Papers and Proceedings of the Royal Society of , Volume 129, 1995 53

WEATHERING AND MORPHOSTRATIGRAPHIC EVIDENCE FOR FOUR GLACIATIONS IN THE HENTY RIVER BASIN, WESTERN TASMANIA by Paul C. Augustinus and Eric A. Calhoun

(with two tables and three rext-figures)

AucGSTIMJS, P.C. & Cor.HOUN, E.A., 1995 (30:vi): Weathering and morphostratigraphic evidence for four glaciations in the Henry River Basin, western Tasmania. Pap. Proc. R. Soc. Tasm. 129: 53-58. https://doi.org/10.26749/rstpp.129.53 ISSN 0080-4703. Department of Physical Sciences, University of Tasmania, PO Box 1214, Launceston, Tasmania, 7250 (PCA); Departmem of Geography, University of Newcastle, Callaghan, NSW, Australia 2'.308 (EAC).

The Quaternary glacial histories of the Picman and Hemy River Basins have largely been developed by using post-depositional weathering criteria to identify and map the drift sheets. Application of relative dating techniques, such as percentage absorption and specific gravity of Cambrian volcanic clasts, forces a reinrerprctation of palaeomagnetic evidence previously offered for the age of the Henty Drift, so that at least three pre-Last Glacial Maximum (LGM) drift sheets can be recognised. The deeply weathered and most extensive (but with limited exposure) Berry Creek Drift is possibly a correlative of the Bulgobac/Linda Drift and can be clearly distinguished from the slightly less extensive but more widely exposed Henty Drift. A minimally weathered and spatially restricted drift, termed the Julia Drift, is exposed beyond and pre-dates the LGM end moraines. Keywords: glaciation, Henty Basin, relative dating, correlation, driftsheets, Tasmania.

INTRODUCTION Correlations between the sequences from each of the basins have necessarily been made on the basis of post­ The Late Cenozoic glacial history of western Tasmania has depositional weathering criteria, such as thickness of been reviewed in a number of studies (Colhoun 1985a, weathering rinds developed on clasts of Jurassic dolerite, Colhoun & Fitzsimons 1990, Kiernan 1990a, Fitzsimons et and relative degrees of till matrix alteration and soil al. 1993). Although the essential elements of glacial histories development. Use of these relative dating (RD) techniques of each major drainage basin have been elucidated (Pieman enabled the discrimination of the sediments deposited by Basin - Augusrinus & Colhoun 1986, Augustinus et al, differentglacial events. However, use of RD criteria (such 1994; Henry Basin - Colhoun, 1985a,b; the King Basin - weathering rind thickness variations) for correlation of Fitzsimons & Colhoun 1991, Fitzsimons et al. 1993), the glacial sequences between even adjacent drainage basins is timing of the identified glacial events from each basin is hampered by the general absence of dolerite in the glacial poorly constrained, due to the general absence of datable sediments from some valleys. In addition, climatic and material. As a consequence of this uncertainty, correlations topographic drainage differencespertaining between basins between glacial sequences, even from adjacent valleys are influencethe rate of weathering and, hence, the validity of problematical. correlations made on the basis of such criteria (Kiernan 19906). .. This paper assesses the use of several post-depositional 145' 30' Mt Murc I weathering criteria in mapping the glacial drifts of the '- ���� Mt Read '- Moxon ',,�, ( Henty River Basin (fig. 1) and attempts to correlate the 2.5 km .. \Saddle ',, ° drift sheets defined with those mapped in the adjacent Study� 40 s Area Pieman Basin. Since Jurassic dolerite is absent from most er0 14S E of the Henty Basin Drifts,alternative RD criteria had to be used, including percentage absorption and specific gravity variations in Cambrian volcanic clasts from glacial diamictons. We combine the RD information with palaeo­ magnetic evidence to construct a correlation of the glacial sequences between the basins that is more reliable than any hitherto.

Huntley GLACIAL HISTORY OF THE HENTY RIVER BASIN

The glacial history of the Henry Basin was first studied by Bowden (1974), and glacial landforms were mapped as the product of one glaciation. The firstevidence that more than one glaciation had occurred in the Henry Basin came from 14C dates on lacustrine organic clays overlying weathered till at Henty Bridge (fig. 1) (Colhoun 19856). Dates of23 640 FIG. I-Location map, Henty River Basin ± 1030 yrs (GaK 5597) to >34 600 yrs BP (GaK 5595) 54 PC and EA. Colhoun confirmed that the USED IN STUDY

The of

discovered interglacial at indicated a mixed wer forest at the site llleans that The site IS located 1 km so successful in the mapping of dsewhere in western Tasmania (i.e. Kiernan 1983, Kiernan 1989) was not here. The degree of of Cam brian volcanic clasts from the Pieman Basin was & Colhoun (1')86) to discriminate bet\ivecn drift of different relative ages 'AliIh d Llnge of other C:an1LJrjan volcdlilcs

I-Ience l vve of Cam brian volcanic clast alre;'ation identification of drift sheets of Basin, and from both the of

date

removed. The variations in %Ab and SC reflect development, and indicate the host drifr 198 Where possible, clast samples Selina were taken from a uniform 1 m in the drift exposures. The than the from last interstadial to the commencement of of clast alteration deterioration and onset of the LGM has been inferred at depth probably 25 000 YIS SF at Tullabardine & van de Geer influence on the One to two 1 with the end of the LGM at Lake Selina dated at of 4-8 crn di3meter Cambrian volcanic BP variations between the

criteria used for the include till matrix did not penetrate west of the West in the d.rift, dolerite Coast Range, unlike the older and far more extensive Boco rind thicknesses appropriate) and tile Boco 2 Glaciation & Calhoun moraine morphology,. The derails of these techniques were , Bobaciil (formerly the Boco 1 Glaci,uion of in Augustinus & Colhoun (1 and are used here & Calhoun 1986) and Bulgobac ill the .to drift: separation based on observations Pieman Basin & C:olhouu 1 Augustinus of degrees of drift m od.ification. et ai. 1994). Sharp-crested LGM end moraines are found below Mrs Read and Murchison and at the southern end of the '1' Plateau (the Hamilton End Moraine) (fig. RESULTS AND DISCUSSION

CV'UC=l1LC {'or an early Last Glacial advance, such as that the King Valley (Chamouni Glaciarion) by Observations on the relative degree of weathering of the Fitzsimons & Colhoun (1991), has been recognised from drifts fi'om the Henty River Basin suggest the development the Henty Basin. of at least three distinct drift sheets of differing relative age beyond the LGM limits. The most useful methods for separating the drift sheets were quantitative evaluation of state of clast alteration, degree of iron panning and clast Fe oxide topographic position and degree of modification morphology (table 1). in the River Basin 55

TABLE 1 Late Cenozoic glacial stratigraphy, Henty River Basin, western Tasmania

-~-.----.---.. ---- of events isation acid rind thickness Fe oxide front max. (mm)

18.8 none Interstadial=; Julia Drift minor Fe <2 Langdon >41.9 7 Drift >34.G~ 15 2.10- 2,A-') thin Fe: 2. ·4 >783 Creek

Drift reversed" 3nu l4.4-20.8 1.95--2.05 thick not >4

" ~ dolerite. :t Colhoun et al. (i § Calhoun et al. ~ Calhoun ( ( 199

Creek and Creek contains up to 40% dolerite clasts with rims of up to 3 em but

ary, and no morame Calhoun (1 the Glaciation.

Grid Reference 5SG CP 80161 limit ofthe LGM moraines on the and the >33 950 yrs BP date from 145" 30' at. 1 indicates that the N fVl1 Read

is the most 2.5 km drift sheets -~ the 1 erate, with the former degree of clast alteration Drift. No dolerite was

ll"-l1LtHJ.C: is COllllnOfl, vilth on volcanic clasts 1). reddish brown a continuous cover is maintained, without clear end mor::lines. This drift sheet is best displayed at the and at the Darn Road- intersection and is considered to represent thc Henty described Calhoun (l985a, b), Con- sequently, the name Henty is maintained for the drift at 42"S this site, and Henty remains the type site. One kilometre north Henty Bndge, on the , the drift exposures moderate degrees of ~ Giacicr flow direction Margaret CHacintion limits weatberingway to a deeply altered diamicton containing 0, Organic site -- Julia Creek Glaciation totally decomposed Cambrian volcanic clasts of up to Henly Glaciation CirquefneadwaU 0.25 m diameter. The till matrix is weathered to bright """""""~,, Berry Creole Glaciation yellowish brown (1 and strong Fe oxide panning 1S (table The depth to the weathering front in this drift. The drift exposure at Tom HG. 2-Glacial RiIJer Basin 56 (md E.A. Colhoun

here t:he Berry Creek afrel' its type sirc north of Creek 011 the Zeehan 3635 -- 1 :25 000; UGR SSG (1985) altered clasrs in rill at Professor Creek (fig. and suggested (hat the at this sile pre-d ate the till at Henty with the Ltnda from the King similar gross stdte of clast lienee. the altered tiU from [his site probably correlates with the Berry Creek Drift on 'the . Zeehan and further supports the subdivision the Hentv least two glacial drift sheets of .

SUPPORT DRIFT SHEET USING PERCENTAGE D SPECIFIC GRAVITY 30 • Borry Creek Drift

Augustinus & Colhoun (1 %Ab SG El Li!lda Drill c .. values from Cambrian volcanic in the 0 20 Basin as a scatter graph, and obtained a line e-o relationship with three identifiable populations. These group­ (fJ .n ings supported other post-depositional weathering evidence « 10 for at least three distinct drift sheets (Augusrinus & Colhoun eR- 1986). The %Ab and SG data showed that the penultimate Boco Glaciation Drift could be discriminated clearly from 0 the much more extensively altered Bulgobac Drift. The 1.6 1.8 2.0 2.2 2.4 2.6 2.8 latter displays reversed detrital magnetisation and is hence > 783 kyrs old (Pollington 1991, Baksi etal. 1992), although B Specific Gravity the overlap between the SG and % Ab results from the Bobadil Drift (formerly termed Boco I in Augustinus & FIG. 3 - Plots Colhoun 1986) and the Blligobac and Boco Drifts suggests volcanic clasts: that the method will only allow the confident discrimination sites; (B) clast values from the of drifts separated by a long weathering interval equal to at delineating fields representing range least one glacial-interglacial cycle. The discrimination power absorption valuesfor the Pieman Basin indicated: solid of the method declines with variability in degree of clast lines = Bulgobac Drift, dashed lines = Bow Drift. alteration between sites. Figure 3A shows the %Ab versus SG plot for all sites from the Pieman Basin, whilst figure 3b shows both the Henty and Pieman River Basin drift sheet data. Examination of figure 3B indicates that the Henty Drift data correlate DATING AND INTER-BASIN CORRELATION well with the Bulgobac and Bobadil Drift results from the OF THE DRIFT SHEETS Pieman Basin, with the Boca sites generally having significantly lower %Ab and higher SG values~ This suggests Measurement of SG and %Ab of volcanic clasts from the a greater relative age for the widespread Henry Drift exposed glacial drift sheets in the Pieman and Henty River Basins along the Anthony Dam Road than for the Boco Drift supports the mapping of several distinct drift sheets using from the Pieman Basin. several other post-depositional weathering criteria. In the The Berry Creek Drift displays a very high degree of Pieman Basin, clear differentiation between the Boco and weathering and clast alteration with many of the Cambrian Bulgobac Drifts is possible, although overlap with both volcanic clasts of up to 0.25 m diameter weathered to occurs when volcanic clasts from the Bobadil Drift are ghosts. Volcanic clasts selected from six sites in this deeply examined. The two extensive glacial drifts from the Henty altered diamicton were examined; their %Ab and SG values Basin can be more confidently mapped on the basis of are plotted in figure 3b. At all but one site, the clasts dis-­ differences in %Ab and SG and assigned to drift sheets of playa significantly greater degree of weathering than than different rdati ve age. do those from the Henty Drift, and the degree of alteration The most extensive and deeply weathered drift from the is comparable ro that of the Bulgobac Drift from the Henry Basin displays a degree of clast alteration comparable Pieman Basin. This is supported by the degree of weather­ with that of the Bulgobac Drift from the Pieman Basin, ing of dolerite clasts from these sites. Dolerite clasts comprise although there is significant overlap between the groupings, up to 40% of the ice-proximal coarse gravei exposed as would be expected given the variability in clast immediately south of Tom Creek (fig. 2) and weathering mineralogy, topography and microclimate at each site. rinds up to 30 mm thick were measured, although most of Hence, the older, deeply weathered drift from the Henry the clasts have been disturbed so that the "true" dolerite Basin (Berry Creek Drift) is tentatively correlated with the weathering rind thicknesses have probably not been Bulgobac Drift from the Pieman Basin (table 2). Dolerite measured. is common in the Berry Creek Drift but not the Henry Evidence for four glaciations in the Henty River Basin 57

Drift: it was probably derived from supraglacial debris The Henty Drift has the most widespread exposure of added to the lateral margins of the extensive Berry Creek any of the Henty Basin drifts and displays clast alteration piedmont glacier when the western margin abutted the values comparable with those of the Bobadil Drift from the dolerite-capped Mt Dundas. Pieman Basin; a tentative basin-wide correlation between Presence of reversed detrital remanent magnetisation the two drifts is suggested here (table 2). This correlation (DRM) in Henty Drift glaciolacustrine sediments from is supported by an amino-acid assay on wood from the Tyndall Creek suggested to Colhoun & Fitzsimons (1990) Langdon River organic site, which is developed on the that the Henty Glaciation may pre-date the Brunhes­ floor of a meltwater channel lined with laminated silts and Matuyama boundary. However, re-examination of the site clays associated with the Henty Drift. The amino-acid by the present authors indicates that the rhythmites contain assay suggests Isotope Stage 7 age for the organic site (B.]. ice-push structures and display clear evidence that the basal Pillans, pers. comm. 1988) and at least Stage 8 for the rhythmites, from which the reversed DRM results were Henty Glaciation. The Bobadil Glaciation has been dated obtained, were overridden. The basal diamicton displays a to Isotope Stage 8 on the basis of bracketing U/Th dates of degree of clast alteration comparable to that of the Berry 166 +41, -33 kyrs (LH 2460) and 279 +100, -65 kyrs (LH Creek Drift, whilst the upper diamicton displays a 2461)(Augustinus eta!' 1994). If the correlations between significantly lower degree of clast alteration and is identical basins based on post-depositional weathering criteria are to the Henty Drift mapped elsewhere on the Henty Surface. correct, the Henty Glaciation can be more securely assigned Hence, the site is interpreted as representing Berry Creek to Isotope Stage 8 than to Stage 6, as suggested previously Drift (including glaciolacustrine rhythmite) with an age (Colhoun 1985a,b). > 783 kyrs, which was later overridden by the Henty Glacier, The Bobadil Drift from the Pieman Basin displays normal resulting in deformation of the older rhythmites. This DRM and is <783 kyrs (Pollington 1991, Augustinus et al. phase terminated with the deposition of indurated basal till in press). U/Th dating of ferricrete bands associated with of Henty Glaciation affinity. No rhythmite has been found Bobadil Glaciation sediments from the Boco Plain indicates associated with the Henty Drift in the Henty Valley. a probable Isotope Stage 8 age for the glacial event

TABLE 2 Late Cenozoic glaciations of Pieman and Henty River Basins

Series Pieman Basin* Henry Basin subdivision

Margaret Glaciation Margaret Glaciation Late max. - 19 kyrs t Pleistocene Interstadial 25-44 kyrs :j: Interstadial >34 kyrs §

Oulia Ck Glaciation ?)

Pieman Interglacial amino acid Isotope Stage 5 #

Boco Glaciation Oulia Ck Glaciation ?) >78, <166 kyrs

Langdon Interglacial Isotope Stage 7 #

Middle Bobadil Glaciation Henty Glaciation Pleistocene Core AKI >13 m >34.6 kyrs t >166, <279 kyrs

Animal Ck Glaciation Core AKI >41 m >279 kyrs

B/M Boundary 783 kyrs BulgobaclQue Berry Ck Glaciation Early Glaciations reversed Pleistocene reversed magnetisation f magnetisation f

* Augustinus et al. (1994). t Colhoun (1985a). :j: Colhoun & van de Geer (1986). § Colhoun etaL (1993). ~ Colhoun & Fitzsimons (1990). # Colhoun etaL (1988). f Pollington (1991). AKI refers to Pasminco drill core from Boco Plain, with depths to U/Th dated horizons. 58 P. C. Augustinus and E.A. Calhoun

(Augustinus et al. 1994). The Robadil Drift is clearly of AUG\TSTINUS, P.e. & COLllOCN, LA., 1986. Glacial history of middle Pleistocene age, and correlation between the Robadi! the upper Pieman and Boco western Tasmania. and Henry Drifts seems reasonable on the basis of the Aust. j. Earth Sci. 33: 181·19l. amino-acid assays and clast weathering data from the Henry AUGUSTINUS, P.e., SHORT, S.A. & COIHOC:-';, F..A., 1994: Pleisto­ cene stratigraphy of the Boco Plain, western Tasmania. Basin (table 2). However, it is not possible to exclude Au]t. j. Earth Sci. 41: 581-591. correlation of [he Henty Drift with the younger Boco Drift AUGUSTINUS, P.e., POLl.JNCTON, M.J. & COI.HOlJN, E.A., in (Isotope Stage 6) from the Pieman Basin without better press: Magnetostratigraphy of the Lare Cenozoic glacial dating of the Henty Basin sequences. sequence, Pie man River basin, western Tasmania. A us!. The Julia Drift extends beyond the LGM end moraines j. Earth Sci. and displays a distinctly lower degree of post-depositional BAKSI, A.K., HSlJ, V., MCWlI.1.IAMS, M.O. & FARRAR, E., 1992: alteration than the much more extensive and more 40Ar/39 Ar dating of the Brunhes-Matuyama geomagnetic weathered Henty Drift (table 1). Dating of this drift is nor field reversal. Geology 256: possible at this time. It could represent either an early Last BOWDEN, A.R., 1974: The glacial geomorphology of the Tyndall Glacial ice advance, similar to that postulated for the Mountains, western Tasmania. U npubl. BSc fIons Thesis, Department of of T,,,;mania. Chamouni Formation from the King Valley (Fitzsimons & Co I.H01.JN , F..A., 1979: EXCURSION TO Calhoun 1991), or it could belong to Isotope Stage 6. In NOR IHWE<; l' AND WEST TASMANIA. Department the absence of firm dating of these drifts, it is not possible of Geography, of Tasmania, Hoban. to determine the age with confidence. COLH01.JN, E.A., 1985a: Glaciations of the , Tasmania. Res. 24: 39-59. COIHOGN, E.A., 1985b: Pre-Last Glacial Maximum vegetation CONCLUSIONS history at Henty Bridge. New Phytol. 100: 681-690. COLHOUN, E.A., 1992: Late Glacial and Holocene vegetation Re-evaluation of the glacial stratigraphy of the Henty Basin history at Poets Hill Lake, western Tasmania. Aust. Geogr. has enabled the mapping of at least three drift sheets of pre­ 23: 1123. COLH01.JN, E.A. & FITZSI!vIONS, S.J., ·1990: Late Cainozoic Last Glacial Maximum age, using relative degree of clast glaciation in western Tasmania, Australia. Quatern. Sci. alteration, Fe panning, degradation of the moraine morph­ Rev. 9: 199-216. ology and topographic position. The Berry Creek Drift COU-101.JN, E.A. & VAN DE GEER, G., 1986: Holocene to Middle represents the oldest and most extensive glaciation identified Last Glaciation vegetation history at Tullabardine Dam, in the basin, although its distribution is fragmentary, as is western Tasmania. Proc. R. Soc. London B229: 177-·207. the case with the similarly extensive Bulgobac Glaciation COLHOCN, E.A., VAN DE GEER, G., HILL, R.S. & BIRD, T., 1988: from the Pieman Basin. Interglacial pollen and plant macrofossils from the The Henty Drift is much more widespread, is tentatively Langdon River, western Tasmania. New Phytol. 111: correla.ted with the Bobadil Drift from the Pieman Basin 531-548. and is dated to the middle Pleistocene. The Henty Drift COLHOUN, E.A., BENCER, S.N., FITZSIMONS,S.]., VAN DE GEER, G. & HILl., RS., 1993: Quaternary organic deposit from pre-dates Isotope Stage 7, and, presuming that the Newton Creek Valley, western Tasmania. Aust. Geogr. weathering evidence for correlation between the Henty Stud. 31: 26-38. and Bobadil Drifts is valid, an Isotope Stage 8 age can be COLI-101.JN E.A. & FITZSIMONS, S.]., 1990: Late Cainozoic assigned to both drifts. The Julia Drift is restricted in glaciation in western Tasmania. Quatern. Sci. Rev. 9: extent, displays minimal clast and matrix alteration and is 199-216. located well beyond the distinct LGM (Margaret Glaciation) FITZSIMONS,S.]. & COLHOlJN, E.A., 1991: Pleistocene Glaciation end moraines below Mt Read, Mt Murchison and the of the King Valley, western Tasmania, Australia. Quat. Tyndall Plateau. Dating of this drift is problematical, since Res. 36: 135~156. weathering evidence suggests an age younger than the Henty FrrZSI!vIONS, S.J" COLHOUN, E.A., VAN DE GEFR, G. & POLUNGTON, Drift, but no datable material has been found to test this M.J., 1993: The Quaternary geology and glaciation of the King Valley. Bul!. Geo!. Surv. Tasm. 68. interpretation. HAMMOND, A., 1985: Soils and glacial geomorphology of the Although the measurement of clast SG and %Ab does Central West Coast Range, Tasmania. Unpub!. BSc Hons not permit discrimination between drifts, unless one has thesis, Geography Department, University of Tasmania. been exposed to prolonged periods of weathering equivalent KIERNAN, K. W., 1983: Weathering evidence for an additional to at least the duration of one interglacial episode, it does glacial stage in Tasmania. Aust. Geogr. Stud. 21: 197-·220. allow clear separation of the middle Pleistocene Hentyl K1ERNAN, K.W., 1989: Multiple glaciation of the upper Franklin Bobadi! from the early Pleistocene Berry Creek/Bulgobac Valley, western Tasmania Wilderness World Heritage Drifts. Although the drifts from the Henry Basin are mostly Area. Aust. Geogr. Stud. 27: 208-233. devoid of dolerite clasts, it has been shown that these KIERNAN, K.W., 199(la: The extent of Late Cenozoic glaciation methods, when used with at least one other relative-dating in the Central Highlands of Tasmania, Australia. Arctic & Alpine Res. 22: 341-354. test, provide the best means of differentiation and correlation KIERNAN, KW., 1990b: Post-depositional modification as an aid between the drift sheets. to dating Tasmanian glacial deposits. Aust. Geogr. 21: 1- 17. MCGREGOR, G.R., 1981: Weathering characteristics of Late REFERENCES Pleistocene tills. NZ/Geol. Geophys. 24: 107-113. POLl.lNGTON, M.J., 1991: Magnetostratigraphy of glacial lake ANKER, S.A., 1991: Pollen and vegetation history from Lake sediments and dating of Pleistocene glacial deposits in Johnson in western Tasmania. Unpub!. BSc Hons thesis, Tasmania. Unpub!. MSc thesis, University of Tasmania, Department of Geography, University of Newcastle. Hobart. (accepted 13 December 1994)