Evaluating isothermal thermoluminescence and thermally transferred optically stimulated luminescence for dating of Pleistocene sediments in Amazonia

Quaternary Geochronology 36 (2016) 28e37

Contents lists available at ScienceDirect

Quaternary Geochronology journal homepage: www.elsevier.com/locate/quageo

Evaluating isothermal thermoluminescence and thermally transferred optically stimulated luminescence for dating of Pleistocene sediments in Amazonia
Oliveira Sawakuchi, Thays Desire Mineli, Fabiano do Nascimento Pupim*, Andre Luciana Nogueira ~o Paulo, Rua do Lago, 562 e Sa ~o Paulo, SP, 05508-080, Brazil Institute of Geosciences, University of Sa

a r t i c l e i n f o

a b s t r a c t

Article history: Received 8 January 2016 Received in revised form 1 August 2016 Accepted 8 August 2016 Available online 9 August 2016

The paleogeography of Amazonia lowlands during the Pleistocene remains hampered by the lack of reliable absolute ages to constrain sediment deposition in the hundred thousand to few million years timescales. Optically stimulated luminescence (OSL) dating applied to quartz has provided important chronological control for late Quaternary sediments, but the method is limited to the last ~150 ka. In order to extend the age range of luminescence dating, new signals from quartz have been investigated. This study tested the application of isothermal thermoluminescence (ITL) and thermally transferred optically stimulated luminescence (TT-OSL) signals of quartz for dating of fluvial terraces from eastern Amazonia. ITL and TT-OSL signals measured in a modern fluvial sediment sample have shown small residual doses (4 and 16 Gy), suggesting adequate bleached sediments for the target dose range (>150 Gy). This sample responded well to dose recovery test, which showed that the ITL and TT-OSL signals grow to higher doses compared to the doses estimated by the conventional OSL signal. The ITL signal saturated for doses significantly lower than doses reported in the literature. Most dating samples were beyond the ITL saturation doses and only TT-OSL signals were suitable to estimate equivalent doses. Burial ages ranging from 107 to 340 ka were estimated for the fluvial terraces in the lower Xingu River. The main ages uncertainties are related to dose rate changes through time. Despite the uncertainties, these ages should indicate a higher channel base level during the Middle Pleistocene followed by channel incision, possibly due to episodes of increased precipitation in the Xingu watershed. © 2016 Elsevier B.V. All rights reserved.

Keywords: Luminescence dating TT-OSL ITL Amazonia paleogeography

1. Introduction The modern Amazonia landscape and biodiversity are strongly related with the evolution of the Amazonian fluvial system during the Neogene and Quaternary (Hoorn and Wesselingh, 2010; Ribas et al., 2011). However, our understanding on the paleogeography of Amazonian rivers is restricted to the last few hundred thousand years (Soares et al., 2010; Rossetti et al., 2015) due to the lack of reliable burial ages to constrain changes in fluvial sedimentation during the Middle to Early Pleistocene. Optically stimulated luminescence (OSL) dating based on the fast OSL component of quartz (Murray and Wintle, 2000) has provided important chronological control for late

* Corresponding author. E-mail address: [email protected] (F.N. Pupim). http://dx.doi.org/10.1016/j.quageo.2016.08.003 1871-1014/© 2016 Elsevier B.V. All rights reserved.

Quaternary sedimentation, but the method is usually limited to the last 100e200 ka due to OSL signal saturation (Wintle and Murray, 2006). The luminescence dating of potassium feldspar, using post-infrared-infrared (pIRIR) signals, has been shown to be suitable to estimate equivalent doses up to 1 kGy. However, its intrinsic higher dose rate limits the age range to around 500 ka (Buylaert et al., 2012). In order to extend the age range of luminescence dating, new signals from quartz have been investigated. Isothermal thermoluminescence (ITL) (Jain et al., 2007) and thermally transferred optically stimulated luminescence (TTOSL) (Wang et al., 2006) are the most promising signals to extend the age limit of luminescence dating of quartz. The origin of the ITL signal has been linked to the 325 and 375
C TL peaks of quartz (Aitken, 1985; Jain et al., 2007). It is a thermally stable signal, relatively easy to optically reset when exposed to light and its saturation occurs for higher doses than for the OSL signal. The TT-OSL signal is composed of a signal dependent on the

F.N. Pupim et al. / Quaternary Geochronology 36 (2016) 28e37

radiation dose (recuperated OSL e Re-OSL) and a thermally transferred signal independent of the accumulated dose (basic transfer BT-OSL) (Wang et al., 2006). The source of charge for the TT-OSL signal is still not well characterized, but two mechanisms have been suggested: a double transfer (Wang et al., 2006) and a single transfer (Adamiec et al., 2008). Adamiec et al. (2010) described that in the double transfer mechanism, the electrons from the trap responsible for the fast OSL signal are released and stored in a refuge trap before being thermally transferred back into the fast OSL trap for subsequent measurement as TT-OSL; and in the single transfer mechanism, the electrons are thermally transferred from a source trap into the fast OSL trap and then measured as TT-OSL. The single aliquot regenerative dose (SAR) protocol using ITL and TTOSL signals of quartz has been tested through laboratory experiments (e.g. Jain et al., 2005; Huot et al., 2006; Stevens et al., 2009; Adamiec et al., 2010). So far, few studies have tested the dating of sediments using ITL (e.g. Choi et al., 2006; Barham et al., 2011) or TT-OSL (e.g. Jacobs et al., 2011; Thiel et al., 2012; Demuro et al., 2015) signals. The results show that both signals have potential to dating sediments older than 1 Ma (Wang et al., 2006; Jain et al., 2007). Besides the specific geomorphological questions related to dating of Amazonian sediments, the characterization of ITL and TTOSL signals of quartz from different geological contexts is necessary to test and develop protocols to extend the age range of luminescence dating. Brazilian sediments are commonly heavily weathered and enriched in quartz with high luminescence sensitivity (Guedes et al., 2013) and buried in relatively low dose rates environments (~0.5e1.0 Gy/ka). The first aim of this study was to test the application of dating protocols based on ITL and TT-OSL signals to extend the age limit for dating of fluvial sediments from eastern Amazonia. Moreover, this is a critical issue to understand changes in the Amazonian fluvial system during the Quaternary and its role as biogeographical barriers for terrestrial fauna or pathways for migration of aquatic or flooded-forest species.

2. Material and methods 2.1. Study site and samples This study focused on fluvial sediments outcropping in terraces of the lower Xingu River, eastern Amazonia. The Xingu River is a clear water river (low concentration of suspended sediments; Sioli, 1985) representing the easternmost tributary of the southern margin of the Amazon River, Brazil (Fig. 1). The lower Xingu River has a modern fluvial channel with approximately 10 km wide and 15e25 m deep. Through recent time, the river has incised by about 3e10 m into its older floodplain, leaving its former fluvial sediments exposed in terraces. Terrace sediments are mainly composed of gravel, coarse to fine sand and mud, suggesting fining upward sequences formed by change from bottom channel (sandy gravel) to sand bar and floodplain muds, prior to renewed downcutting. Muddy portions would be related to overbank deposits, despite the possibility of authigenic minerals formation and clay or fine silt infiltration during long-time tropical weathering. Eight samples were taken from two sediment profiles in the left margin of the lower Xingu River (Fig. 1). The samples for luminescence measurements were collected in aluminum tubes in order to avoid sunlight exposure. Sediments within a radius of 30 cm from the luminescence sampling point were collected separately to estimate radiation dose rates. Additionally, a sample of shallow buried sands (XNG-47-2) from a fluvial bar was used as a modern analogue to evaluate residual ITL and TT-OSL signals.

29

2.2. Sample preparation The preparation of quartz concentrates were conducted under subdued red light conditions. Samples were wet-sieved to isolate the fraction 180e250 mm. Afterward, the target fraction was treated with hydrogen peroxide (H2O2 27%) and hydrochloric acid (HCl 3.7%) to remove organic material and carbonates, respectively. Lithium metatungstate solution was used to remove heavy minerals (>2.75 g/cm3) and feldspar grains (<2.62 g/cm3). Then, quartz concentrates (2.62e2.75 g/cm3) were etched with hydrofluoric acid (HF 40%) for 40 min to further purify the quartz and also etch a ~10 mm outer layer of quartz grains, thus removing the alpha particles contribution from the dose rate. 2.3. Dose rate calculation Radionuclides concentrations for dose rate calculations were determined by gamma ray spectrometry using a high-purity germanium (HPGe) detector with a 55% of relative efficiency, 2.1 keV of energy resolution at 1332 keV and encased in an ultralow background shield (Canberra Industries). The samples for gamma ray spectrometry were weighed and dried to estimate the water content. Each sample was packed in sealed plastic containers and stored for at least 28 days to allow radon to reach equilibrium with its parent radionuclides, prior to gamma spectrometry measurement. Beta and gamma radiation dose rates were determined using radionuclides concentrations (U, Th and K) and conversion factors
rin et al. (2011). The cosmic dose rate contribution outlined by Gue was calculated following Prescott and Hutton (1994). Water saturation was determined by the ratio between water weight and dry sample weight. Considering that the samples are under the influence of changes in water table, the effect of water content over dose rate was investigated through dose rate calculation using a minimum and maximum water content assuming constant concentrations of radionuclides. The water content measured in the collected samples was considered as minimum water content estimate since the samples were collected during the dry season. For estimation of maximum water content, each pre-dried sample was water saturated in the laboratory, until we observed that all pore spaces were occupied by water. The samples were weighed both before and after water saturation, the difference was assumed to the maximum water content for each the sample. 2.4. Luminescence measurements Aliquots of quartz grains (100e200 grains) were mounted on stainless steel cups. All luminescence measurements were carried out on an automated Risø TL/OSL DA-20 reader system, equipped with a 90Sr/90Y beta source delivering a dose rate of 0.11 ± 0.03 Gy/s and blue LEDs (470 ± 20 nm) operated at 90% power (~40 mW/cm) for stimulation. The near UV emissions were measured with a bialkali PM tube (Thorn EMI 9635QB) coupled with Hoya U-340 detection filters (290e340 nm). The calibration quartz distributed by Risø National Laboratory (Hansen et al., 2015) was used for the calibration of the beta source using stainless steel cups at the same measurements conditions used in this study. Heating rate of 5
C/s was used in all heat treatments and luminescence measurements. The ITL and TT-OSL signals of quartz are relatively hard to bleach compared to the OSL signal (Jain et al., 2007; Duller and Wintle, 2012). Studied sediments from the modern fluvial bar and terraces are characterized by bright quartz grains with rapidly decaying OSL signal (first 0.8s of light emission). Thus, the sample (XNG-47-2) collected from the modern sediment bar at the Xingu River was used to evaluate both its luminescence (OSL, ITL and TTOSL) bleaching characteristics and its dosimetric performance with

30

F.N. Pupim et al. / Quaternary Geochronology 36 (2016) 28e37

Fig. 1. A) Digital elevation model highlighting the sampling sites (red circles) along a fluvial terrace at the western margin of the Xingu River. Image from Shuttle Radar Topography Mission (SRTM 3 arc-seconds). B) Sedimentary facies succession of the sampling profiles. Arrows and numbers are indicating each sampling location and TT-OSL age estimation. C) Sampling profile ALC-01, the top is 15 m.a.s.l.. D) Sampling profile ALC-02, the top is 6 m.a.s.l. (meters above sea level - m.a.s.l.). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

a dose recovery test. The use of modern samples is an advantage for dose recovery experiments using ITL and TT-OSL because there is no need to apply laboratory procedures like long time artificial bleaching before delivering the initial laboratory dose, what might induce adverse sensitivity changes (Huot et al., 2006; Tsukamoto et al., 2008; Hernandez et al., 2012). Also, modern samples allow to check the bleaching capacity in the studied sedimentary system.

Table 1 SAR-OSL protocol used for equivalent dose estimation of the studied samples. Step

Procedure

1 2 3 4 5 6 7 8

Dose (Di)a Preheat at 200
C for10s OSL at 125
C for 40s (Li) Test dose Heat at 160
C OSL at 125
C for 40s (Ti) Blue LED bleach at 280
C for 40s Return to step 1

a For the natural sample, Di ¼ 0 Gy; regeneration doses Di: D1>D2>D3>D4; D5 ¼ 0 Gy, D6 ¼ D1; D7 ¼ D6, with additional infrared stimulation before blue LED stimulation for OSL measurement of D7. The test dose kept constant throughout the SAR sequence. A corrected luminescence signal was calculated through the ratio between Li and Ti. OSL signal was calculated using the integral of the first 0.8s of light emission with subtraction of the normalized last 10 s s as background.

The SAR protocol using the OSL signal (Table 1; Murray and Wintle, 2000) was applied to determine the natural dose and burial age of the sample XNG-47-2. Artificial bleaching was performed only in aliquots used in the dose recovery test for the OSL signal of sample XNG-47-2. These aliquots were bleached in a solar simulator for 2e3 h. For the OSL signal, the dose recovery test was performed for a given dose of 33 Gy and preheat temperature of 200
C. ITL protocols proposed by Jain et al. (2007) and Vandenberghe et al. (2009) and TT-OSL protocols proposed by Wang et al. (2006) and Adamiec et al. (2010) (Table 2) were tested with a dose recovery test using six aliquots per protocol and given dose of 250 Gy. Recycling ratio and recuperation tests were used to evaluate data quality of equivalent dose (De) measurements, with tolerance of 0.90e1.10 and < 5%, respectively, as suggested by Murray and Wintle (2000, 2003). For all aliquots used in luminescence measurements, the purity of the quartz was checked by infrared depletion ratio (Duller, 2003), in which all measured aliquots showed ratio within 10% of unity, indicating insignificant contribution from feldspar signals. The saturation characteristics of each luminescence protocol were compared with their respective characteristic dose (D0), with the aid of a single saturating exponential growth curve equation (Murray and Wintle, 2003). The De of all studied samples was estimated using OSL to check signal saturation. If it was saturated, the ITL and TT-OSL protocols with the best performance in dose recovery tests were used for De

F.N. Pupim et al. / Quaternary Geochronology 36 (2016) 28e37

31

Table 2 ITL and TT-OSL dating protocols used for equivalent dose estimation of the studied samples. Step

ITL 310
C (Jain et al., 2007)a

ITL 270
C (Vandenberghe et al., 2009)b

TT-OSL 1 (Adamiec et al., 2010)c

TT-OSL 2 (Wang et al., 2006)d

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

Dose (Di) Preheat at 260
C for 10s Heat at 310
C and hold for 500s (ITL-Li) Test dose (60 Gy) Preheat at 260
C for 10s Heat at 310
C and hold for 500s (ITL-Ti) Blue LED bleach at 280
C for 40s

Dose (Di) Preheat at 300
C for 10s Heat at 270
C and hold for 500s (ITL-Li) Test dose (60 Gy) Preheat at 300
C for 10s Heat at 270
C and hold for 500s (ITL-Ti) Blue LED bleach at 280
C for 40s

Dose (Di) Preheat at 260
C for 10s LM-OSL at 125
C for 200s Preheat at 260
C for 10s OSL at 125
C for 100s (LTT) Test dose (60 Gy) Preheat at 220
C for 10s OSL at 125
C for 100s (TTT) Heat at 350
C for 10s

Dose (Di) Preheat at 260
C for 10s OSL at 125
C for 300s Preheat at 260
C for 10s OSL at 125
C for 100s (LTT) Test dose (60 Gy) Preheat at 220
C for 10s OSL at 125
C for 100s (TTT) Heat (annealing) at 260
C for 10s OSL at 125
C for 100s Preheat at 260
C for 10s OSL at 125
C for 100s (LBT) Test dose (60 Gy) Preheat at 220
C for 10s OSL at 125
C for 100s (TBT)

ITL310
C signal was calculated by integrating of the first 5 s of light emission, and subtracting a background over the last 24 s. ITL270
C signal was calculated by integrating of the first 10 s of light emission, and subtracting a background over the last 50 s. c TT-OSL 1 signal was calculated by integrating of the first 1.75 s of light emission, and subtracting a background over the last 10 s. d TT-OSL 2 signal was calculated subtracting BT-OSL (steps 9 to 15) from TT-OSL signal (steps 1 to 8), while both were integrated over the first second of measurement and the background after 100 s removed. Regeneration doses Di: were 75 Gy (D1), 150 Gy (D2), 300 Gy (D3), 600 Gy (D4) and 1200 Gy (D5). Test dose was 60 Gy. A D6 dose of 0 Gy was used to evaluate recuperation and a D7 ¼ D1 was used to calculate the recycling ratio. a

b

estimation. De was calculated using the central age model (Galbraith et al., 1999).

3. Results 3.1. Dose recovery tests The OSL decay curve (Fig. 2) obtained from the modern sediments (XNG-47-2) shows a signal dominated by the fast OSL component (0.8 s corresponding to ~25% of light emission), indicating quartz with high sensitivity and suggesting a relatively long sedimentary history (Pietsch et al., 2008; Sawakuchi et al., 2011; Fitzsimmons, 2011; Lü et al., 2014). Dose recovery test showed

Fig. 2. Dose response curve and natural OSL decay curve (inset) for quartz aliquot of XNG-47-2 sample. The signal of the 0 Gy dose was used to fit the curve, since the low natural dose makes difficult the administration of a lower D1 regeneration dose. The linear dose response ensures the reliability of the natural dose estimate. Ln/Tn is the corrected luminescence signal for the equivalent radiation dose (De).

calculated-to-given dose ratio of 0.93 ± 0.02 for a given dose of 33 Gy (Table 3). Average values of recycling ratio and recuperation were 0.99 and 0.03%, respectively. The natural dose calculated for sample XNG-47-2 was 0.31 ± 0.01 Gy (24 aliquots), with De distribution showing overdispersion of around 10%, suggesting wellbleached sediment and absence of sediment mixing after deposition (Arnold and Roberts, 2009). The OSL age calculated for XNG47-2 sample is 204 ± 5 years, for a dose rate of 1.51 ± 0.02 Gy/ka. This young age is reliable, given that the sample is near the surface (0.35 m depth) and covered by an open riparian vegetation and an absence of significant soil development. Dose recovery tests using ITL and TT-OSL signals were carried out in quartz aliquots of sample XNG-47-2. The modern sediment sample shows low residual doses for all tested signals (OSL 0.31 ± 0.01 Gy; ITL 4 ± 1 Gy; TT-OSL 16 ± 6 Gy). Thus, these residual doses are negligible in a dose recovery test, where the given dose is 250 Gy. The ITL 310
C signal shows high intensity and relatively fast decay (Fig. 3A) compared to the ITL 270
C signal (Fig. 3B). Vandenberghe et al. (2009) already reported that ITL signal intensity is correlated with heating temperature, but even so the ITL 270
C of the studied samples is clearly distinguished from the background level. The TT-OSL signals for both tested protocols have a significant rapid decay to ~50% of the initial signal reached after the first second of stimulation. The signal continues to decay until it reaches a relatively constant level of 15e20% of the initial signal after ~3e5 s of stimulation (Fig. 3C and D). This suggests the dominance of a rapidly decaying luminescence signal during the first second of stimulation, which can be isolated by subtraction of the last part of the TT-OSL decay curve (background) (Tsukamoto et al., 2008; Jacobs et al., 2011). All ITL and TT-OSL signals grow with dose and their dose response curves are well described by fitting of a saturating exponential function (Fig. 3). In dose recovery tests, the average calculated-to-given dose ratios for six aliquots were 1.06 ± 0.07, 0.92 ± 0.05, 1.10 ± 0.02 and 0.90 ± 0.04 for the ITL 310
C, ITL 270
C, TT-OSL 1 and TT-OSL 2 signals, respectively (Table 3 and Fig. 1S). All these luminescence signals performed well (recycling ratio and recuperation tests), except for TT-OSL 2, which presented high recuperation (average of 30%). In addition, their 2D0 values were in the range reported in the literature (Wang et al., 2006; Jain et al., 2007). The ITL 310
C and TT-OSL 1 signals showed higher 2D0, with average values of

32

F.N. Pupim et al. / Quaternary Geochronology 36 (2016) 28e37

Table 3 Data summary of dose recovery tests for sample XNG-47-2.

Calculated/given dose Average 2D0 (Gy) Average recycling ratio Average recuperation (%)

OSL

ITL 310
C

ITL 270
C

TT-OSL 1

TT-OSL 2

0.93 ± 0.02 95 ± 0.5 0.99 0.1

1.06 ± 0.07 1335 ± 239 1.00 0.6

0.92 ± 0.05 696 ± 34 1.04 4.2

1.10 ± 0.02 1639 ± 162 1.00 0.3

0.90 ± 0.04 1112 ± 123 0.98 30.5

The given dose was 33Gy for OSL signal and 250 Gy for all ITL and TT-OSL signals. There were measured 9 aliquots for OSL and 6 aliquots for each ITL and TT-OSL protocols. Equivalent doses calculated through the central age model (Galbraith et al., 1999).

Fig. 3. Dose response curves and natural luminescence decay curves (inset) for the ITL 310
C (A), ITL 270
C (B), TT-OSL 1 (C) and TT-OSL 2 (D) signals measured in quartz aliquots of XNG-47-2 sample. The laboratory applied doses were 250 Gy. Ln/Tn is the corrected luminescence signal for the equivalent radiation dose (De).

1335 ± 239 Gy and 1639 ± 162 Gy, respectively (Table 3). Therefore, our experiments demonstrate that these ITL and TT-OSL protocols are suitable to recover radiation doses applied in the laboratory and that these fluvial sediments display favorable bleaching and luminescence characteristics for dating. 3.2. Dose rates The studied samples presented dose rates for quartz ranging from 1.49 ± 0.11 to 3.10 ± 0.24 Gy/ka (Table 4). These dose rates are significantly higher than the typical dose rates (1.0e1.5 Gy/ka) reported for sandy sediments from southern and northeastern Brazil (Guedes et al., 2013). The relatively higher dose rates are attributed

to the elevated content of fine silt and clay in the sediments from the Xingu River terraces. Similar dose rates were also reported for Pleistocene fluvial deposits along the Madeira River in southwestern Amazonia (Rossetti et al., 2015). For the studied terraces, significant uncertainties regarding the dose rates can be attributed to changes in water saturation and clay illuviation through time. Water saturation values measured from samples under field conditions (dry season) ranged from 1 to 12% (Table 4), while the maximum water saturation values simulated in the laboratory ranged from 25 to 60%. The maximum water saturation simulated in the laboratory would represent the wet season condition, when samples are below water table. Low values of water saturation were measured in samples rich in iron oxide cements. The field and

F.N. Pupim et al. / Quaternary Geochronology 36 (2016) 28e37

33

Table 4 Radionuclides concentrations and dose rate data for the studied samples. Total dose rate data are shown in Table 6. Sample

Elev. (m)

Depth (m)

Water (%)

U (ppm)

XNG-47-2 ALC-01-1 ALC-01-2 ALC-01-3 ALC-01-4 ALC-02-1 ALC-02-2 ALC-02-3 ALC-02-4

4 10 10 10 10 4 4 4 4

0.35 2.5 3.3 4.7 5.1 0.3 1.4 1.8 2.2

6±2 5±2 5±2 2±2 5±2 6±2 12 ± 2 2±2 1±2

1.87 1.84 1.21 1.30 3.23 2.81 3.21 2.33 1.65

± ± ± ± ± ± ± ± ±

0.01 0.08 0.05 0.05 0.12 0.10 0.16 0.08 0.06

Th (ppm)

K (%)

5.82 ± 0.01 16.19 ± 0.55 13.74 ± 0.45 12.64 ± 0.37 27.24 ± 0.75 12.47 ± 0.42 17.18 ± 0.74 14.19 ± 0.43 10.71 ± 0.33

0.57 0.18 0.14 0.17 0.43 0.33 0.69 0.43 0.18

maximum water saturation can induce changes in dose rates in the order of 18e32% (Table 1S).

± ± ± ± ± ± ± ± ±

Gamma dose rate (Gy/ka) 0.02 0.01 0.01 0.01 0.02 0.02 0.03 0.02 0.01

0.59 0.96 0.78 0.78 1.67 0.93 1.22 1.02 0.74

± ± ± ± ± ± ± ± ±

0.01 0.10 0.08 0.09 0.18 0.10 0.13 0.12 0.08

0.75 0.69 0.54 0.57 1.28 0.84 1.18 0.94 0.59

± ± ± ± ± ± ± ± ±

0.02 0.08 0.06 0.07 0.15 0.10 0.13 0.12 0.07

Cosmic dose rate (Gy/ka) 0.17 0.15 0.14 0.11 0.11 0.20 0.17 0.16 0.16

± ± ± ± ± ± ± ± ±

0.01 0.01 0.01 0.01 0.01 0.04 0.01 0.01 0.01

Table 5 Maximum doses (2D0) that can be estimated using the ITL 310
C and TT-OSL 1 signals measured in quartz aliquots from samples of profiles ALC-01 and ALC-02. Sample

3.3. Equivalent doses and burial ages The SAR-OSL protocol was used to estimate the De from Xingu Terrace samples, of which only one was below saturation, with equivalent dose suitable for measurement by the OSL protocol (ALC-02-1). All the others had a natural signal significantly higher than 2D0 signal. For the samples with saturated OSL signal, the ITL 310
C and TT-OSL 1 signals were used to estimate the natural doses. The ITL 310
C signal has 2D0 values up to 400 Gy for samples from ALC-01 and ALC-02 profiles, which is significantly lower than the 2D0 estimated for sample XNG-47-2 (Fig. 4A and Table 3). Thus, most quartz aliquots were saturated for the ITL 310
C, with calculated De higher than 2D0, hindering its use to estimate the natural doses of profiles ALC-01 and ALC-02. However, ITL 310
C allowed to obtain minimum equivalent doses (equal to 2D0) for the samples of profiles ALC-01 (Table 5). The TT-OSL1 signal shows higher saturation doses, with 2D0 higher than 1000 Gy. Considering the saturation of the ITL 310
C signal for most of the studied samples, the TT-OSL 1 protocol was used for natural doses estimation. The natural doses values calculated for the Xingu River terraces using the TT-OSL 1 protocol ranged from 275 ± 33 to 651 ± 27 Gy (Table 6). The recuperation values are less than 2% in all cases and most samples presented recycling ratios between 0.90 and 1.10 (Table 6). The only exceptions were samples ALC-02-2 and

Beta dose rate (Gy/ka)

ITL 310
C a

XNG-47-2 ALC-01-1 ALC-01-2 ALC-01-3 ALC-01-4 ALC-02-2 ALC-02-3 ALC-02-4 a b

TT-OSL 1

N

2D0 (Gy)

Na

2D0 (Gy)

2 2 2

1082 ± 34 369 ± 6 321 ± 5 N/Ab 382 ± 6 N/A N/A N/A

2 6 6 6 6 6 6 6

1710 1707 3490 3204 2880 2461 4745 1174

2

± ± ± ± ± ± ± ±

63 33 337 198 162 141 444 62

N is the number of aliquots. N/A ¼ not available.

ALC-02-4, which showed average recycling ratios of 1.25 ± 0.07 and 0.86 ± 0.07, respectively. TT-OSL De and dose rates provide the first attempt to determine the burial ages for Pleistocene fluvial terraces from eastern Amazonia. The ages of ALC-01 profile were 209 ± 28, 340 ± 38, 259 ± 25 and 210 ± 21 ka; and for ALC-02 profile were 1.6 ± 0.2, 107 ± 15, 234 ± 43 ka and 200 ± 55 ka, from top to bottom (Tables 4 and 5). The De estimated by ITL 310
C (~4 Gy) and TT-OSL-1 (~16 Gy) from modern sample yield apparent burial ages of 2.6 and 10.6 ka, respectively, which are older than the true burial age estimated by OSL (204 ± 5 years). These residual doses are relatively low and correspond around of 1.5e6% of the De values calculated for the

Fig. 4. Dose response curves and equivalent doses obtained for the ITL 310
C (A) and TT-OSL 1 (B) signals measured in quartz aliquots of ALC-01-4 sample. Ln/Tn is the corrected luminescence signal for the equivalent radiation dose (De). The equivalent dose based on ITL signal is unreliable, since it is above the maximum dose that can be estimated (2D0 ¼ 420 Gy).

34

F.N. Pupim et al. / Quaternary Geochronology 36 (2016) 28e37

Table 6 Equivalent doses based on the TT-OSL 1 signal, dose rates and burial ages for quartz aliquots of samples from profiles ALC-01 and ALC-02. Sample

Na

De (Gy)b

O.D. (%)

RRc

ALC-01-1 ALC-01-2 ALC-01-3 ALC-01-4 ALC-02-1e ALC-02-2 ALC-02-3 ALC-02-4

6 6 6 6 18 6 6 6

381 ± 31 507 ± 29 388 ± 14 651 ± 27 3.1 ± 0.3 275 ± 33 499 ± 81 296 ± 78

26 18 11 13 37 29 39 64

± ± ± ± ± ± ± ±

0.99 1.05 1.05 1.00 1.02 1.25 0.96 0.86

a b c d e

8 5 4 4 10 9 12 19

Rc (%)d ± ± ± ± ± ± ± ±

0.03 0.05 0.03 0.03 0.01 0.07 0.07 0.07

0.75 1.15 1.57 0.79 0.23 1.17 1.15 1.58

± ± ± ± ± ± ± ±

Total dose rate (Gy/ka) 0.2 0.3 0.2 0.2 0.04 0.9 0.8 0.8

1.82 1.49 1.50 3.10 1.97 2.57 2.13 1.48

± ± ± ± ± ± ± ±

0.13 0.11 0.11 0.24 0.15 0.18 0.16 0.11

Age (ka) 209.3 ± 28.3 340.3 ± 38.1 258.7 ± 25.5 210.0 ± 21.3 1.6 ± 0.2 107.0 ± 15.5 234.3 ± 42.6 200.0 ± 54.9

N is the number of aliquots. Equivalent doses calculated through the central age model (Galbraith et al., 1999). RR is the average values of recycling ratio. Rc is the average values of recuperation. Equivalent dose estimated using the OSL signal measured with blue stimulation.

fluvial terraces using the TT-OSL 1 signal, which are negligible for ages in the range of hundreds of thousands years.

4. Discussion 4.1. Modern sediment: residual doses and dose recovery test Bleaching studies have demonstrated that quartz ITL and TT-OSL signals decrease in a much slower rate than have been observed for the OSL signal, and both ITL and TT-OSL signals do not reduce to zero level (Jain et al., 2007; Tsukamoto et al., 2008). Therefore, it is important to be sure that the quartz grains had been exposed to sunlight as much time as necessary to be sufficiently well bleached in a way that the residual signal is negligible for dating (Jain et al., 2007; Duller and Wintle, 2012). The studied modern fluvial sediment sample (XNG-47-2; 204 ± 5 years) shows residual De values up to ~4 and ~16 Gy for the ITL and TT-OSL signals, respectively, which are low considering that the target sediments for dating have equivalent doses in the 100 Gy range. Even if fluvial sediments are considered typically poorly bleached, the low overdispersion observed in equivalent dose distributions suggest adequate conditions for complete bleaching in the Xingu River, similar to conditions observed in aeolian and beach sands. Previous studies in modern sediments showed that residual doses for the ITL signal ranged from 1 to 5 Gy in aeolian and coastal sands and reach up to 15 Gy in fluvial sands (Huot et al., 2006; Vandenberghe et al., 2009; Barham et al., 2011). The TT-OSL signal usually has higher residual doses, ranging from 5 to 20 Gy in aeolian and coastal sands (Wang et al., 2006; Tsukamoto et al., 2008; Jacobs et al., 2011) and up to around 40 Gy in fluvial sediments (Demuro et al., 2015). In addition, Hu et al. (2010) reported an extremely high residual dose (~380 Gy) for the TT-OSL measured in modern fluvial sands from the Yellow River, indicating that TT-OSL signal is not appropriated for sediments transported by hyperconcentrated flows. However, ITL and TT-OSL measurements in older fluvial deposits have shown ages in right stratigraphic position and that overlap independent ages determined by other dating methods (OSL and pIRIR ages and paleomagnetism) (e.g. Kim et al., 2010; Zhao et al., 2010; Barham et al., 2011; Demuro et al., 2015). The presence of significant residual signals led Duller and Wintle (2012) to suggest that TT-OSL signal should be applied only to well-bleached aeolian or shallow marine deposits. These authors also argue that there must be some concern about the use of TT-OSL signal to determine burial ages of fluvial sediments because the slow rate to reduce the latent signal by exposure to daylight might be insufficient to full bleaching at deposition (Duller and Wintle, 2012). However, when compared to published data, our modern sediment shows small residual doses (~4 and ~16 Gy),

suggesting adequate bleaching prior to deposition. This could be explained by the fact that the Xingu River is a clear water river (Sioli, 1985), with low concentration of suspended sediments and shallow water rapids transporting sands for hundreds of kilometers before reaching the study sites. These conditions would provide long sunlight exposure time of quartz grains before their last deposition event. The dose recovery experiments using the ITL and TT-OSL 1 protocols applied to the studied fluvial sediments show calculatedto-given dose ratio, recycling ratio and recuperation results accurate as previously reported for aeolian and coastal sediments in other studies (e.g. Jain et al., 2007; Vandenberghe et al., 2009; Jacobs et al., 2011; Brown and Forman, 2012; Hernandez et al., 2012; Thiel et al., 2012). These results confirm the adequate conditions for bleaching of quartz grains in the Xingu River as well as their high luminescence sensitivity, endorsing their reliability as natural dosimeters using ITL and TT-OSL protocols. Best results for dose recovery tests were obtained using the ITL 310
C (Jain et al., 2007) and TT-OSL 1 (Adamiec et al., 2010) protocols, which were applied for dating of the Xingu River terraces.

4.2. Equivalent doses and burial ages Samples of the studied profiles were saturated for the ITL signal and only the TT-OSL signal was suitable to estimate equivalent doses. The ITL signal measured in samples from ALC-01 and ALC-02 profiles saturates at doses significantly lower than the values measured for the modern sediment bar sample (XNG-47-2) and values reported in Jain et al. (2007). This suggests that the dosimetric characteristics of quartz ITL is more varied compared to quartz OSL, whose saturation dose usually varies in a narrower range (100e200 Gy). The maximum dose that could be estimated (2D0) using the ITL 310
C signal had average values around 380 Gy for some samples from ALC-01 profile (Table 5). This indicates that the ITL 310
C signal is limited to determine burial ages up to 250 ka, considering dose rates of 1.2e3.2 Gy/ka (Table 4). The TT-OSL signal shows 2D0 from around 1700 to 4700 Gy in the samples from ALC-01 and ALC-02 profiles, suggesting that the TT-OSL signal has a potential to extend the dating range to a significant portion of the Quaternary period. Most samples from ALC01 and ALC-02 profiles shows TT-OSL ages overlapped by errors (Table 6). An age inversion occurs between samples ALC-01-2 (340.3 ± 38.1 ka) and ALC-01-3 (258.7 ± 25.5 ka), which is possibly related to dose rate changes through time rather than post-depositional mixing or insufficient sunlight bleaching since both samples show equivalent dose distributions with low overdispersion (11e18%). Although the reduced number of aliquots, the De distributions

F.N. Pupim et al. / Quaternary Geochronology 36 (2016) 28e37

for samples from the ALC-01 profile have low overdispersion values (11e26%), pointing out to a single population of grains with doses centered on the weighted mean De value. Three samples from the ALC-02 profile show moderate (29e39%) to high (64%) overdispersion values. Therefore, most samples presents overdispersion values lower than 35%, which is consistent with the commonly reported range for well-bleached sediments not affected by postdepositional mixing (e.g., Arnold and Roberts, 2009; Demuro et al., 2015). Sawakuchi et al. (2016) show that quartz-rich fluvial sediments in Brazil are well bleached as indicated by equivalent dose distributions with low overdispersion. These sediments have low feldspar content and are dominated by quartz grains with high luminescence sensitivity. This suggests that overdispersion in equivalent dose distributions is strongly related with microdosimetric heterogeneities. Therefore, sediments in different Brazilian rivers have adequate bleaching and are suitable for luminescence dating using multigrain aliquots. High overdispersion observed in equivalent dose distributions of some studied samples can result from microdosimetric effects in sediments with more heterogeneous texture and/or composition. Laboratory experiments have suggested that the lifetime of TTOSL trap might vary between hundred thousands and few million years (Li and Li, 2006; Adamiec et al., 2010; Shen et al., 2011; Pagonis et al., 2011; Thiel et al., 2012). Adamiec et al. (2010) calculated a thermal lifetime of 4.5 Ma at a storage temperature of 10
C and 580 ka at 20
C for the ReOSL signal used in the TT-OSL 1 protocol, where a 10% age underestimation, for a 1 Ma burial, would be expected. Thiel et al. (2012) presented a lifetime of 0.7 Ma at 19
C for the TT-OSL signal of sediments from Cap Bon coast (Tunisia). According to Thiel et al. (2012), TT-OSL ages were 23e43% underestimated when compared to ages in the 200e700 ka range obtained through the pIRIR at 290
C signal. Results from numerical modeling indicate that natural doses from TT-OSL signal can be measured in the range 0e400 Gy with an accuracy of ~1e5%, but it could underestimate doses above 400 Gy (Pagonis et al., 2011). This factor can hinder the use of this signal for dating of samples beyond 1 Ma. Brown and Forman (2012) used TT-OSL signal to date earlymid Pleistocene loess-paleosol sequences at the Mississippi and Missouri Rivers valleys. The TT-OSL ages for the youngest units (~52e63 ka, 66 ka, and 133e192 ka) agreed well with previously published thermoluminescence (TL) and infrared stimulated luminescence (IRSL) ages, but the ages for the oldest units (~167e200 ka) were underestimated by 20%. Thus, we cannot discard the possibility that all of our TT-OSL ages are underestimating the real burial age. In this case, the TT-OSL ages would indicate minimum ages for the formation of terraces along the lower Xingu River. Considering TT-OSL ages underestimation up to around 45% reported in Thiel et al. (2012) for the same age range of the studied samples, the Xingu River terraces would have ages up to 700 ka.

35

the profile (Anderson and Schaetzl, 2005). The low content of K (from 0.17 to 0.69%) and relatively high Th/U ratio (ALC-01 profile from 8.4 to 11.4; ALC-02 profile from 4.4 to 6.5) suggest sediments produced under intense weathering rates (Table 4). The Th/U ratio in most upper crustal rocks is typically 3.8 (Taylor and McLennan, 1985). In most cases, weathering and sedimentary recycling result in loss of U, leading to an elevation in the Th/U ratio (Asiedu et al., 2000; Carmichael et al., 2014). Thus, Th/U values ranging from 4.4 to 11.4 in the studied samples would indicate sediments with different origins regarding weathering and source areas. The tropical weathering typical of Amazonia combined with cratonic areas partially covered by old sedimentary rocks in the Xingu River basin would favor the formation of sediments with high Th/U ratio. Despite the possibility of clay illuviation could promote postdepositional changes in sediment texture and composition, this process would produce minor shifts in the primary sedimentologic characteristics of the studied profiles. Changes in water saturation since sediment deposition and due to seasonal climate (dry versus wet season) can promote variation of radiation dose rate through time (Aitken, 1985). Considering that the studied fluvial sediments were deposited below water table (completely water saturated), and that they are significantly above the present water table due to channel incision, we interpret a significant decrease of water content through time. Based on the water saturation range estimated for the studied samples and considering constant radionuclides concentrations, water content variation can promote decrease in dose rate and increase in burial ages in the order of 18e32% (Table 1S). The maximum water saturation estimates allow to determine a boundary scenario for sediment burial ages of the Xingu River terraces, in which the maximum ages for ALC-01 and ALC-02 profiles are around 446 ± 44 and 326 ± 56 ka, respectively. Changes in the clay content through time due to weathering of coarser sediment grains can also affect dose rate due to variation in water saturation promoted by the higher capacity of clay minerals in keep moisture. Pleistocene climate changes recorded for Amazonia (Cheng et al., 2013) and their influence over weathering processes can also provoke loss or uptake of radionuclides through mineral leaching or authigenesis. All those factors result in dose rate changes through time, then any attempt at applying luminescence dating should be problematic. Thus, dose rate heterogeneities and temporal variations in dose rate could contribute to the observed age inversions. According to the sedimentary facies interpretation, the profiles would correspond to a succession of sediment accretion in a fluvial channel covered by sand bar and floodplain deposits. In this way, the overlap of burial ages is expected, considering that age errors of tens of thousands of years avoid the discrimination of depositional events in timescales of fluvial channel dynamics. 4.4. Paleogeography implications

4.3. Dose rate uncertainties The dose rates calculated for the studied samples show significant variation in both sedimentary profiles. Samples collected in layers with higher concentration of silt and clay present radionuclides concentration two times higher than in coarse sand layers (see details in Fig. 1 and Tables 4 and 6). The studied profiles comprise fining upward intervals and sedimentary facies separated by clear surfaces, indicating sediment accretion in channel, sand bar and floodplain environments. In this way, the observed grain size pattern can be mainly attributed to depositional processes. However, the development of clay-enriched sand layers in the Xingu River terraces could be related partially to clay translocation downward by percolating water and subsequent deposition into

Despite age uncertainties, the studied terraces represent a phase of higher river level during the Middle Pleistocene, which represent a maximum age for the Xingu River channel under similar position and dimensions as today. Moreover, the Xingu River channel has depth around 15e25 m in most zones of the studied sector. Thus, TT-OSL ages would indicate a river base level 15e25 m higher than the present level during the Middle Pleistocene, considering the interpretation of the bottom sediment beds of the profiles as channel deposits. After this time, increase in precipitation and river discharge resulted in vertical channel incision and development of the modern Xingu River valley, whose flooding would allow the formation of its ria-type channel (Archer, 2005) before reaching the Amazon River. The younger deposits on the top of ALC-02 profile

36

F.N. Pupim et al. / Quaternary Geochronology 36 (2016) 28e37

correspond to a Holocene sand layer covering the Pleistocene terraces and it might be related to more recent extreme flood events. 5. Conclusions The first evaluation of ITL and TT-OSL signals of quartz from modern Xingu River sediments achieved satisfactory results for dose recovery tests, showing their suitability to estimate equivalent doses. Both signals have high sensitivity and are well bleached in nature. The best dose recovery results were achieved by protocols based on the ITL 310
C (Jain et al., 2007) and TT-OSL signals (Adamiec et al., 2010), which can recover equivalent doses up to 1082 ± 34 Gy and 1710 ± 63 Gy, respectively. When applied to calculate De for the Xingu River terraces, the maximum dose that can be estimated using the ITL 310
C signal is around 380 Gy, which is significantly lower than observed for the modern analogue sample and previous studies. This limited its application to dating the studied sedimentary sequences. Further efforts are necessary to investigate the factors controlling the huge variation in saturation doses of ITL signal as observed in quartz from the Xingu River sediments. TT-OSL signal provided De up to 651 ± 27 Gy and ages between 107.0 ± 15.5 and 340.3 ± 38.1 ka for the Xingu River terraces. With the exception of one sample, all other studied fluvial sediments samples showed equivalent dose distributions with low to moderate overdispersion (11e39%), similar to values observed in coastal or aeolian sediments. However, possible instability of TT-OSL and uncertainties regarding dose rate changes through time are the major concerns regarding the estimation of accurate burial ages. So far, innovation efforts in dose rate modeling are necessary to enhance the accuracy of burial ages. The TT-OSL signal shows 2D0 up to 4745 ± 444 Gy in the samples from Xingu terrace profiles, which would suggest the potential of TT-OSL signal to determine burial ages for the whole Quaternary of Amazonia, considering average dose rates of 1.5e2.0 Gy/ka. This would allow to better constrain changes in the Amazonian river system during the Quaternary, shedding light into the role of rivers for biodiversity in Amazonia. Acknowledgments
n Huot and an anonymous The authors sincerely thank Sebastie reviewer for valuable efforts and detailed suggestions, which ~o Paulo Research Foungreatly improved the manuscript. The Sa dation funded the Post-doctorate fellowship to FNP (grant 2014/ 23334-4) and the set-up of the Luminescence and Gamma Spec~o Paulo (grant 2009/ trometry Laboratory (LEGaL) at University of Sa 53988-8). AOS is supported by National Council for Scientific and Technological Development (CNPq grant 3009223/2014-8). This research had the financial support of the project “Structure and evolution of the Amazonian biota and its environment: an integrative approach”, a collaborative Dimensions of BiodiversityBIOTA grant supported by FAPESP (grant 2012/50260-6), NSF and NASA. We are grateful to the field team Mauricio Parra, Bettina Bozzi, Jandessa de Jesus, Nelson and Tonho (boat pilots); and
Adriano Giorgi (UFPA) for Leandro Sousa, Tatiana Pereira and Jose the logistics support during the fieldwork preparation in Altamira city. Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.quageo.2016.08.003.

References Adamiec, G., Bailey, R.M., Wang, X.L., Wintle, A.G., 2008. The mechanism of thermally transferred optically stimulated luminescence in quartz. J. Phys. D Appl. Phys. 41, 135503. Adamiec, G., Duller, G.A.T., Roberts, H.M., Wintle, A.G., 2010. Improving the TT-OSL SAR protocol through source trap characterisation. Radiat. Meas. 45, 768e777. Aitken, M.J., 1985. Thermoluminescence Dating. Academic Press, London. Anderson, S., Schaetzl, R., 2005. Soils Genesis and Geomorphology. Cambridge University Press, New York. Archer, A.W., 2005. Review of amazonian depositional systems. In: Blum, M.D., Marriott, S.B., Leclair, S.F. (Eds.), Fluvial Sedimentology VII. Blackwell Publishing Ltd. Oxford, UK, pp. 17e39. Arnold, L.J., Roberts, R.G., 2009. Stochastic modeling of multi-grain equivalent dose (De) distributions: implications for OSL dating of sediment mixtures. Quat. Geochronol. 4, 204e230. Asiedu, D.K., Suzuki, S., Nogami, K., Shibata, T., 2000. Geochemistry of lower cretaceous sediments, inner zone of southwest Japan. Constraints on provenance and tectonic environment. Geochem. J. 34 (2), 155e173. Barham, L., Phillips, W.M., Maher, B.A., Karloukovski, V., Duller, G.A., Jain, M., Wintle, A.G., 2011. The dating and interpretation of a mode 1 site in the Luangwa Valley, Zambia. J. Hum. Evol. 60 (5), 549e570. Brown, N.D., Forman, S.L., 2012. Evaluating a SAR TT-OSL protocol for dating finegrained quartz within Late Pleistocene loess deposits in the Missouri and Mississippi river valleys, United States. Quat. Geochronol. 12, 87e97. Buylaert, J.P., Jain, M., Murray, A.S., Thomsen, K.J., Thiel, C., Sohbati, R., 2012. A robust feldspar luminescence dating method for Middle and Late Pleistocene sediments. Boreas 41 (3), 435e451. Carmichael, S.K., Waters, J.A., Suttner, T.J., Kido, E., DeReuil, A.A., 2014. A new model for the kellwasser anoxia events (late devonian): shallow water anoxia in an open oceanic setting in the central asian orogenic belt. Palaeogeogr. Palaeoclimatol. Palaeoecol. 399, 394e403. Cheng, H., Sinha, A., Cruz, F.W., Wang, X., Edwards, R.L., d’Horta, F.M., Ribas, C.C., Vuille, M., Stott, L.D., Auler, A.S., 2013. Climate change patterns in Amazonia and biodiversity. Nat. Commun. 4, 1411. Choi, J.H., Murray, A.S., Cheong, C.S., Hong, D.G., Chang, H.W., 2006. Estimation of equivalent dose using quartz isothermal TL and the SAR procedure. Quat. Geochronol. 1, 101e108.
s, J.M., Sala, R., 2015. Extended-range luminescence Demuro, M., Arnold, L.J., Pare chronologies suggest potentially complex bone accumulation histories at the
scar-1 (Guadix-Baza Early-to-Middle Pleistocene palaeontological site of Hue basin, Spain). Quat. Int. 389, 191e212. Duller, G.A.T., 2003. Distinguishing quartz and feldspar in single grain luminescence measurements. Radiat. Meas. 37 (2), 161e165. Duller, G.A.T., Wintle, A.G., 2012. A review of the thermally transferred optically stimulated luminescence signal from quartz for dating sediments. Quat. Geochronol. 7, 6e20. Fitzsimmons, K.E., 2011. An assessment of the luminescence sensitivity of Australian quartz with respect to sediment history. Geochronometria 38, 199e208. Galbraith, R.F., Roberts, R.G., Laslett, G.M., Yoshida, H., Olley, J.M., 1999. Optical dating of single and multiple grains of quartz from Jinmium rock shelter, northern Australia: Part I. Experimental design and statistical models. Archaeometry 41, 339e364. Guedes, C.C.F., Sawakuchi, A.O., Giannini, P.C.F., Dewitt, R., Aguiar, V.P., 2013. Luminescence characteristics of quartz from Brazilian sediments and con^ncias (Impresso) 85, 1303e1316. straints for OSL dating. An. Acad. Bras. Cie
rin, G., Mercier, N., Adamiec, G., 2011. Dose-rate conversion factors: update. Anc. Gue TL 29 (1), 5e8. Hansen, V., Murray, A., Buylaert, J.P., Yeo, E.Y., Thomsen, K., 2015. A new irradiated quartz for beta source calibration. Radiat. Meas. 81, 123e127. Hernandez, M., Mauz, B., Mercier, N., Shen, Z., 2012. Evaluating the efficiency of TTOSL SAR protocols. Radiat. Meas. 47 (9), 669e673. Hoorn, C., Wesselingh, F.P., 2010. Introduction: Amazonia, landscape and species evolution. In: Hoorn, C., Wesselingh, F.P. (Eds.), Amazonia: Landscape and Species Evolution: a Look into the Past. John Wiley & Sons, pp. 1e6. Hu, G., Zhang, J.-F., Qiu, W.-L., Zhou, L.P., 2010. Residual OSL signals in modern fluvial sediments from the Yellow River (HuangHe) and the implications for dating young sediments. Quat. Geochronol. 5, 187e193. Huot, S., Buylaert, J.-P., Murray, A.S., 2006. Isothermal thermoluminescence signals from quartz. Radiat. Meas. 41, 796e802. Jacobs, Z., Roberts, R.G., Lachlan, T.J., Karkanas, P., Marean, C.W., Roberts, D.L., 2011. Development of the SAR TT-OSL procedure for dating Middle Pleistocene dune and shallow marine deposits along the southern Cape coast of South Africa. Quat. Geochronol. 6, 491e513. Jain, M., Bøtter-Jensen, L., Murray, A.S., Denby, P.M., Tsukamoto, S., Gibling, M.R., 2005. Revisiting TL: dose measurement beyond the OSL range using SAR. Anc. TL 23, 9e24. Jain, M., Duller, G.A.T., Wintle, A.G., 2007. Dose response, thermal stability and optical bleaching of the 310
C isothermal TL signal in quartz. Radiat. Meas. 42, 1285e1293. Kim, J.C., Duller, G.A.T., Roberts, H.M., Wintle, A.G., Lee, Y.I., Yi, S.B., 2010. Re-evaluation of the chronology of the palaeolithic site at Jeongokri, Korea, using OSL and TT-OSL signals from quartz. Quat. Geochronol. 5, 365e370. Li, B., Li, S.-H., 2006. Studies of thermal stability of charges associated with thermal

F.N. Pupim et al. / Quaternary Geochronology 36 (2016) 28e37 transfer of OSL from quartz. J. Phys. D Appl. Phys. 39, 2941e2949. Lü, T.Y., Sun, J., Li, S.-H., Gong, Z., Xue, L., 2014. Vertical variations of luminescence sensitivity of quartz grains from loess/paleosol of Luochuan section in the central Chinese Loess Plateau since the last interglacial. Quat. Geochronol. 22, 107e115. Murray, A.S., Wintle, A.G., 2000. Luminescence dating of quartz using an improved single-aliquot regenerative-dose protocol. Radiat. Meas. 32, 57e73. Murray, A.S., Wintle, A.G., 2003. The single aliquot regenerative dose protocol: potential for improvements in reliability. Radiat. Meas. 37, 377e381. Pagonis, V., Adamiec, G., Athanassas, C., Chen, R., Baker, A., Larsen, M., Thompson, Z., 2011. Simulations of thermally transferred OSL signals in quartz: accuracy and precision of the protocols for equivalent dose evaluation. Nucl. Instrum. Methods Phys. Res. B 269, 1431e1443. Pietsch, T.J., Olley, J.M., Nanson, G.C., 2008. Fluvial transport as a natural luminescence sensitiser of quartz. Quat. Geochronol. 3, 365e376. Prescott, J.R., Hutton, J.T., 1994. Cosmic ray contributions to dose rates for luminescence and ESR dating: large depths and long-term time variations. Radiat. Meas. 23 (2e3), 497e500. Ribas, C.C., Aleixo, A., Nogueira, A.C.R., Miyaki, C.Y., Cracraft, J., 2011. A palaeobiogeographic model for biotic diversification within Amazonia over the past three million years. Proc. R. Soc. B Biol. Sci. 279 (1729), 681e689.
Rossetti, D.F., Cohen, M.C., Tatumi, S.H., Sawakuchi, A.O., Cremon, E.H., Mittani, J.C., Bertania, T.C., Munitae, C.J.A.S., Tudelaf, D.R.G., Yeec, M., Moya, G., 2015. MidLate Pleistocene OSL chronology in western Amazonia and implications for the transcontinental Amazon pathway. Sediment. Geol. 330, 1e15. Sawakuchi, A.O., Blair, M.W., DeWitt, R., Faleiros, F.M., Hyppolito, T., Guedes, C.C.F., 2011. Thermal history versus sedimentary history: OSL sensitivity of quartz grains extracted from rocks and sediments. Quat. Geochronol. 6, 261e272. Sawakuchi, A.O., Mendes, V.R., Pupim, F.N., Mineli, T.D., Ribeiro, L.M.A.L., Zular, A., Guedes, C.C.F., Giannini, P.C.F., Nogueira, L., Sallun Filho, W., Assine, M.L., 2016.

37

Optically stimulated luminescence and isothermal thermoluminescence dating of high sensitivity and well bleached quartz from Brazilian sediments: from Late Holocene to beyond the Quaternary? Braz. J. Geol. 46 (Suppl. 1), 209e226. Shen, Z.X., Mauz, B., Lang, A., 2011. Source-trap characterization of thermally transferred OSL in quartz. J. Phys. D Appl. Phys. 44 (29), 295405. ^nia: Fundamentos de Ecologia da Maior Regi~ Sioli, H., 1985. Amazo ao de Florestas
polis. Tropicais. Editora Vozes, Petro Soares, E.A.A., Tatumi, S.H., Riccomini, C., 2010. OSL age determinations of Pleistocene fluvial deposits in Central Amazonia. Ann. Braz. Acad. Sci. 82 (3), 691e699. Stevens, T., Buylaert, J.-P., Murray, A.S., 2009. Towards development of a broadly applicable SAR TT-OSL dating protocol for quartz. Radiat. Meas. 44, 639e645. Taylor, S.R., McLennan, S.M., 1985. The Continental Crust: its Composition and Evolution. Blackwell, Oxford. Thiel, C., Buylaert, J.-P., Murray, A.S., Elmejdoub, N., Jedoui, Y., 2012. A comparison of TT-OSL and pot-IR IRSL dating of coastal deposits on Cap Bon peninsula, northeastern Tunisia. Quat. Geochronol. 10, 209e217. Tsukamoto, S., Duller, G.A.T., Wintle, A.G., 2008. Characteristics of thermally transferred optically stimulated luminescence (TT-OSL) in quartz and its potential for dating sediments. Radiat. Meas. 43, 1204e1218. Vandenberghe, D.A.G., Jain, M., Murray, A.S., 2009. Equivalent dose determination using a quartz isothermal TL signal. Radiat. Meas. 44 (5), 439e444. Wang, X.L., Lu, Y.C., Wintle, A.G., 2006. Recuperated OSL dating of fine-grained quartz in Chinese loess. Quat. Geochronol. 1, 89e100. Wintle, A.G., Murray, A.S., 2006. A review of quartz optically stimulated luminescence characteristics and their relevance in single-aliquot regeneration dating protocols. Radiat. Meas. 41, 369e391. Zhao, H., Lu, Y.C., Wang, C.M., Chen, J., Liu, J.F., Mao, H.L., 2010. ReOSL dating of aeolian and fluvial sediments from Nihewan Basin, northern China and its environmental application. Quat. Geochronol. 5, 159e163.