Sutikna T.,University of Wollongong |
Tocheri M.W.,Lakehead University |
Tocheri M.W.,Smithsonian Institution |
Morwood M.J.,University of Wollongong |
And 24 more authors.
Nature | Year: 2016
Homo floresiensis, a primitive hominin species discovered in Late Pleistocene sediments at Liang Bua (Flores, Indonesia), has generated wide interest and scientific debate. A major reason this taxon is controversial is because the H. floresiensis-bearing deposits, which include associated stone artefacts and remains of other extinct endemic fauna, were dated to between about 95 and 12 thousand calendar years (kyr) ago. These ages suggested that H. floresiensis survived until long after modern humans reached Australia by ∼50 kyr ago. Here we report new stratigraphic and chronological evidence from Liang Bua that does not support the ages inferred previously for the H. floresiensis holotype (LB1), ∼18 thousand calibrated radiocarbon years before present (kyr cal. bp), or the time of last appearance of this species (about 17 or 13-11 kyr cal. bp). Instead, the skeletal remains of H. floresiensis and the deposits containing them are dated to between about 100 and 60 kyr ago, whereas stone artefacts attributable to this species range from about 190 to 50 kyr in age. Whether H. floresiensis survived after 50 kyr ago - potentially encountering modern humans on Flores or other hominins dispersing through southeast Asia, such as Denisovans - is an open question. © 2016 Macmillan Publishers Limited. All rights reserved. Source
The 2007–2014 excavations at Liang Bua, including eight 2 × 2 m and two 1 × 2 m areas (referred to as Sectors), proceeded in 10 cm intervals (referred to as spits) while following stratigraphic units. Timber shoring of the baulks was installed after ~2.5 m depth for safety. In situ findings (for example, bones, artefacts and charcoal) were plotted in three dimensions and the sediments from each spit were sieved by hand, followed by wet sieving (2 mm mesh). All recovered findings were cleaned, catalogued and transported to Pusat Penelitian Arkeologi Nasional (Jakarta, Indonesia) for curation and further study. Bulk samples of four tephras were wet sieved at >32 μm and then dry sieved into 32–63 μm, 63–125 μm, 125–250 μm, 250–500 μm and 500 μm–1 mm fractions. Depending on the grain-size distribution of the sample, the dominant dry fraction containing the most glass material was mounted into an epoxy resin. Glass from bubble-wall shards or vesiculated pumice fragments was analysed as individual grains (>63 μm) using an electron microprobe. Major element determinations were made using a JEOL Superprobe (JXA-8230) housed at Victoria University of Wellington, New Zealand, using the ZAF correction method31. Analyses were performed with 15 kV accelerating voltage, 8 nA beam current, and an electron beam defocused to 10–20 μm. Standardization was achieved by means of mineral and glass standards. Rhyolitic glass standard ATHO-G (ref. 32) was routinely used to monitor calibration in all analytical runs and to evaluate any day-to-day differences in calibration. All analyses were normalized to 100% (by weight) anhydrous and total iron was calculated as FeO (Supplementary Information section 1). The major element compositions of the four tephras and the Youngest Toba Tuff33 are displayed as bivariate plots in Extended Data Fig. 3c–e. Sediment samples were collected using opaque plastic tubes hammered horizontally into cleaned stratigraphic sections and wrapped in black plastic after removal. Field measurements of the gamma dose rate at each sample location were made using a portable gamma-ray detector (Exploranium GR-320) inserted into the emptied tube holes. Sediment samples were also collected from the tube holes for laboratory determinations of water content and beta dose rate at the University of Wollongong. The environmental dose rate of each sample was calculated as the sum of the beta dose rate (estimated from beta-counting of dried and powdered sediment samples using a Risø low-level beta multicounter system and allowing for beta-dose attenuation and other factors34), the in situ gamma dose rate, and the estimated cosmic-ray dose rate35. The latter took account of the burial depth of each sample, the thickness of cave roof overhead, the zenith-angle dependence of cosmic rays, and the latitude, longitude and altitude of Liang Bua. Each of these external dose rate contributors were adjusted for long-term water content. The measured (field) water contents of the seven samples collected in 2012 range from 26 to 35%, but higher and lower field water contents have been reported for other sediment samples collected at Liang Bua: 15–30% (ref. 2), 3–22% (ref. 7) and 24–39% (Supplementary Information section 2). We used a value of 20 ± 5% as a mid-range estimate of the long-term water content, with the standard error sufficient to cover (at 2σ) most of the field measurements. The total dose rate of each sample also includes an estimate of the internal beta dose rate due to the radioactive decay of 40K and 87Rb. This estimate was based on a K concentration of 12 ± 1%, determined from energy- and wavelength-dispersive X-ray spectroscopic measurements of individual feldspar grains36, 37 (see Supplementary Information section 2), and an assumed Rb concentration of 400 ± 100 μg g–1 (ref. 38). Potassium-rich feldspar (K-feldspar) grains of 90–212 or 180–212 μm diameter were extracted from the sediment samples under dim red illumination and prepared using standard procedures39, including the use of a sodium polytungstate solution of 2.58 g cm–3 density to separate the feldspar grains from heavier minerals. The separated grains were also washed in 10% hydrofluoric acid for 40 min to clean their surfaces and reduce the thickness of the alpha-irradiated outer layer by ~15 μm, which was taken into account in the dose rate calculation. For each sample, single aliquots composed of a few hundred grains were measured using an automated Risø TL/OSL reader equipped with infrared (875 nm) light-emitted diodes (LEDs) for stimulation40 and a calibrated 90Sr/90Y source for beta irradiations. We also made measurements of individual K-feldspar grains (180–212 μm diameter) from two samples collected in 2012 (LB12-OSL3 and -OSL4) and the two original sediment samples (LBS7-40a and LBS7-42a) collected from above and alongside the remains of LB1 (ref. 2). Infrared stimulation of individual K-feldspar grains was achieved using a focused laser beam (830 nm)40. The IRSL emissions from the single aliquots and single grains were detected using an Electron Tubes Ltd 9235B photomultiplier tube fitted with Schott BG-39 and Corning 7-59 filters to transmit wavelengths of 320–480 nm. To estimate the equivalent dose (D ) for each aliquot or grain39, 41, we initially used the multiple elevated temperature post-infrared IRSL (MET-pIRIR) regenerative-dose procedure21, 30, 42, in which the IRSL signals are measured by progressively increasing the stimulation temperature from 50 °C to 250 °C in 50 °C increments. The samples yielded very dim signals, however, so we used a two-step pIRIR procedure21, 43 to improve the signal-to-noise ratio. Grains were preheated at 320 °C for 60 s before infrared stimulation of the natural, regenerative and test doses at 50 °C for 200 s. The pIRIR signal was then measured at 290 °C for either 200 s (single aliquots) or 1 s (single grains), followed by an infrared bleach at 325 °C for 100 s at the end of each regenerative-dose cycle. Example pIRIR decay and dose–response curves are shown in Extended Data Fig. 7a–c. Performance tests of the regenerative-dose procedure and details of residual dose measurements21, 44 and anomalous fading tests45, 46 are described in Supplementary Information section 2. We estimated the pIRIR age of each sample from the weighted mean D (calculated using the central age model47, 48) divided by the environmental dose rate, and applied corrections for residual dose and anomalous fading. For the latter corrections, we used the weighted mean fading rate (0.9 ± 0.3% per decade) measured for five of the seven samples collected in 2012. Sediment samples were taken in the same manner as the IRSL samples, as were the in situ gamma dose rate measurements (made using an ORTEC digi-DART gamma-ray detector). Additional sediment samples were also collected for determinations of water content and beta dose rate at Macquarie University. The environmental dose rate of each sample consists of three external components, each adjusted for long-term water content (20 ± 5%): the beta dose rate (estimated from beta-counting and allowing for beta-dose attenuation), the in situ gamma dose rate, and the estimated cosmic-ray dose rate. An assumed internal dose rate of 0.03 ± 0.01 Gy kyr–1 was also included in the total dose rate. Quartz grains were separated from the sediment samples under dim red illumination and prepared using standard procedures39, including mineral separations using 2.70 and 2.62 g cm–3 density solutions of sodium polytungstate to isolate the quartz and a 40% hydrofluoric acid etch for 45 min to remove the alpha-irradiated outer ~20 μm of each grain. Aliquots composed of about 10,000 quartz grains were measured using a dual-aliquot regenerative-dose TL protocol22 to isolate the light-sensitive red emissions49. This procedure was originally developed for application at Liang Bua2 and requires two aliquots of each sample: one to estimate the D associated with the heat-reset TL traps and the other to measure the total TL signal, from which the D associated with the light-sensitive TL traps is estimated by subtraction. Measurements were made on a Risø TL/OSL reader fitted with an Electron Tubes Ltd 9658B photomultiplier tube and Schott BG-39 and Kopp 2-63 filters designed to transmit red emissions (peak transmission in the 600–620 nm range, with minimum and maximum wavelengths of 580 and 670 nm), and cooled to –22 °C to reduce the background count rate. Bleaching was performed using a halogen lamp and light guide integrated into the reader, and laboratory doses were given using a 90Sr/90Y source mounted on the reader. The quartz grains were first heated to 260 °C at a heating rate of 5 K s–1 and then held at this temperature for 1,000 s to induce an isothermal TL signal and reduce the unwanted glow from incandescence. Following ref. 22, D values were estimated from the 20–30 s interval of isothermal TL (which was shown to be light-sensitive) and the final 160 s was used as background. Two of the samples contained abundant quantities of quartz, so problems associated with inter-aliquot variability50, 51 were reduced by measuring 12 replicates of each pair of aliquots (Supplementary Information section 2). As prolonged sunlight exposure is required the empty the light-sensitive TL traps, the ages obtained should be regarded as maximum estimates of the time elapsed since sediment deposition22. Samples of bone from three specimens of Homo floresiensis (LB1, LB2 and LB6), one Homo sapiens and eight Stegodon florensis insularis were analysed using laser ablation uranium-series isotope measurement analysis procedures and instruments similar to those described previously23, 52. Cuts were made perpendicular to the bone surface using a rotatory tool equipped with a thin (100 μm wide) diamond saw blade. The cut samples were mounted into aluminium cups, aligning the cross-sectioned surfaces with the outer rim of the sample holder to position the samples on the focal plane of the laser in the sampling cell. Sequential laser spot analyses were undertaken along 1–5 parallel tracks per sample, starting from the interior of each cross-sectioned bone (Extended Data Figs 8 and 9 and Supplementary Information section 3). Uranium and thorium isotopes were measured at the Australian National University using a Finnigan MAT Neptune multi-collector inductively-coupled plasma mass spectrometer (MC-ICP-MS) equipped with multiple Faraday cups and ion counters. Two ion counters were set to masses of 230.1 and 234.1, while the Faraday cups measured the masses 232, 235 and 238. Samples were ablated using a Lambda Physik LPFPro ArF excimer laser (193 nm) coupled to the MC-ICP-MS through a custom-designed Helex ablation cell. The analyses were made at regular spacing (typically 2–3 mm) along each track, using a laser spot size of 265 μm and a 5 Hz pulse rate. The samples were initially cleaned for 5 s and ablation pits were measured for 50 s. These measurements were bracketed by analyses of reference standards to correct for instrument drift. Semi-quantitative estimates of uranium and thorium concentrations were made from repeated measurements of the SRM NIST-610 glass standard (U: 461.5 μg g–1, Th: 457.2 μg g–1) and uranium-isotope (activity) ratios from repeated measurements of rhinoceros tooth dentine from Hexian sample 1118 (ref. 53). Apparent 234U/230Th ages were estimated for each track using a diffusion–adsorption–decay (DAD) model24, which uses the entire set of isotope measurements made along the track to calculate the rate of diffusion of 238U and 234U into the bone. Bones behave as geochemically ‘open’ systems, so the diffusion of uranium into a bone may have occurred soon after it was deposited or much later. The original isotopic signature may also be overprinted by secondary uranium uptake that is hard to recognize. As such, uranium-series ages for bones are most likely to be minimum estimates of time since deposition, with the extent of age underestimation being very difficult or impossible to evaluate23. The modelled ages were calculated after rejecting data points associated with detrital contamination (U/Th elemental ratios of ≤ 300) and data points at the surface of the bones where secondary overprinting was suspected. Results are tabulated in Supplementary Information section 3, with all errors given at 2σ. Samples of calcite deposited as speleothems (flowstones and a 1 cm-tall stalagmite) were collected during excavation. Unlike bone, speleothems commonly behave as geochemically ‘closed’ systems, with no loss or gain of uranium after calcite crystallization54. The cleanest portion of each sample was selected, ground to the size of a rice grain, cleaned ultrasonically and then handpicked, avoiding pieces that appeared to be porous or contaminated. Age determinations were made at the University of Queensland using a Nu Plasma MC-ICP-MS. Uranium and thorium separation procedures, MC-ICP-MS analytical protocols, details of spike calibrations, blank and ‘memory’ assessments, and repeat measurements of standards have been described previously55, 56. The 234U/230Th ages were calculated using Isoplot 3.75 (ref. 57) and half-lives of 245.25 kyr (234U) and 75.69 kyr (230Th) (ref. 58). As most samples consist of impure calcite with some degree of detrital contamination, a correction for non-radiogenic 230Th was applied to the measured 230Th/232Th activity ratio of each sample using an assumed bulk-Earth 230Th/232Th activity ratio of 0.825 (with a relative error of ± 50% and assuming secular equilibrium in the 238U–234U–230Th decay chain), as is typical of most other studies59. This non-radiogenic 230Th correction reduces the calculated ages of the samples and increases the age uncertainties by an amount dependent on the extent of detrital contamination. All ten dated samples have measured 230Th/232Th activity ratios of >20 (Supplementary Information section 4), so the detritally-corrected 234U/230Th ages are only slightly younger than the uncorrected (measured) ages. Crystals of hornblende were obtained for the 100–150 and 150–250 μm size fractions of the T1 tephra using standard heavy liquid and magnetic separation techniques. Crystals were loaded into wells in 18 mm-diameter aluminium sample discs for neutron irradiation, along with the 1.185 million year-old Alder Creek sanidine26 as the neutron fluence monitor. Neutron irradiation was carried out for 4 min in the cadmium-shielded CLICIT facility at the Oregon State University TRIGA reactor. Argon isotopic analyses of gas released from 6 hornblende aliquots during CO laser step-heating, including a final fusion step (‘fuse’ in Supplementary Information section 5), were made on a fully automated, high-resolution, Nu Instruments Noblesse multi-collector noble-gas mass spectrometer, using procedures documented previously25. One set of four experiments (Lab IDs 2598 and 2599 in Supplementary Information section 5) consisted of strong degassing of 10 mg hornblende aliquots in square pits in the laser disc, using a beam integrator lens that gives a ‘top hat’ 6 × 6 mm energy profile at the focal plane. The laser was then operated at 32 W and this high-temperature, pre-fusion step measured (‘D’ in Supplementary Information section 5). The fusion step was performed using a conventional focus lens. In the second set of experiments, 20 mg hornblende aliquots were loaded into 5 mm-wide channels in the laser disc and a defocused laser beam was programmed to raster in 50 traverses, each 30 mm in length and separated by 0.1 mm. Low power (<1 W) steps were performed initially to remove loosely trapped argon. At 1 W of laser power (step ‘C’), the hornblende crystals began to glow and release significant amounts of 39Ar. This and subsequent steps, including the final fusion, are reported in Supplementary Information section 5 as Lab IDs 2584-03C to -03O (fuse) and 2584-04C to -04I (fuse). Sample gas clean-up was through an all-metal extraction line, equipped with a –130 °C cold trap (to remove H O) and two water-cooled SAES GP-50 getter pumps (to absorb reactive gases). Argon isotopic analyses of unknowns, blanks and monitor minerals were carried out in identical fashion during a fixed period of 400 s in 14 data acquisition cycles. 40Ar and 39Ar were measured on the high-mass ion counter, 38Ar and 37Ar on the axial ion counter and 36Ar on the low-mass ion counter, with baselines measured every third cycle. Measurement of the 40Ar, 38Ar and 36Ar ion beams was carried out simultaneously, followed by sequential measurement of 39Ar and 37Ar. Beam switching was achieved by varying the field of the mass spectrometer magnet and with minor adjustment of the quad lenses. Data acquisition and reduction was performed using the program ‘Mass Spec’ (A. Deino, Berkeley Geochronology Center). Detector intercalibration and mass fractionation corrections were made using the weighted mean of a time series of measured atmospheric argon aliquots delivered from a calibrated air pipette25. The accuracy of the primary air pipette measurements was verified by cross-referencing to data produced from a newly charged second air pipette. Decay and other constants, including correction factors for interference isotopes produced by nucleogenic reactions, are as reported in ref. 25. An isotope correlation (inverse isochron) plot of the data for all 28 aliquots is shown in Extended Data Fig. 3f. The age determined from the inverse isochron is 85 ± 13 kyr for all 28 aliquots, or 79 ± 12 kyr (errors at 1σ) if the data point on the far right-hand side of the plot is excluded. In both cases, the 40Ar/36Ar intercept value is statistically indistinguishable from the atmospheric ratio of 298.6 ± 0.3 (ref. 60), indicating the absence of significant excess 40Ar in the hornblende crystals. Three charcoal samples recovered during excavation were sent to DirectAMS Radiocarbon Dating Service in Bothell, Washington (Supplementary Information section 6). Samples were pretreated using acid–base–acid (ABA) procedures and the 14C content was measured using accelerator mass spectrometry61. Conventional 14C ages in radiocarbon years before present (bp) were converted to calendar-year age ranges at the 68% and 95% confidence intervals using the SHCal13 calibration data set62 and CALIB 7.1 (http://calib.qub.ac.uk/calib/). One of these three samples yielded an ‘old’ radiocarbon age (>40 kyr bp), so we submitted the remaining charcoal to the Oxford Radiocarbon Accelerator Unit (ORAU) for a harsher cleaning protocol known as ABOx-SC63, 64. This pretreatment is known to improve charcoal decontamination and has been shown repeatedly to produce more reliable results for ‘old’ charcoal. However, the harshness of the ABOx-SC procedure often results in large material loss and sample failure, so a new preparative method (AOx-SC) has been developed and tested at the ORAU (K. Douka, personal communication). The AOx-SC procedure does not include a NaOH step and produces identical results to ABOx-SC, but with much higher yields and reduced sample failure. For the Liang Bua sample, ~100 mg and ~50 mg of hand-picked charcoal underwent ABOx-SC and AOx-SC pretreatments, respectively. Only the AOx-treated charcoal survived the wet chemistry, yielding sufficient carbon for stepped combustion, first at 630 °C and then at 1,000 °C, with the latter fraction collected for graphitization and measurement by accelerator mass spectrometry.
Ali M.S.M.,National University of Malaysia |
Ramli Z.,National University of Malaysia |
Rahman N.H.S.N.A.,National University of Malaysia |
Samian A.L.,National University of Malaysia |
And 2 more authors.
Journal of Food, Agriculture and Environment | Year: 2015
The aim of this study was to determine whether the ancient bricks from Candi SEG II (Lempeng) are made from local raw material or otherwise. Candi SEG II located in cultivation area has unearthed various interesting artefacts like fragment of pottery, rouletted pottery, glass beads, animal bones, skeletons and inscriptions with Palava characters. The main construction materials used to build the temple consisted of bricks and limestone was used as a mortar. The upper part of the temple is believed to have been built using wooden structures and the roof used palm leaves. Scientific analysis on the bricks shows that local raw material was used to produce these bricks. Scientific analysis using the X-ray fluorescence technique and X-ray diffraction technique can determine the chemical composition of the bricks, among others the mineral content of the bricks as well as the major element and trace element contents. The usage of local raw material also demonstrated the local wisdom in temple construction technology and also technique in producing bricks that had existed. © 2015, World Food Ltd. and WFL Publishers. All rights reserved. Source