Geomorphic evidence of major sea-level fluctuations during marine isotope substage-5e, Cape Cuvier, Western Australia

Geomorphology 102 (2008) 595–602

Contents lists available at ScienceDirect

Geomorphology j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / g e o m o r p h

Geomorphic evidence of major sea-level fluctuations during marine isotope substage-5e, Cape Cuvier, Western Australia M.J. O'Leary a,b,⁎, P.J. Hearty c, M.T. McCulloch d a

School of Earth and Environmental Science, James Cook University, Townsville QLD 4811 Australia Department of Environmental and Geographical Sciences, Manchester Metropolitan University, Manchester, M1 5GD, United Kingdom GeoQuest Research Centre and School of Earth and Environmental Sciences, University of Wollongong, Wollongong, NSW 2522, Australia d Research School of Earth Sciences, Australian National University, Canberra, ACT, 0200, Australia b c

A R T I C L E

I N F O

Article history: Received 31 December 2007 Received in revised form 30 May 2008 Accepted 2 June 2008 Available online 12 June 2008 Keywords: Sea-level changes Coral reefs Marine terraces U-series MIS 5e Cape Cuvier

A B S T R A C T A detailed geomorphologic and morphostratigraphic investigation of raised marine terraces at Cape Cuvier, Western Australia, reveals two morphologically distinct units. A lower, well-developed accretional reef terrace between 3 and 5.5 m above MLWS (mean low-water springs; hereafter denoted as “+”) represents an extended interval of stable sea level. An upper erosional terrace and incipient coralgal rim between +8.5 to 10.5 m represents a brief sea-level stillstand at this higher elevation. These features suggest the lower and upper terraces developed during discrete sea-level events. In an attempt to better define the timing of emplacement of each marine unit, 20 coral samples collected along vertical and lateral reef growth axis from both terraces were analysed with U-series dating. Unfortunately, all coral samples exhibited elevated δ234Uinitial values, suggesting that pervasive uptake of 234U-enriched uranium and 230Th thorium had occurred. Despite the shortcomings of absolute dating, a succession of events can be resolved though morphostratigraphic relationships. Comparison of the facies relationships, coral growth, and morphostratigraphic features between the lower and upper terraces indicates that an early to mid MIS 5e stillstand at + 3 to 5 m was followed by a late rise to +8.5 to 10.5 m. This agrees with an emerging global view of MIS 5e sealevel history derived from stable carbonate platforms, rejecting the hypothesis that these higher sea-level benchmarks are an artefact of localized tectonic processes. Crown Copyright © 2008 Published by Elsevier B.V. All rights reserved.

1. Introduction Cape Cuvier (24° 13.45′ S 113° 23.42′ E) is located on the central west coast of Western Australia (WA) and is one of three prominent headlands situated along the 180 km long north–south trending coastal Quobba Ridge (Fig. 1). Fringing reefs line the modern coastline and broaden northward into the Ningaloo Reef tract, Cape Range. Immediately south of Cape Cuvier an emergent fringing reef complex provides stratigraphic and geomorphic evidence for fluctuating sea levels during the peak of the last interglacial, marine isotope substage (MIS) 5e. Two distinct marine terraces were previously identified and described by Denman and Van de Graff (1978). They observed a lower well-developed coralgal reef terrace (“lower reef terrace” Fig. 2A) at + 3 to 5.5 m, and an upper incipient erosional terrace at about + 10 m with in situ corals and incipient coralgal rim at + 9.4 m (“upper terrace/rim”, Fig. 2B). Based on the same field evidence, Veeh et al. (1979) considered that the two marine terraces might be the result of two distinct highstand events. Alternatively, they postulated that the ⁎ Corresponding author. Department of Environmental and Geographical Sciences, Manchester Metropolitan University, Manchester, M1 5GD, United Kingdom. Tel.: +44 161 247 1529. E-mail address: [email protected] (M.J. O'Leary).

difference in elevation between lower and upper reef terraces could be an artefact of tectonic uplift. In addition, significant Quaternary warping and faulting have been cited at other coastal sites including; Cape Range to the north (Van de Graff et al., 1976), Shark Bay to the south (Playford, 1990), and Lake Macleod to the east (Denman and Van de Graff, 1976). Notwithstanding these observations, well-developed early MIS 5e terrace features are well documented along coastal WA (Zhu et al., 1993; Collins et al., 1993a; Stirling et al., 1995; Stirling et al., 1998; Hearty, 2003; Hearty and O'Leary, 2008), record a single phase of high sea-level and reef development between ~ 128 to ~ 121 ka (Stirling et al., 1998; O'Leary et al., 2008) and show little variation in elevation across 2500 km of coastline. Although there is a paucity of coral U-series ages younger than 121 ka, a palaeoreef site at Mangrove Bay (Ningaloo) WA, documents a regressive phase from the MIS 5e highstand at around ~ 116 ka (Stirling et al., 1998). At Cape Cuvier, α-counting U-series dating (Veeh et al., 1979), returned ages of ~ 123 ± 11 ka for the lower reef terrace and between 120 and 140 ka for the upper terrace/rim. More recently, at a locality several kilometres north of Cape Cuvier, near the wreck of the Korean Star, Stirling et al. (1998) document a reef terrace between ~+ 7 and 9 m. Thermal ionisation mass spectrometry coral ages of 128 ± 0.7 ka and 126 ± 0.5 ka (Stirling et al., 1998), are considered unreliable on the

0169-555X/$ – see front matter. Crown Copyright © 2008 Published by Elsevier B.V. All rights reserved. doi:10.1016/j.geomorph.2008.06.004

596

M.J. O'Leary et al. / Geomorphology 102 (2008) 595–602

Fig. 1. A map and aerial photograph showing the location of Cape Cuvier. The modern fringing reef and lower emergent terrace can be seen in the areal photograph, the survey transect 500 m south of the Cape is not shown.

basis of their elevated back-calculated δ234U initial values, which in pristine samples, should reflect the uranium isotopic composition of seawater at the time the corals grew. The present day marine δ234U of ca. 146.6‰ (Robinson et al., 2004; Andersen et al., 2008) is commonly adopted as the reference composition for seawater, assuming the marine 234U/238U has remained constant throughout the Quaternary. However, there is a growing body of evidence suggesting that up to 15 ‰ variability in 234U/238U has occurred on glacial–interglacial timescales (Henderson, 2002). 1.1. Aims In the context of predicted increases in global sea-levels, the difference between coral reefs tracking near-future sea-level rise (the

‘keep-up’ mode) and increasing water depths over reefs (as reefs switch from ‘keep-up’ to ‘catch up’; sensu Neumann and Macintyre, 1985) represents an important issue in tropical coastal geomorphology. Thus, a key question in contemporary reef geomorphology is to what extent and how rapidly will reefs be able to respond positively to future increases in sea level? The utilization of palaeoreef geomorphology, stratigraphy and high-resolution geochronology provides a unique and powerful tool for placing climate change impacts affecting modern reefs into a temporal context (Pandolfi et al., 2002; Greenstein and Pandolfi, 2008). However, before we can attempt to answer this question we must first, using Cape Cuvier as a type section, address the eustatic versus tectonic issues regarding elevation of these emergent marine terraces and the resolution to which these deposits can be dated.

M.J. O'Leary et al. / Geomorphology 102 (2008) 595–602

597

2. Methods

2.2. Sample preparation and analytical procedures

2.1. Surveying, site descriptions and sample collection

Coral samples were sectioned and micro-sampled with a dental drill to an approximate weight of 200 mg. Due to the open nature of the A. humilis skeleton, it proved very difficult to identify or remove any potential secondary mineralization or detrital contamination. As a result, bulk coral samples were analysed. Mechanical cleaning involved soaking in MilliQ water and sonication. Samples were progressively dissolved in distilled water by step addition of 10 M HNO3. All remaining organic material was removed, dried and weighed. The dissolved samples were spiked with a 50 mg ‘U-2' 233 Th/235U isotope solution and evaporated to a minimum solution. A few drops of H2O2 were added to oxidise any remaining organic material. Samples were redissolved in 3 ml of 2 M HNO3, then transferred to bio-spin ®Tru.spec columns for separation of uranium and thorium isotopes. A 0.1 normal solution of HF/HCl was then passed though the columns to collect and concentrate uranium and thorium. The solution was evaporated to dryness then redissolved with 2 ml 2% HNO3 prior to injection. U-series measurements were performed using a Neptune MCICPMS at the Research School of Earth Sciences, Australian National University. Detailed descriptions of the analytical methodologies, in-

A theodolite survey at a site 500 m south of the Cape was conducted and a topographic profile was constructed (Fig. 3). This transect incorporates the broadest section of the lower reef terrace perpendicular to the coast, to include the upper terrace/rim. The surface of the adjacent modern fringing reef (Fig. 2A), representing MLWS, was assigned the 0 m benchmark. The upper terrace/rim is exposed intermittently and was traced from windward southern side to the protected northern side of the Cape. A section of the actively eroding 3 m high scarp, forming the seaward edge of the lower reef terrace was also measured and logged in detail (Fig. 2C). Coral samples of the species Acropora humilis (common to upper reef slopes and reef flats) were collected in situ, for U-series dating, using a petrol-powered rock drill. A total of 15 coral cores, with an average penetration of between 15 and 20 cm, were taken in stratigraphic succession from the 3 m scarp of the lower reef terrace. Seven cores were drilled across the surface of the lower terrace. Due to the limited exposures only two growth-position corals were collected using a chisel and hammer from the upper terrace/rim.

Fig. 2. A) Modern and emergent reef platform at Cape Cuvier (looking north) at low tide, coralline algae can be seen encrusting the fossil reef in the foreground, cliffs in the background host the upper coralgal rim. B) Wave cut terrace at an elevation of +10 m capped by a coralline algal encrusted conglomerate. C) 3.2 m high erosional scarp at the seaward edge of the lower reef terrace, thick coral (A. humilis) plates can be observed, becoming more numerous higher in the section. D) A large favid coral colony exposed on the surface of the lower terrace, it appears truncated and capped by coralline algae. E) Seaward looking view of the lower terrace reef flat, a drainage gutter displays similar morphology to the modern reef flat in the background. F) A thin rim of coralline algae can be seen encrusting the palaeosea–cliff mid-way up photo, between arrows. G) Fossil beach at 7.4 m.

598

M.J. O'Leary et al. / Geomorphology 102 (2008) 595–602

Fig. 3. Surveyed transect of the emergent reef terrace 500 m south of Cape Cuvier, not shown in Fig. 1.

house uranium (SRM 960) standard solution performance, sensitivities and precision, partly adopted from Stirling et al. (1995) TIMS procedures, are reported in McCulloch and Mortimer (in press). Briefly, 234 U is measured on an ion counting system, while simultaneously 233 U, 235U and 238U are analysed on the Faraday array. Th and U isotope measurements are performed in four sequential steps utilising the high abundance energy filter (RPQ) in the central channel. 229Th and 230 Th isotopes are measured in the central ion counter (with RPQ) concurrently with 232Th in the Faraday cup (L2) and 238U in Faraday cups (cups H3 and H4). The multi-collection measurements in Faraday cups reduce errors created by ion beam instabilities. 230Th and 238U concentrations are calculated using the enriched 229Th and 233U tracers, respectively. These protocols allow for accurate high-precision measurements of 230Th/238U and 234U/238U ratios, reproducible to precisions of ~ 1–2‰ (McCulloch and Mortimer, in press). However, despite this several order-of-magnitude increase in analytical precision over α-counting, isotopic exchange processes, secondary mineral precipitation and detrital contamination have been shown to affect the accuracy and reliability of U-series ages. Thus, following the procedures of earlier workers (Chen et al., 1991; Stirling et al., 1995, 1998; Robinson et al., 2004) corals were screened for potential U and Th loss or gain based on the following criteria:

landward at about 1.5° over 50 m. (Fig. 2D,E). Large massive corals exposed at the surface are commonly truncated, suggesting a shift from vertical reef flat accretion to marine planation (Fig. 2D,). An inflection point at +5.1 m sees reef gradient steepen to 3.3° landward over the next 50 m up to 7.4 m and (Fig. 3) probably represents the maximum height of the palaeoshoreline during lower reef terrace formation. Reefal (predominantly coralline algae) deposits across this steeper gradient section are typically 30 to N10 cm thick, thinning landward over a basal eolianite surface. The erosional scarp exposes a full sequence of vertical reef framework development for the lower reef terrace. At the measured section (Fig. 3) overall terrace thickness to the basal eolianite is between ~ 3 to 3.5 m, suggesting vertical growth was limited by a narrow window of accommodation space. This together with its proximity to the shoreface would indicate that reef development occurred in a shallow-water high-wave-energy environment. Style and reef framework composition confirm this with lower half of the section composed of an alternating sequence of coralgal bindstone and skeletal carbonate grainstone, while an encrusting A. humilis framestone and coralgal bindstone dominate the upper half. The sequence is then capped by a thick N10 cm coralgal bindstone crust. 3.2. Upper terrace/rim morphology and stratigraphy

• Gallup et al. (1994) observed that a 4‰ shift in δ234U initial corresponded to a 1 ka shift in the U-series age in Barbados corals, and defined their limit for “acceptable” δ234U initial by considering these shifts in the context of the errors associated with their U-series ages. However, taking into account uncertainties in marine δ234U during MIS 5e, this reliability criterion has been extended to ±10‰ for “acceptable” δ234U initial (Stirling et al., 1995, 1998). • The total uranium concentration of fossil corals should approximate modern coral values of about 3 ± 0.5 ppm of uranium. • Fossil corals should be free of allochthonous 230Th, as indicated by the absence of detrital 232Th (b1 ppb). 3. Results

The upper terrace/rim extends discontinuously from a site 100 m north of the Cape where an in situ coral head at +9.4 m encrusts a palaeosea cliff, to the south, where is forms a narrow erosional (wavecut) terrace at +8 to 10 m overlain by ~ 0.3 to 0.5 m thick coralline algal encrusted conglomerate (cobble beach) (Fig. 2B). An irregular 0.1 to 0.2 m rim of encrusting coralgal material with the poorly preserved non-framework corals is exposed along the windward south side of the Cape (Fig. 2F). Loosely cemented coarse carbonate sands with shallow, seaward dipping planar beds at +7.4 and +12 m are interpreted to be a beach facies (Fig. 2G). It is unfortunate however that significant sections of the upper terrace/rim have either been covered by scree, resulting from the construction of salt-loading facilities at the Cape, or buried by talus or active Holocene coastal dunes.

3.1. Lower reef terrace morphology and stratigraphy 3.3. U-series coral ages The lower reef terrace is shore-attached fringing reef and developed atop an older erosional wave-cut platform; this basal unit is composed of Tertiary carbonate eolianites. The lower terrace is positioned adjacent to the modern fringing reef platform, with which it bears many geomorphic similarities; including reef aspect and platform width, and surface features including drainage gutters and small rills, which run perpendicular to the shoreface (Fig. 2A). An actively eroding scarp at +3.2 m represents the seaward edge of the lower terrace (Fig. 3). This feature clearly indicates that the original width of the lower terrace has been greatly reduced during the Holocene, a result of Holocene erosive processes. The lower terrace surface is covered predominantly in crustose coralline algae and rises

The results of 27 U-series coral dates from Cape Cuvier are shown in Table 1 and Fig. 4. Uranium concentrations for all corals except LCV2c_4 fall within the 2.5–3.5 ppm range common for modern corals (Edwards et al., 1988; Eisenhauer et al., 1993; Stirling et al., 1995). Of the 27 analysed corals, 20 had 232Th concentrations b1 ppb suggesting minimal non-radiogenic 230Th contamination. Seven corals with suspected detrital contamination had 232Th concentrations ranging between 1.27 and 18.92 ppb, and yielded U-series ages that fall outside the known duration of MIS 5e highstand. Due to the uncertainties regarding the degree to which non-radiogenic 230Th may have affected age accuracy; these samples were excluded from age

M.J. O'Leary et al. / Geomorphology 102 (2008) 595–602

599

Table 1 U-series data Summary

U ppma

230

Th pptb

Sea cliff corals LCCg_C LCC1i_C LCCk_C LCC1m_C LCC1n_C LCC1o_C LCC1r_C LCCs_C LCCt_C

(lower terrace) 3.01 39.30 3.38 43.25 3.16 40.76 3.04 38.83 3.11 40.02 3.18 41.08 3.28 42.88 3.26 43.33 3.21 41.22

Reef flat corals LCV7 T1 LCV7 T2a LVC7 T2b LCV7 T3 LCV7 T4 LCV7 T5 LCV7 T6

(lower terrace) T1 seaward to T6 landward 2.69 34.21 0.16 3.13 42.64 0.08 3.16 42.17 0.10 2.21 28.88 0.06 3.44 42.37 0.31 3.14 40.36 0.13 3.22 40.83 0.75

Coralgal rim corals (upper terrace) LCV_1e_3 3.35 44.07 LCV_2e_1 3.17 42.81 LCV_2e_4 4.37 56.01 Detrital 232Th LCCa_C LCCb_C LCC1j_C LCCp_C LCCq_C LCCw_C LCV_1e_2 LCV_2e_3 a

N1 ppt (rejected) 3.05 42.01 2.90 37.65 2.86 38.52 3.47 46.98 3.68 43.63 3.39 44.03 2.20 28.28 3.88 50.91

232

Th ppbc

δ234U Meas.d

± 2σ

230

124.80 120.18 123.60 119.54 122.49 134.44 130.74 124.52 121.59

1.84 1.34 0.87 0.97 1.13 1.26 0.64 0.82 1.00

128.60 143.32 139.27 124.98 124.62 119.20 112.43

0.15 0.10 0.75

5.39 3.59 18.92 2.01 12.48 1.27 11.63 1.46

0.16 0.25 0.16 0.18 0.07 0.47 0.32 0.17 0.34

Th/238Ue

± 2σ

[230Th/232Th]

Age (ka)f

± 2σ

Initial δ234Ug

± 2σ

0.8067 0.7897 0.7978 0.7891 0.7942 0.7991 0.8080 0.8224 0.7928

0.005 0.004 0.005 0.010 0.005 0.004 0.007 0.003 0.002

46,520 32,419 48,364 40,107 113,409 16,419 25,335 46,761 22,683

132.40 128.30 130.00 128.30 129.00 127.70 131.30 137.50 128.90

1.55 1.35 1.45 2.85 1.50 1.10 2.25 0.97 0.69

181.60 172.90 178.60 172.00 176.60 193.10 189.70 183.80 175.20

2.50 1.85 1.30 1.80 1.60 1.70 1.45 1.20 1.30

1.03 0.85 1.21 0.68 1.04 1.02 1.01

0.7837 0.8384 0.8208 0.8021 0.7577 0.7917 0.7803

0.002 0.003 0.003 0.002 0.002 0.002 0.002

39,635 94,188 76,512 89,622 25,669 56,496 10,213

124.57 137.53 133.06 130.93 118.13 129.18 125.02

0.61 0.88 0.86 0.59 0.66 0.73 0.53

183.07 211.66 203.08 181.14 174.19 171.92 174.50

1.41 1.20 1.68 0.90 1.33 1.38 1.31

114.10 133.64 126.80

0.84 0.76 0.84

0.8091 0.8309 0.7887

0.002 0.003 0.002

54,759 78,065 13,939

135.98 137.73 126.46

0.68 0.90 0.53

167.76 197.47 181.48

1.17 1.09 1.10

117.45 115.49 121.89 128.56 136.55 120.14 111.69 130.34

1.33 1.34 1.21 1.14 0.85 0.64 0.63 0.80

0.8500 0.8021 0.8330 0.8359 0.7317 0.8013 0.7874 0.8076

0.005 0.003 0.005 0.003 0.003 0.002 0.001 0.003

1462 1965 381 4377 655 6511 454 6441

149.00 133.40 141.70 140.80 109.00 131.70 129.74 131.25

1.70 1.10 1.60 0.89 0.79 0.73 0.37 0.87

179.20 168.50 182.10 191.60 186.00 175.70 161.34 189.10

2.00 1.80 1.80 1.55 1.10 0.91 0.79 1.15

Uranium concentrations are measured in parts per million (ppm). Th concentrations measured in parts per trillion (ppt). Th concentrations measured in parts per billion (ppb). 234 δ U = {[(234U/238U) / (234U/238U)eq] − 1} × 103. (234U/238U)eq is the atomic ratio at secular equilibrium and is equal to λ238/λ234 = 5.4891 × 10− 5,, where λ238 and λ234 are the decay constants for 238U and 234U, respectively, adopting half-lives of Cheng et al. (2000). [230Th/238U]act = (230Th/238U) / (λ238/λ230). U-series ages are calculated iteratively using: 1 − [230Th/238U]act = exp− λ230T − (δ234U(0) / 1000)(λ230/(λ230 − λ234))(1 − exp(λ–234λT230) where T is the age in years and λ230 is the decay constant for 230Th. λ238 = 1.551 × 10− 10 y− 1; λ234 = 2.826 × 10− 6 y− 1; λ230 = 9.158 × 10− 6 y− 1. Strictly reliable ages will have 234Uinitial 146.6 ± 10‰. The initial value is given by δ234Ui = δ234U(0)exp(λT234) where T is the age in years.

b 230 c 232 d

e f

g

interpretations. All corals analysed had elevated δ234Uinitial values ranging from 168 to 211‰ (Fig. 4), which are considered to be unreliable based on the above screening criteria. However, despite this geochemical uncertainty with regards to age accuracy, it is possible to further test U-series age reliability against the morphostratigraphic developmental history of the reef. Five corals from the surface of the lower terrace with elevated δ234Uinitial values of 171.92 to 211.66‰ yielded ages between 118 ka to 138 ka. Unfortunately, the succession of measured coral ages across the reef flat are inconsistent with the seaward-younging succession (Table 1) we would expect from a laterally prograding fringing reef system (Kennedy and Woodroffe, 2002). The seaward-most corals LCV T1,2,3 generally returned older ages compared to landward examples LCV T4,5,6. Ten in situ coral samples drilled from the 3 m scarp provide the fundamental test of stratigraphic superposition for the U-series ages (Fig. 5). All ten corals returned ages ranging from ~ 128 to ~132 ka with one outlier at 137 ka (Fig. 5). A vertical age isochron such as this is indicative of laterally accreting reef systems, however, these ages suggest coral growth (at the present seaward edge of the terrace) occurred during the very initial phase of the MIS 5e highstand. Due to the poor preservation of corals collected from the upper terrace, the result of surface weathering and exposure, they were less than ideal material for dating. Only 2 of the five corals analysed, LCV1e_3 and LCV2e_3 returned equivalent modern uranium concentrations.

These two corals returned unreliable U-series ages of 136.0 ± 0.7 and 131.3 ± 0.9 ka, which would suggest the upper terrace/rim development occurred at the initiation of the MIS 5e highstand (Table 1; Fig. 4). 4. Discussion 4.1. U-series coral ages and palaeoreef development 4.1.1. Lateral reef accretion The presence of a marine planation surface beneath parts of the lower reef terrace indicates that coastal erosion took place prior to reef development, with coral growth occurring close to the shoreline in a shallow-water high-energy setting. The resulting lack of vertical accommodation space would have resulted in predominant laterally prograding reef sequence (Kennedy and Woodroffe, 2002). Lateral progradation should produce age isochrons parallel to the reef front, such that progressively older corals are exposed on the reef surface landward of the reef crest (e.g., Fig. 3). However, despite this broad style of reef development, reef flats remain geomorphically active, being a zone of both accretion and erosion probably influenced by minor oscillations in sea level and local changes in the intensity of mechanical and bioerosive processes (Fig. 2A,D,E). As a result, age isochrons across the reef flat surface are likely to be more variable. Analyses of coral age across the reef flat (LCVT1 to 6) do not display any particular age pattern or succession, however, whether this is

600

M.J. O'Leary et al. / Geomorphology 102 (2008) 595–602

Fig. 4. Initial δ234U at time of coral growth versus 230Th-age for MIS 5e Cape Cuvier corals. For strictly reliable measurements, initial δ234U values in MIS 5e examples should approximate a modern seawater value of 146 ± 10‰ (shaded horizontal area). As values fall above this area all measured ages are considered unreliable.

primarily due to differential erosion or geochemical contamination is uncertain. 4.1.2. Erosional scarp retreat The present seaward limit of the lower reef terrace at Cape Cuvier does not represent the true maximum lateral extent of the palaeoreef flat. Holocene erosive processes within the mid- and upper littoral zone have resulted in the notching and retreat of the lower reef terrace. It is difficult to calculate erosion rates on limestone coasts, requiring the quantification of a number of variables including: 1) diversity and density of bioeroding species; 2) the chemical action of seawater; 3) the mechanical action of wave-laden sediment; 4) substrate density (Spencer and Viles, 2002) and; 5) freshwater dissolution. Erosion rates along tropical carbonate coastlines can range anywhere from 2 to 15 mm/yr and up to 33 mm/yr (Playford, 1988). Even using a conservative rate of 4 mm/yr (Trudgill, 1983) would suggest at least 30 m or almost half of the original terrace has been eroded during the Holocene. The fossil corals now exposed at the shoreface should approximate an interval mid-way through the reefs progradational development or a period approximating mid MIS 5e (125–120 ka). Although all ten corals returned near equivalent ages up-section, corresponding to a laterally accreting reef model (Fig. 5), their average age of 130.5 ± 3 ka is too old to correlate with a mid MIS 5e growth period. It is through the use of geomorphology, superposition, and morphostratigraphic succession it is possible to confirm the geochemical uncertainties with regard to U-series age accuracy. 4.1.3. Coralgal rim The incipient nature of geomorphic features, which constitute the upper terrace/rim points to a brief stillstand of sea level at between +8 to +10 m. In the Bahamas, Neumann and Hearty (1996) argued on the basis of bioerosion rates creating a 1.5 m deep notch at +6 m that the late MIS 5e sea level event may have lasted for less than 600 years.

This is reasonable when applied to Cape Cuvier, as this short period was insufficient to promote abundant coral growth at this +10 m upper terrace elevation, or for the lower reef terrace to fill the newly available accommodation space. The short duration of this peak sealevel event should result in a brief period of coral growth and consequently a tight cluster of coral ages. Unfortunately, due to the exposed nature of the deposit all samples were subject to U/Th exchange, expressed in elevated δ234Uinitial values and ages ranging from between 126.46 ± 0.5 and 137.73 ± 0.9 ka. However, on the basis of stacked MIS 5e coral sequences in Bahamas and Hawaii, the highest highstand of MIS 5e is documented to have occurred at the end of the interglacial (Hearty et al., 2007) so these WA ages are untenable. 4.2. Palaeoecology and palaeoreef development The reef framework assemblage dominated by coralgal bindstone and encrusting coral framestone, is representative of a reef crest biofacies, and further evidence that the lower reef terrace developed in a shallow-water high-energy setting. Additionally, a very depauperate coral fauna with a sole species of encrusting A. humilis, common to exposed upper reef slope and flat environments, being the predominant reef framework builder, is suggestive of very marginal coral growth conditions, resulting from high-energy wave related stress. Much of the reef flat is mantled in a thick (N10 cm) crust of coralline algae. This facies transition from coral to coralline algae may represent reef response to a range of interrelated environmental parameters including; sea-level change, loss-gain in accommodation space, and increasing–decreasing wave-energy. Where corals outcrop on the reef flat they generally appear truncated suggesting additional accommodation space was created for massive and encrusting corals to grow. The timing of such an event is likely to correlate with a late MIS 5e peak in sea level. Since there was no substantial vertical accretion of the entire reef platform, we infer the duration of higher sea level was very brief, or conditions were not

M.J. O'Leary et al. / Geomorphology 102 (2008) 595–602

601

(Fig. 2 B,F), late in the MIS 5e highstand. This period is characterised by erosional rather than constructional features. Constructional features include decimetre thick, narrow rim of encrusting coralline algae, which tend to be more resistant to higher wave energies. Corals found encrusting the rim of coralline algae tend to be isolated and non-reef forming, with larger colonies confined to the betterprotected northern side of the Cape. 4.3. Cape Cuvier and other global MIS 5e sites The most recent and detailed contribution to our understanding of sea level during MIS 5e is presented in Hearty at al. (2007). This global study documents intervals of transition and stability, suggesting sea level during this period was far more complex than portrayed in previous models and interpretations. During the early half of the highstand (~132–125 ka) sea level remained relatively stable at +2.5 m, and characterised by broad terraces and reef flats; followed by a minor regression, sea level then rose again to +3 to 4 m. The end of the interglacial period (ca. 120–118 ka) was marked by abrupt shifts of sea level between +6 and +9 m, forming multiple notches and narrow rubble benches (Hearty et al., 2001; Hearty et al., 2007). Although Cape Cuvier does not exhibit the geomorphic complexity of such sea-level oscillations, it does however importantly document an extended sea level stillstand followed by a brief late MIS 5e transition in sea level. Although localized tectonics have previously been used to explain the presence of these higher marine units, we consider this explanation to be incorrect. There is now an established, independent, and reliable palaeosea-level datum of +2.5 ± 1 m, representing an extended period of stable sea level between ~ 132 and 125 ka (Hearty et al., 2007), and characterised by a geomorphically distinct fossil strandline (i.e., the lower reef terrace). This then allows for a correction of palaeosea-level datums, along coasts that may have affected by isostatic adjustments or neotectonism. That there exists, at Cape Cuvier, a broad emergent lower reef terrace at an elevation consistent with other MIS 5e stable margin terraces would suggest that this site has not been subject to tectonic displacement. The upper terrace, which formed during the same MIS 5e highstand, was the result of, based on geomorphic evidence; 1) a rapid rise sea level from ~+3 to +8.5 m, 2) a brief hiatus at this new elevation, and 3), to best preserve these incipient geomorphic features of the upper terrace, a rapid regression to MIS 5d interstadial (Neumann and Hearty, 1996; Hearty and Neumann, 2001). We also suggest a prevailing view of stable sea-levels during MIS 5e from Western Australia may have led to inferences of localized tectonism, whereby these upper terrace features were interpreted to have also formed during the general stable sea-level period between ~ 132 and 120 ka. 5. Conclusions

Fig. 5. Measured erosional scarp section showing an alternating sequence of A. humilis, coralline algae and detrital sediment packages. U-series coral ages return an average upsection age of 130.5 ± 3.0 ka. These ages, coinciding with the initiation of the MIS 5e highstand are considered to be too old, based on the position of the measured section and the inferred developmental history of the reef.

conducive for reef development during this transitional sea level phase. The subsequent erosion/planation of these large coral heads probably occurred during a rapid regression of sea level during the final phase MIS 5e, with the reef flat environment passing through subtidal to intertidal, increasing both wave energy and bioerosive processes. The underdeveloped nature of the upper terrace/rim is interpreted to be a consequence of a brief sea-level stillstand at this new position

Cape Cuvier is characterised by two distinct marine geomorphic units, 1) a lower, well-developed accretional reef terrace at + 3 to + 5.5 m, representing an extended interval of stable sea level, and 2) an upper erosional terrace and incipient coralgal rim between + 8.5 to + 10.5 m represents a brief sea-level stillstand at this higher elevation. Collection and dating of corals to resolve timing of deposition prove inadequate despite the highest level of analytical precision, due to pervasive diagenetic alteration. Despite the shortcomings of absolute dating, a succession of events can be resolved through morphostratigraphic relationships. Comparison of the facies relationships, coral growth, and morphostratigraphic features between the lower and upper terraces indicate that an early MIS 5e stillstand at + 3–5 m was followed by a late rise to + 8.5 to 10.5 m. This agrees with an emerging global view of MIS 5e sealevel history derived from stable carbonate platforms, and thus precludes the effects of localized tectonics. Finally, in the context of

602

M.J. O'Leary et al. / Geomorphology 102 (2008) 595–602

modern reef response to possible future sea-level rise events, this study has determined that (1) limited reef growth occurred through direct re-initiation on the pre-existing reef flat surfaces, with no evidence for any re-established coral communities being able to fully utilize the newly available accommodation space due to the brevity of this event, and (2) that re-establishment of coral communities and most incipient reef development primarily took place at new upslope and/or further inland locations. Acknowledgements We would like to thank Graham Mortimer and Les Kinsley who provided technical support and a special thanks to Dave Bauer who provided logistical support while in the field. We thank Jody Webster for an earlier review of this manuscript, and constructive comments by Claudine Stirling whose detailed editorial contribution greatly improved the manuscript. Research funding was made available through an Australia Research Council Discovery Grant (ID: DP0209059) to M. McCulloch, P. Hearty and A. Halliday, and grants from Kanagawa Museum, and an internal JCU merit research grant.

References Andersen, M.B., Stirling, C.H., Potter, E-K., Halliday, A.N., Blake, S.G., McCulloch, M.T., Ayling, B.F., O'Leary, M.J., 2008. High-precision U-series measurements of more than 500,000 year old fossil corals. Earth and Planetary Science Letters 265 (1–2), 229–245. Chen, J.H., Curran, H.A., White, B., Wasserburg, G.J., 1991. Precise chronology of the Last Interglacial period: 234U230Th data from fossil coral reefs in the Bahamas. Geological Society of America Bulletin 103, 82–97. Cheng, H., Edwards, R.L., Hoff, J., Gallup, C.D., Richards, D.A., Asmerom, Y., 2000. The half-lives of uranium-234 and thorium-230. Chemical Geology 169, 17–33. Collins, L.B., Zhu, Z.R., Wyrwoll, K.H., Hatcher, B.G., Playford, P.E., Chen, J.H., Eisenhauer, A., Wasserburg, G.J., 1993a. Late Quaternary evolution of coral reefs on a cool-water carbonate margin: the Abrolhos carbonate platforms, Southwest Australia. Marine Geology 110, 203–212. Denman, P.D., Van de Graaff, W.J.E., 1976. Emergent Quaternary marine deposits in the Lake MacLeod area, Western Australia. Western Australia Geological Survey Annual Report 1977, 32–36. Eisenhauer, A., Zhu, Z.R., Collins, L.B., Wyrwoll, K.-H., Eichstatter, R., 1993. The Last Interglacial sea level change: new evidence from the Abrolhos Islands, West Australia. Geoligische Rundschau 85, 606–614. Edwards, R.L., Taylor, F.W., Wasserburg, G.J., 1988. Dating earthquakes with highprecision thorium-230 ages of very young corals. Earth and Planetary Science Letters 90, 371–381. Gallup, C.D., Edwards, R.L., Johnson, R.G., 1994. The timing of high sea levels over the past 200,000 years. Science 263, 796–800. Greenstein, B.J., Pandolfi, J.P., 2008. Escaping the heat: range shifts of reef coral taxa in coastal Western Australia. Global Change Biology 14, 1–16. Hearty, P.J., Neumann, A.C., 2001. Rapid sea level and climate change at the close of the Last interglaciation (MIS 5e): evidence from the Bahamas Islands. Quaternary Science Reviews 20, 1881–1895.

Hearty, P.J., 2003. Stratigraphy and ages of fossil carbonate deposits of Rottnest Island, Western Australia. Quaternary Research 60, 211–222. Hearty, P.J., O'Leary, M.J., 2008. Carbonate eolianites, quartz sands, and Quaternary sealevel cycles, Western Australia: a chronostratigraphic approach. Quaternary Geochronology 3, 26–55. Hearty, P.J., Neumann, A.C., Hollin, J.T., O'Leary, M.J., McCulloch, M.T., 2007. Global sealevel fluctuations during the last interglaciation (MIS 5e). Quaternary Science Reviews 26, 17–18 2090–2112. Henderson, G.M., 2002. Seawater (234U/238U) during the last 800 thousand years. Earth and Planetary Science Letters 199, 97–110. Kennedy, D.M., Woodroffe, C.D., 2002. Fringing reef growth and morphology: a review. Earth-Science Reviews 57, 255–277. McCulloch, M.T., Mortimer, G., in press. Applications of the 238U–230Th decay series to dating of fossil and modern corals using MC-ICPMS, Australian Journal of Earth Science. Neumann, A.C., Macintyre, I., 1985. Reef response to sea level rise: keep-up, catch-up or give-up. In: Gabrie, C., Toffart, J.L., Salvat, B. (Eds.), Proceedings Of The Fifth International Coral Reef Congress, Tahiti, 27 May –1 June. Symposia and Seminars (A), vol. 3, pp. 105–110. Neumann, A.C., Hearty, P.J., 1996. Rapid sea-level changes at the close of the last interglacial (substage 5e) recorded in Bahamian island geology. Geology 24 (9), 775–778. O'Leary, M.J., Hearty, P.J., McCulloch, M.T., 2008. U-series evidence for widespread reef development in Shark Bay during the last interglacial. Palaeogeoraphy Palaeoclimatology Palaeoecology 259, 424–435. Pandolfi, J.M., 2002. Coral community dynamics at multiple scales. Coral Reefs 21, 13–26. Playford, P.E. 1988. Guidebook to the Geology of Rottnest Island. Excursion Guidebook No. 2, Western Australia Division of the Geological Society of Australia: Perth; 67 pp. Playford, P.E., 1990. Geology of the Shark Bay area, Western Australia. In: Berry, P.F., Bradshaw, S.D., Wilson, B.R. (Eds.), Research in Shark Bay. Report of the FranceAustrale Bicentenary Expedition Committee. Western Australian Museum, Perth, pp. 13–31. Robinson, L.F., Belshaw, N.S., Henderson, G.M., 2004. U and Th concentrations and isotope ratios in modern carbonates and waters from the Bahamas. Geochimica et Cosmochimica Acta 68, 1777–1789. Spencer, T., Viles, H., 2002. Bioconstruction, bioerosion, and disturbance on coral reefs and rocky carbonate coasts. Geomorphology 48, 23–50. Stirling, C.H., Esat, T.M., McCulloch, M.T., Lambeck, K., 1995. High-precision U-series dating of corals from Western Australia and implications for the timing and duration of the Last Interglacial. Earth and Planetary Science Letters 135, 115–130. Stirling, C.H., Esat, T.M., Lambeck, K., McCulloch, M.T., 1998. Timing and duration of the Last Interglacial: evidence for a restricted interval of widespread coral reef growth. Earth and Planetary Science Letters 160, 745–762. Trudgill, S.T., 1983. Preliminary estimates on intertidal limestone erosion, One Tree Island, Southern Great Barrier Reef, Australia. Earth Surface Processed and Landforms 8, 189–193. Van de Graaff, W.J.E., Denman, P.D., Hocking, R.M., 1976. Emerged Pleistocene marine terraces on Cape Range, Western Australia. Geological Survey of Western Australia Annual Report 62–69. Veeh, H.H., Schwebel, D., van de Graaff, W.J.E., Denman, P.D., 1979. Uranium series ages of coralline terrace deposits in Western Australia. Australian Journal of Earth Science 26, 285–292. Zhu, Z.R., Wyrwoll, K.-H., Collins, L.B., 1993. High precision U-series dating of Last Interglacial events by mass spectrometry: Houtman Abrolhos Islands, Western Australia. Earth and Planetary Science Letters 118, 281–293.