IAEA-SM-354/99 XA9951925 THE USE OF NATURAL 14C AS A TRACER TO IDENTIFY THE INCORPORATION OF YOUNGER MATERIAL INTO THE ORGANIC COMPONENT OF SEDIMENTS FROM THE CARPENTARIA BASIN, AUSTRALIA HEAD1, M. J. and P. De DECKKER Research School of Earth Sciences and Dept. of Geology, Faculty of Science, The Australian National University, Canberra, ACT 0200, Australia E. M. LAWSON ANTARES AMS Facility, Division of Physics, ANSTO, Lucas Heights, NSW 2234, Australia Abstract A "chemically inert" organic fraction was isolated from samples taken from a sediment core that was collected from the Gulf of Carpentaria, lying between Australia and Papua New Guinea. Accelerator mass spectrometry (AMS) 14C ages obtained from this fraction become younger with depth, the maximum 14C concentration being 107% M, then become older again. The shape of the plot of apparent 14C age with depth can be compared to a typical concentration plot of a compound being passed through a chromatographic column. The shape of the so-called natural chromatogram has a relatively sharp leading edge, then gradually tails off. Stable carbon isotope values indicate that the material is most probably a non polar compound (or mixture) of yet unknown origin. 1. INTRODUCTION The Gulf of Carpentaria is a large, shallow (<70m) epicontinental sea lying between Australia and Papua New Guinea. It is connected to the Arafura Sea to the west across a 53m sill located below sea level, and to the Coral Sea to the east across the 12m deep Torres Strait (see Fig IB). During a large part of the Quaternary period (viz. the last 2 million years of the geological record), the Gulf was disconnected from the ocean, and a large lake was formed periodically in this large depression. The lake, when it last existed prior to the last sea-level transgression that commenced after the Last Glacial Maximum (= LGM) ca. 20 cal ka and that culminated in its flooding by marine waters at approximately 9 cal ka, may have had a surface area of some 150,000 km2. The existence of this lake had been predicted by Nix and Kalma [1] based on Phipps' [2] early bathometric reconstructions for the Gulf. Torgersen [3] conducted an extensive marine cruise in the Gulf in 1982, and this resulted in the collection of some 35 gravity cores with good seismic control. Most cores were taken below 60m of water depth. Investigations in the Gulf that followed the cruise are presented in [3] and [4], and references therein. When full around the time of the LGM, "Lake Carpentaria" (see Fig. IB) retained fresh to slightly saline water and was less than 10 m deep; for more information, refer to [4-6]. Torgersen et al., [4] have summarised the stratigraphic units recovered in the 1 Current address: State Key Laboratory of Loess and Quaternary Geology, Chinese Academy of Sciences, PO Box 17, Xi'an 710054, China. 226 cores and have provided a 14C chronology for the top 220 cm of the most representative core GC-2. The stratigraphic section has been divided into 5 units which are recognised in most cores. These units are schematised graphically in Fig. 1C. IIO'E A GC-2 CORE 4 d Gulf of IK- -« s 12°31.18'S 140°21.14'E ^Carpentaria STRATIGRAPHIC DEPTH UNIT (cm) ( \: Poo! near Hutt Lagoon ) r*—\A -36' Little Dip Lake ^"N -100 , SOOkm . LIVING Cyprideis LOCALITIES 110'E ISJ' B |-. •, | sand shell hash consisting mainly of bivalves | ~i_ | mud Y///A mottled Y///A sediment © shell hash consisting \—~1 laminations mainly of gastropods FIG. 1. A. Location map showing the Gulf of Carpentaria positioned between Australia and Papua New Guinea. B. is a more detailed map showing the extent of "Lake Carpentaria" during periods of low sea levels, as well as the position of core GC-2 which is located more than 60 km from the present-day coastline of the Gulf. C. is a simplified stratigraphic section of core GC-2 showing the five principal lithological Units (I-V) discussed in the text. The top unit I (0-30 cm) consists of grey-green, siliciclastic to bioclastic sediments which contain abundant marine molluscs (including pteropods), bryozoans, ostracods and foraminifers. This unit was deposited under truly marine conditions, but it contains some reworked faunal elements from the previous underlying units (for further details, refer to ref 5). It is very likely that the sediments representing the present-day deposition at the bottom of the Gulf are missing from the core as the gravity coring technique frequently compresses or loosens the often 'soupy' layers at the bottom of lakes or the oceans. Unit II consists of firm dark grey siliciclastic sediments containing ostracods and foraminifers deposited under lacustrine conditions. It is the thickest unit in core GC-2 and extends down to 150 cm. Radiocarbon ages give the time of deposition for this unit between 9 and 25 cal ka . Unit III is characterised by fine laminations with layers consisting of small calcite crystals (10-20 Lim long) alternating with siliciclastic sediments. This unit was deposited in a stratified lake, and this is further confirmed by evidence of bleaching and partial dissolution 227 of biogenic carbonates such as ostracods and foraminifers and the presence of pyrite crystals indicative of dysaeorobic conditions. This unit in core GC-2 was deposited between 25 and 30.5 cal ka. The underlying Unit IV consists of shelly siliciclastic sediments with abundant molluscs, ostracods and foraminifers. Some horizons in this Unit are best described as consisting of a shell hash. Postulated conditions [ref. 5] for this Unit are those of a very productive freshwater lake during the 30.5- 41.9 cal ka period. Dates for this Unit have been extrapolated through core GC-10A which contains the best representation of this Unit. The bottom of the core is characterised by mottled, siliciclastic sediments that have undergone some pedogenesis/aerial exposure; this is typically Unit V which contains reworked marine faunas mixed with some freshwater/low salinity ones, and for which an age >41 cal ka is postulated. All the dates referred to for this section were measured on bulk carbonates [see ref 4]. The purpose of this study has been to compare 14C ages of "chemically inert" organic carbon recovered from core GC-2 with those of foraminifera and ostracods (biogenic carbonates) from the same core collected at regular intervals, using the technique of accelerator mass spectrometry, with the aim of determining the best possible material for dating sedimentary sequences formed under a marginal environment which registers marine/non-marine transitional facies. 2. METHODS CO2 from each foram or ostracod sample was obtained by the standard acid evolution technique using dry HP3O4. For organic extractions, each sediment sample was subjected to a series of three solvent extractions using solvents of different polarity, in an ultrasonic bath. A full explanation of the procedure has been given previously [6]. From the lack of colour in the solutions, it was assumed that very little material was extracted from the organic component of the sediment samples. The samples were then treated with hot 2% NaOH solution, and it was found that the NaOH solutions remained only slightly coloured, indicating that there was very little NaOH soluble material in the sediment. The NaOH liquor was then separated by filtration, and the residue was treated with hot 10% HC1, rinsed and dried. CO2 was obtained from portion of the NaOH insoluble component of each sample using the standard sealed tube technique with CuO. The procedure used for preparation of sample graphite for 14C AMS determinations is as described by Fifield et al., [7]. The AMS determinations were carried out at the ANT ARES AMS Facility at the Australian Nuclear Science and Technology Organisation (ANSTO), Lucas Heights, NSW. Stable isotope determinations were carried out on portion of the CO2 samples prepared for AMS graphite target preparation. The samples were analysed using a Europa Scientific 20-20 mass spectrometer. 3. RESULTS Figure 2 indicates a plot of organic and inorganic paired 14C ages using the AMS technique and facility at ANSTO. It shows that the inorganic carbon results form a reasonable sequence with depth, but the organic fraction ages actually become younger with depth down to 136 cm (107% M), then become older again. The shape of the organic age plot with depth can be compared with a typical concentration plot of a compound moving through a chromatographic column, having a relatively sharp leading edge and a long tail. From this observation there is a strong possibility that young (post atmospheric nuclear testing) non- polar organic material is moving down the profile in a similar manner to a chemical compound moving through a chromatographic column. At this stage no analytical work has 228 been carried out to identify the contaminant. Table 1 lists 513C values (with respect to the PDB standard) of the organic fractions with 14C ages. Gulf of Carpentaria GC-2 AMS + conventional analyses o *^~—bulk carbonates -20- -40- iogenic carbonate -60- I -80- •£ -100 - ft*Or JSi - -140- -160- B -180- \ games -200 0 10000 20000 30000 40000 50000 radiocarbon age (uncalibrated) FIG. 2. Plot of all the organic as well as inorganic paired results for all the samples prepared and analysed for AMS. TABLE I. 8I3C VALUES AND 14C AGES OF ORGANIC FRACTIONS FROM LAKE CARPENTARIA. ANSTO No. Submitter's Code 813C (%o) 14CAge OZB203 GC2 57-58 -17.29 ± 0.2 4020 ± 80 OZB202 GC2 66-67 -13.39 + 0.2 2820 ± 70 OZB204 GC2 82-83 -11.69 ±0.2 2790 ± 50 OZB205 GC2 96-97 -13.22 ± 0.2 1980 ± 40 OZB322 GC2 127-128 -15.98 ±0.2 107.0 ±1.0 OZB201 GC2 155-156 -15.79 ±0.2 6150 ±60 OZB199 GC2 174-175 -20.83 ± 0.2 14,800 ± 530 At this stage, given that each fraction consists of a young component plus a much older, possibly original component, there seems to be no apparent pattern in the results.