 + 20)e7(0.042 + 0.003)H (Heimsath While weathering geochronology provides a robust et al. 2001). Based on the measured soil depths in the approach for solving significant questions in surficial area, Heimsath et al. (2001) estimated soil-production processes, it is not a panacea. The superposition of rates of 15–25 m/Ma for the area and inferred that the weathering fronts through time, the irregular advance soil mantle in the area could have been generated in of weathering solutions, the complex precipitation– 530 000 years, and perhaps as fast as 10 000 years. dissolution–reprecipitation of supergene minerals in In contrast to the measurements of Heimsath et al. weathering profiles, and the simultaneous action of (2000, 2001), which showed that soil production de- Downloaded By: [University of Queensland] At: 05:12 12 September 2008 weathering and erosion prevent some basic questions creases exponentially with soil depth, Wilkinson et al. from being answered, such as how long it takes to form a (2005) concluded, from a similar cosmogenic isotope soil profile. Cosmogenic isotopes provide a promising study, that soil production in ferruginised sandstones in avenue in this line of research (Heimsath et al. 1997). the Marrangaroo Creek, Blue Mountains, does not show significant depth dependence. The soil-production rates measured for a forested plateau (10–16 m/Ma) and for Soil-production functions from cosmogenic-isotope heath-mantled spurs (13–24 m/Ma) is compatible with exposure chronology erosion rates estimated for the area, and the absence of Cosmogenic isotopes, particularly 10Be and 26Al formed depth dependence in the soil-production function is in quartz in bedrocks, saprolites, soils and sediments, attributed to lithological controls (Wilkinson et al. 2005). provide the means to quantify soil-production rates, and This lithological control on erosion rates is also erosion rates, as formulated by Heimsath et al. (1997). observed in other studies of Cenozoic erosion in The change in the concentration of a cosmogenic isotope Australia.

(C) at a depth zx at time t is given by: dC ¼ P emzx lC dt 0 CHRONOLOGY OF CENOZOIC EROSION

where P0 is the production rate at a flat surface, m is the The absence of significant orogenic processes in Aus- absorption coefficient (m ¼ r/L), i.e. the material density tralia since the Early Mesozoic has led to the proposal in g/cm3, r, divided by the mean attenuation length for that the Australian landscape is very ancient (Twidale cosmic rays, L, *165 g/cm2); l is the decay constant of 1994). An exception is the eastern margin, where 884 P. M. Vasconcelos et al.

Figure 8 (a) Measurements of concentrations of cosmogenic isotopes (10Be and 26Al) to quantify soil-production rates (e) from a bedrock of density rr overlain by a soil mantle of thickness H and density rs at a depth Z below the ground surface with a slope q was applied to a soil profile in the coastal plains of eastern Australia, in the Bega River catchment, New South Wales, as outlined in Heimsath et al. (2001). The results (b) yield a soil production function e(H) ¼ (53 + 3)e7(0.020 + 0.001)H, yielding maximum erosion rate of 53 + 3 m/Ma on bare rocks and a minimum of 7 m/Ma under a local soil maximum depth of 100 cm. Also shown in (b) are a stream sediment sample, showing a catchment average erosion rate of 51.5 + 4.8 m/Ma and bedrock erosion rates measured on tors, as illustrated in (c). The diagram in (d) plots the measured 10Be and 26Al concentrations against predicted values assuming both steady-state erosion (continuous line) and a two-stage erosion history (dashed line) showing that the distribution of cosmogenic nuclides is compatible with steady-state erosion, but not compatible with a two-

stage erosion history. The measured profile matches a predicted profile with es ¼ 25 m/Ma and er ¼ 8 m/Ma, where es and er are

Downloaded By: [University of Queensland] At: 05:12 12 September 2008 the soil and bedrock (tor) erosion rates, respectively (modified from Heimsath et al. 2001).

pronounced relief was interpreted as indicative of Australia. It is also a technique where contributions significant uplift and denudation in the Late Cenozoic by Australians to fundamental developments in the (Figure 9). In order to test some of these ideas, methodology, analytical approaches, and theoretical chronological tools suitable for measuring rates of solutions have led the field. In a series of classic papers, erosion are necessary. Four such tools have provided issues such as the analysis of the qualitative (Green important insight into the Cenozoic evolution of the et al. 1986), quantitative (Laslett et al. 1987) and variable Australian landscape: K–Ar, AFTT, (U–Th)/He thermo- temperature behaviour (Duddy et al. 1988) of thermal chronology, and cosmogenic-isotope studies. The K–Ar annealing of fission tracks in apatite resulted in a methodology has been discussed above, and its applica- quantitative modelling technique (Green et al. 1989) that tion in landscape evolution studies is detailed below, in permitted prediction of fission-track parameters in the Discussion, and will not be reviewed here. geological situations from laboratory-based descriptions of annealing kinetics. The most significant consequence AFT and ZFT thermochronology of these developments was the recognition that fission- track thermochronology should not rely solely on a After K–Ar dating of volcanic rocks, apatite fission-track fission-track age, but that a more complete thermal thermochronology (AFTT) is clearly the most widely history could be extracted from combining information applied geochronological approach used in unravelling obtained from a sample’s age, single-grain ages and the Cenozoic tectonic and paleoclimatic history of track-length distributions (Gleadow et al. 1986a, b; Green Geochronology of the Australian Cenozoic 885

Figure 9 Digital Elevation Model, plotted from NASA SRTM 3’’ data (from Sandiford 2008), showing the contrast between the higher-elevation eastern seaboard, which hosts most of Australia’s Cenozoic volcanism and is characterised by relatively young fission-track ages, and the low-elevation remainder of Australia, characterised by older fission-track ages and a distinct lack of volcanism. Also shown are the locations of weathering profiles reported to record Paleozoic pole positions (large black-filled circles) in Western Australia, Northern Territory and New South Wales (Pillans et al. 1999; Pillans 2004), which suggest the preservation of weathering profiles dating as far back as the Paleozoic. Downloaded By: [University of Queensland] At: 05:12 12 September 2008 et al. 1989). Galbraith et al. (1990) and Gallagher (1995) (2002a, b) and Kohn et al. (2002). In addition, Kohn et al. took advantage of the annealing-kinetics models for- (2002) also presented an approach, recently revised in mulated by Laslett et al. (1987) and generated computa- Kohn et al. (2005), for interpolating thermal histories tional methods for taking full advantage of fission-track and reconstructing continental-scale paleorelief and data to derive thermal histories. A significant assump- denudation histories through time. These results, sum- tion in fission-track thermochronology studies of ex- marised in Figures 10 and 11, reveal an episodic posed land surfaces is that cooling equates with erosion erosional history for Australia and illustrate that the of overlying cover. In order to estimate erosion, another Early Cenozoic (50–40 Ma) corresponds to a significant practical assumption in AFTT and ZFTT is necessary: period of denudation, when average continental denu- that geothermal gradients have been constant through dation rates reached 20–25 m/Ma. The regional distribu- time. This assumption is necessary because measuring tion of AFT ages illustrated in Figure 10 clearly shows paleo-geothermal gradients is not possible unless ver- that this erosive event, interpreted from the spatial tical sections through the crust are available, which is distribution of low apatite fission-track ages, is not not the case for most continental study sites. evenly distributed throughout Australia, and that it is Regionally extensive studies, applying this more particularly strong along the eastern and southeastern comprehensive approach to the interpretation of margin of the continent, and in Tasmania. The anom- fission-track thermochronology data, were employed to alous areas in South Australia and the Gawler Craton unravel the denudation history of Australia recorded in (Figure 10) may be associated with thermal anomalies in the fission-track ages and track-length distributions. these areas, and not erosive events (Kohn et al. 2002). The cumulative knowledge generated in decades of Modelling track-length distributions also reveals sig- fission-track studies is summarised in Gleadow et al. nificant Cenozoic erosive events in relatively stable Downloaded By: [University of Queensland] At: 05:12 12 September 2008 886 .M Vasconcelos M. P. tal. et

Figure 10 Apatite fission-track age map of Australia illustrates interpolated results from the regional distribution of apatite central ages (in Ma). A noticeable feature in the map is the narrow belt of young AFT ages along the eastern coastal plains and a few interior locations (e.g. Flinders Ranges and central South Australia) (from Gleadow et al. 2002a, b). The younger results along the coast may reveal more active denudation in the recent past in these areas, while the interior sites identify areas of possible anomalous heat flow. Geochronology of the Australian Cenozoic 887

Figure 11 Smoothed average denudation chronology for Australia, as illustrated and discussed in Kohn et al. (2002). Note the maximum denudation peak in the Cenozoic, possibly driven by enhanced tectonic activity driven by (1) spreading phase in Andare Trough and (2) initiation of fast spreading in Southern Ocean leading to the final separation of Australia and Antarctica. It is worth noticing that the Australian Plate collision with Papua New Guinea (3) is not associated with increasing denudation in the continent.

cratonic areas, where no Cenozoic ages appear (i.e. Mt method more than a century ago (Rutherford 1905; Isa Block) (Spikings et al. 2002). The erosional history of Strutt 1908), and revisited and re-abandoned in the Australia, summarised in Figure 11, suggests that two 1950s (Hurley 1954), its potential as a thermochron- Cenozoic tectonic events (i.e. the possible spreading ometer was fully realised by the isothermal step-heating phase in the Adare Trough and the commencement of analyses carried out by researchers in Australia (Zeitler fast spreading in the Southern Ocean, leading to the et al. 1987). Despite this significant discovery and the full final separation of Australia and Antarctica) may be recognition of the potential utility of the method, the important in promoting accelerated continental erosion. practical utility of (U–Th)/He was not properly realised However, the collision between northern Australia and until Farley et al. (1996) provided the mathematical Papua New Guinea does not lead to any apparent framework for quantitative evaluation of the alpha- denudation rate increase, at least as recorded in the stopping effects on (U–Th)/He ages, and Wolf et al. (1996) AFT annealing history (Kohn et al. 2002) (Figure 11). demonstrated that apatites from distinct sources yielded An important conclusion arising from the compre- remarkably similar diffusion properties and closure Downloaded By: [University of Queensland] At: 05:12 12 September 2008 hensive results obtained from the AFT data is that temperatures at *75 + 58C. Collectively, these studies erosion rates in Australia, although not high by global confirmed the suitability of apatite for high-precision standards, are not as low (0–2 m/Ma) as some paleogeo- chronometry of low-temperature geological events. graphic reconstructions propose (Stewart et al. 1986). Since then, (U–Th)/He laboratories have proliferated, The AFTT rates are also significantly higher than and the application of (U–Th)/He dating to a variety of the short-term rates (*0.7 m/Ma) obtained from geological processes has seen a dramatic resurgence. some cosmogenic-isotope studies (Bierman & Turner Despite the fact that modern (U–Th)/He thermochro- 1995). Another important conclusion of the continental- nology was developed in Australia, very few studies scale analysis of fission-track data is the realisation have taken advantage of the technique to address that shallow-crustal processes, not detectable through geological problems on Australian soil. A notable and AFT results, may have contributed to the erosional important exception is the application of (U–Th)/He history of Australia. Thermochronological methods thermochronology to test proposed models for the sensitive at lower temperatures are required to address evolution of the Great Escarpment (Persano et al. this issue. 2002). This study provided (U–Th)/He ages for surface samples collected along a traverse from the coast, (U–Th)/He thermochronology throughout the coastal plains, up the escarpment, and across the highlands in southern New South Wales A method potentially suited to provide thermochrono- (Figure 12). The results reveal a bimodal distribution of logical information at a lower temperature window is ages, where most of the samples collected along the coast (U–Th)/He, particularly apatite (U–Th)/He thermochron- cluster around 100 Ma (mean age of 102 + 9 Ma), while ology. Originally considered as a potential dating the samples on the highlands yield (U–Th)/He ages in 888 P. M. Vasconcelos et al.

Figure 12 Schematic east–west cross-section (after Persano et al. 2002), showing bimodal distribution of Early Mesozoic apatite (U–Th)/He ages at the crest of the escarpment in the southeastern highlands in New South Wales and Late Mesozoic ages along the coastal plain, suggesting rapid development of the escarpment and coastal lowlands after rifting and continental breakup at ca 100–85 Ma, followed by limited erosion and landscape stability throughout the Cenozoic.

the 247–183 Ma range. These results are interpreted to erosion rates were discussed above. Bierman & Turner indicate that the coastal plains must have formed by (1995) pioneered these applications in Australia by rapid downwearing or scarp retreat in the mid-Cretac- showing that high activities of cosmogenic 10Be and eous and, when combined with published fission-track 26Al in bedrock samples from the Eyre Peninsula, South results for the same area, suggest the removal of 2–4 km Australia, suggested exceptionally low rates of bedrock of overlying crust in a short interval (*30 million years erosion (0.7 + 0.1 m/Ma) (Figure 13). In a more compre- duration) immediately after the opening of the Tasman hensive study, Bierman & Caffee (2002) showed, based Sea. The results are also interpreted to indicate that the on the activities of 10Be and 26Al in 61 surface bedrock uplands west of the escarpment witnessed erosion of samples, that all samples (except one) had been con- only 1–2 km of overlying crust in the Early Mesozoic tinuously eroding at model erosion rates ranging from and have remained relatively stable ever since. Persano 0.3 + 0.1 to 5.7 + 1.0 m/Ma. Nuclide activity decreases et al. (2006) have applied similar approaches [combined from the sites in the south (Eyre Peninsula) to sites in AFT and (U–Th)/He thermochronology] to study the the north (Northern Territory), and the decrease in landscape evolution history for the Bathurst region, nuclide concentration is proportional to rainfall, sug- New South Wales, arriving at similar [but controver- gesting a climate-controlled increase in erosion rates sial (Green & Duddy 2007; Brown 2007; Gibson 2007)] from south to north. The results also show that the conclusions about the evolution of the Great Escarp- Australian landscape is undergoing a slow but measur- ment in this region. Despite the higher sensitivity of the able rate of erosion. The slow rates indicate that the (U–Th)/He method, the interpretation of thermochrono- landscape in Australia, although evolving, preserves logical results is still evolving, and differing views on forms that may date as far back as the Mesozoic. The cooling histories and the effect of re-heating events will rates of erosion measured for bare inselbergs is different have to be resolved through carefully designed field from the erosion rates measured for inselbergs with Downloaded By: [University of Queensland] At: 05:12 12 September 2008 studies, modelling, and integration of complementary bouldery tops. Boulders and surface tops on bouldery geochronological and thermochronological approaches inselbergs have lower rates of nuclide activity than bare with geological observations. inselbergs. This observation is consistent with geo- The only other tentative use of (U–Th)/He to deter- morphic interpretations suggesting that bouldery insel- mine rates of erosion in Australia (Belton et al. 2004) bergs evolve by weathering and formation of deep revealed that, in some cases, excess 4He associated with saprolites surrounding core stones; when the saprolite implanted 4He or the presence of micro-inclusions of is quickly eroded, the core stones are exposed at the monazite in apatite yields uninterpretable results. surface. Eventually, the boulders disintegrate faster Despite these problems, widespread application of than the surface of the inselbergs, exposing the more (U–Th)/He is likely to contribute to a better under- resistant bare inselberg surfaces. The results also show standing of the Cenozoic erosional history of Australia. that saprolites surrounding inselbergs erode faster than However, studies targeting the very recent history (10–5 the exposed bare rocks. Extrapolating relative erosion Ma) of erosion require techniques sensitive to the rates for saprolite and inselberg back in time, taking removal or thin layers (metres to 100 m scale) of cover. into account the height of inselbergs above the sur- In these cases, in situ-produced cosmogenic isotopes rounding plains, reveals that some of the inselberg provide the most reliable approach. landscapes in Australia may date as far back as 70 Ma, confirming suggestions that Australia hosts some of the Cosmogenic isotopes most ancient landscapes on Earth (Twidale 1993). Belton et al. (2004) showed, through the application of Basic principles in the application of cosmogenic cosmogenic isotopes at the Davenport Range, Northern isotopes in exposure dating and quantification of Territory, that the Ashburton Surface has been eroding Geochronology of the Australian Cenozoic 889

Figure 13 Nuclide-ratio diagram illustrating that all samples col- lected for Eyre Peninsula insel- bergs reveal a history of constant exposure or exposure under a thin (0.5 m) soil layer. These results reveal a simple and slow history of erosion for this part of Austra- lia, with average erosion rates of *0.7 m/Ma (from Bierman & Turner 1995).

at a rate of 0.34 + 0.03 m/Ma, while dissected valleys model erosion rate calculated from the soil-production have been eroding at a rate of 2.45 + 0.39 to 3.93 + 0.15 function. These results reveal that the short-term erosion m/Ma for the past 0.5–1 million years. These results rates for the coastal plains in eastern Australia, although suggest that the landscape in the Davenport Ranges is faster than average erosion rates estimated from AFTT undergoing slow but active denudation and that local studies, are only one tenth of the rates of erosion landforms, although ancient from a global perspective, estimated for this region from scarp retreat models are not likely to date as far back as the Cambrian, as (Weissel & Seidl 1998). Heimsath et al. (2000) interpreted previously proposed (Stewart et al. 1986). It is more this discrepancy as an indication that the landscape in the likely that those landscapes represent re-exhumation of region has evolved from processes (e.g. landslides) other ancient unconformities, as suggested by Twidale (1994). than soil production and creep. The soil-production studies of Heimsath et al.(2000, The application of the same approach to a site on the 2001) and Wilkinson et al. (2005) provided important upland area of the Great Escarpment shows quite information on erosion rates for the eastern margin of distinct results, with a soil-production function of Australia. Soil-production functions estimated by cosmo- e(H) ¼ (143 + 20)e7(0.042 + 0.003)H, and the indication from genic nuclides assume that soil formation and erosion the cosmogenic-isotope profile on a tor that erosion rates rates are equal (steady-state landscape). The steady-state were not in steady state: it appears that the region soil-production function of Heimsath et al. (2000), suffered a recent episode of accelerated denudation e(H) ¼ (53 + 3)e7(0.020 + 0.001)H, reveals maximum erosion (Heimsath et al. 2001). Combining measured soil depths Downloaded By: [University of Queensland] At: 05:12 12 September 2008 on bare rocks, while erosion rates decrease to a minimum with the soil-production function above suggests an of 7 m/Ma under a local soil maximum depth of 100 cm. average erosion rate of 25–35 m/Ma for the plateau at the Heimsath et al. (2000) used exposed tors embedded in the Frogs Hollow site. On the other hand, the average soil profiles to test the steady-state assumption. By denudation, measured from the concentration of cosmo- measuring cosmogenic-isotope concentrations at various genic isotopes on stream (15 m/Ma) and catchment (16 positions along the tor’s surface (Figure 8) and plotting m/Ma) sediments, is lower than that predicted from the the measured 10Be and 26Al concentrations against soil-production function. The lower rates measured predicted values assuming both steady-state erosion and from stream sediments, stream incision into bedrock a two-stage erosion history (slow erosion accelerating (9 m/Ma), and bare-rock erosion rates measured from after a time interval) (Figure 8), the authors showed that tors (3.8 m/Ma) suggest that the landscape may have the distribution of cosmogenic nuclides is compatible undergone rapid stripping in response to recent, possi- with steady-state erosion, but not compatible with a two- bly climatic-driven, changes. The effects of these stage erosion history. The measured profile matches a changes in the Australian landscape are best revealed

predicted profile with es ¼ 25 m/Ma and er ¼ 8m/Ma, by the chronology of sedimentation. where es and er are the soil and bedrock (tor) erosion rates, respectively. These results reveal that tors form when the saprolite around them is eroded faster than corestones CHRONOLOGY OF CENOZOIC SEDIMENTATION embedded in the saprolites. Heimsath et al. (2000) also AND DIAGENESIS measured cosmogenic nuclide concentrations in a stream sediment sample, showing a catchment average erosion The application of geochronological techniques to the rate of 51.5 + 4.8 m/Ma, compatible with their maximum study of sedimentation in Australia is extremely diverse 890 P. M. Vasconcelos et al.

and rapidly evolving. A great motivation for the develop- & Freeman (1999) concluded that the ‘peneplanation’ that ment and application of novel geochronological ap- reduced the Yilgarn Block to its present form occurred proaches to unravel the evolution of the Cenozoic before the Mesozoic, and since that time, denudation in sedimentary cover in Australia is the fact that these Western Australia must have been minimal. In contrast, sediments host important information, not derived from Cawood et al. (2003) showed that detrital zircons from other geological features, about the climatic evolution alluvial sands along the Frankland River, Western Aus- and the subtle but active tectonism that have contributed tralia, reveal changes in age populations that correspond to slowly shaping the Australian landscape. This sedi- to the underlying basements. These authors concluded mentary cover also hosts valuable mineral deposits and that renewed river incision, driven by a recent drop in potentially conceals even more valuable ones. Finally, base level, is currently eroding the local basement, and Cenozoic sediments host the remains of the first inhabi- suggested that some sediments currently derived from the tants of Australia, providing crucial information about Yilgarn Block may be temporarily stored in upstream the arrival of humans in this part of the globe. Some reservoirs. If correct, this may explain the paucity of geochronological techniques used to study the Cenozoic detrital Archean zircons in some marginal deposits. sedimentation (e.g. SHRIMP U–Pb dating of zircons or AFFT) provide information on sediment provenance. 40 39 ZFT and AFT dating of detrital zircons and Other techniques (K–Ar, Ar/ Ar, (U–Th)/He, U-series, apatites OSL, TL, etc.) yield information on the depositional or post-depositional history of the sediments. A similar and important approach in provenance studies was the realisation that apatites are very U–Pb dating of sedimentary zircons resistant to chemical weathering and that detrital apatites contain information about the thermal history SHRIMP U–Pb dating of sedimentary zircons provides of their source areas (Gleadow & Lovering 1974). The information on the Mesozoic–Cenozoic erosional history retention of fission tracks and the stability of apatite of Australia. Although the zircons themselves are not during weathering permit the minimum ages of sedi- necessarily Cenozoic, the weathering and erosional ments in deposits that have not undergone significant processes that delivered them to the depositional sites post-depositional re-heating to be determined. Where are crucial for our understanding of the Cenozoic sediments are deeply buried and re-heated, AFFT surficial history of the continent. The approach, pio- provides constraints on the thermal evolution of sedi- neered by Pell et al. (1997, 1999), is based on comparing mentary basins, a valuable tool in petroleum exploration detrital zircon ages with zircon ages from suspect (Gleadow & Lovering 1974; Gleadow et al. 1983). Detrital protoliths to determine the source areas for sedimentary sphene and zircons have a higher retention temperature deposits. The high resolution and analytical speed than apatite and are more resistant to chemical and possible with SHRIMP geochronology provide statisti- physical processes during weathering and erosion, cally valid databases on age populations for detrital providing an even better medium for fission-track zircon. This approach allowed Pell et al. (1997, 1999) to provenance studies (McGoldrick & Gleadow 1978). The fingerprint the sources of sand for three Australian use of sedimentary zircons in tracking potential sources deserts: the Great Victoria Desert, the Simpson Desert for detrital sapphires in eastern Australia was reviewed and the Mallee Dunefield. These initial studies showed above in the chronology of Cenozoic volcanism. that various individual protosources provided material for each individual desert; that some protosources Magnetostratigrapy were local while others were as far as 850 km away; Downloaded By: [University of Queensland] At: 05:12 12 September 2008 that most of the protosources no longer yield detrital In magnetostratigraphy, a measured pattern of magnetic material, suggesting significant changes in the climatic polarities in a stratigraphic sequence is correlated with regime in Australia; and that long-range transport of the known geomagnetic polarity time-scale (Cande & Kent eolian material is limited, and most deserts derive their 1995). Neogene sequences are often the subject of magneto- sand locally from reworked fluvial and marine deposits. stratigraphic investigations because analytical uncertain- Sircombe (1999) and Sircombe & Freeman (1999) ties that are inherent in radiometric dating generally limit expanded this application to study the provenance of application to sequences younger than about 5 million zircons in coastal placer deposits in eastern and western years. That is, radiometric dating, notably K–Ar dating of Australia. The age distribution for east Australian zircons basalts, is used to calibrate the geomagnetic polarity time- showed that nearby protosources are not necessarily the scale; beyond 5 Ma, the errors in K–Ar ages approach the major contributors of detrital zircons for coastal sand typical duration of polarity intervals. The Brunhes/ dunes, and that modern coastal sediments were derived Matuyama polarity transition at 0.78 Ma marks the last from tectonic blocks no longer exposed on the continent. major reversal in the Earth’s magnetic field, from They also illustrated the importance of sediment recycling. reversed to normal polarity, and its precise calibration Zircon populations in central eastern Australia, for at 780 + 10 ka (Spell & McDougall 1992) provides a valuable example, were recycled from Triassic sandstones from chronostratigraphic marker. Similarly, the demarcation the Sydney Basin. Similarly, zircons from placer deposits of the Jaramillo normal polarity subchron from 1010 to in Western Australia were probably derived from Proter- 915 ka (Spell & McDougall 1992) contributes to high- ozoic orogenic sources eroded during the Jurassic– resolution magnetostratigraphic calibrations. Cretaceous rifting event, and not the nearby Magnetostratigraphy has played an important role in Archean Yilgarn Craton, as previously assumed. Sircombe unravelling the Neogene climatic, faunal, and floral Geochronology of the Australian Cenozoic 891

evolution of Australia. Idnurm & Cook (1980) provided yield zero ages for the authigenic minerals. The fine tentative correlations between beach deposits, tempo- spatial resolution of the 40Ar/39Ar and (U–Th)/He rally calibrated through magnetostratigraphy, with methods should overcome the contamination problem insolation maxima, and suggest that the cyclicity in and provide a more comprehensive record for chemical dune formation is consistent with sea-level rise and sedimentation and diagenesis in salt lake deposits. drop associated with Milankovitch cycles. Singh et al. Australian salt lakes are prime research targets waiting (1981) and Chivas et al. (1986) employed magnetostrati- for detailed geochronological resolution. graphy to study the sedimentation history of saline lakes in Australia. The time-calibrated sedimentologi- (U–Th)/He of channel iron deposits cal, mineralogical and fossiliferous records provide information on the aridification of the interior of In addition to partially ferruginised channels feeding Australia throughout the Late Cenozoic. This approach salt lake deposits, Western Australia hosts unusual has also been used to date major hydrological and occurrences of ferruginised paleo-river deposits which climatic changes in the continent (An et al. 1986; Chen & today constitute major iron-ore resources (Ramanaidou Barton 1991; Zheng et al. 1998; Pillans & Bourman 2001). et al. 2003). The origin of these paleo-channel deposits is For example, results from five different sites across controversial, and the high degree of ferruginisation southern Australia indicate that the timing of onset of and iron-cementation of the detrital phases is a unique full arid conditions post-dated the Brunhes/Matuyama Australian occurrence (with one minor exception in polarity transition (An et al. 1986; Pillans & Bourman Kazakhstan). (U–Th)/He dating of authigenic goethite 1996; Zheng et al. 1998). Arid conditions appear to have cements in these channel deposits suggests that channel set in slightly earlier (around 900 ka) in the Amadeus aggradation may have occurred as far back as ca 36 Ma Basin, central Australia (Chen & Barton 1991), while (Heim et al. 2006) and that ferruginisation is a protracted alternating wet and dry episodes occurred in northeast process that may have spanned the entire Neogene (see Queensland on the Great Dividing Range (Lake Bucha- the weathering section above). Further application of nan) during the past *1.6 million years (Chivas et al. this approach will undoubtedly provide a clearer view 1986). These records provide conclusive evidence for the on the genesis of these unique sediments. influence of global Cenozoic climatic deterioration on the progressive aridification of Australia. An important Exposure ages of periglacial deposits and observation is the progressive drop in lake levels and desert pavements increase in evaporitic phases towards the present, and the several excursions towards more humid conditions Cosmogenic isotopes also provide a means for dating associated with glacial–interglacial cycles. Cenozoic sedimentation, such as direct dating of perigla- cial deposits. Barrows et al. (2001) measured in situ- 10 K–Ar and 40Ar/39Ar of salt lake deposits produced Be in boulders deposited in glacial moraines to identify at least two distinct glaciations: the Early Large saline lakes (4100 km2) in internally draining Kosciuszko (59 300 + 5400 a) and the Late Kosciuszko areas of tectonically quiescent central and western (consisting of three glacier advances, at 32 000 + 2500, 19 Australia (De Deckker 1983) contain a variety of 100 + 1600 and 16 800 + 1400 a). A similar study in chemical and detrital sediments that appear to have Tasmania, also using in situ-produced 10Be, showed that changed throughout the Cenozoic as Australia became boulders from moraines deposited during the Last Glacial progressively more arid. Magnetostratigraphic studies Maximum yield a tight range of ages in the 20–17 ka range have been useful in studying the most recent (since 2 (Barrows et al. 2002). However, 10Be exposure ages of Downloaded By: [University of Queensland] At: 05:12 12 September 2008 Ma) history of sedimentation in these lakes. However, boulders from the Hamilton moraine, previously inter- these deposits host a potentially useful record of preted to be deposited during the Last Glacial Maximum, continental chemical and detrital sedimentation as old yield ages in the 350–190 ka range, revealing a much as the Paleogene (De Deckker 1983). These lakes are also earlier glaciation event in this region. 36Cl exposure ages fed by now completely aggraded paleodrainage systems for four different locations in southeastern Australia which host important channel gold and calcrete ura- (Snowy Mountains, New South Wales; Victorian Alps, nium deposits. The chronology of sedimentation in Victoria; and Ben Lamond Plateau and Mt Wellington, these lakes and in the associated drainage may poten- Tasmania) also reveal intense periglacial activity during tially provide a long-term record for the paleoclimatic the Last Glacial Maximum between 23 and 16 ka (Barrows history of Australia and the formation of surficial et al. 2004), consistent with previous 10Be studies and a mineral deposits in the interior of the continent. K–Ar proposed age of 19 000 + 600 a for the innermost moraine and 40Ar/39Ar dating of authigenic alunite, jarosite, and preserved at Schnells Ridge, Tasmania (Kiernan et al. Mn-oxides potentially provide useful constraints on 2004); these results indicate that the Last Glacial Max- their sedimentary or diagenetic histories. (U–Th)/He imum left a strong imprint on the Australian landscape. dating of authigenic goethite would also be suitable to Moreover, the numerical results from these studies date diagenetic processes in these lakes. Currently, only correlate well with glacier advances in New Zealand limited K–Ar dates have been published for authigenic and South America, suggesting that the glacial deposits minerals in these salt lakes (Bird et al. 1990a). Unfor- record hemisphere climate changes during the last glacial tunately, the authigenic alunites in these deposits cycle (Barrows et al. 2001). Interestingly, the periods of contained large proportions of detrital micas, and the glaciation in the highlands correspond to peak lake levels results, after correction for the mica contamination, and river discharge at lower elevations (Barrows et al. 892 P. M. Vasconcelos et al.

2001), demonstrating the important effects of glacial– at 128 + 1 ka, and its termination at 116 + 1 ka. However, interglacial periods on the Australian climate. reef growth was restricted to a narrow interval between An additional recent and innovative application of 128 and 121 ka, suggesting that optimal ocean-surface cosmogenic isotopes to the study of climatic change in temperatures or a stable sea-level highstand conducive to Australia is the application of in situ-generated 21Ne and reef growth only existed at these times. 10Be to date the formation of stony deserts in central McCulloch & Esat (2000) combined high-resolution Australia (Fujioka et al. 2005). Gibber plains, a typical TIMS with Sr/Ca ratios from corals in Western Aus- feature of arid Australia and interpreted to reflect a tralia and the Huon Peninsula, Papua New Guinea, to transition towards more arid conditions in a recent past, evaluate Last Interglacial sea-levels and sea-surface are composed of single layers of varnished stones, mostly temperatures. They concluded that sea-surface tempera- fragments of silcrete, overlying alluvium or aeolian tures at these study sites during the Last Interglacial clays. Measured in situ-generated cosmogenic 21Ne (with were similar to that of the present day, suggesting careful corrections for in situ nucleogenic, trapped asymmetric warming of the Earth, and that northern crustal, and trapped atmospheric 21Ne) and 10Be yield hemisphere insolation during the Last Interglacial may exposure ages revealing that desert pavements in Aus- have been responsible for the extensive (and possibly tralia began to develop at ca 4 Ma and became abundant catastrophic) melting of mainly northern hemisphere- at ca 3 Ma, coincident with Neogene global cooling and based ice sheets. The high sea-levels measured for the the initiation of global glacial cycles (Fujioka et al. 2005). Last Interglacial (*3 m above present) and the rapid changes in sea-level measured from the Huon Peninsula U-series disequilibria corals suggest a rapid rise in sea-level (30–50 m/ka) at 130 + 1 ka. The importance of Late Cenozoic climate changes to the The effects of glacial–interglacial cycles on continen- present landscape and environment in Australia is clear. tal climates was investigated by Zhao et al. (2001), who Determining when, how and how fast these changes also used high-precision TIMS 230Th–238U ages to deter- occurred is important in predicting environmental con- mine the rates of stalagmite growth in the Newdegate ditions during future global cycles. U-series dating of Cave, Tasmania. Their results revealed fast stalagmite carbonates (corals and speleothems), combined with growth (61.5 mm/ka) from 129.2 + 1.6 to 122.1 + 2.0 ka, temperature proxies, provides the high-resolution re- coincident with the period of fast coral growth in cords necessary to test causal relationships between Western Australia, and suggested that the interglacial global temperature records, sea-level changes, insolation period corresponded to a period of increased precipita- and orbital forcing of climate changes. Stirling et al. tion on the continent. Xia et al. (2001) also provided a (1995) showed that improved methodologies (using charge high-resolution, multi-proxy record (d18O, d13C, growth collection) in U and Th analysis by thermal ionisation rate, initial 234U/238U, and colour-transparency-porosity) mass spectrometry (TIMS) increase the precision of the calibrated by high-precision TIMS 230Th–238U ages that measurements, reducing the age uncertainty in analyti- yielded detailed information on climatic variations cal methods by a factor of four. This increased precision throughout the Holocene. also permits identification of samples that have under- TIMS U-series dating of speleothems was combined gone diagenetic changes (from precise 234Umeasure- with ESR in the study of the rich and diverse fossilifer- ments) and exclusion of the altered samples from age ous deposits from the Naracoorte Caves, South Australia calculations. Their improved analytical approach was (Gru¨n et al. 2001). U-series and ESR dating of flowstones used to date coral growth from coastal sites in Western delimiting the sedimentary sequences in the cave Australia containing extremely well-preserved reefs, sediments provided a compatible, reliable and high- Downloaded By: [University of Queensland] At: 05:12 12 September 2008 which showed minimal levels of post-deposition altera- resolution record for the history of sedimentation in the tion (possibly due to currently arid climate). Moreover, cave. ESR and U-series dating of bones and teeth also because they are located at the original growth position provided useful constraints on the age of the local fauna. (due to subdued tectonic activity in the area), the reefs The records show good agreement between the U-series provide reliable records of sea-levels at the time of coral and ESR methods, revealing that most of the fauna in the growth without requiring corrections for tectonic uplift Naracoorte Cave is twice as old as previously assumed. or subsidence. The distance between the Western Aus- The results also show very little change in the mega- tralian sampling sites and the ice sheets of the Penulti- faunal assemblages from early Middle to early Late mate Glacial Maximum also indicate that this area was Pleistocene but reveal the sudden disappearance of the not affected by glacial unloading, and the chronological megafauna after the Last Interglacial, suggesting a information provides accurate sea-level information at possible causal link between the megafauna extinction the time of coral growth. Stirling et al. (1995) measured with the human occupation of Australia. reliable ages for coral growth from 127.3 + 1.0 to Another significant advance is the demonstration by 121.8 + 0.8 ka and interpreted the results as evidence that Woodhead et al. (2006) that under ideal circumstances either the Last Interglacial only lasted from 127 to 122 ka, (high initial U and/or Th and low common Pb), shorter than previously proposed (130–117 ka), or condi- speleothems can be dated by the U–Pb method through tions conducive to reef-building only lasted for a short a combination of low-blank preparation and multi- period during the Last Interglacial. Stirling et al. (1998) collector inductively coupled plasma mass spectrometry expanded on that approach and showed that coral growth (MC-ICPMS) procedures. Extending the age resolution along the entire Western Australian coastline was for the growth of speleothems beyond ca 600 ka possible contemporaneous, fixing the onset of the Last Interglacial with the U-series method, and into the realm of the U–Pb Geochronology of the Australian Cenozoic 893

method, will increase our ability to date the precise method for dating old groundwaters in Australia, the timing and the environmental consequences of Neogene complexity of the analytical procedures and the large un- climate change in Australia (Woodhead et al. 2006). certainties in the ages obtained suggest that the method must be refined before it can be widely applied. 36 TL, OSL and ESR dating of sediments On the other hand, Cl, a cosmogenic isotope generated from spallation of 40Ar in the atmosphere, is Luminescence techniques (TL and OSL) are widely used ideal for dating old groundwaters because of its suitable in dating recent sedimentary cover in Australia. Entire half-life (3.01 6 105 a), simple geochemistry, conserva- special volumes of journals have been dedicated to the tive behaviour and absence of significant subsurface technical developments and applications in this area sources (Bentley et al. 1986). However, a potential (Gru¨ n & Roberts 2007). Two areas where these techni- complication is that in addition to spallation of 40Ar, ques have been particularly important are in dating 36Cl is also produced by cosmic-ray spallation reactions sedimentary horizons containing evidence of human on K- and Ca-bearing rocks exposed at the Earth’s occupation (reviewed below) and in dating fault reacti- surface; by neutron activation of 35Cl; and by atmo- vation and paleoseismicity (Crone et al. 2003). ESR has spheric nuclear tests (Bentley et al. 1986). Therefore, 36Cl often been used in Australia together with 14C, U-series, dating of groundwaters must account and correct for TL and OSL to date recent sediments, particularly those these other sources. Bentley et al. (1986) provided the sediments associated with important fossil-bearing sites basic calculations for these corrections, and illustrated (as discussed above) and sites hosting evidence of that for an aquifer where the hypogene production of human occupation. A notable example is ESR dating of 36Cl from neutron activation of 35Cl is in secular sediments associated with Lake Mungo (Thorne et al. equilibrium, it is possible to calculate a 36Cl age (t, in 1999, discussed below). In general, ESR results are years) for the groundwater through: compatible with, although not as precise as, results 1 R R0 obtained from U-series and OSL dating. However, ESR t ¼ ln has been instrumental in providing reliable information l36 R0 Rse on the ages of human occupation sites. 36 76 71 where l36 is the Cl decay constant ( ¼ 2.302 6 10 a ); 36 36 14 R is the measured Cl/Cl ratio; R0 is the initial Cl/Cl C dating of sediments 36 ratio; and Rse is the secular equilibrium Cl/Cl ratio Modern 14C dating, where AMS permits the analysis of due to hypogene production of 36Cl. Procedures for

small samples, and aggressive washing techniques calculating Rse are provided by Bentley et al. (1986). ensure the removal of inorganic and modern organic In addition to improved methodological understand- carbon, has been instrumental in resolving strati- ing of 36Cl geochemistry, advances in accelerator mass graphic questions in key fossil sites, particularly those spectrometry were also instrumental in making the associated with human occupation of Australia (Fifield routine analysis of 36Cl in groundwater possible (Bentley et al. 2001; Turney et al. 2001), as discussed below. A et al. 1986). Australian capabilities in this area were particularly useful approach has been the combination implemented in the 1980s, when the 36Cl measurement of TL, OSL, ESR and 14C in the same sites (Roberts et al. program at the Australian National University provided 1990, 1994; Thorne et al. 1999; Fifield et al. 2001; Turney increased capacity for studying Australian groundwater et al. 2001; Sherwood et al. 2004). reservoirs (Fifield et al. 1987). The application of 36Cl dating of groundwater in Australia focused initially on the Great Artesian Basin. The measurement of 36Cl/Cl in Downloaded By: [University of Queensland] At: 05:12 12 September 2008 36 GROUNDWATER CHRONOLOGY 26 groundwater samples reveals Cl groundwater ages ranging from less than 100 ka to over 1 Ma (Figure 14), The progressive aridification of Australia throughout reasonable results when compared with hydrodynamic the Cenozoic, and the near-complete absence of peren- ages calculated for the long-range transport in the Great nial rivers and freshwater lakes on most of the continent, Artesian Basin (Bentley et al. 1986). Torgersen et al. makes subsurface water reservoirs vital. The practical (1991) provided further 36Cl evidence for the longevity of importance of subsurface water reservoirs, and the need groundwaters in the Great Artesian Basin, but also for data on water flow and replenishment rates for the identified an area of groundwater with relatively young sustainable exploitation of aquifers, has promoted the 36Cl ages, which they attributed to a source at the basin development and application of cosmogenic-isotope margin between the Simpson Desert and Mt Isa. techniques suitable for dating groundwaters (mostly Bird et al. (1989) illustrated the combined application of 36Cl, but also 14C and 81Kr). This approach is based on the 14C and 36Cl to confirm that some Australian ground- assumption that the subsurface waters acquire most of waters are very old (41 Ma), but that other reservoirs, their cosmogenic-isotope loads in the form of meteoric such as the Murray Basin, have a more complex and water, and that once these waters enter the subsurface recent history of recharge. Davie et al. (1989) used 36Cl as a reservoir, the only process affecting their cosmogenic- tracer and dating tool to show significant recharge into isotope contents is radioactive decay. Complexities the Murray Group aquifer by downwards leakage of associated with carbon exchanges in groundwater rainfall over much of the Mallee region. Cresswell et al. reservoirs make the use of 14C as a dating tool proble- (1999b) used 36Cl/Cl ratios to identify two distinct matic. Similarly, while Collon et al. (2000) presented sources for shallow regional groundwaters in the arid results suggesting that 81Kr may provide a suitable southwestern part of the Northern Territory. They 894 P. M. Vasconcelos et al.

Figure 14 Application of 36Cl/Cl dating of groundwaters from confined aquifers in the Great Artesian Basin (a, b) reveals that water ages increase southwestward (along X–Y cross-section) and reach ages of ca 1 Ma (c), attesting to the longevity of recharge in these reservoirs. The 36Cl/Cl ages obtained are compatible with estimated hydrodynamic ages (c) for the waters (after Bentley et al. 1986). Downloaded By: [University of Queensland] At: 05:12 12 September 2008

identified Holocene recharge and a groundwater with processes have been instrumental in shaping our mean residence time of 80–100 thousand years, which current understanding of Australia. While the applica- they interpreted to reflect major recharge during the tion of these methods has provided a more realistic view last interglacial. 36Cl/Cl ratios for groundwaters in the of the age and longevity of geological features and the Amadeus Basin, central Australia, also reveal old (ca 400 nature of processes shaping the continent since its ka) recharge ages, interpreted as evidence that Austra- separation from Antarctica, many questions about the lian aquifers only undergo significant recharge during history of volcanism, weathering, erosion and sedimen- periods of relatively high rainfall and possibly low evapo- tation in the Australian Cenozoic remain unanswered. ration associated with interglacial periods (Cresswell et al. 1999a). These studies demonstrate the importance of A volcanic framework for the Australian Late Cenozoic global climatic changes, particularly Cenozoic glacial–interglacial cycles, on the recharge of subsurface water reservoirs in Australia. An understanding of the distribution and chronology of Cenozoic volcanism in eastern Australia underpins attempts to resolve a wide range of fundamental RECONSTRUCTING THE AUSTRALIAN CENOZOIC geological problems, including the plate-tectonic history AND BEYOND of Australia, the timing and driving forces for the formation and evolution of the eastern highlands, and As outlined above, the development and application of the assessment of volcanic hazards. In addition, volca- geochronological tools suitable for studying Cenozoic nic geochronology places important constraints on the Geochronology of the Australian Cenozoic 895

age of geological indicators of past climatic conditions ing between Antarctica and Australia (Le Pichon & and fossil sites, including human occupation. While a Heirtzler 1968; Weissel & Hayes 1971; Cande & Mutter comprehensive assessment of the origin of Cenozoic 1982) and appears to track the northward motion of the magmatism in eastern Australia is beyond the scope of Australian Plate over a fixed melting anomaly (or this paper (see Johnson 1989), we nevertheless attempt anomalies) in the underlying asthenospheric mantle to condense the current understanding that has devel- (Wellman & McDougall 1974a). Similar age-progressive oped from the geochronological evidence outlined above volcanism is recorded by the parallel seamount chains of in order to explore some future research directions that the Tasman Sea (Vogt & Conolly 1971; McDougall & may address several outstanding issues. Towards this Duncan 1988) and leucitite volcanism in the central New end, we first review the distribution and age of east South Wales and Victoria (Cundari et al. 1978; Cohen et al. Australian volcanism in relation to regional tectonics. 2008). This allows an examination of various lines of investi- This impressive geochronological record, as well as gation, which rely heavily on the volcanic record, into those obtained from the ocean floor by Australian the history of Australia’s northward journey away from geochronological laboratories (McDougall 1971), played Antarctica. The chronological review also serves to an important role in the early investigation into the use underpin the important realisation that not all of the of hotspot magmatism as a tool for tracking plate volcanic activity in eastern Australia is well understood motions relative to the mantle. A significant outcome and is by no means over. Improving our ability to of these studies is that calculations of hypothetical determine the age and eruptive patterns of young motions of lithospheric plates can now be made, if it is volcanism is therefore of critical social and economic assumed that hotspots do not move relative to one importance. another. Although this assumption has been questioned The general picture that emerges from the geochron- (Steinberger & O’Connell 1998; Tarduno et al. 2003), as ological record is one in which volcanism is connected has the existence of plumes altogether (Foulger et al. to the breakup of Gondwanaland, thinning and rifting of 2005), the debate over these issues and their implica- Australia’s eastern margin, upwelling of hot deep tions for plate kinematics and mantle dynamics are yet mantle and the onset of fast northward drift of the to be resolved. Notwithstanding these issues, compar- Australian Plate (Johnson 1989). In Late Mesozoic time, ison of the record of age-progressive volcanism in Australia, Antarctica and New Zealand were grouped eastern Australia (Wellman & McDougall, 1974a) with together in eastern Gondwanaland at a high latitudinal the hypothetical motion of the Australian Plate, as position (Figure 1). Australia and Antarctica began to calculated by Duncan & McDougall (1989b), indicates separate slowly at about 95 Ma at a rate of about 6 mm/a that the K–Ar record, while providing important insight (Veevers 1986). The Tasman Sea began to open about into the time-averaged history of the Australian Plate, the same time, but the main phase of rifting and does not resolve the subtle variation in velocity and generation of new oceanic crust along the eastern direction of the Australian Plate predicted to have margin of Australia took place between about 84 and occurred over the past 40 million years (Figure 15). 53 Ma (Figure 2b) (Gaina et al. 1998). It was during this The solution to this problem lies in the increased period that mafic lava-field activity commenced in accuracy, precision and speed of analysis provided by eastern Australia (ca 70 Ma). However, the major pulse advances in fully automated, laser incremental-heating of activity, which largely occurred along the east 40Ar/39Ar geochronology. As discussed earlier and Australian highlands (Figure 2a), was between about illustrated in Figure 3, southward propagation rates 55 and 35 Ma (Figure 2c). Significantly, about the time for central-volcano activity in eastern Australia can lava-field activity began to wane, Australia began its now be resolved, using the 40Ar/39Ar method, over time Downloaded By: [University of Queensland] At: 05:12 12 September 2008 rapid northward migration away from Antarctica windows of only a few million years (Cohen et al. 2007). (Figure 1), and the earliest bimodal central volcano Thus, we are now in a position to identify and directly erupted on the continent (Figure 2c). These volcanoes track rapid changes in plate velocity (both rate and hold one of the keys to our understanding of this direction) during Australia’s northward migration by important period in the Cenozoic plate-tectonic evolu- accurate and precise dating of the volcanic record tion of Australia. (Knesel et al. in press). In turn, such high-resolution speedometry for the Australian Plate will be critical for both testing and informing computer-based reconstruc- HOTSPOT VOLCANISM AND AUSTRALIA’S NORTHWARD tions and models of the Cenozoic evolution of the TRAJECTORY southeast Asia – southwest Pacific region where plate In contrast to the lava-field provinces, which display no boundaries and motions appear to have changed obvious age-progressive spatial patterns, the central- abruptly, probably because of collisional events, at volcano activity migrated southward starting at about about 45, 25 and 5 Ma (Hall 2002). 34 Ma at Cape Hillsborough on the coast in central Despite the apparent success of the hotspot model in Queensland and arriving at Mt. Macedon in central accounting for the age-progressive central-volcano and Victoria at about 6 Ma (Figure 2). The average rate of leucitite activity in eastern Australia, uncertainty migration of this volcanism (*65 mm/a), derived remains regarding the number and geometry of melting from K–Ar dating of 13 volcanoes covering a north– anomalies or hotspots (Johnson 1989). To account for the south distance of about 1700 km (Wellman & McDougall broad zone of central-volcano magmatism, which is 1974a; Duncan & McDougall 1989b and references about 660 km wide (Figure 2a), two (Wellman & McDou- therein), is consistent with the record of seafloor spread- gall 1974a) or more (Sutherland 1981) hotspots have been 896 P. M. Vasconcelos et al.

eastern Australia is the presence of significant silicic volcanic rocks (Johnson 1989). The implication is that the blanketing effect of thick continental crust, com- bined with concentrated high flux of hot plume-derived magma, ultimately resulted in development of crustal plumbing systems in which crystallisation and variable crustal–magma interactions lead to strongly differen- tiated derivative magmas and, locally, formation of crustally derived rhyolites (Ewart 1982). It is the age and duration of the silicic rocks that are best documen- ted and provide the most precise constraints on migra- tion of volcanism with time: where volcanic centres have been studied in detail, the picture emerges that silicic lavas and plugs were inevitably emplaced over relatively brief time windows towards the end of the mafic activity (Cohen 2007). Further detailed sampling, particularly with attention to structural features and both latitudinal and longitudinal control, coupled with high-resolution 40Ar/39Ar geochronological study of mafic lavas, as well as silicic rocks, is critical to resolving this longstanding issue of the number of hotspots beneath eastern Australia.

LAVA FIELDS AND VOLCANIC HAZARDS The origin of the lava-field volcanism is more proble- matic. These provinces comprise mainly alkaline mafic Figure 15 Calculated motion of the Australian Plate, relative lavas with lesser volumes of tholeiitic basalt. The lava to the hotspot reference frame for African Plate hotspots fields have a wide variety of form and size (Wellman & (after Duncan & McDougall 1989a), illustrates hypothetical McDougall 1974a; Johnson 1989 and references therein). changes in rate (thick line) and azimuth (thin line) during They are typically thin and often characterised by Australia’s northward migration following breakup of east- ern Gondwanaland. Although not apparent in the K–Ar remnant patches of lava and plugs scattered over a wide record of hotspot volcanism in eastern Australia, recent area, but thick piles of lava, which may represent studies (Knesel et al. in press) show that accurate and former shield volcanoes, are preserved in some cases precise dating by the 40Ar/39Ar method provides a powerful (e.g. Barrington and Liverpool volcanoes in New South high-resolution geochronologic tool for verifying and track- Wales). The young lava fields in Victoria and northern ing such changes in the motion of the Australian Plate. Queensland are widespread and, in addition to hosting Motion of the Australian Plate was calculated for a fixed extensive lava flows, are characterised by maars and point in eastern Australia (Canberra) from rotation poles at small scoria and lava cones. Unlike the central volca- five million year intervals for two models of Indo-Australia/ noes and leucitites, the lava fields have much less Antarctic relative motion (Weissel et al. 1977; Stock & systematic time–space distributions (Wellman & Molnar 1982). McDougall 1974a) (Figure 2b). However, it is important Downloaded By: [University of Queensland] At: 05:12 12 September 2008 to note that while most of this activity took place before proposed. The resolution of the current geochronologi- the onset of central-volcano activity at any given cal data does not allow unequivocal discrimination latitude, lava-field volcanism has continued into the between potential models. Nevertheless, application of Holocene (Figure 2b; Appendix 1*). Occam’s Razor suggests that perhaps the simplest Perhaps because of less systematic time–space his- interpretation (i.e. that central volcano activity is tory, there is little agreement regarding the origin of simply the manifestation of a broad zone of magmatism lava-field volcanism in eastern Australia (see Johnson associated with a single hotspot) may be the best. This 1989 for a review). Nonetheless, a recurrent theme is the scenario is supported by observations that, in compar- requirement for a tensional environment permitting the ison with narrow oceanic tracks that form on thin ascent of mantle-derived magma. Lava-field volcanism lithosphere, well-defined continental hotspot traces are took place along the ‘hot’ Tasman Fold Belt, and this rare. This scarcity may arise because plume-derived volcanic activity may be variably linked to reactivated melts are perforce unable to easily penetrate and deep fold-belt structures, processes and structures traverse thick continental lithosphere. associated with detachment faulting in the Late Cretac- In eastern Australia, the (diffuse) east–west distribu- eous, as well as those related to subsequent opening of tion of many volcanoes is localised on, and hence the Tasman and Coral Seas, and finally structures presumably controlled by, sets of parallel north–south produced by ongoing intraplate deformation, perhaps faults (Johnson 1989) that mirror the margin-parallel related to plate-tectonic reorganisations during the mid- fabric of the Tasman Fold Belt. In this regard, it is also to Late Cenozoic (Johnson 1989). A further complicating important to note that, paradoxically, the defining factor is that in some cases, hotspot-related volcanism characteristic of (mantle-derived) hotspot activity in passed through areas of prior or even contemporaneous Geochronology of the Australian Cenozoic 897

lava-field activity; in such cases the products of the two units would provide a well-dated and well-calibrated types of activity may be difficult to distinguish. Estab- APWP for Australia. The original APWP derived for lishment of a comprehensive high-resolution geochro- Australia (Figure 4), based on magnetic pole positions nological record will be critical to clarifying the details measured on dated igneous rocks (Wellman et al. 1969; of time–space distributions of individual lava-fields, as Wellman 1973; McElhinny et al. 1974; McElhinny & well as the lava-field volcanism as a whole and its Embleton 1974), was questioned because it was incon- relationship to central-volcano activity, and will ulti- sistent with the APWP derived for India (Klootwijk & mately help resolve uncertainties about their origins. Peirce 1979). To resolve this discrepancy, an alternate The socio-economic importance of this goal becomes APWP was obtained from measurements of the mag- evident with the realisation that volcanism in eastern netic properties of laterites and weathering profiles in Australia has not yet come to an end; lava-field volcan- Australia (Embleton & McElhinny 1982). Through the ism has continued throughout the Quaternary in Victor- application of a 17-point digital-filter analysis, a smooth- ia, South Australia, and southeast and north ing process that according to the authors ‘is not of Queensland. Fortunately, there are abundant opportu- course statistically rigorous,’ Embleton & McElhinny nities for future geochronological studies to refine the (1982, p. 145) generated a smoothed Cenozoic APW path, history of Quaternary volcanism in eastern Australia compatible with the APWP for India, that they believed and thus better inform hazard assessment and predic- to represent a better description of the most likely tion. There are two main reasons for these new APWP than that obtained from the volcanic data opportunities. First, as discussed above, many of the (Figure 4). Idnurm (1985a) further refined an alternative classic techniques have undergone technological ad- APWP derived from magnetic pole positions retrieved vances but have not yet been widely applied to Austra- from six sedimentary deposits, one volcanic unit, one lian volcanoes. A prime example is 14C, where the AMS lake sediment and six undated weathering profiles technique allows much smaller volumes of material to be (Figure 4). analysed, thus reducing the potential of contamination, The reasoning for abandoning the APWP derived and allowing a far greater possibilities for sample selec- from dated igneous rocks was the large scatter in pole tion. Similarly, the greater precision of the 40Ar/39Ar data believed to result from incomplete averaging out of technique also enables intermediate-to-mafic samples geomagnetic secular variations (McElhinny & Embleton younger than 50 ka to be analysed reliably (Gamble et al. 1974; Idnurm 1985a). An attempt to revive the apparent 2003). Second, newer techniques, such as cosmogenic polar wonder path for the last 100 million years, isotopes, OSL and ESR, which have been widely applied calibrated by the remanence directions of volcanic elsewhere in the world, also have abundant potential to rocks through a more robust statistical treatment of date Australian Quaternary volcanism. the data (Musgrave 1989) (Figure 4), was not well Refining the history of Quaternary volcanism is also received (Idnurm 1990). The consequence is that there a worthwhile exercise because well-dated volcanic are currently three competing and incompatible APW events and units provide essential chronostratigraphic paths for Australia (Figure 4), one derived from dated markers for archaeological (Sherwood et al. 2004), volcanic units (Wellman et al. 1969; Wellman 1973; paleontological (Mackness et al. 2000) and climatic McElhinny et al. 1974; McElhinny & Embleton 1974), (Haberle 2005) studies. However, as highlighted above, one derived from sedimentary sequences and weath- any future attempts to achieve these ends should have ering profiles (Idnurm 1985a) and one derived from a an integrated approach [such as recently undertaken by statistical treatment of paleomagnetic data from igneous Sherwood et al. (2004) at Tower Hill volcano in Victoria], rocks and weathering profiles (Musgrave 1989). to avoid the current confusion in the literature regard- Idnurm (1985a, b, 1986, 1990, 1994) argued convin- Downloaded By: [University of Queensland] At: 05:12 12 September 2008 ing estimates of eruption ages, which has arisen from cingly that the record derived from the sedimentary mixtures of techniques and different laboratories ap- rocks and weathering profiles is less subject to secular plied to the same volcano, often with conflicting results. variations because these rocks acquire their chemical remanent magnetisation during a relatively long period of time (1–2 million years) after their deposition, which PALEOMAGNETIC VOLCANOLOGY AND AUSTRALIA’S averages out any geomagnetic secular variations. How- APW PATH ever, the exact time when the sedimentary rocks and In principle, geochronology of volcanism in eastern weathering profiles acquired their main chemical Australia should provide additional constraints on the magnetisation is only estimated, but not accurately absolute motion of the Australian Plate by supplying the known. The suitability of this information in providing necessary record for calibrating the Cenozoic Austra- an age-calibrated APWP is, thus, problematic. Further- lian apparent polar wander path (APWP) (Wellman et al. more, recent weathering geochronological results sug- 1969; Wellman 1971; McElhinny et al. 1974; McElhinny & gest that the precipitation of iron-rich minerals in Embleton 1974). Given the extensive distribution of weathering profiles may span tens of millions of years Cenozoic volcanic rocks in eastern Australia, the four (Shuster et al. 2005; Bernal et al. 2006). If so, some of the decades of geochronology applied to these units, the chemical remanent magnetisation derived from weath- large age span of the volcanic units, the numerous ering profiles (and iron-cemented sediments), as is the individual flows that occur in each volcanic edifice and case in Idnurm’s favoured APWP, may average out the relatively large age span (*3–5 Ma) represented by magnetisation acquired during tens of millions of years, single volcanic centres, one would conclude that mea- further increasing the uncertainty of the APWP derived surement of thermal magnetisation from these volcanic from these materials. 898 P. M. Vasconcelos et al.

Renewed interest in using the APWP to date weather- ing profiles in Australia (Schmidt & Ollier 1988; Acton & Kettles 1996; Pillans 1998, 2004; Pillans et al. 1999) indicates that the refinement of the APWP for the Cenozoic is a worthwhile exercise. If a time-averaged field direction is necessary in a geomagnetic study, Tauxe (2005) stated that a series of well-dated volcanic units, spanning *100 thousand years, need to be sampled in sufficient density to provide a statistically valid pole position. Recent detailed 40Ar/39Ar geochro- nology of several volcanic centres in eastern Australia provides well-dated sequences each spanning *3–5 Ma (Waltenberg 2006; Cohen et al. 2007). Further refinement of the chronology of magma extrusion in these systems by high-resolution 40Ar/39Ar dating should provide the necessary geochronological constraints for the identifi- cation of a sequence of flows spanning 100 000 years. To generate a time-averaged pole position, Tauxe (2005) also suggested sampling about 100 sites (each with nine to 10 samples) from the flows spanning the 100 000 years interval. This would be a large but fruitful undertaking that would greatly increase the accuracy of the Aus- tralian Cenozoic APWP. This time-calibrated path Figure 16 Schematic illustration of early models for the would, in turn, provide a more robust and rigorous development of the highlands in southeast Australia (after way of dating, by paleomagnetism, soils, weathering Wellman 1979). Prior to widespread K–Ar dating of volcanic profiles and sedimentary deposits of unknown age. units in eastern Australia, most models advocated signifi- cant uplift during the Late Cenozoic. Uplift models are from: a, Andrews (1911); b, Craft (1933); c, King (1959); d, Browne Cenozoic tectonic uplift, denudation and (1969); e, Hills (1975); f, Wellman (1979). origin of the Eastern Highlands Despite its lengthy northward trajectory, paleogeo- graphic reconstructions indicate that most of Australia underwent only mild vertical motion without signifi- cant folding and faulting during the Mesozoic and Cenozoic (Gurnis et al. 1998; Sandiford 2007). A possible exception is the eastern and southeastern margins of the continent, where Mesozoic and Cenozoic magmatism is spatially, and postulated to be temporally, associated with uplift of the Eastern Highlands. Evidence for a causal link between volcanism and uplift is not con- clusive, however, and the timing and mode of uplift of the Eastern Highlands are still controversial issues in Downloaded By: [University of Queensland] At: 05:12 12 September 2008 the Cenozoic history of Australia (Wellman 1987). Early models of uplift, as reviewed in Wellman (1979), proposed that the Eastern Highlands were relatively Figure 17 Cross-section of the Shoalhaven River gorge in youthful features of the Australian landscape and that New South Wales (after Nott et al. 1996), showing ca 31–29 significant uplift had occurred only in the Neogene, Ma basalt flows descending 50–70 m below the rim of the mostly from the Pliocene to the present (Figure 16). 500 m deep gorge, providing clear evidence of significant uplift of the highlands in this region by, or prior to, the However, the distribution of K–Ar-dated volcanic units Oligocene. in the region led researchers to realise that significant relief must have existed along the eastern margin before the Eocene. For example, based on the widespread interpreted renewed channel incision after lava infilling distribution of Eocene (Monaro, Nerriga, and Mittagong as evidence for additional uplift during or after erup- provinces) and Paleocene to Oligocene lava flows tion. As a result, Wellman & McDougall (1974b) identi- (Hunter Valley) overlying a relatively dissected and flat fied two major periods of uplift for the highlands in New erosion surface, Wellman & McDougall (1974b) con- South Wales: mid-Cretaceous to Oligocene and middle cluded that the Eastern Highlands had the lowest relief Miocene to Present. In addition, on the basis of in the Late Mesozoic. The presence of Oligocene to geomorphological reconstructions derived from the Holocene lava flows in paleoriver channels (Figure 17) distribution of the dated basalts, they also inferred that was interpreted as evidence that most of the tectonic the highland was uplifted differentially both along and uplift that created the east Australian escarpment had transverse to the axis. already occurred by the Oligocene, probably in the Late Wellman (1974) employed a similar approach Cretaceous to Paleogene. However, these authors also to conclude that the Eastern Highlands in Victoria Geochronology of the Australian Cenozoic 899

Figure 18 Cross-sections through Sturgeon lava field in far (after Coventry et al. 1985), showing evidence for river incision and erosion over the past 5 million years. Similar features in were attributed to recent uplift of the Eastern Highlands in this region (Wyatt & Webb 1970; Griffin & McDougall 1975).

underwent a three-stage uplift history, each correspond- main period of volcanism and had a quiescent tectonic ing to *300 m of uplift. The first (Late Mesozoic) history after basalt extrusion. Taylor et al. (1985) also resulted in the cutting of broad U-shaped highland interpreted the age and spatial distribution of basalts in valleys; the second (Oligocene) promoted the incision the Canberra–Monaro region as evidence that the sub- of narrow V-shaped valleys cut into the wide highland basaltic and present topography are very similar, valleys; and the third promoted further downcutting of suggesting minimal changes in topography and drai- the narrow valleys in the Late Cenozoic. The distribu- nage patterns during the Cenozoic. The authors con- tion of basalts dated by K–Ar indicates that the high- cluded that ‘the drainage system in the region was lands already possessed 1000 m of relief by the established well before the Eocene, perhaps as far back Oligocene, with a drainage similar to that of the present, as the early Cretaceous, and has remained substantially implying that most of the highland uplift occurred in the unaltered since’ (Taylor et al. 1985 p. 69), suggesting that Late Cretaceous or Early Cenozoic. Nevertheless, young either no significant uplift had occurred during the basalt flows from the Newer Volcanics of Victoria Cenozoic or that the drainage system in the region had show evidence for fault movement (up to 250 m: Aziz-ur- responded extremely slowly to a change in base level. Rahman & McDougall 1972) and warping subsequent They also concluded that ‘the Divide between the coastal to their eruption, suggesting that tectonic activity and inland drainage pre-dates the basalts and is continued during the past four million years (Wellman probably pre-Cenozoic in origin’ (Taylor et al. 1985 p. 69). 1974). Bishop et al. (1985) arrived at a similar conclusion. K–Ar dating of far northern Queensland basalt flows They used the distribution of Miocene basaltic lavas also suggests a history of volcanism, uplift and river in- that flowed into river systems in central eastern New cision that continued well into the Pleistocene (Wyatt & South Wales to reconstruct the characteristics of Webb 1970; Griffin & McDougall 1975) (Figure 18). These the Early Miocene drainage (Figure 19). By deducing authors interpreted, from K–Ar results, that the Eastern the characteristics of the Miocene drainage from the Downloaded By: [University of Queensland] At: 05:12 12 September 2008 Highlands in far northern Queensland were actively morphology of the surface preserved beneath the basalt rising at the time of volcanic activity (4.5–1.1 Ma). River in the catchment, they showed that the Miocene valley incision rates calculated from dated basalts in far of the Lachlan River had a gradient of 0.0018, similar to northern Queensland (10–35 m/Ma) are, on average, its present gradient of 0.0019 (Figure 19). Variation in greater than those calculated for southeastern Australia gradients in the Miocene Upper Lachlan drainage is (Wyatt & Webb 1970) for the same time interval, a associated with variations in catchment lithology, possible result of different climatic conditions or more similar to what occurs at present (Figure 19). These active uplift history in this region during the Late observations suggest that uplift is not required to Cenozoic. Support for the latter comes from the work of explain river morphology in the area. In addition, these Griffin & McDougall (1975), who calculated average authors showed that the rate of incision (8 m/Ma) since erosion rates of 10.2 m/Ma in far northern Queensland the extrusion of the basalts indicates that the passing of for the period between 7.8 and 2.3 Ma; 21.5 m/Ma for the 20 million years resulted in little change in the land- period from 2.27 + 0.22 to 0.41 + 0.1 Ma; and 24.4 m/Ma scape; much more than 100 million years would be since 0.41 Ma. The authors tentatively interpreted this required to reduce the eastern flanks of the Eastern increase in erosion rates (actually, river incision rates) Highlands to a low-relief planation surface (Figure 19). as the result of general Late Cenozoic uplift or, These observations suggest that the low-relief coastal alternatively, localised erosional controls. plains in central New South Wales are ancient landscape Sutherland et al. (1977), on the other hand, calculated features and are not associated with Cenozoic processes. erosion rates of 3–5 m/Ma for the Bowen–St Lawrence Young & McDougall (1993) provided further evidence hinterland, north Queensland, concluding that the that most of the landscape features in the Eastern northern Bowen Basin underwent uplift prior to the Highlands were already developed in the Mesozoic and 900 P. M. Vasconcelos et al.

bimodal distribution, with apparent ages dropping from 360–260 Ma on the plateaus to 175–150 Ma along the coastal lowlands, with a minimum age of 80 Ma south of the Sydney Basin (Moore et al. 1986) (Figure 10). Track- length distribution suggests partial to complete anneal- ing of tracks by a widespread thermal event at 100–90 Ma (Moore et al. 1986). Taking into consideration present thermal gradients (258C/km), estimated past thermal gradients (38–538C/km), and AFT ages and track-length distribution, Moore et al. (1986) estimated rapid erosion of 1.5–2.5 km of cover along the coast quickly after the breakup. Further evidence for the bimodal distribution of AFT ages was provided by Dumitru et al. (1991), who proposed that the coastal rocks in southeastern Aus- tralia were heated above *1208C in the mid-Cretaceous and cooled relatively quickly at 100–80 Ma, by rapid erosion of a relatively thick (1.5–3 km) cover. However, Nott & Purvis (1995) questioned the 1.5– 3 km of denudation interpreted from AFFT because of the presence of Early Cretaceous lavas and hypabyssal rocks of the inferred Tilba-Tilba Lake Caldera in the Mt Dromedary igneous complex, which they interpreted as surface flows that were never buried (Nott & Purvis Figure 19 Long profiles of the Early Miocene and present (a) 1995). Yet, the presence of a 98.5 + 0.8 Ma zoned Lachlan River and (b) Crookwell River–Wheeo Creek monzonite pluton at Mt Dromedary (Spell & McDougall valleys illustrating that the gradients in these rivers have 2003), which underwent differentiation by fractional remained relatively unchanged since extrusion of basalts in the Miocene. In addition, present gradients are controlled crystallisation (Smith et al. 1988), is inconsistent with by lithology, similarly to Miocene gradients. These observa- the interpretation that these rocks were at the surface at tions have been interpreted as suggesting that no significant ca 100 Ma. Shallow but unconstrained depths of tectonic uplift has occurred in these regions in the Neogene emplacement for the Mt Dromedary pluton (Smith (from Bishop et al. 1985). et al. 1988) allow for several kilometres of overburden, and as such are consistent with the eroded cover estimated from the AFTT results. that very little, if any, active uplift occurred during the The (U–Th)/He thermochronology results of Persano Cenozoic. In a comprehensive study of the spatial and et al. (2002) are also compatible with the formation of the temporal distribution of Miocene basaltic magmas coastal plains in southeastern Australia by rapid filling river valleys in southern New South Wales, they erosion of a significant cover immediately after con- were able to reconstruct the pre-Miocene landscape, tinental breakup at 100–85 Ma (Figure 12). According to infer paleotopography and quantify post-extrusion ero- Persano et al. (2002), this period of rapid erosion sion rates. They concluded that plateau surfaces were occurred within 28–50 million years of breakup, and lowered by only about 2–5 m/Ma; that average river involved a minimum vertical erosion rate of 130 m/ incision since the Miocene ranged from 5 to 18 m/Ma, million years along the coast or a minimum mean scarp Downloaded By: [University of Queensland] At: 05:12 12 September 2008 with a maximum incision rate of 30 m/Ma; and that retreat rate of 1.5 km/million years. The (U–Th)/He river gradients remained essentially unchanged in the results were also interpreted to be consistent with either past 20 million years. By matching steep and flat a downwearing or scarp-retreat model of passive margin sections on the pre- and post-Miocene stream profiles, evolution, but not with a downwarped rift-shoulder they concluded that evolution of stream profiles is model. Persano et al. (2002) also argued that the rapid mostly controlled by stream power rather than head- exhumation of 2 km or more of crust immediately after ward nickpoint propagation. The authors argued that breakup is consistent with the preservation of subaerial the matching of Miocene and modern stream profiles is lavas (Nott & Purvis 1995) and Late Mesozoic weathering not consistent with the interpretation that post-Miocene profiles (Bird & Chivas 1993) on the coastal plains. incision results from a period of uplift after basalt However, if rifting occurred at 100–80 Ma, as currently extrusion. They concluded, instead, that denudation accepted (Moore et al. 1986; Veevers 1986), and the rapid had been essentially constant throughout the Cenozoic, erosion that created the coastal plain occurred within with a minor quiescent period during basalt extrusion, 28–50 million years of rifting, it is not feasible to expect and that no Cenozoic isostatic uplift is required to the preservation of ca 98 Ma subaerial lavas along the explain the geomorphological features in the Eastern plain. The lavas must have been under a significant Highlands. thickness of overburden, as discussed above. The K–Ar evidence suggesting that uplift of the Despite the emerging consensus that most of the Eastern Highlands occurred primarily during the Me- topographic features of the Eastern Highlands were sozoic has been substantiated by AFTT (Moore et al. already developed by the end of the Cretaceous, some 1986; Dumitru et al. 1991). AFT ages and track-length interpretations of lava distribution and ages still call for distributions for southeastern Australia show a clearly more recent periods of uplift, faulting and denudation. Geochronology of the Australian Cenozoic 901

Sharp (2004) used the distribution and ages of basalt (Heimsath et al. 2000, 2001) reveal maximum average flows and associated sediments in the Kiandra–Cabra- erosion rates of *50 m/Ma for the coastal plains, and 15– murra and Adaminaby–Cooma areas to identify a 25 m/Ma for the plateaus during the Quaternary (Figure complex pattern of drainage rearrangement resulting 8). While the coastal-plain erosion rates are only one- from Cenozoic faulting and folding, which he interpreted tenth of the rates of erosion estimated from scarp retreat to be associated with two major periods of differential models (Weissel & Seidl 1997), these rates are faster than tectonic uplift and volcanism in the Oligocene and Early expected if we accept that the retreat of the escarpment Miocene. Similarly, K–Ar dating of Oligocene (27.4 + 0.3 occurred rapidly after the opening of the Tasman Sea at and 28.0 + 0.3 Ma) basalt flows overlying fluvial sedi- ca 100–85 Ma and also accept the suggestions (Nott & ments deposited along the New South Wales coastal Purvis 1995) that the coastal plains have remained plains suggests a diversion of the Clyde River drainage unchanged since retreat of the escarpment. Similarly, near Bateman’s Bay (Spry et al. 1999). The ages for the erosion rates estimated from soil-production functions basalts, combined with the interpretation that the fluvial and from measurement of cosmogenic nuclides from sediments underlying the basalts were deposited by a stream sediments in the plateaus (15–25 m/Ma) are not south-flowing paleochannel of the Clyde River before its compatible with long-term estimates of erosion rates diversion to the present east-flowing direction, led Spry (510 m/Ma) from other studies (Wellman & McDougall et al. (1999) to postulate a possible Oligocene tectonic 1974b; Bishop et al. 1985). Heimsath et al. (2001) uplift for the Mogo area. attributed the higher rates measured through cosmo- An additional observation worth pointing out in this genic isotopes as evidence that the pace of landscape context is that, in several areas along the eastern evolution in the eastern margin of Australia has margin, the Great Escarpment is actually carved on accelerated in the Quaternary, possibly as a response Miocene volcanoes (Ollier 1982). The uneroded portion to changing climates, and that the accelerated pace of the volcanoes constitutes the highlands, while the may coincide with the penultimate glaciation at ca dissected plains were formed by removal of Miocene 150 ka. volcanic units (Figure 20). In the Ebor area, for example, Late Cenozoic climatic changes and tectonic reacti- the coastal plains extend for 20 km and were formed by vation have also promoted uplift and more active the erosion of at least 1500 m of the volcanic edifice erosion elsewhere in Australia, but in a much more and basement rocks in the past 19 million years, subtle way. Reactivation of major continental structures implying an average rate of scarp retreat of 1 km/Ma is evident from the active seismicity of areas such as the or a vertical erosion rate of 79 m/Ma. These rates are Flinders Ranges and southwestern Western Australia. much faster than rates measured for the same time Cenozoic fault reactivation in some of these areas is interval in the southern margin, and indicate that difficult to ascertain because of the absence of datable portions of the Great Escarpment are young and volcanic units and the fact that the geological features actively forming today. offset by fault movement have not been, until recently, Further evidence for recent activity arrives from easily dated. OSL studies of historically aseismic cosmogenic-isotope studies. Soil-production functions Quaternary faults in South Australia and southwestern for landscapes in the coastal plains and highland Western Australia (Crone et al. 2003) reveal episodic plateaus in the southeastern margin of Australia reactivation of these faults in 10 000–100 000 year cycles Downloaded By: [University of Queensland] At: 05:12 12 September 2008

Figure 20 Digital elevation model of the heavily dissected ca 19 Ma Ebor shield volcano in northeastern New South Wales illustrating the effects of significant erosion and scarp retreat during the Late Cenozoic in this area, and supporting models of differential development of the Great Escarpment in space and time. Erosion during the past *19 million years has exposed the core of the volcano, which comprises a high-level mafic to felsic intrusion, known as The Crescent Complex (Ashley et al. 1995). 902 P. M. Vasconcelos et al.

during the recent past. A combination of cosmogenic- ignoring uncertainties associated with thermochronolo- isotope erosion-rate measurements with OSL study of gical models in the partial annealing zone (Vasconcelos recent sediments reveals that both recent climatic 1999a; Gunnell 2003), the fission-track results require the changes and fault reactivation have promoted recent removal of at least several kilometres of overburden stream incision and increased erosion and sedimenta- from ca 300 Ma to the present day. Although erosion is tion in the Flinders Ranges, South Australia (Quigley unlikely to have been continuous, these results reveal et al. 2004). Lilley (1997) and Vasconcelos (1998) inter- that much of Western Australia must have been several preted the offset of silcrete horizons along the Pilgram kilometres below the surface in the Paleozoic. These Fault, Mt Isa region, as evidence for reactivation of that observations are not consistent with the continuous fault in the Cenozoic. The 40Ar/39Ar ages of manganese exposure and preservation of weathering profiles since oxides replacing fractured silcrete along the Pilgram the Proterozoic. Fault reveals Miocene tectonic activity. These examples Recently, two lines of geochronological evidence, clearly demonstrate that, despite the relatively slow both based on relative dating methods (stable isotopes rates of landscape development in Australia, the con- and paleomagnetism), have been used to infer that some tinent is still undergoing active tectonism (Sandiford Australian weathering profiles have been intermittently 2007) and climate-driven intensification of erosion. exposed to weathering since the Paleozoic. The stable Further quantification of these processes is essential isotope record identifies clay minerals with light d18O for assessing the effects of climatic changes on the signatures in several currently exposed weathering future of Australia. It is also important in assessing profiles and, based on the calibration curve of Bird & natural hazards: a complacent attitude based on the Chivas (1988), assigns a Permian age for these profiles false impression that ‘not much happens in Australia’ (Bird & Chivas 1993). The chemical remanent magneti- has potential human and financial consequences. Lastly, sation measured for profiles in Western Australia, in order to develop a realistic picture of the pace of Northern Territory and New South Wales purportedly Australian geological evolution, it is necessary to also records Paleozoic pole positions (Pillans et al. 1999; reconcile the evidence for recent activity with the Pillans 2004 p. 240). If interpreted pole positions are view promoting extreme antiquity of the Australian correct, these measurements provide geochronological landscape. evidence for a history of weathering that started in the Paleozoic. Pillans et al. (1999) compared pole positions Weathering and erosion and the antiquity measured on 27 samples from four pink saprolites at of the Australian landscape Northparkes with two distinct Paleozoic APWPs (those of Lackie & Schmidt 1993 and Klootwijk 1996) and Australia is a red and ‘sunburnt country’ (Mackellar tentatively interpreted the paleomagnetic results to 1908). Reddening of the continent reflects its long history ‘infer that weathering of regolith at Northparkes mine of exposure to weathering-prone and transport-limited has been an ongoing process, although not necessarily conditions. However, the longevity of this history is continuous, since at least Paleozoic time.’ O’Sullivan poorly constrained. Some authors propose that parts of et al. (2000) combined the paleomagnetic results of Australia, particularly Western Australia, have been Pillans et al. (1999) with modelling of AFTT results to potentially continuously exposed to weathering and infer that the 30 m thick saprolite overlying the erosion since the Proterozoic (Butt 1988; Anand & Paine Northparkes deposit formed in the Paleozoic was buried 2002). Evidence for such assertions is scarce, and there and re-heated to 4*110–1208C in the Late Paleozoic, are no geochronological results to support such propo- partially cooled to below 608C by the Triassic, reheated sals. The flatness of the Western Australian landscape to 60–908C by burial under 1.4–2.5 km of sediments Downloaded By: [University of Queensland] At: 05:12 12 September 2008 and the smooth flat-lying unconformities between during the Mesozoic, and the same 30 m-layer of Proterozoic sedimentary rocks, at the margins of the saprolite re-exposed at the surface by rapid cooling craton, and the flattened western shield are presented as (denudation) during the 60–40 Ma period, and it has been evidence that the surface exposed today is a Proterozoic exposed to further weathering ever since. If correct, unconformity (Butt 1988). Further evidence for Proter- burial of the saprolite to depths 43 km did not affect the ozoic weathering and continuous exposure is based on supergene minerals carrying a chemical remanent the interpretation that some enriched orebodies in the magnetisation attributed to Paleozoic weathering, but Hamersley Province resulted from supergene enrich- it did cause total and partial annealing of apatite fission ment and that this enrichment occurred at 1.8 Ga (Butt tracks. 1988). However, Butt (1988) questioned the likelihood of Absolute direct-dating geochronological methods, on preservation of widespread Proterozoic weathering the other hand, have, so far, failed to measure weathering remnants (except for the Hamersley region), and argued ages older than ca 80 Ma for any supergene mineral in that most previously weathered material would have Australia (Appendix 2*) (Figure 21). The oldest K–Ar been removed by Permian glaciations, resetting the results measured by Bird et al. (1990a) ca 62 Ma for start of the weathering history to ca 250 Ma. alunites from Springsure, Queensland—are substantially The models proposing that some tectonic blocks in younger than ages obtained from relative dating meth- Australia have been continuously exposed since the ods. The greatest K–Ar ages (ca 12 Ma) obtained for Proterozoic are not substantiated by recent thermo- alunite nodules in the Kingscote profile, Kangaroo Island, chronological measurements. AFTT results reveal that contrasts with the estimated Permian d18O age for that rocks exposed at the surface in Western Australia today profile (Bird & Chivas 1993). The greatest weathering age record apatite fission-track ages of 300–200 Ma. Even obtained for profiles in Western Australia (ca 60.9 Ma) is Geochronology of the Australian Cenozoic 903

Figure 21 Probability density dis- tribution for all currently pub- lished weathering geochronology results for Australia (Appendix 2*) showing that the weathering history spans as far back as the Late Cretaceous and that the his- tory of mineral dissolution–repre- cipitation has not stopped in the Neogene, despite clear evidence for the progressive aridification of the continent.

also lower than inferred ages of weathering in that The answer to the first question must lie in the region. Bird et al. (1990a) and Bird & Chivas (1993) composition of the sediments in Antarctic marginal interpreted the lower ages obtained by K–Ar dating of basins, and the chemical composition of the ocean. The alunites as indicating that sulfates form late in the answer to the second question has important implica- history of weathering. K–Ar and 40Ar/39Ar dating of tions to paleoclimates and the use of weathering profiles regionally distributed supergene Mn-oxides in Western as climatic indicators. Observations that weathering Australia and the Northern Territory (Dammer et al. profiles have formed at extra-tropical latitudes in the 1996, 1999) reveal that the greatest age is ca 52 Ma. past (Goldich 1938; Fitzpatrick 1963; Dury 1971; Dury & Comprehensive single crystal 40Ar/39Ar geochronology Knox 1971; Hall 1985) but are not forming today have led results for Mn-oxides, alunites and jarosites from researchers to propose that humid and possibly warm throughout Queensland (Mt Isa, Lawn Hill, Charters climates extended to higher latitudes in the Mesozoic Towers, Central Coast, Mary Valley, southeastern and Early Cenozoic. However, Bird et al. (1990b) and Bird Queensland plateaus and coastal plains), South Australia & Chivas (1993) provided an alternative explanation for (Coober Pedy and Andamooka), New South Wales (Light- weathering at high latitudes. They proposed that higher

ening Ridge and Northparks) and Western Australia atmospheric CO2 in the past would result in enhanced (Yilgarn, Hamersley) reveal that the oldest weathering biological activity, even at high latitudes, which could mineral yet identified is ca 80 Ma (Heim 2007) (Figure 21). have generated weathering solutions sufficiently aggres- Similarly, (U–Th)/He ages for regional studies in Wes- sive and reactive to promote weathering at relatively low Downloaded By: [University of Queensland] At: 05:12 12 September 2008 tern Australia and Mt Isa yield results consistently (558C) temperatures. If correct, this interpretation has younger than ca 70 Ma. Thus, the longevity of Australian significant implications for the use of weathering weathering profiles determined from direct dating meth- profiles as climatic indicators. Techniques suitable for ods contrasts with that obtained from relative dating extracting age and paleoclimatic information from methods. This discrepancy needs resolution. minerals precipitated in weathering profiles provide Notwithstanding the uncertainties above, the geo- ample opportunity for future research. chronological record obtained so far does suggest that Ages obtained for weathering profiles are also the formation and preservation of continuously exposed important to unravel the more recent climatic changes weathering profiles in Australia, if not as ancient as the that have shaped the Australian continent since it Proterozoic or maybe Paleozoic, certainly record a started its rapid northward migration at ca 55 Ma. history extending into the Mesozoic (Figure 21). The Paleontological evidence reveals that Australia has implication is that weathering profiles in Australia evolved towards more arid conditions throughout the were forming when Australia was at very high latitudes, Cenozoic (White 1994). The Miocene and Pliocene epochs attached to Antarctica. This conclusion raises three appear to record the major transitions towards arid interesting questions. What happened to the weathering conditions. Geochronological constraints for the exact profiles that once must have blanketed Antarctica? time of these transitions are still missing. Nevertheless, What conditions in the past supported the formation of 40Ar/39Ar and (U–Th)/He weathering geochronology weathering profiles at high latitudes? What does the results suggest regional water-table drawdown through- formation and preservation of continuously exposed out the Miocene and Pliocene (Vasconcelos & Conroy weathering profiles teach us about the climatic and 2003; Heim et al. 2006); 21Ne dating of desert pavements tectonic history of Australia? suggests a major arid phase started at ca 4–2 Ma 904 P. M. Vasconcelos et al.

(Fujioka et al. 2005). Magnetostratigraphic studies also assemblages interpreted as indicative of continental suggest that the trend towards aridification intensified cooling may be as old as Pliocene, suggesting that throughout the Pleistocene (An et al. 1986; Zhisheng continental cooling may have started earlier than et al. 1986; Chen & Barton 1991; Zheng et al. 1998; Pillans previously interpreted. Fossil marsurpial Glaucodon & Bourman 2001). In addition, U-series dating of corals ballaratensis from sediments overlying a ca 2.1 Ma and speleothems and cosmogenic-isotope geochronology basalt flow indicates that this fauna is latest Pliocene or of glacial deposits reveal that glacial–interglacial cycles younger, while fossil crocodilian remains in sediments strongly controlled continental climates in Australia below the same basalts reveal that the climate in the (Stirling et al. 1998; McCulloch & Esat 2000; Zhao et al. region must have been significantly warmer before the 2001). Laser-ablation U-series dating of pedogenic iron Pliocene (Aziz-Ur-Rahman & McDougall 1972). K–Ar oxides (Bernal et al. 2006) and 40Ar/39Ar dating of dating (3.6 + 0.1 Ma) of basalt flows overlying the manganese oxides (Feng & Vasconcelos 2001, 2007) also fossiliferous units in the Allingham Formation reveals confirm the orbital forcing on Australian climates. that the Buff Downs Local Fauna provides a minimum Thus, further high-resolution studies at geographically age for this key locality in Australian vertebrate distributed key sites in Australia would greatly enhance biostratigraphy (Mackness et al. 2000). our understanding of long- and short-term climate Despite the knowledge provided by K–Ar constraints cycles on the biological evolution of the continent, and on fossiliferous sites, most important paleontological would provide the necessary information to predict the sites in Australia do not contain datable volcanic units likely effects of future climate changes. delimiting their ages. ESR, TL, OSL, 14C and U-series disequilibria are the current methods of choice to provide Australia’s Cenozoic climatic history, biological age constraints on some of these sites. Combined ESR and evolution and human occupation U-series dating of the fossiliferous horizons in the Naracoorte Caves, South Australia (Gru¨n et al. 2001; Unravelling the climatic history of Australia is essential Prideaux et al. 2007) clearly indicate that better geochro- to trace the environmental changes that controlled nological constraints will increase our understanding of biological evolution on the continent. Much that is the biological evolution in Australia and the relative roles known about Cenozoic climatic change and biological of climatic changes and human occupation in driving evolution comes from fossiliferous sediments whose important extinction events. However, many fossiliferous ages are constrained by underlying and/or overlying deposits remain undated. For example, time constraints volcanic rocks. These dated sites yield a consistent on the Riversleigh fossil site in northwest Queensland are history showing the progressive cooling and aridifica- only tentatively determined. Geochronological techni- tion of Australia throughout the Cenozoic. Wellman & ques suitable for dating these sites would greatly improve McDougall (1974b) showed that the K–Ar ages of our understanding of faunal, floral, and climatic evolu- Cenozoic volcanic rocks in New South Wales impose a tion of Australia. (U–Th)/He dating of fossil apatite, 45–16 Ma limit on distribution of the Cinnamomum flora although problematic due to poor He retentivity, may in this region, consistent with a transition toward more provide usable geochronological constraints. (U–Th)/He arid and possibly seasonal climate during the Miocene. and laser-ablation U-series dating of iron oxides and K–Ar dating (22.4 + 0.5 Ma) of basalt flows interpreted to hydroxides precipitated during the fossilisation process overlie a fossiliferous travertine deposit in Tasmania may also provide useful age constraints on the fossil sites. suggests that marsurpials have resided in Australia Finally, 40Ar/39Ar dating of supergene manganese oxides, since the Late Oligocene–Early Miocene (Tedford et al. which often coat fossilised bones, may also provide 1975). Present-day representatives of these fossil mar- complementary age information about these sites. Downloaded By: [University of Queensland] At: 05:12 12 September 2008 surpials only occur in northern Australia and Papua An important example where the combined applica- New Guinea (Tedford et al. 1975). Their modern dis- tion of modern geochronological tools has enhanced our tribution, together with evidence suggesting that the present knowledge is in the study of the human marsurpial-bearing travertine accumulated in a warm- occupation of Australia. The timing of first human temperate to subtropical environment, indicates that settlement in Australia provides crucial constrains on the climatic conditions in Tasmania, then at 528S, were the timings of human evolution and migration globally. much warmer at that time than at present (Tedford et al. Until the early 1990s, the earliest human occupation 1975). K–Ar dating (26.7 + 0.3 Ma) of basalt flows capping in Australia and adjacent islands (Tasmania and New fluviolacustrine sediments provides chronological con- Guinea) was considered to have occurred between 40 straints on palynofloras in continental Tasmania during and 35 ka. 14C chronology of volcanic flows was the Paleogene (Macphail & Hill 1994). The results instrumental in dating some of these important sites. confirm that the latest Eocene–earliest Oligocene was a For example, aboriginal artefacts are found above and time of significant cooling, and that the Eocene mega- below the Tower Hill Tuff; therefore, establishing an age therm–mesotherm rainforest in Bass Strait did not for this volcano was a major goal of several archeologi- survive into the Oligocene. These results are consistent cal, paleontological and geological investigations. Initial with global cooling trends, and they reveal that con- efforts to date materials above and below the tuff yielded tinental glacial deposits may have been present in many disputed dates, but it was generally accepted that elevated topographic areas in Tasmania at the time the eruption occurred sometime between ca 9 and 6 ka (Macphail & Hill 1994). (Gill 1967; Coutts 1981; Edney et al. 1985). Subsequent Similarly, K–Ar dating of Pliocene and Quaternary dating of organic materials from basal sediments in basalt flows in the Newer Volcanics reveals that fossil Tower Hill Lake yields older results (between 23 Geochronology of the Australian Cenozoic 905

260 + 2540 and 9980 + 140 a), suggesting a greater age for eruption age, revealing that the aboriginal artefacts at the Tower Hill volcano (Head et al. 1991). This inter- Tower Hill are older than previously considered. pretation is supported by the results of Sherwood et al. The application of modern AMS 14C techniques has (2004), who undertook conventional 14C on woody plant been important in pushing back the record of human material preserved beneath tuff, AMS 14C on basal occupation elsewhere in Australia. For example, Fifield material recovered from a lake sediment core, and et al. (2001) used AMS 14C at the Devils Lair site in thermoluminescence on quartz sands beneath tuff. The southwest Australia to provide evidence for human concordance of dates from these three methods suggests occupation by at least 44 ka, and probably as early as 47– that 35 + 3 ka is the best estimate for the Tower Hill 46 ka (Figure 22). The pre-treatment technique used by Downloaded By: [University of Queensland] At: 05:12 12 September 2008

Figure 22 Several sites in Australia (a) hosting evidence for early human occupation. Accurate ages for these sites are necessary to determine the first arrival of humans on the continent and the pattern of human occupation since arrival. Careful stratigraphic studies combined with the application of modern 14C (b); combined TL, OSL, and 14C (c); combined TL and 14C (d); and OSL dating of confining sediments combined with direct dating of fossilised remains by ESR and U-series disequilibria (e) show that humans had arrived and spread throughout Australia by ca 60 ka. Results in (b) Devil’s Lair site in southwestern Australia, from Fifield et al. (2001) and Turney et al. (2001); (c) Nauwalabila I, after Roberts et al. (1994) and Roberts & Jones (1994); (d) Malakunanja II, after Roberts & Jones (1994); and (e) Lake Mungo, from Thorne et al. (1999). 906 P. M. Vasconcelos et al.

Fifield et al. (2001), known as ABOX-SC (acid–base-wet and weathered volcanic rocks suggest that Australia at oxidation followed by stepped combustion), enables the the time was a green continent, containing abundant removal of non-original carbon from ancient carbon and vegetation and a rich fauna, despite its high latitude allows reliable dating of altered samples. position. Dated volcanic rocks reveal that the Cenozoic Another important approach that greatly increases marks the time when Australia started its long and the confidence in such geochronologic results is the rapid (*70 km/Ma) northward trajectory. As it moved simultaneous applications of different but complemen- northward, it travelled over hotspots (as indicated by tary techniques combined with careful stratigraphic the ages of central volcanos), it underwent compression studies to ensure that the material undergoing analysis and extension (as potentially suggested by the ages of confidently constrains the age of the fossil or artefact of lava fields), and it slowed down and changed paths (as interest. Notable examples of the power of this approach indicated by the time–space distribution of central are the evidence for even earlier human occupation of volcanos and seamounts) probably due to collisions at Australia provided by U-series and luminescence (TL, its northern margins. While it moved northward, OSL, or ESR) studies, which yield ages ranging between Australia became colder and drier. These conditions 60 and 50 ka for several sites on the continent (Roberts drove a progressive change in flora and fauna, and et al. 1990, 1994; Thorne et al. 1999; Turney et al. 2001; promoted the reddening of the continent. Water-tables Fifield et al. 2001) (Figure 22). In the case of the Lake withdrew, rivers dried, and salt-lake deposits devel- Mungo skeleton (Thorne et al. 1999), material from the oped. These conditions were enhanced by the cyclical skeleton was dated directly by both U-series and ESR glaciations at the end of the Cenozoic, creating techniques, yielding a weighted mean age of 62 + 6ka further pressures on the Australian fauna and flora. for these human remains. This age for the Lake Mungo These pressures were enhanced by the arrival of skeleton suggests that humans existed in southeastern humans at more than 60 ka, which may have further Australia by ca 60 ka. However, OSL results yield contributed to massive extinctions and environmental younger results and suggest that humans were only changes. present at Lake Mungo by 50–46 ka (Bowler et al. 2003). The search to understand the nature of time and its Overall, the balance of evidence provided by several effect on the Australian continent has never been more different studies using luminescence or U-series meth- important, as the demand and exploration for new ods (Roberts et al. 1990, 1994; Thorne et al. 1999; Fifield mineral and energy resources accelerate, and concerns et al. 2001; Turney et al. 2001) (Figure 22) suggests that about the human impact on the natural environment humans arrived in Australia no later than 60 ka. An increase. To address these pressing issues, Australia important problem still remaining to be solved is the has embarked on a new nationwide effort to improve pattern of human dispersion and the possible influence the quality of the spatial information on the continent. of human occupation on the extinction of the megafauna That effort is accompanied by investment in the on the continent (Brook et al. 2007; Prideaux et al. 2007; development and application of computational tools to Wroe & Field 2007). Only a more complete record of make the earth more transparent to scrutiny by human occupation will permit these issues to be humans. These efforts will be greatly enhanced if we resolved. also continue to generate the high-resolution chronolo- gical record necessary to build and calibrate computa- tional models and translate spatial information into CONCLUDING REMARKS geological reconstructions.

The spatial distribution of Earth’s materials (rocks, Downloaded By: [University of Queensland] At: 05:12 12 September 2008 sediments, soil, water, air and biota) creates a geogra- ACKNOWLEDGEMENTS phical snapshot of the planet; the spatial distribution of these materials through time creates a geological movie We express our deepest appreciation to John De Laeter of our planet. It is the direction of time’s arrow that for the invitation to contribute to this thematic issue of permits the static geographical image to be translated the Australian Journal of Earth Sciences and for his into a dynamic geological history. Without time, there is outstanding contributions to mass spectrometry and to no geological record. With poor time information, the science, education and industry in Australia. We have record is blurry. As time resolution increases, the had helpful discussions with numerous people, includ- record becomes clearer, more vivid and dynamic. Our ing: Tony Ewart, Ken Farley, Andy Gleadow, Trevor view of Australia has become clearer, more vivid and Ireland, Barry Kohn, Gordon Lister, Ian McDougall, dynamic thanks to the geochronologists who dedicated Mike Sandiford, David Schuster, John Stone, Lin time to develop, perfect and widely apply geochronolo- Sutherland, Simon Turner and Alan Chivas. Their gical tools to unravel the evolutionary history of this comments and discussions have contributed to our continent. knowledge of the Australian Cenozoic and the diverse Geochronology paints a colourful and active motion field of geochronology, although they may not necessa- picture; as a continent, Australia came into being in the rily share our views. We thank Brad Pillans for detailed Mesozoic, with separation from Antarctica and New and helpful comments and suggestions on the contents Zealand through opening of the Southern Ocean and the of this manuscript, and his corrections to the sections Tasman Sea, respectively. This separation started on paleomagnetism applied to weathering processes. We reluctantly, and Australia initially moved away from also thank Ian McDougall for his thorough review of the Antarctica slowly (6 km/Ma). Dated weathering profiles manuscript. We thank Maria Lima for help with the Geochronology of the Australian Cenozoic 907

references; Maria Lima, Marinho Morais Neto, and BIRD M. I., AYLIFFE L. K., FIFIELD L. K., TURNEY C. S. M., CRESSWELL Hevelyn Monteiro for drafting many of the figures; and R. G., BARROWS T. T. & DAVID B. 1999. Radiocarbon dating of ‘old’ charcoal using a wet oxidation-stepped combustion procedure. Kathryn Waltenberg for helping compiling the informa- Radiocarbon 41, 127–140. tion in Appendices 1* and 2*. Our work on the Cenozoic BIRD M. I. & CHIVAS A. R. 1988. Stable-isotope evidence for history of Australia would not be possible without the low-temperature kaolinitic weathering and post-formational support, to UQ-AGES, from the Australian Research hydrogen-isotope exchange in Permian kaolinites. Chemical Geology 72, 249–265. Council Grants (ARC Large Grant A39531815, A39701507, BIRD M. I. & CHIVAS A. R. 1989. Stable-isotope geochronology of the DP06666925). Australian regolith. Geochimica et Cosmochimica Acta 53, 3239– 3256. BIRD M. I. & CHIVAS A. R. 1993. Geomorphic and palaeoclimatic implications of an oxygen-isotope chronology for Australian REFERENCES deeply weathered profiles. Australian Journal of Earth Sciences 40, 345–358. ACTON G. D. & KETTLES W. A. 1996. Geologic and palaeomagnetic BIRD M. I., CHIVAS A. R. & MCDOUGALL I. 1990a. An isotopic study of constraints on the formation of weathered profiles near Inverell, surficial alunite in Australia: 2. Potassium–argon geochronol- Eastern Australia. Paleogeography, Palaeoclimatology, Palaeoe- ogy. Chemical Geology 80, 133–145. cology 126, 211–225. BIRD M., FYFE B., CHIVAS A. & LONGSTAFFE F. 1990b. Deep weath- AN Z., BOWLER, J. M., OPDYKE, N. D., MACUMBER P. G. & FIRMAN J. B. ering at extra-tropical latitudes: a response to increased atmo-

1986. Palaeomagnetic stratigraphy of Lake Bungunnia: Plio- spheric CO2. In: Bouwman A. F. ed. Proceedings of the Pleistocene precursor of aridity in the Murray Basin, south- International Conference, Soils and the Greenhouse Effect, eastern Australia. Paleogeography, Palaeoclimatology, Palaeoe- pp. 383–389. Wiley, Chichester. cology 54, 219–239. BISHOP P., YOUNG R. W. & MCDOUGALL I. 1985. Stream profile change ANAND R. R. & PAINE M. 2002. Regolith geology of the Yilgarn Craton, and longterm landscape evolution: Early Miocene and modern Western Australia: implications for exploration. Australian rivers of the East Australian Highland Crest, Central New South Journal of Earth Sciences 49, 3–162. Wales, Australia. Journal of Geology 93, 455–474. ANDREWS E. C. 1911. Geographical unity of eastern Australia in late BLACKBURN G., ALLISON G. B. & LEANEY F. W. J. 1982. Further and post Tertiary time, with applications to biological problems. evidence on the age of the tuff at Mt Gambier, South Australia. Journal and Proceedings of the Royal Society of New South Wales Transactions of the Royal Society of South Australia 106, 163–167. 44, 420–480. BOWLER J. M., HARVEY J., OLLEY J. M., PRESCOTT J. R., ROBERTS R. ASHLEY P. M., DUNCAN R. A. & FEEBREY C. A. 1995. Ebor Volcano and G., SHAWCROSS W. & SPOONER N. A. 2003. New ages for human crescent complex, northeastern New South Wales: age and occupation and climatic change at Lake Mungo, Australia. geological development. Australian Journal of Earth Sciences Nature 421, 837–840. 42, 471–480. BROOK B. W., BOWMAN D. M. J. S., BURNEY D. A., FLANNERY T. F., AZIZ-UR-RAHMAN &MCDOUGALL I. 1972. Potassium–argon ages of the GAGAN M. K., GILLESPIE R., JOHNSON C. N., KERSHAW P., MAGEE J. Newer Volcanics of Victoria. Proceedings of the Royal Society of W., MARTIN P. S., MILLER G. H., PEISER B. & ROBERTS R. G. 2007. Victoria 85, 61–70. Would the Australian megafauna have become extinct if humans BARROWS, T. T., STONE J. O. & FIFIELD L. K. 2004. Exposure ages for had never colonised the continent? Comments on ‘A review of Pleistocene periglacial deposits in Australia. Quaternary Science the evidence for a human role in the extinction of Australian Reviews 23, 697–708. megafauna and an alternative explanation’ by S. Wroe and J. BARROWS, T. T., STONE,J.O.,FIFIELD L. K. & 2001. Late Pleistocene Field. Quaternary Science Reviews 26, 560–564. glaciation of the Kosciuszko Massif, Snowy Mountains, Austra- BROWN M. C. 2007. Discussion. Apatite (U–Th)/He age constraints on lia. Quaternary Research 55, 179–189. the Mesozoic and Cenozoic evolution of the Bathurst region, New BARROWS, T. T., STONE, J. O., FIFIELD L. K. & CRESSWELL R. G. 2002. South Wales: evidence for antiquity of the continental drainage The timing of the last glacial maximum in Australia. Quaternary divide along a passive margin. Australian Journal of Earth Science Reviews 21, 159–173. Sciences 54, 1009–1020. BARTON C. E. & MCELHINNY M. W. 1980. Ages and ashes in lake floor BROWN M. C., MCQUEEN K. G. & TAYLOR G. 1992. A core through the sediment cores from Valley Lake, Mt Gambier, South Australia. Monaro basalt: Bega (BMR) no. 7. Australian Journal of Earth Transactions of the Royal Society of South Australia 104, 161–165. Sciences 39, 555–559. BELTON, D. X., BROWN, R. W., KOHN, B. P., FINK D. & FARLEY K. A. BROWNE W. R. 1969. Geomorphology of New South Wales. Journal of Downloaded By: [University of Queensland] At: 05:12 12 September 2008 2004. Quantitative resolution of the debate over antiquity of the the Geological Society of Australia 16, 559–569. central Australian landscape: implications for the tectonic and BUTLER R. F. 1992. Paleomagnetism: Magnetic Domains to Geologic geomorphic stability of cratonic interiors. Earth and Planetary Terranes. Blackwell Scientific Publications, Oxford. Science Letters 219, 21–34. BUTT C. R. M. 1988. Major uranium provinces: Yilgarn Block and BENTLEY, H. W., PHILLIPS,F.M.,DAVIS,S.N.,HABERMEHL,M.A., Gascoyne Province, Western Australia. In: Recognition of AIREY,P.L.,CALF,G.E.,ELMORE D., GOVE H. E. & TORGERSEN T. Uranium Provinces, pp. 273–304. International Atomic Energy 1986. Chlorine 36 dating of very old groundwater: 1, The Great Agency, Vienna. Artesian Basin, Australia. Water Resources Research 22, 1991–2001. CANDE S. C. & KENT D. V. 1995. Revised calibration of the BERNAL, J. P., EGGINS S. M. & MCCULLOCH M. T. 2005. Accurate in geomagnetic polarity timescale for the late Cretaceous and situ 238U–234U–232Th–230Th analysis of silicate glasses and iron Cenozoic. Journal of Geophysical Research 100, 6093–6095. oxides by laser-ablation MC-ICP-MS. Journal of Analytical CANDE S. C. & MUTTER J. C. 1982. A revised identification of the Atomic Spectrometry 20, 1240–1249. oldest sea-floor spreading anomalies between Australia and BERNAL, J. P., EGGINS, S. M., MCCULLOCH, M. T., GRU¨ N R. & Antarctica. Earth and Planetary Science Letters 58, 151–160. EGGLETON R. A. 2006. Dating of chemical weathering processes CAWOOD P. A., NEMCHIN A. A., FREEMAN M. & SIRCOMBE K. 2003. by in situ measurement of U- supergene Fe-oxy/hydroxides Linking source and sedimentary basin: detrital zircon record of using LA-MC-ICPMS. Chemical Geology 235, 76–94. sediment flux along a modern river system and implications for BIERMAN P. R. & CAFFEE M. 2002. Cosmogenic exposure and erosion provenance studies. Earth and Planetary Science Letters 210, history of Australian bedrock landforms. Geological Society of 259–268. America Bulletin 114, 787–803. CHEN X. Y. & BARTON C. E. 1991. Onset of aridity and dune-building 10 26 BIERMAN P. & TURNER J. 1995. Be and Al evidence for exceptionally in Central Australia: sedimentological and magnetostrati- low rates of Australian bedrock erosion and the likely existence graphic evidence from Lake Amadeus. Palaeogeography, Palaeo- of pre-Pleistocene landscapes. Quaternary Research 44, 378–382. climatology, Palaeoecology 84, 55–73. BIRD, J. R., CALF, G. E., DAVIE, R. F., FIFIELD, L. K., OPHEL,T.R., CHIVAS A. R., DE Deckker P., NIND M., THIRIET D. & WATSON G. 1986. EVANS W. R., KELLETT J. R. & HABERMEHL M. A. 1989. The role of The Pleistocene palaeoenvironmental record of Lake Buchanan: 36Cl and 14C measurements in Australian groundwater studies. an atypical Australian playa. Palaeogeography, Palaeoclimatol- Radiocarbon 31, 877–883. ogy, Palaeoecology 54, 131–152. 908 P. M. Vasconcelos et al.

CHUKHROV F. V., SHANIN L. L. & YERMILOVA L. P. 1966. Feasi- DAVIE R. F., KELLETT J. R., FIFIELD L. K., EVANS W. R., CALF G. E., bility of absolute-age determination for potassium-carrying BIRD J. R., TOPHAM S. & OPHEL T. R. 1989. Chlorine-36 measure- manganese minerals. International Geology Review 8, 278– ments in the Murray Basin; preliminary results from the 280. Victorian and South Australian Mallee region. BMR Journal of CHUKHROV F. V., YERMILOVA L. P. & SHANIN L. L. 1969. Age of alunite Australian Geology & Geophysics 11, 261–272. from certain deposits. Transactions (Doklady) of the USSR DE DECKKER P. 1983. Notes on the ecology and distribution of non- Academy of Sciences: Earth Science Sections 185, 49–51. marine ostracods in Australia. Hydrobiologia 106, 145–153. COENRAADS R. R., SUTHERLAND F. L. & KINNEY P. D. 1990. The origin DICKIN A. P. 1995. Radiogenic Isotope Geology. Cambridge University of sapphires: U–Pb dating of zircon inclusions sheds new light. Press, Cambridge. Mineralogical Magazine 54, 113–122. DUDDY I. R., GREEN P. F. & LASLETT G. M. 1988. Thermal annealing of COHEN B. E. 2007. High-resolution 40Ar/39Ar geochronology of fission tracks in apatite—3. Variable temperature behavior. Intraplate Volcanism in Eastern Australia. PhD thesis. Univer- Chemical Geology 73, 25–38. sity of Queensland, Brisbane. (unpubl.). DUFF M. C., COUGHLIN J. U. & HUNTER D. B. 2002. Uranium co- COHEN B. E., KNESEL K. M., VASCONCELOS P. M., THIEDE D. S. & precipitation with iron oxide minerals. Geochimica et Cosmochi- 40 39 HERGT J. M. 2008. Ar/ Ar constraints on the timing mica Acta 66, 3533–3547. and origin of Miocene leucitite volcanism in southeastern DUMITRU T. A., HILL K. C., COYLE D. A., DUDDY I. R., FOSTER D. A., Australia. Australian Journal of Earth Sciences 55, 391–402. GLEADOW A. J. W., GREEN P. F., KOHN B. P., LASLETT G. M. & COHEN B. E., VASCONCELOS P. M. & KNESEL K. M. 2007. 40Ar/39Ar O’SULLIVAN A. J. 1991. Fission track thermochronology; applica- constraints on the timing of Oligocene intraplate volcanism in tion to continental rifting of south-eastern Australia. APEA southeast Queensland. Australian Journal of Earth Sciences 54, Journal 31, 131–142. 105–125. DUNCAN R. A. & MCDOUGALL I. 1989a. Plate tectonic setting. In: COLLON P., KUTSCHERA W., LOOSLI H. H., LEHMANN B. E., PURTSCHERT Johnson R. W. ed. Intraplate Volcanism in Eastern Australia and R., LOVE A., SAMPSON L., ANTHONY D., COLE D., DAVIDS B., New Zealand, pp. 13–18. Cambridge University Press, Cambridge. MORRISSEY D. J., SHERRILL B. M., STEINER M., PARDO R. C. & PAUL DUNCAN R. A. & MCDOUGALL I. 1989b. Volcanic time–space relation- M. 2000. 81Kr in the Great Artesian Basin, Australia: a new ships. In: Johnson R. W. ed. Intraplate Volcanism in Eastern method for dating very old groundwater. Earth and Planetary Australia and New Zealand, pp. 43–54. Cambridge University Science Letters 182, 103–113. Press, Cambridge. COMPSTON W., WILLIAMS I. S. & MEYER C. 1984. U–Pb geochronology DURY G. H. 1971. Relict deep weathering and duricrusting in relation of zircons from lunar Breccia 73217 using a sensitive high mass- to the palaeoenvironments of middle latitudes. Geographical resolution ion microprobe. Journal of Geophysical Research 89, Journal 137, 511–521. 525–534. DURY G. H. & KNOX J. C. 1971. Duricrusts and deep-weathering COOPER J. A., RICHARDS H. C. & WEBB A. W. 1963. Some potassium– profiles in southwestern Wisconsin. Science 174, 291–292. argon ages in New England, New South Wales. Journal of the DURY G. H., LANGFORD-SMITH T. & MCDOUGALL I. 1969. A minimum Geological Society of Australia 10, 313–316. age for the duricrust. Australian Journal of Science 31, 362–363. COUTTS P. J. F. 1981. Coastal archaeology in Victoria. Part 2: EDNEY P., KERSHAW A. P., PETERSON J. A. & COUTTS P. J. F. 1985. Adaptation, technology, and volcanism. Proceedings of the Royal Evidence of a Pleistocene age for Tower Hill, Western Victoria. Society of Victoria 93, 15–22. Search 16, 302–303. 238 COVENTRY R. J. 1979. Late Cainozoic geology, soils, and landscape EGGINS S., GRU¨ N R., PIKE A. W. G., SHELLEY M. & TAYLOR L. 2003. U, evolution of the Torrens Creek area, north Queensland. Journal 232Th profiling and U-series isotope analysis of fossil teeth by of the Geological Society of Australia 25, 415–427. laser ablation – ICPMS. Quaternary Science Reviews 22, 1373– COVENTRY R. J., STEPHENSON P. J. & WEBB A. W. 1985. Chronology 1382. of landscape evolution and soil development in the upper EMBLETON B. J. J. 1981. A review of the paleomagnetism of Australia Flinders River area, Queensland, based on isotope dating of and Antarctica. In: McElhinny M. W. & Valencio D. A. eds. Cainozoic basalts. Australian Journal of Earth Sciences 32, 433– Paleoreconstruction of the Continents, pp. 77–92. American 447. Geophysical Union Geodynamics Series 2. CRAFT F. A. 1933. The coastal tablelands and streams of New South EMBLETON B. J. J. & MCELHINNY M. W. 1982. Marine magnetic Wales. Proceedings of the Linnean Society of New South Wales 58, anomalies, paleomagnetism and the drift history of Gondwana- 437–460. land. Earth and Planetary Science Letters 58, 141–150. CRESSWELL R. G., JACOBSON G., WISCHUSEN J. & FIFIELD L. K. 1999a. EWART A. 1982. Petrogenesis of the Tertiary anorogenic volcanic Ancient groundwaters in the Amadeus Basin, central Australia: series of southern Queensland, Australia, in the light of trace evidence from the radio-isotope 36Cl. Journal of Hydrology 223, element geochemistry and O, Sr and Pb isotopes. Journal of Downloaded By: [University of Queensland] At: 05:12 12 September 2008 212–220. Petrology 23, 344–382. CRESSWELL R. G., WISCHUSEN J., JACOBSON G. & FIFIELD L. K. 1999b. EWART A., CHAPPELL B. W. & LE MAITRE R. W. 1985. Aspects of the Assessment of recharge to groundwater systems in the arid mineralogy and chemistry of the intermediate–silicic Cainozoic southwestern part of Northern Territory, Australia, using volcanic rocks of eastern Australia. Part 1: introduction and chlorine-36. Hydrogeology Journal 7, 393–404. geochemistry. Australian Journal of Earth Sciences 32, 359–382. CRONE A. J., DE MARTINI P. M., MACHETTE M. N., OKUMURA K. & EXON N. F., LANGFORD-SMITH T. & MCDOUGALL I. 1970. The age and PRESCOTT J. R. 2003. Paleoseismicity of two historically quiescent geomorphic correlations of deep-weathering profiles, silcrete, faults in Australia: implications for fault behavior in stable and basalt in the Roma-Amby region, Queensland. Journal of the continental regions. Bulletin of the Seismological Society of Geological Society of Australia 17, 21–30. America 93, 1913–1934. EXON N. F., MILLIGAN E. N. & DAY R. W. 1967. Discussion: age of the CUNDARI A., RENARD J. G. R. & GLEADOW A. J. W. 1978. Uranium– duricrust in southern Queensland. Australian Journal of Science potassium relationship and apatite fission-track ages for a 30, 110. differentiated leucitite suite from New South Wales (Australia). FARLEY K. A. 2002. (U–Th)/He dating: techniques, calibrations, and Chemical Geology 22, 11–20. applications. In: Porcelli D., Ballentine C. J. & Wieler R. eds. DAMMER D., CHIVAS A. R. & MCDOUGALL I. 1996. Isotopic dating Noble Gases in Geochemistry and Cosmochemistry, pp. 819–844. of supergene manganese oxides from Groote Eylandt Reviews in Mineralogy and Geochemistry 47. deposit, Northern Territory, Australia. Economic Geology 91, FARLEY K. A., WOLF R. A. & SILVER L. T. 1996. The effects of long 386–401. alpha-stopping distances on (U–Th)/He ages. Geochimica et DAMMER D., MCDOUGALL I. & CHIVAS A. R. 1999. Timing of weath- Cosmochimica Acta 60, 4223–4229. ering-induced alteration of manganese deposits in Western FAURE G. 1991. Principals and Applications of Geochemistry (2nd Australia: evidence from K/Ar and 40Ar/39Ar dating. Economic edition). Prentice-Hall, Englewood Cliffs, NJ. Geology 94, 87–108. FAURE G. & MENSING T. M. 2005. Isotopes; Principles and Applica- DASCH E. J. & MILLAR D. J. 1977. Age and strontium-isotope tions. Wiley, New York. geochemistry of differentiated rocks from the Newer Volcanics, FENG Y. & VASCONCELOS P. 2001. Quaternary continental weathering Mt Macedon area, Victoria, Australia. Journal of the Geological geochronology by laser-heating 40Ar/39Ar analysis of supergene Society of Australia 24, 195–201. cryptomelane. Geology 29, 635–638. Geochronology of the Australian Cenozoic 909

FENG Y. X. & VASCONCELOS P. M. 2007. Chronology of Pleistocene GLEADOW A. J. W., DUDDY I. R. & LOVERING J. F. 1983. Fission track weathering process, southeast Queensland, Australia. Earth and analysis: a new tool for the evaluation of thermal histories and Planetary Science Letters 263, 275–287. hydrocarbon potential. APEA Journal 23, 93–102. FIFIELD L. K., BIRD M. I., TURNEY C. S. M., HAUSLADEN P. A., SANTOS GLEADOW A. J. W. & LOVERING J. F. 1974. Effect of weathering on fission- G. M. & DI TADA M. L. 2001. Radiocarbon dating of the human track dating. Earth and Planetary Science Letters 22, 163–168. occupation of Australia prior to 40 ka BP: successes and pitfalls. GLEADOW A. J. W. & OLLIER C. D. 1987. The age of gabbro at The Radiocarbon 43, 1139–1145. Crescent, New South Wales. Australian Journal of Earth Sciences FIFIELD L. K., OPHEL T. R., BIRD J. R., CALF G. E., ALLISON G. B. & 34, 209–212. CHIVAS A. R. 1987. The chlorine-36 measurement program at the GODWIN H. 1962. Radiocarbon dating; fifth international conference. Australian National University. Nuclear Instruments and Meth- Nature 195, 943–945. ods in Physics Research 29, 114–119. GOLDICH S. S. 1938. A study in rock weathering. Journal of Geology FITZPATRICK E. A. 1963. Deeply weathered rock in Scotland, its 46, 17–58. occurrence, age, and contribution to the soils. Journal of Soil GRAHAM I., HOLLIS J. D., SUTHERLAND F. L. & JOYCE E. B. 2003. Science 14, 33–43. Insights into the Newer Volcanics Province of Victoria. Geolo- FOULGER G. R., NATLAND J. H., PRESNALL D. C. & ANDERSON D. L. gical Society of Australia, Specialist Group in Geochemistry, 2005. Plates, plumes, and paradigms. Geological Society of Mineralogy and Petrology, Field Guide 6. America Special Paper 388. GREEN P. F. & DUDDY I. R. 2007. Discussion. Apatite (U–Th)/He age FRAKES L. A. & BOLTON B. R. 1984. Origin of manganese giants: sea- constraints on the Mesozoic and Cenozoic evolution of the level change and anoxic–oxic history. Geology 12, 83–86. Bathurst region, New South Wales: evidence for antiquity of the FUJIOKA T., CHAPPELL J., HONDA M., YATSEVICH I., FIFIELD K. & continental drainage divide along a passive margin. Australian FABEL D. 2005. Global cooling initiated stony deserts in central Journal of Earth Sciences 54, 1009–1020. 21 10 Australia 2–4 Ma, dated by cosmogenic Ne- Be. Geology 33, GREEN P. F., DUDDY I. R., GLEADOW A. J. W. & LOVERING J. F. 1989. 993–996. Apatite fission track analysis as a palaeotemperature indicator GAINA C., MU¨ LLER R. D., ROYER J-Y., STOCK J., HARDEBECK J. & for hydrocarbon exploration. In: Naeser N. D. & McCulloh T. W. SYMONDS P. 1998. The tectonic history of the Tasman Sea: eds. Thermal History of Sedimentary Basins—Methods and Case puzzle with 13 pieces. Journal of Geophysical Research 103, Histories, pp. 181–195. Springer, New York. 14413–12433. GREEN P. F., DUDDY I. R., GLEADOW A. J. W., MEGARTY K. A. & GALBRAITH R. F., LASLETT G. M., GREEN P. F. & DUDDY I. R. 1990. LOVERING J. F. 1986. Constraints on thermal evolution of Apatite fission-track analysis—geological thermal history ana- sedimentary basins from apatite fission-track analysis. AAPG lysis based on a 3-Dimensional random process of linear Bulletin 70, 596. radiation-damage. Philosophical Transactions of the Royal So- GRIFFIN T. J. & MCDOUGALL I. 1975. Geochronology of the Cainozoic ciety of London 332A, 419–438. McBride volcanic province, northern Queensland. Journal of the GALLAGHER K. 1995. Evolving temperature histories from apatite Geological Society of Australia 22, 387–396. fission-track data. Earth and Planetary Science Letters 136, 421– GRIMES K. G. 1979. The stratigraphic sequence of old land surfaces in 435. northern Queensland. BMR Journal of Australian Geology & GAMBLE J. A., PRICE R. C., SMITH I. E. M., MCINTOSH W. C. & Geophysics 4, 33–46. DUNBAR N. W. 2003. 40Ar/39Ar geochronology of magmatic GRU¨ N R. MORIARTY K. & WELLS R. 2001. Electron spin resonance activity, magma flux and hazards at Ruapehu volcano, Taupo dating of the fossil deposits in the Naracoorte Caves, South Volcanic Zone, New Zealand. Journal of Volcanology and Australia. Journal of Quaternary Science 16, 49–59. Geothermal Research 120, 271–287. GRU¨ N R. & ROBERTS R. G., eds. 2007. LED 2005. 11th International GEYH M. A. & SCHLEICHER H. 1990. Absolute Age Determination. Conference on Luminescence and Electron Spin Resonance Physical and Chemical Dating Methods and their Application. Dating. Quaternary Geochronology 2(1–4). Springer, Berlin. GRU¨ N R., STRINGER C., MCDERMOTT F., NATHAN R., PORAT N., GIBSON D. L. 2007. Discussion. Apatite (U–Th)/He age constraints on ROBERTSON S., TAYLOR L., MORTIMER G., EGGINS S. & MCMULLOCH the Mesozoic and Cenozoic evolution of the Bathurst region, New M. 2005. U-Series and ESR analyses of bones and teeth relating to South Wales: evidence for antiquity of the continental drainage the human burials from skhul. Journal of Human Evolution 49, divide along a passive margin. Australian Journal of Earth 316–334. Sciences 54, 1009–1020. GUNNELL Y. 2003. Radiometric ages of laterites and constraints on GILL E. D. 1967. Evolution of the Warrnambool – Port Fairy long-term denudation rates in West Africa. Geology 31, 131–134. Coast and the Tower Hill eruption, western Victoria. In: GURNIS M., MU¨ LLER R. D. & MORESI L. 1998. Cretaceous vertical Jennings J. N. & Mabbutt J. A. eds. Landform Studies from motion of Australia and the Australian–Antarctic Discordance. Downloaded By: [University of Queensland] At: 05:12 12 September 2008 Australia and New Guinea, pp. 64–340. Australian National Science 279, 1499–1504. University Press, Canberra. HALL A. M. 1985. Cenozoic weathering covers in Buchan, Scotland, GILL E. D. 1972. Eruption date of Tower Hill volcano, western and their significance. Nature 315, 392–395. Victoria, Australia. Victorian Naturalist 89, 188–192. HALL R. 2002. Cenozoic geological and plate tectonic evolution of SE GILL E. D. & ELMORE L. K. M. 1973. Radiocarbon dating of Mount Asia and the SW Pacific: computer-based reconstructions, model Napier eruption, western Victoria, Australia. Victorian Natur- and animations. Journal of Asian Earth Sciences 20, 353–431. alist 90, 304–306. HABERLE S. G. 2005. A 23,000-yr pollen record from Lake Euramoo, GILL E. D. & GIBBONS F. R. 1969. Radiocarbon date related to wet tropics of NE Queensland, Australia. Quaternary Research volcanism and lake deposits in western Victoria. Australian 64, 343–356. Journal of Science 32, 109–110. HARE A. G., CAS R. A. F., MUSGRAVE R. & PHILLIPS D. 2005. Magnetic GLEADOW A. J. W., BELTON D. X., KOHN B. P. & BROWN R. W. 2002a. and chemical stratigraphy for the Werribee Plains basaltic lava Fission track of phosphate minerals and the thermochronology flow-field, Newer Volcanics Province, southeast Australia: of apatite. Tectonophysics 349, 5–21. implications for eruption frequency. Australian Journal of Earth GLEADOW A. J. W., BELTON D. X., KOHN B. P. & BROWN R. W. 2002b. Sciences 52, 41–57. Fission track of phosphate minerals and the thermochronology HEAD L., D’COSTA D. & EDNEY P. 1991. Pleistocene dates for volcanic of apatite. In: Hugles J., Kohm M. & Rakovan J. eds. Phosphates, activity in Western Victoria and implications for Aboriginal Geochemical, Geobiological and Materials Importance, pp. 579– occupation. In: Williams M. A. J., De Deckker P. & Kershaw A. P. 630. Reviews in Mineralogy and Geochemistry, 48. eds. The Cainozoic of Australia: A Re-appraisal of the Evidence. GLEADOW A. J. W., DUDDY I. R., GREEN P. F. & HEGARTY K. A. 1986a. pp. 15–43. Geological Society of Australia Special Publication 18. Fission-track lengths in the apatite annealing zone and the HEIM J. 2007. Geochronology of weathering and landscape evolution, interpretation of mixed ages. Earth and Planetary Science Letters Hamersley Iron Province, Australia. PhD Thesis, University of 78, 245–254. Queensland, Brisbane. (unpubl.). GLEADOW A. J. W., DUDDY I. R., GREEN P. F. & LOVERING J. F. 1986b. HEIM J., VASCONCELOS P. M., SHUSTER D. L., FARLEY K. A. & Confined fission track lengths in apatite—a diagnostic tool for BROADBENT G. 2006. Dating palaeochannel iron ore by (U–Th)/ thermal history analysis. Contributions to Mineralogy and He analysis of supergene goethite, Hamersley province, Austra- Petrology 94, 405–415. lia. Geology 34, 173–176. 910 P. M. Vasconcelos et al.

HEIMSATH A. M., CHAPPELL J., DIETRICH W. E., NISHIIZUMI K. & KLOOTWIJK C. T. 1996. Phanerozoic configuration of greater Aus- FINKEL R. C. 2000. Soil production on a retreating escarpment in tralia: evolution of the North West Shelf. Part 2. Australian southeastern Australia. Geology 28, 787–790. Geological Survey Organisation Record 1996/52. HEIMSATH A. M., CHAPPELL J., DIETRICH W. E., NISHIIZUMI K. & KLOOTWIJK C. T. & PEIRCE J. W. 1979. India’s and Australia’s pole FINKEL R. C. 2001. Late Quaternary erosion in southeastern path since the late Mesozoic and the India-Asia collision. Nature Australia: a field example using cosmogenic nuclides. Quatern- 282, 605–607. ary International 83/85, 169–185. KNESEL K. M., COHEN B. E., VASCONCELOS P. M. & THIEDE D. S. Rapid HEIMSATH A. M., DIETRICH W. E., NISHIIZUMI K. & FINKEL R. C. 1997. change in the drift of the Australian Plate records collision with The soil production function and landscape equilibrium. Nature the Ontong Java Plateau. Nature (in press). 388, 358–361. KOHN B. P., GLEADOW A. J. W., BROWN R. W., GALLAGHER K., HILLS E. S. 1975. The Physiography of Victoria. Whitcombe & Tombs, LORENCALK M. & NOBLE W. P. 2005. Visualizing thermotectonic Melbourne. and denudation histories using apatite fission track thermo- HOURIGAN J. K., REINERS P. W. & BRANDON M. T. 2005. U–Th zonation- chronology. In: Reiners P. & Ehlers T. eds. Thermochronology, dependent alpha-ejection in (U–Th)/He chronometry. Geochimi- pp. 527–565. Reviews in Mineralogy and Geochemistry 58. ca et Cosmochimica Acta 69, 3349–3365. KOHN B. P., GLEADOW A. J. W., BROWN R. W. GALLAGHER K., HURFORD A. J. & GREEN P. F. 1983. The zeta age calibration of fission- O’SULLIVAN P. B. & FOSTER D. A. 2002. Shaping the Australian track dating. Chemical Geology 41, 285–317. crust over the last 300 million years: insights from fission track HURLEY P. M. 1954. The helium age method and the distribution thermotectonic imaging and denudation studies of key terrenes. and migration of helium in rocks. In: Faul H. ed. Nuclear Australian Journal of Earth Sciences 49, 697–717. Geology, pp. 301–328. Wiley, New York. LACKIE M. A. & SCHMIDT P. W. 1993. Drill core orientation using IDNURM M. 1985a. Late Mesozoic and Cenozoic palaeomagnetism of palaeomagnetism. Exploration Geophysics 24, 609–613. Australia: I. A redetermined apparent polar wander path. LAL D. 1991. Cosmic ray labeling of erosion surfaces: in situ nuclide Geophysical Journal of the Royal Astronomical Society 83, 399–418. production rates and erosion models. Earth and Planetary IDNURM M. 1985b. Late Mesozoic and Cenozoic palaeomagnetism of Science Letters 104, 424–439. Australia: II. Implications for geomagnetism and true polar LANGFORD-SMITH T., DURY G. H. & MCDOUGALL I. 1966. Dating the wander. Geophysical Journal of the Royal Astronomical Society duricrust in southern Queensland. Australian Journal of Science 83, 419–433. 29, 78–80. IDNURM M. 1986. Late Mesozoic and Cenozoic palaeomagnetism of LASLETT G. M., GREEN P. F., DUDDY I. R. & GLEADOW A. J. W. 1987. Australia: III, Bias-corrected pole paths for Australia, Antarctica Thermal annealing of fission tracks in apatite. 2. A quantitative and India. Geophysical Journal of the Royal Astronomical Society analysis. Chemical Geology 65, 1–13. 86, 277–287. LEANEY F. W. J., ALLISON G. B., DIGHTON J. C. & TRUMBORE S. 1995. IDNURM M. 1990. Comments on ‘A weighted least-squares fit of the The age and hydrological history of Blue Lake, South Australia. Australian apparent polar wander path for the last 100 Myr’ by Palaeogeography, Palaeoclimatology, Palaeoecology 118, 111–130. R. J. Musgrave: Discussion. Geophysical Journal International LE PICHON X. & HEIRTZLER J. R. 1968. Magnetic anomalies in the 102, 191–193. Indian Ocean and sea-floor spreading. Journal of Geophysical IDNURM M. 1994. New Late Eocene pole for Australia, time-averaging Research 73, 2101–2117. of remanence directions, and palaegeographic reference sys- LI J-W. & VASCONCELOS P. M. 2002. Cenozoic continental weathering tems. Geophysical Journal International 117, 827–833. and its implications for the palaeoclimate: evidence from 40 39 IDNURM M. & COOK P. J. 1980. Palaeomagnetism of beach ridges in Ar/ Ar geochronology of supergene K–Mn oxides in Mt Tabor, south Australia and the Milankovitch theory or ice ages. Nature central Queensland, Australia. Earth and Planetary Science 286, 699–702. Letters 200, 223–239. IDNURM M. & SCHMIDT P. V. 1986. Paleomagnetic dating of weathered LIBBY W. F. 1955. Radiocarbon Dating (2nd edition). University of profiles. Geological Survey of India Memoir 120, 79–88. Chicago Press, Chicago. IDNURM M. & SENIOR B. R. 1978. Palaeomagnetic ages of Late LILLEY G. 1997. Regolith Distribution and Genesis, Tick Hill Region, Cretaceous and Tertiary weathered profiles in the Eromanga Northwest Queensland. Honours thesis, University of Queens- Basin, Queensland. Paleogeography, Palaeoclimatology, Palaeoe- land, Brisbane (unpubl.). cology 24, 263–277. LIPPOLT H. J., BRANDER T. & MANKOPF N. R. 1998. An attempt to IMBRIE J., HAYS J. D., MARTINSON D. G., MCINTYRE A., MIX A., determine formation age of goethites and limonites by MORLEY J. J., PISIAS N. G., PRELL W. & SHACKLETON N. J. 1984. (U þ Th)–4He dating. Neues Jahrbuch Fur Mineralogie-Monat- The orbital theory of Pleistocene climate: support from a revised shefte 200, 505–528. chronology of the marine oxygen isotope record. In: Berger A. L., MACKELLAR D. 1908. Core of My Heart. Spectator, London. Downloaded By: [University of Queensland] At: 05:12 12 September 2008 Imbrie J., Hays J., Kukla G. & Saltzman B. eds. Milankovitch and MACKNESS B. S., WHITEHEAD P. W. & MCNAMARA G. C. 2000. New Climate: Understanding the Response to Astronomical Forcing potassium–argon basalt date in relation to the Pliocene Bluff (NATO Science Series C), pp. 269–305. D. Reidel, Boston. Downs Local Fauna, northern Australia. Australian Journal of IRELAND T. R., CLEMENT S., COMPSTON W., FOSTER J. J., HOLDEN P., Earth Sciences 47, 807–811. JENKINS B., LANC P., SCHRAM N. & WILLIAMS I. S. 2008. MACPHAIL M. K. & HILL R. S. 1994. K–Ar dated palynofloras in Development of SHRIMP. Australian Journal of Earth Sciences Tasmania 1: Early Oligocene Proteacidites tuberculatus Zone 55, 937–954. sediments, Wilmot Dam, northwestern Tasmania. Papers and JAQUES A. L., WEBB A. W., FANNING C. M., BLACK L., PIDGEON R. T., Proceedings of the Royal Society of Tasmania 128, 1–15. FERGUSON J., SMITH C. B. & GREGORY G. P. 1984. The age of the MCCULLOCH M. T. & ESAT T. 2000. The coral record of last diamond-bearing pipes and associated leucite lamproites of the interglacial sea levels and sea surface temperatures. Chemical West Kimberley region, Western Australia. BMR Journal of Geology 169, 107–129. Australian Geology & Geophysics 9, 1–7. MCDOUGALL I. 1961. Determination of the age of a basic igneous intru- JOHNSON R. W. 1989. Intraplate Volcanism in Eastern Australia and sion by the potassium–argon method. Nature 190, 1184–1186. New Zealand. Cambridge University Press, Cambridge. MCDOUGALL I. 1966. Precision methods of potassium–argon isotopic KAPLAN G., FAURE D., ELLOY R. & HEILLAMER, R. 1967. Contribution a age determination on young rocks. In: Runcorn S. K. ed. Methods l’etude de l’origine des lamproites. Bulletin du Centre de and Techniques in Geophysics Volume 2, pp. 279–304. Wiley, New Recherches Exploration-Production Elf-Aquitaine 1, 153–159. York. KARNER D. B., RENNE P. R., MCDOUGALL I. & BECKER T. A. 2001. The MCDOUGALL I. 1971. Volcanic island chains and sea floor spreading. viability of leucite for 40Ar/39Ar dating as a Quaternary Nature 231, 141–144. 40 39 standard. Chemical Geology 177, 473–482. MCDOUGALL I. 1974. The Ar/ Ar method of K–Ar age determina- KIERNAN K., FIFIELD L. K. & CHAPPELL J. 2004. Cosmogenic nuclide tion of rocks using Hifar reactor. Atomic Energy in Australia 17, ages for Last Glacial Maximum moraine at Schnells Ridge, 9–18. Southwest Tasmania. Quaternary Research 61, 335–338. MCDOUGALL I., ALLSOPP H. L. & CHAMALAUN F. H. 1966. Isotopic KING L. C. 1959. Denudation and tectonic relief of southeastern dating of the newer volcanics of Victoria, Australia, and Australia. Transactions and Proceedings of the Geological Society geomagnetic polarity epochs. Journal of Geophysical Research of South Africa 62, 113–138. 71, 6107–6118. Geochronology of the Australian Cenozoic 911

MCDOUGALL I. & DUNCAN R. A. 1988. Age progressive volcanism in PELL S. D., WILLIAMS I. S. & CHIVAS A. R. 1997. The use of protolith the Tasmantid Seamounts. Earth and Planetary Science Letters zircon-age fingerprints in determining the protosource areas for 89, 207–220. some Australian dune sands. Sedimentary Geology 109, 233–260. MCDOUGALL I. & GILL E. D. 1975. Potassium–argon ages from the PERSANO C., BISHOP P. & STUART F. M. 2006. Apatite (U–Th)/He age Quaternary succession in the Warrnambool–Port Fairy area, constraints on the Mesozoic and Cenozoic evolution of the Victoria, Australia. Proceedings of the Royal Society of Victoria Bathurst region, New South Wales: evidence for antiquity of the 87, 175–178. continental drainage divide along a passive margin. Australian MCDOUGALL I. & HARRISON T. M. 1999. Geochronology and Thermo- Journal of Earth Sciences 53, 1041–1050. 40 39 chronology by the Ar/ Ar Method (2nd edition). Oxford PERSANO C., STUART F. M., BISHOP P. & BARFORD D. N. 2002. Apatite University Press, New York. (U–Th)/He age constraints on the development of the Great MCDOUGALL,I.&ROKSANDIC Z. 1974. Total fusion 40Ar/39Ar ages Escarpment on the southeastern Australian passive margin. using Hifar reactor. Journal of the Geological Society of Australia Earth and Planetary Science Letters 200, 79–90. 21, 81–89. PETKOVIC P. & BUCHANAN C. 2002. Australian Bathymetry and MCELHINNY M. W. & EMBLETON B. J. J. 1974. Australian palaeomag- Topography Grid (Digital Dataset). Geoscience Australia, netism and the Phanerozoic plate tectonics of eastern Gondwa- Canberra. naland. Tectonophysics 22, 1–29. PIDGEON R. T., BRANDER T. & LIPPOLT H. J. 2004. Late Miocene (U– MCELHINNY M. W., EMBLETON B. J. J. & WELLMAN P. 1974. Th)/He ages of ferruginous nodules from lateritic duricrust, A synthesis of Australian Cenozoic palaeomagnetic results. Darling Range, Western Australia. Australian Journal of Earth Geophysical Journal of the Royal Astronomical Society 36, 141– Sciences 51, 901–909. 151. PILLANS B. 1997. Soil development at snail’s pace: evidence from a 6 MCGOLDRICK P. J. & GLEADOW A. J. W. 1978. Fission-track dating of Ma soil chronosequence on basalt in north Queensland, Aus- Lower Palaeozoic sandstones at Tatong, north central Victoria. tralia. Geoderma 80, 117–128. Journal of the Geological Society of Australia 24, 461–464. PILLANS B. 1998. Regolith dating methods. A guide to numerical MCQUEEN K. G., GONZALEZ O. R., ROACH I. C., PILLANS B. J., dating techniques. Cooperative Research Centre for Landscape DUNLAP W. J. & SMITH M. L. 2007. Landscape and regolith Evolution and Mineral Exploration, Perth. features related to Miocene leucitite lava flows El Capitan PILLANS B. 2004. Geochronology of the Australian Regolith. Co- northeast of Cobar, New South Wales. Australian Journal of operative Research Centre for Landscape Environments and Earth Sciences 54, 1–17. Mineral Exploration. Online 5http://crcleme.org.au/RegLan- MEESTERS A. & DUNAI T. J. 2002a. Solving the production-diffusion dEvol/Geochron_of_%20Aust_Regolith.pdf4 equation for finite diffusion domains of various shapes Part 1. PILLANS B. 2005. Geochronology of the Australian regolith. In: Implications for low-temperature (U–Th)/He thermochronology. Anand R. R. & de Broekert P. eds. Regolith–Landscape Chemical Geology 186, 333–344. Evolution Across Australia, pp. 41–61. Cooperative Research MEESTERS A. & DUNAI T. J. 2002b. Solving the production-diffusion Centre for Landscape Environments and Mineral Exploration, equation for finite diffusion domains of various shapes Part 2. Perth. Application to cases with alpha-ejection and nonhomogeneous PILLANS B. & BOURMAN R. 1996. Brunhes/Matuyama polarity distribution of the source. Chemical Geology 186, 57–73. transition (0.78 Ma) as a chronostratigraphic marker in Aus- MERRIHUE C. & TURNER G. 1966. Potassium–argon dating by tralian regolith studies. AGSO Journal of Australian Geology & activation with fast neutrons. Journal of Geophysical Research Geophysics 16, 289–294. 71, 2852–2857. PILLANS B. & BOURMAN R. 2001. Mid Pleistocene arid shift in MITCHELL J. G. 1968. The argon-40/argon-39 method for potassium– southern Australia, dated by magnetostratigraphy. Australian argon age determination. Geochimica et Cosmochimica Acta 32, Journal of Soil Research 39, 89–98. 781–790. PILLANS B., TONUI E. & IDNURM M. 1999. Paleomagnetic dating of MOORE M. E., GLEADOW A. J. W. & LOVERING J. F. 1986. Thermal weathered regolith at Northparkes Mine, NSW. In: Taylor G. & evolution of rifted continental margins: new evidence from Pain C. eds. New Approaches to an Old Continent, 3rd Australian fission tracks in basement apatites from southeastern Australia. Regolith Conference Proceedings, Regolith ’98, pp. 237–242. Co- Earth and Planetary Science Letters 78, 255–270. operative Research Centre for Landscape Evolution and Mineral MORTLOCK A. J. 1977. Progress in thermoluminescence dating. Exploration, Perth. Memorial of the Victorian Archaeological Survey 1, 10–17. POLACH H. A., HEAD M. J. & GOWER J. D. 1978. ANU radiocarbon date MURRAY-WALLACE C. V., BELPERIO A. P. & CANN J. H. 1998. list VI. Radiocarbon 20, 360–385. Quaternary neotectonism and intra-plate volcanism: the Coorong PRACEJUS B. 1989. Nature and formation of supergene manganese to Mt Gambier Coastal Plain, southeastern Australia: a review. deposits on Groote Eylandt, N. T., Australia. PhD thesis. Downloaded By: [University of Queensland] At: 05:12 12 September 2008 In: Stewart I. S. & Vita-Finzi C. eds. Coastal Tectonics, pp. 255–267. University of Adelaide, Adelaide. (unpubl.). Geological Society of London Special Publication, 146. PRIDEAUX G. J., LONG J. A., AYLIFFE L. K., HELLSTROM J. C., PILLANS MUSGRAVE R. 1989. A weighted least-squares fit of the Australian B., BOLES W. E., HUTCHINSON M. N., ROBERTS R. G., CUPPER M. L., apparent polar wander path for the last 100 Myr. Geophysical ARNOLD L. J., DEVINE P. D. & WARBURTON N. M. 2007. An arid- Journal 96, 231–243. adapted middle Pleistocene vertebrate fauna from south-central NOTT J. & PURVIS A. C. 1995. Geomorphic and tectonic significance Australia. Nature 445, 422–425. of Early Cretaceous lavas on the coastal plain, southern PRIDER R. T. 1960. The leucite lamproite of the Fitzroy Basin, New South Wales. Australian Journal of Earth Sciences 42, Western Australia. Journal of the Geological Society of Australia 145–149. 6, 71–118. NOTT J. F., YOUNG R. W. & MCDOUGALL I. 1996. Wearing down, QUIGLEY M. C., SANDIFORD M., CUPPER M. & ALIMANOVIC A. 2004. wearing back, and gorge extension in the long-term denudation Quantifying South Australian landscape evolution. Geological of a highland mass: quantative evidence from the Shoalhaven Society of America Abstracts with Programs 36, 306. catchment, Southeast Australia. Journal of Geology 104, 224– RAMANAIDOU E. R., MORRIS R. C. & HORWITZ R. C. 2003, Channel iron 232. deposits of the Hamersley Province, Western Australia. Austra- OLLIER C. D. 1981. A buried soil at Mt Eccles, western Victoria, and lian Journal of Earth Sciences 50, 669–690. the date of eruption. Victorian Naturalist 98, 195–199. REINERS P. W., SPELL T. L., NICOLESCU S. & ZANETTI K. A. 2004. Zircon OLLIER C. D. 1982. Geomorphology and tectonics of the Dorrigo (U–Th)/He thermochronometry: He diffusion and comparisons Plateau, NSW. Journal of the Geological Society of Australia 29, with 40Ar/39Ar dating. Geochimica et Cosmochimica Acta 68, 431–435. 1857–1887. O’SULLIVAN P. B., GIBSON D. L., KOHN B. P., PILLANS B. & PAIN C. F. RENNE P. R., SWISHER C. C., DEINO A., KARNER D. B., OWENS T. L. & 2000. Long-term landscape evolution of the Northparkes region DE PAOLO D. J. 1998. Intercalibration of standards, absolute ages of the Lachlan fold belt, Australia: constraints from fission track and uncertainties in 40Ar/39Ar dating. Chemical Geology 145, and paleomagnetic data. Journal of Geology 108, 1–16. 117–152. PELL S. D., CHIVAS A. R. & WILLIAMS I. S. 1999. The Great Victoria ROBERTS R. G. & JONES R. 1994. Luminescence dating of sediments: Desert: development and sand provenance. Australian Journal of new light on the human colonization of Australia. Australian Earth Sciences 46, 289–299. Aboriginal Studies 2, 2–17. 912 P. M. Vasconcelos et al.

ROBERTS R. G., JONES R. & SMITH M. A. 1990. Thermoluminescence SINGH G., OPDYKE N. D. & BOWLER J. M. 1981. Late Cainozoic dating of a 50,000-year-old human occupation site in northern stratigraphy, palaeomagnetic chronology and vegetational his- Australia. Nature 345, 153–156. tory from Lake George, NSW. Journal of the Geological Society of ROBERTS R. G., JONES R., SPOONER N. A., HEAD M. J., MURRAY A. S. & Australia 28, 435–452. SMITH M. A. 1994. The human colonization of Australia: optical SIRCOMBE K. N. 1999. Tracing provenance through the isotope ages of dates of 53,000 and 60,000 years bracket human arrival at Deaf littoral and sedimentary detrital zircon, eastern Australia. Adder Gorge, Northern Territory: Quaternary Science Reviews Sedimentary Geology 124, 47–67. 13, 575–583. SIRCOMBE K. N. & FREEMAN M. J. 1999. Provenance of detrital zircons ROBERTSON A. D. 1985. Cainozoic volcanic rocks in the Bundaberg– on the Western Australia coastline—implications for the geolo- Gin–Pialba area, Queensland. Papers of the Department of gic history of the Perth basin and denudation of the Yilgarn Geology, University of Queensland 11, 72–92. craton. Geology 27, 879–882. ROBERTSON A. D., SUTHERLAND F. L. & HOLLIS J. D. 1985. Upper SMITH I. E. M., WHITE A. J. R., CHAPPELL B. W. & EGGLETON R. A. 1988. mantle xenoliths and megacrysts and the origin of the Brigooda Fractionation in a zoned monzonite pluton; Mount Dromedary, basalt and breccia, near Proston, Queensland. Papers of the southeastern Australia. Geological Magazine 125, 273–284. Department of Geology, University of Queensland 11, 58–71. SMITH V. W. & PRESCOTT J. R. 1987. Thermoluminescence dating of ROBERTSON G. B., PRESCOTT J. R. & HUTTON J. T. 1996. Thermo- the eruption at Mt Schank, South Australia. Australian Journal luminescence dating of volcanic activity at Mt Gambier, South of Earth Sciences 34, 335–342. Australia. Transactions of the Royal Society of South Australia SPELL T. L. & MCDOUGALL I. 1992. Revisions to the age of the 120, 7–12. Brunhes–Matuyama boundary and the Pleistocene geomagnetic RUTHERFORD E. 1905. Present problems in radioactivity. Popular polarity timescale. Geophysical Research Letters 19, 1181–1184. Science (May), pp. 1–34. SPELL T. L. & MCDOUGALL I. 2003. Characterization and calibration 40 39 SANDIFORD M. 2007. The tilting continent: a new constraint on the of Ar/ Ar dating standards. Chemical Geology 198, 189–211. dynamic topographic field from Australia. Earth and Planetary SPIKINGS R. A., FOSTER D. A., KOHN B. P. & LISTER G. S. 2002. Post- Science Letters 261, 152–163. orogenic (51500 Ma) thermal history of the Palaeo–Mesoproter- SANDIFORD M. 2008. SRTM Landform atlas. Online 5http://jaeger. ozoic, Mt. Isa Province, NE Australia. Tectonophysics 349, 327–365. earthsci.unimelb.edu.au/Images/Landform/landform.htm4 SPRY M. J., GIBSON D. L. & EGGLETON R. A. 1999. Tertiary evolution of SCHMIDT P. W., CURREY D. T. & OLLIER C. D. 1976. Sub-basaltic the coastal lowlands and the Clyde River palaeovalley in weathering, damsites, paleomagnetism, and the age of lateriti- southeast New South Wales. Australian Journal of Earth zation. Journal of the Geological Society of Australia 23, 367– Sciences 46, 173–180. 370. STEIGER R. H. & JA¨ GER E. 1977. Subcommission on geochronology: SCHMIDT P. W. & EMBLETON B. J. J. 1976. Paleomagnetic results from convention on the use of decay constants in geo- and cosmo- sediments of Perth Basin, Western Australia, and their bearing chemistry. Earth and Planetary Science Letters 36, 359–362. on timing of regional lateritization. Palaeogeography, Palaeocli- STEINBERGER B. & O’CONNELL R. J. 1998. Advection of plumes in matology, Palaeoecology 19, 257–273. mantle flow: implications for hot spot motion, mantle viscosity SCHMIDT P. W. & OLLIER C. D. 1988. Palaeomagnetic dating of Late and plume distribution. Geophysical Journal International 132, Cretaceous to Early Tertiary weathering in New England, NSW, 412–434. Australia. Earth-Science Reviews 25, 363–371. STEPHENSON P. J., GRIFFIN T. J. & SUTHERLAND F. L. 1980. Cainozoic SCHMIDT P. W., TAYLOR G. & WALKER P. H. 1982. Palaeomagnetic volcanism in northeastern Australia. In: Henderson R. A. & dating and stratigraphy of a Cenozoic lake near Cooma, NSW. Stephenson P. J. eds. The Geology and Geophysics of Northeastern Journal of the Geological Society of Australia 29, 49–53. Australia, pp. 349–374. Geological Society of Australia, Queens- SEWARD D. & KORSCH R. J. 1989. Geological Note. Fission track of land Division, Brisbane. alluvial zircons from Inverell sapphire field, New England STEWART A. J., BLAKE D. H., & OLLIER C. D. 1986. Cambrian river Orogen. Australian Journal of Earth Sciences 36, 475–477. terraces and ridgetops in central Australia: oldest persisting SHARP K. R. 1994. Discussion. A core through the Monaro basalt: landforms?Science 233, 758–761. Bega (BMR) No. 7. A reinterpretation and discussion. Australian STIRLING C. H., ESAT T. M., LAMBECK K. & MCCULLOCH M. T. 1998. Journal of Earth Sciences 41, 69–70. Timing and duration of the last interglacial: evidence for a SHARP K. R. 2004. Cenozoic volcanism, tectonism and stream restricted interval of widespread coral reef growth. Earth and derangement in the Snowy Mountains and northern Monaro Planetary Science Letters 160, 745–762. of New South Wales. Australian Journal of Earth Sciences 51, STIRLING C. H., ESAT T. M., MCCULLOCH M. T. & LAMBECK K. 1995. 67–83. High-precision U-series dating of corals from Western Australia SHERWOOD J., OYSTON B. & KERSHAW A. P. 2004. The age and and implications for the timing and duration of the Last Downloaded By: [University of Queensland] At: 05:12 12 September 2008 contemporary environments of Tower Hill volcano, southwest Interglacial. Earth and Planetary Science Letters 135, 115–130. Victoria, Australia. Proceedings of the Royal Society of Victoria STOCK J. & MOLNAR P. 1982. Uncertainties in the relative positions of 116, 71–78. the Australia, Antarctica, Lord Howe, and Pacific plates since the SHORT S. A., LOWSON R. T., ELLIS J. & PRICE D. M. 1989. Thorium– Late Cretaceous. Journal of Geophysical Research 87, 4697–4714. uranium disequilibrium dating of Late Quaternary ferruginous STONE J., PETERSON J. A., FIFIELD L. K. & CRESSWELL R. G. 1997. concretions and rinds. Geochimica et Cosmochimica Acta 53, Cosmogenic chlorine-36 exposure ages for two basalt flows in the 1379–1389. Newer Volcanics Province, western Victoria. Proceedings of the SHULMEISTER J., SHORT S. A., PRICE D. M. & MURRAY A. S. 1993. Royal Society of Victoria 109, 121–131. Pedogenic uranium/thorium and thermoluminescence chronol- STRUTT R. J. 1908. Helium and radioactivity in rare and common ogies and evolutionary history of a coastal dunefield, Groote minerals. Proceedings of the Royal Society of London A80, 572– Eylandt, northern Australia. Geomorphology 8, 47–64. 594. SHUSTER D. L., EHLERS T. A., RUSMORE M. E. & FARLEY K. A. 2005. STRUTT R. J. 1910. On the accumulation of helium in geological time Rapid glacial erosion at 1.8 Ma revealed by 4He/3He thermo- II. Proceedings of the Royal Society of London A83, 96–99. chronometry. Science 310, 1668–1670. SUTHERLAND F. L. 1981. Migration in relation to possible tectonic and 4 3 SHUSTER D. L. & FARLEY K. A. 2004. He/ He thermochronometry. regional controls in eastern Australian volcanism. Journal of Earth and Planetary Science Letters 217, 1–17. Volcanology and Geothermal Research 9, 181–213. SHUSTER D. L. & FARLEY K. A. 2005. Diffusion kinetics of proton- SUTHERLAND F. L. 1993. Late thermal events based on zircon fission induced Ne-21, He-3, and He-4 in quartz. Geochimica et Cosmochi- track ages in northeastern New South Wales and southeastern mica Acta 69, 2349–2359. Queensland: links to Sydney Basin seismicity? Australian SHUSTER D. L., FARLEY K. A., SISTERSON J. M. & BURNETT D. S. 2004. Journal of Earth Sciences 40, 461–470. Quantifying the diffusion kinetics and spatial distributions of SUTHERLAND F. L. 2003. ‘Boomerang’ migratory intraplate Cenozoic radiogenic He-4 in minerals containing proton-induced He-3. volcanism, eastern Australian rift margins and the Indian- Earth and Planetary Science Letters 217, 19–32. Pacific mantle boundary. In: Hillis R. R. & Mu¨ ller R. D. eds. SHUSTER D. L., VASCONCELOS P. M., HEIM J. A. & FARLEY K. A. 2005. Evolution and Dynamics of the Australian Plate, pp. 203–221. Weathering geochronology by (U–Th)/He dating of goethite. Geological Society of Australia Special Publication 22 and Earth and Planetary Science Letters 69, 659–673. Geological Society of America Special Paper 372. Geochronology of the Australian Cenozoic 913

SUTHERLAND F. L., COLCHESTER D. M. & WEBB G. B. 2005. An apparent VASCONCELOS P. M. 1998. Geochronology of weathering in the Mount diatreme source for gem corundums and zircons, Gloucester Isa and Charters Towers regions, northern Queensland. CRC River, New South Wales. Journal and Proceedings of the Royal LEME Report 68R (restricted) (released in 2002 as CRC LEME Society of New South Wales 138, 77–84. Open-file Report 139). SUTHERLAND F. L. & FANNING C. M. 2001. Gem-bearing basaltic VASCONCELOS P. M. 1999a. K–Ar and 40Ar/39Ar geochronology of volcanism, Barrington, New South Wales; Cenozoic evolution, weathering processes. Annual Reviews of Earth and Planetary based on basalt K–Ar ages and zircon fission track and U–Pb Science 27, 183–229. 40 39 isotope dating. Australian Journal of Earth Sciences 48, 221–237. VASCONCELOS P. M. 1999b. Ar/ Ar geochronology of supergene SUTHERLAND F. L., GRAHAM I., POGSON R. E., SCHWARZ D., WEBB G. B., processes in ore deposits. Reviews in Economic Geology 12, 73–113. COENRAADS R. R., FANNING C. M., HOLLIS J. D. & ALLEN T. C. 2002. VASCONCELOS P. M., BECKER T., RENNE P. & BRIMHALL G. 1992. K–Ar The Tumbarumba basaltic gem field, New South Wales: in and Ar/Ar dating of supergene K-bearing sulphates and Mn relation to sapphire–ruby deposits of eastern Australia. Records oxides; paleoclimatic implications from the time-stratigraphic of the Australian Museum 54, 215–248. evolution of deep weathering profiles. In: 29th International SUTHERLAND F. L., POGSON R. E. & HOLLIS J. D. 1993. Growth of the Geological Congress Abstracts, p. 627 central New England basaltic gemfields, New South Wales, based VASCONCELOS P. M., BRIMHALL G. H., BECKER T. A. & RENNE P. R. on zircon fission track dating. In: Flood P. G. & Aitchison J. C. eds. 1994a. 40Ar/39Ar analysis of supergene jarosite and alunite: New England Orogen: Eastern Australia, pp. 483–491. Department implications to the paleoweathering history of the western USA of Geology and Geophysics University of New England, Armidale. and West Africa. Geochimica et Cosmochimica Acta 58, 401–420. SUTHERLAND F. L., STUBBS D. & GREEN D. C. 1977. K–Ar ages of VASCONCELOS P. M. & CONROY M. 2003. Geochronology of weathering Cainozoic volcanic suites, Bowen–St Lawrence Hinter- and landscape evolution, Dugald River Valley, NW Queensland, land, north Queensland (with some implications for petrologic Australia. Geochimica et Cosmochimica Acta 67, 2913–2930. models). Journal of the Geological Society of Australia 24, 447– VASCONCELOS P. M., RENNE P. R., BRIMHALL G. H. & BECKER T. A. 460. 1994b. Direct dating of weathering phenomena by 40Ar/39Ar and TAGAMI T. & O’SULLIVAN P. B. 2005. Fundamentals of fission-track K–Ar analysis of supergene K–Mn oxides. Geochimica et Cosmo- thermochronology. Mineralogy and Geochemistry 58, 19–47. chimica Acta 58, 1635–1665. TARDUNO J. A., DUNCAN R. A., SCHOLL D. W., COTTRELL R. D., VASCONCELOS P. M., WENK H-R. & ECHER C. 1994c. In-situ study of the STEINBERGER B., THORDARSON T., KERR B. C., NEAL C. R., FREY F. thermal behavior of cryptomelane by high-voltage and analytical A., TORII M. & CARVALLO C. 2003. The Emperor Seamounts: electron microscopy. American Mineralogist 79, 80–90. southward motion of the Hawaiian hotspot plume in Earth’s VEEVERS J. J. 1986. Breakup of Australia and Antarctica estimated mantle. Science 301, 1064–1069. as mid-Cretaceous (95 + 5Ma) from magnetic and seismic data at TAUXE L. 2005. Lectures in Paleomagnetism. 5Online http://earth the continental margin. Earth and Planetary Science Letters 71, ref.org/MAGIC/books/Tauxe/2005/4 91–99. TAYLOR G., TAYLOR G. R., BINK M., FOUDOULIS C., GORDON I., VOGT P. R. & CONOLLY J. R. 1971. Tasmantid Guyots, the age of the HEDSTROM J., MINELLO J. & WHIPPY F. 1985. Pre-basaltic Tasman Basin. and motion between the Australia Plate and the topography of the northern Monaro and its implications. mantle. Geological Society of America Bulletin 82, 2577–2584. Australian Journal of Earth Sciences 32, 65–71. WALTENBERG K. 2006. 40Ar/39Ar geochronology of lava flows at TAYLOR G., TRUSWELL E. M., MCQUEEN K. G. & BROWN M. C. 1990. Carnarvon Gorge, Buckland Volcano. BSc (Hons) thesis. Uni- Early Tertiary palaeogeography, landform evolution, and pa- versity of Queensland, Brisbane (unpubl.). laeoclimates of the Southern Monaro, NSW., Australia. Palaeo- WEISSEL J. K. & HAYES D. E. 1971. Asymmetric seafloor spreading geography, Palaeoclimatology, Palaeoecology 78, 109–134. south of Australia. Nature 231, 518–522. TEDFORD R. H., BANKS M. R., KEMP N. R., MCDOUGALL I. & WEISSEL J. K., HAYES D. E. & HERRON E. M. 1977. Plate tectonic SUTHERLAND F. L. 1975. Recognition of the oldest known fossil synthesis: the displacements between Australia, New Zealand, marsupials from Australia. Nature 255, 141–142. and Antarctica since the Late Cretaceous. Marine Geology 25, THORNE A., GRU¨ N R., MORTIMER G., SPOONER N. A., SIMPSON J. J., 231–277. MCCULLOCH M., TAYLOR L. & CURNOE D. 1999. Australia’s earliest WEISSEL J. K. & SEIDL M. A. 1997. Influence of rock strength human remains: age of the Lake Mungo 3 skeleton. Journal of properties on escarpment retreat across passive continental Human Evolution 36, 591–612. margins. Geology 25, 631–634. TORGERSEN T., HABERMEHL M. A., PHILLIPS F. M., ELMORE D., KUBIK WEISSEL J. K. & SEIDL M. 1998. Inland propagation of erosional P., JONES B. G., HEMMICK T. & GOVE H. E. 1991. Chlorine 36 dating escarpments and river profile evolution across the southeast of very old groundwater: 3, Further studies in the Great Artesian Australian passive continental margin. In: Tinkler K. J. & Basin, Australia. Water Resources Research 27, 3201–3213. Wohl E. E. eds. Rivers Over Rock: Fluvial Processes in Bedrock Downloaded By: [University of Queensland] At: 05:12 12 September 2008 TRUMBORE S. 2000. Age of soil organic matter and soil respiration: Channels, pp. 189–206. American Geophysical Union Geophysical radiocarbon constraints on belowground C dynamics. Ecological Monograph, 107. Applications 10, 399–411. WELLMAN P. 1971. The age and palaeomagnetism of the Australian TURNEY C. S. M., BIRD M. I., FIFIELD L. K., ROBERTS R. G., SMITH M., Cenozoic volcanic rocks. PhD thesis. Australian National Uni- DORTCH C. E., GRU¨ N R., LAWSON E., AYLIFFE L. K. & MILLER G. H. versity, Canberra (unpubl.). 2001. Early human occupation at Devil’s Lair, southwestern WELLMAN P. 1973. Early Miocene potassium–argon age for the Australia 50,000 years ago. Quaternary Research 55, 3–13. Fitzroy lamproites of Western Australia. Journal of the Geologi- TWIDALE C. R. 1993. The research frontier and beyond: granitic cal Society of Australia 19, 471–474. terrains. Geomorphology 7, 187–223. WELLMAN P. 1974. Potassium–argon ages on the Cainozoic volcanic TWIDALE C. R. 1994. Gondwanan (Late Jurassic and Cretaceous) rocks of eastern Victoria, Australia. Journal of the Geological palaeosurfaces of the Australian Craton. Palaeogeography, Society of Australia 21, 259–376. Palaeoclimatology, Palaeoecology 112, 157–186. WELLMAN P. 1979. On the Cainozoic uplift of the southeastern TWIDALE C. R. & BOURNE J. A. 1998. The use of duricrusts and Australian highland. Journal of the Geological Society of topographic relationships in geomorphological correlation: con- Australia 26, 1–9. clusions based in Australian experience. Catena 33, 105–122. WELLMAN P. 1987. Eastern Highlands of Australia: their uplift and VAN DER PLICHT J., BECK J. W., BARD E., BAILLIE M. G. L., erosion. BMR Journal of Australian Geology & Geophysics 10, BLACKWELL P. G., BUCK C. E., FRIEDRICH M., GUILDERSON T. P., 277–286. HUGHEN K. A., KROMER B., MC CORMAC F. G., BRONKRAMSEY C., WELLMAN P., CUNDARI A. & MCDOUGALL I. 1970. Potassium–argon REIMER P. J., REIMER R. W., REMMELE S., RICHARDS D. A., ages for leucite-bearing rocks from New South Wales, Australia. SOUTHON J. R., STUIVER M. & WEYHENMEYER C. E. 2004. NOT Journal and Proceedings of the Royal Society of New South Wales CAL04—Comparison/Calibration 14C records 26–50 Cal Kyr BP. 103, 103–107. Radiocarbon 46, 1225–1238. WELLMAN P. & MCDOUGALL I. 1974a. Cainozoic igneous activity in VARENTSOV I. M. & GOLOVIN D. I. 1987. The Groote Eylandt eastern Australia. Tectonophysics 23, 49–65. manganese deposit, northern Australia: potassium–argon age WELLMAN P. & MCDOUGALL I. 1974b. Potassium–argon ages on the of its cryptomelane ore and some aspects of its genesis. Cainozoic volcanic rocks of New South Wales. Journal of the Transactions of the USSR, Earth Sciences Section 294, 149–153. Geological Society of Australia 21, 247–272. 914 P. M. Vasconcelos et al.

WELLMAN P., MCELHINNY M. W. & MCDOUGALL I. 1969. On the polar- YIM W. S., GLEADOW A. J. W. & VAN MOORT J. C. 1985. Fission track wander path for Australia during the Cenozoic. Geophysical dating of alluvial zircons and heavy mineral provenance in Journal of the Royal Astronomical Society 18, 371–395. north-east Tasmania. Journal of the Geological Society of London WHITE M. E. 1994. After the Greening: The Browning of Australia. 142, 351–356. Kangaroo Press, Kenthurst. YOUNG R. W. & MCDOUGALL I. 1993. Long-term landscape evolution: WHITEHEAD P. W., STEPHENSON P. J., MCDOUGALL I., HOPKINS M. S., Early Miocene and modern rivers in Southern New South Wales, GRAHAM A. W., COLLERSON K. D. & JOHNSON D. P. 2007. Temporal Australia. Journal of Geology 101, 35–39. development of the Atherton Basalt Province, north Queensland. ZHAO J., XIA Q. & COLLERSON K. D. 2001. Timing and duration of the Australian Journal of Earth Sciences 54, 691–709. Last Interglacial inferred from high resolution U-series chron- WILKINSON M. T., CHAPPELL J., HUMPHREYS G. S., FIFIELD K., SMITH B. ology of stalagmite growth in Southern Hemisphere. Earth and &HESSE P. 2005. Soil production in heath and forest, Blue Planetary Science Letters 184, 635–644. Mountains, Australia: influence of lithology and palaeoclimate. ZEITLER P. K., HERCZEG A. L., MCDOUGALL I. & HONDA M. 1987. U–Th– Earth Surface Processes and Landforms 30, 923–934. He dating of apatite—a potential thermochronometer. Geochimi- WOLF R. A., FARLEY K. A. & SILVER L. T. 1996. Helium diffusion and ca et Cosmochimica Acta 51, 2865–2868. low-temperature thermochronology of apatite. Geochimica et ZHENG H., WYRWOLL K., LI Z. & POWELL C. MCA. 1998. Onset of Cosmochimica Acta 60, 4231–4240. aridity in southern Western Australia—a preliminary WOODHEAD J., HELLSTROM J., MAAS R., DRYSDALE R., ZANCHETTA G., palaeomagnetic appraisal. Global and Planetary Change 18, DEVINE P. & TAYLOR E. 2006. U–Pb geochronology of speleothems 175–187. by MC-ICPMS. Quaternary Geochronology 1, 208–221. ZHISHENG A., BOWLER J. M., OPDYKE N. D., MACUMBER P. G. & WROE S. & FIELD J. 2007. A review of the evidence for a human role FIRMAN J. B. 1986. Palaeomagnetic stratigraphy of Lake Bun- in the extinction of Australian megafauna and an alternative gunnia: Plio-Pleistocene precursor of aridity in the Murray interpretation. Quaternary Science Reviews 25, 2692–2703. Basin, southeastern Australia. Palaeogeography, Palaeoclimatol- WYATT D. H. & WEBB A. W. 1970. Potassium–argon ages of some ogy, Palaeoecology 54, 219–239. northern Queensland basalts and an interpretation of Late Cainozoic history. Journal of the Geological Society of Australia 17, 39–51. XIA Q., ZHAO J. & COLLERSON K. D. 2001. Early–Mid Holocene climatic variations in Tasmania, Australia: multi-proxy records in a stalagmite from Lynds Cave. Earth and Planetary Science Letters 194, 177–187. Received 16 June 2007; accepted 28 February 2008

SUPPLEMENTARY PAPERS APPENDIX 1: GEOCHRONOLOGICAL RESULTS FOR CENOZOIC VOLCANISM IN AUSTRALIA

APPENDIX 2: GEOCHRONOLOGICAL RESULTS FOR CENOZOIC WEATHERING IN AUSTRALIA Downloaded By: [University of Queensland] At: 05:12 12 September 2008