The upper 2.5m of a gravity core Fr10/95-GC17, taken offshore Cape Range in northwestern Australia, has been studied in great detail and, on average, samples were taken at 2 cm intervals. This core was taken at a water depth of 1,093 m and its location is 22°07.74’S 113°30.11’E. Above the site, salinity in December 1995 was 34.92 ‰, typical of the poleward-flowing Leeuwin Current water. Below that shallow watermass, obvious Indian Central Water [ICW] occurs with a maximum salinity of 35.65, and from 560 m water depth down to the sea floor, Antarctic Intermediate Water [AAIW] prevails with an average salinity of 34.45. Sea-surface temperature at the time of core collection was 25.7°C.

This core was chosen for an extensive study among a selection of 51 other cores taken during 2 oceanic cruises using the national facility RV Franklin along 2 transects between Perth, Java and Darwin. It was chosen because it is conveniently located today at the southwestern margin of the Warm Pool in the Indian Ocean, because it is situated below the Leeuwin Current which is characteristically of low density [ as a result of low salinity and high temperature]. The site also occurs close to the coast and receives airborne pollen and dust from mainland Australia.

Monospecific samples of planktic foraminifers from sixteen horizons from Core Fr10/95-GC17 were radiocarbon-dated by AMS [by Fifield, Santos, Yokoyama and Lawson together with members of ANSTO], thus providing a good chronology for changes recorded in the core spanning the last 45,000 years. In addition, OSL dates have successfully been obtained on individual aeolian quartz grains recover from specific intervals of the core deposited during the period spanning the Last Glacial Maximum. This is the first time OSL dates have been obtained from deep-sea cores, and which permit comparison with radiocarbon dates from deep marine sediments.

The following studies were carried out on the core: faunal counts of planktic foraminifers [by Martinez; see ref.1] and calculation of sea-surface temperature based on those counts [by Barrows; see ref. 2, and in prep.], floral counts of calcareous nannoplankton and estimates of broad temperature signals [by Takahashi; see ref. 3], presence of pteropods [by Martinez], pollen counts [by van der Kaars; MS in prep.], d¹³C and d¹⁸O on planktic foraminifers [see ref.1], total carbon, inorganic and organic carbon [by M. Sloan], XRD analyses [by M. Pryzbylak and D. Isaacs]. Interpretation of data from the core also rely on a calibration against 52 core tops obtained during the 2 Franklin cruises and also some 300 water samples for which stable isotope measurements have been made.

The following results will be presented at the talk:

1. For the last 40,000 years, there have been two climatic and oceanic modes affecting the southwestern margin of the Warm Pool. One, during the glacial period, saw much reduced precipitation [by down to 30-40% from the present- ref.4], enhanced sea-surface salinities [=SSS] and moderately reduced sea-surface temperatures [=SST]. At the time, the extent of the Warm Pool was reduced. The other, which commenced at the onstart of the Holocene, registered broad SSS fluctuations due to an increase in precipitation over the Warm Pool and the throughflow of the Leeuwin Current. Cyclonic activity, which contributes to over 40% of the rainfall over Cape Range today, also did commence at the onstart of the Holocene. This is recognised also by a dramatic switch of the supply of terrigenous material to the coring site. The Holocene sequence is characterised by dark brown sediments compared to the previous sequence which is much richer in carbonates and pale grey in colour.

Those 2 different modes of atmospheric circulation, reflected in the nature of the surface of the ocean of the Warm Pool has already been developed [ ref. 4, 5] and is further substantiated by the work of van der Kaars et al. [ref. 6] from a core from the Banda Sea.

2. Sea-surface temperature conditions for core site GC17 have already been compared against a dozen other sites along a north-south transect between Java and Perth [ref. 1]. Additional levels for the Holocene were examined for the present study to obtain a higher resolution, and we now can provide a good chronology to support the SST reconstructions .With confidence, we can claim that mean SST were in the vicinity of 22°C between 33,000 and 20,000 cal yearsBP. Lowest temperatures were registered around 31,000 and also from 23 to 21,000 cal yearsBP. After that mean SST rose slowly at first up to 16,000 [23°C], followed by a slight drop until 14,000 cal yearsBP by which time SST rose more dramatically until 7,000 cal yearsBP where it reached ~ 26.5°C. This was followed by a temperature drop lasting one millennium, and at 5,500 cal yearsBP it climbed to reach 27.5°C at 4,000 cal yearsBP. After that, SST dropped progressively with minor fluctuations. Of note is that seasonality in SST were at their lowest during the highest temperature reconstructions centred around 4,000 cal yearsBP.

3. The vegetation reconstruction by van der Kaars on core GC17 further confirms the 2 climatic modes. At the Last Glacial Maximum [=LGM], Callitris levels were highest for the last 45,000 years; Eucalyptus percentages were at their lowest. The ‘deglaciation’ [=transition between LGM and Holocene with sea level rapidly rising] was characterised by an increase in eucalypts, the low-shrub Gyrostemon and chenopods. The Holocene vegetation spectra, on the other hand, differ; it could be called the ‘Poaceae period’, to the detriment of chenopods. In addition, eucalypts increase but do fluctuate much, but were never as high in percentages as during the period predating the LGM. Of note also is the predominance of mangrove pollen peaking around 8,400 cal yBP; this is earlier than at other sites in northern Australia, but is consistent with a similar observation from the Banda Sea region [ref. 6]. For the Holocene, characterised by much SST fluctuations, there is no apparent parallelism with vegetational changes. This may be the consequence of the sampling intervals as fewer levels were examined for pollen compared to the levels for planktic forams and d¹⁸O.

4. The calcareous nannoflora in core GC17 has been compared against that of two other cores [ref. 3], one located in the Java Upwelling System, and the other further south offshore Carnarvon. In core GC17, the Holocene is characterised by warmer taxa [as a greater abundance of Gephyrocapsa taxa] than during the LGM and the entire marine isotope stage 2 [-MIS], but differences are subtle compared to the core offshore Carnarvon. Lower SSTs, prior to the middle of MIS2, point out to the diminishing [or absence] of the Leeuwin Current. It is postulated that many of the nannoplankton taxa could have been immersed into a cold ICW watermass during the glacial period. For the Holocene especially, the increase in warm-water indicators [Gephyrocapsa taxa] coincide with the peaks in SST obtained from the foraminifer faunal counts; these independent observations based on different organisms give additional support to the validity of SST reconstructions obtained from the foraminifer faunal assemblages by Barrows [AUSMAT-2, unpubl.].