Figures
Abstract
Tuyuhun and Tubo were two important states that thrived in north-western China during AD 311-900 in parallel with the Han Chinese dynasties of Sui and Tang periods. The Reshui Tomb Cluster located in Dulan County of the north-eastern Tibetan Plateau is an important cultural relic of the Tuyuhun-Tubo age. The official excavations of the Reshui tombs were regarded as top events in archaeology in the 1980s and 1990s in China. The Reshui-1 Tomb is the largest one among the tombs in the area. Since its excavation, there have been debates on whether the owner of the tomb belonged to the Tuyuhun or Tubo ethnicity. Therefore, accurately dating the Reshui-1 Tomb has a critical place in studying the Tubo and Tuyuhun histories. We collected 7 discs and 11 increment cores of Qilian juniper (Juniperus przewalskii Kom.) from the exposed and fallen beams of the roof of the Reshui-1Tomb. The lengths of the 16 tree-ring records are between 69 and 152 years. Based on a previously developed master dating chronology using Qilian juniper samples from the eastern Qaidam Basin, the calendar dates of the 16 specimens were determined by the COFECHA program and visual dating procedure. The average inter-series correlation among the dated sample series is 0.696, indicating good quality of cross-dating. The year of the outermost rings is AD 715 for the 7 discs and 4 out of the 9 increment cores. Moreover, the ring-width variations of the samples are consistent with the existing chronologies from the region. The presence of late-wood of AD 715 in the samples indicated that the Reshui-1 Tomb was completed in late AD 715 or early 716, which means that the Reshui-1 Tomb was finished in the Tubo age. This date provides direct evidence for archaeologists to determine the owner’s ethnicity and identify of the Reshui-1 Tomb.
Citation: Li M, Shao X, Yin Z-Y, Xu X (2015) Tree-Ring Dating of the Reshui-1 Tomb in Dulan County, Qinghai Province, North-West China. PLoS ONE 10(8): e0133438. https://doi.org/10.1371/journal.pone.0133438
Editor: Xiaoyan Yang, Chinese Academy of Sciences, CHINA
Received: March 5, 2015; Accepted: June 26, 2015; Published: August 5, 2015
Copyright: © 2015 Li et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited
Data Availability: All relevant data are within the paper.
Funding: This study was supported by grants from the Chinese National Natural Science Foundation (No. 41471170 and 41071061) and University of San Diego (FRG 2014–2015). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Since the early 1900s, there have been numerous rescue excavations of ancient tombs by governmental archaeologists in China. Even though China has a long documented history, these tombs still provided new information about the tomb owners and various aspects of culture and political and economic conditions for the specific time periods. In 1983, governmental archaeologists excavated the Reshui-1 (also known as Xuewei No.1) Tomb in Dulan County, Qinghai Province in the north-eastern Tibetan Plateau, North-west China (Fig 1), which was claimed as one of the “Top Six Archaeological Discoveries of 1983 in China” [1]. The Reshui Tomb Cluster was one of the most important cultural relics of the Tuyuhun age (AD 329–663) [2] and Tubo age (early 7th century to the second half of the 9th century) [3]. The official excavation of the Reshui Tomb Cluster was regarded as one of the “Top Ten Archaeological Discoveries of 1996 in China” [1]. The Reshui-1 Tomb lies in the center of the Reshui Tomb Cluster and has the largest dimension among all tombs in the area. In fact it is also the largest one among the tombs of the Tuyuhun-Tubo age in Qinghai Province [4], suggesting the owner of the tomb as a high-ranking official or an important aristocrat. Therefore, it has a very important place in studying the Tubo and Tuyuhun histories.
XTT includes four tombs [26] and DRX includes 7 tombs (00DRXM3, 00DRXM8, 00DRXM10, 00DRXM14, 00DRXM19, 00DRXM21, 00DRXM23) [26]. Picture a) Signage of the Reshui Tomb Cluster; b) Full view of the Reshui-1 Tomb (from west); c) Collapsed portion of the tomb roof and the exposed roofing beams.
Tuyuhun was a nomadic tribe of approximately 1700 families that migrated from northeastern China to this region after AD 329 [5]. It reached its peak during the 5th century, with influences strong enough to compete with the Han Chinese dynasties (Sui and Tang) in this region [5]. Around the turn of the 7th century, however, the ancient Tibetan Kingdom, Tubo Kingdom, grew stronger and eventually conquered Tuyuhun in AD 663 and took over all the land of Tuyuhun [5]. Based on the tomb structure and unearthed artifacts, the age of the Reshui-1 Tomb was determined to be from the late 7th century to mid- 8th century [4, 6]. Nonetheless, an exact year could not be designated to the tomb at this moment. As of now, no radiocarbon dates have been published and even if they are available, they could not pinpoint the exact year of tomb construction. Without an exact date of the tomb construction, the tomb owner and his/her ethnicity cannot be determined, which leaves an important gap for the studies of the Tuyuhun and Tubo histories.
In the past decades, tree-rings analysis has been successfully employed to archaeological dating [7–13], including dating pre-historical relic sites [14–18], historical structures [19–26], tombs and burial sites [7, 27], artifacts [28–33], and past events (e.g., shipwrecks) [34].
Qilian juniper (Juniperus przewalskii Kom.), growing in the north-eastern part of the Qinghai-Tibetan Plateau, provides a unique opportunity to develop long tree-ring chronologies in China [35–38]. In the past decades, Qilian juniper has been employed to study the paleoclimate and construct dating chronologies, using samples from both living trees and archaeological wood unearthed from ancient tombs [39–46]. In the meantime, dendrochronology has been used to determine the ages of ancient tombs in the Dulan region. For example, using archaeological Qilian juniper samples, Wang [8] determined that 7 tombs (00DRXM3, 00DRXM8, 00DRXM10, 00DRXM14, 00DRXM19, 00DRXM21, 00DRXM23, Fig 1) of the Reshui Tomb Cluster were built between AD 611 and AD 784. Additionally, Wang et al. [7] presented tree ring dates for three additional tombs in Dulan, including a tomb (DRN3) on the south bank of Reshui (Reshuinanan) dated to AD 784, a tomb (MKL) near the village of Mokeli to AD 783, and a tomb (ZGR) in Zhigari to AD 789. Wang et al. [26] also presented ages for four tombs (XTT) in Xiatatu from AD 756 to AD 790 near Delingha, approximately 120 km north of Dulan (Fig 1). All these dates were based on dendrochronological work summarized in Shao et al. [46].
Although previous studies provided the context and general timeframe for the Reshui-1 Tomb, the exact date of construction of this most important archaeological site of the region is still missing at the time of the current study. As a result, there is not a firm conclusion on the owner’s ethnicity and identify. For example, based on historical documents, it has been speculated that the owner of the Reshui-1 Tomb may be a Tuyuhun Khan or a Tubo minister/governor who died from the late 7th century to the mid-8th century [4, 47, 48]. Therefore, the goal of this study is to determine the exact date for the Reshui-1 Tomb using Qilian juniper tree-ring data by cross-dating the samples from the tomb roof to an existing master dating chronology developed by Shao et al. [46] for the eastern Qaidam Basin. In doing so, we will be able to help archaeologists and historians to identify the potential candidates of the tomb owner to those who died close to the date of tomb construction.
Materials and Methods
2.1 Sampling, preparation and measuring of tree-ring widths
The Reshui-1 Tomb (36.18°N, 98.3°E; 3440m a.s.l.) is situated on the north bank of the Chahanwusu River in Dulan County and is on public land (Fig 1). The tomb itself covers an area of approx. 10,240 m2, with a height of 27.8 m [49]. It was robbed during the period of the Republic of China (AD 1911–1949) [49]. The tomb and its affiliated relics were excavated by the Qinghai Provincial Institute of Cultural Relics and Archaeology during the period of 1982–1985. Unearthed artifacts included animal bones (horses, dogs, and cattle), a large number of silk pieces, and various gold and silver wares [4]. The tomb’s roof was built with small beams (approx. 5–15 cm in diameter) of Qilian junipers covered by loess, sand and gravel, and rocks. During a field excursion in 2013, we discovered that the north-western corner of the roof partially collapsed (Fig 1), and several Qilian juniper beams fell down to the northern base of the tomb.
We collected seven beams fallen from the roof of the Reshui-1 Tomb and sawed one disc from each (numbered RS01-05, 17 and 18) in situ (Fig 2). The first five discs were collected in June 2013 and the rest two were collected in June 2014. We also took eleven increment cores (numbered RS06-16) from the exposed wood on the north-eastern side of the roof in June 2014. All specimens sampled were kept in the Tree-Ring Lab of the Institute of Geographic Sciences and Natural Resources Research, Chinese Academy of Sciences in Beijing, China. We obtained permission for fieldwork from the Qinghai Provincial Institute of Cultural Relics and Archaeology, which is in charge of all activities of archaeological research in Qinghai Province. One of the authors, Xinguo Xu, as the former director of the Qinghai Provincial Institute of Cultural Relics and Archaeology was in charge of the excavation of the Reshui-1 Tomb in the early 1980s [1]. Moreover, great caution was exercised in the field sampling process to avoid any damage to the tomb.
The seven discs were 6.5–12 cm in diameters (Table 1) and had no barks, but based on their appearances they were not processed before they were used to build the tomb (Fig 3). The barks were probably shed when the beams were exposed in the air for a long time. The eleven cores that we sampled with increment borer still had barks attached on the outside (Fig 4). The specimens were first sanded with increasingly fine grades of sand paper, to at least 600-grit so that all cellular details of the annual rings can be seen clearly under microscopes. The tree-ring widths were measured to the nearest 0.01 mm using a Lintab ring-width measurement system. Each series was treated as a floating series with the innermost ring of each series temporarily assigned as AD 1 prior to cross-dating.
2.2 Cross-dating
Cross-dating is the most important principle of dendrochronology. Its application provides a type of “experimental” quality control because it assures the proper placement in time of an annual growth layer [50]. Normally, the skeleton-plot technique is utilized primarily to match narrow rings among the samples to cross-date tree rings in relatively dry regions [50, 51], such as south-western North America and North-east Tibetan Plateau. Then, the COFECHA program is used to assess the quality of cross-dating and measurement accuracy of tree-ring series [52]. The COFECHA program was developed based on the concept that if all the tree ring data are perfectly dated and accurately measured, the correlation coefficients among them should be the highest comparing with series with incorrect dating or measurements [52]. Since the temporal trends contained in the data could exert certain influences on the correlation coefficients, it is preferable to first remove the low-frequency variations from all the series in practice. Then the correlations between each series and the mean series of the remaining series in the corresponding time segments are computed with a moving window shifting back and forth through time to calculate correlation coefficients. These coefficients of correlation offer the hints whether the dating and measurements are accurate. The COFECHA program also has a “cross-dated and undated” function that can be used to cross-date the undated sample series to an existing master chronology from the same region. The absolute date of a sample series is suggested when the correlation coefficient with the master chronology peaks at a given year when a moving window shifts back and forth. Good cross-dating quality is indicated by strong positive correlation coefficients that are highly significant statistically, among the sample series and between the individual sample series and the master chronology.
In this study, we used the cross-dated and undated function of COFECHA after the ring widths of the collected specimens were measured. The master dating chronology we used here was constructed previously using 1438 series from 713 trees, including samples from 22 archaeological sites, 24 living tree sites and 5 standing snags sites in the eastern Qaidam Basin [46] (as Qaidam Chronology (QC) hereafter). In order to best match the undated sample series to the existing dating chronology, we first constructed a master chronology using only specimens from the archaeological sites, known as the Qaidam Archaeological Chronology (QAC). QAC was precisely dated and well-replicated for the period before AD 800 [46], and should serve well as the master chronology for the tomb samples. Then the likely calendar dates of the sample series were assigned according to the COFECHA results. Finally, we dated each ring visually and added missing rings, if any, according to the narrow-wide ring-width patterns of the QAC. When we ran COFECHA, the program default was used, which means that all correlations were calculated after removing long-term trends in the sample series using a 32-year cubic spline function. The segment length is 50 years lagged successively by 25 years (e.g., from year 1 to year 50, and then from year 26 to year 75 until the end of the sample series) and all the series were log-transformed to enhance the influence of narrow rings. We also calculated correlations between the sample series and two more reference chronologies from Dulan [36, 40] as a separate evaluation of the cross-dating quality.
Results
Two of the eleven increment cores (RS7 and RS10) were fragmented so badly that we could not date them. Therefore, they were not included in the following analysis. The number of rings that each specimen contains was between 69 and 152 (Table 1). The maximum width of all rings was 1.89 mm, the minimum was 0.02 mm and the average width was 0.47 mm (Table 1). The average mean sensitivity was 0.352 (Table 1), showing a similar level of year-to-year variation to that of QC (0.37) [46].
As described earlier, the essence of cross-dating for this study is the comparison between each individual sample series and the master dating chronology, Qaidam Archaeological Chronology (QAC). Fig 5 shows the changes in correlation coefficients between the first 50-year segment of sample RS01 and QAC (AD 605–654) as the moving window shifted from -10 to +10 years of a potential target date (year 0) with a one-year lag. Here, the correlations were calculated between RS01 and QAC successively for the 50-year window of the QAC dates from AD 595–644 (year -10) to AD 615–664 (year +10) It can be seen that as the sample series’ date matched with the master chronology at the target year (year 0 corresponding to QAC dates of AD 605–654), the correlation reached its peak value. Table 2 contains examples of COFECHA output of cross-dating for sample series RS01, RS03, and RS13. For RS01, AD 604 was the suggested calendar year prior to the first year of the sample series (Table 2). In other words, adding 604 to the years of the floating series RS01 would convert them to the calendar years, so that the first year of sample RS01 was AD 605. Ten out of the 16 sample series can be dated very easily in similar ways (RS01, 02, 05, 08, 09, 11, 14–17). For RS03, COFECHA indicated that 590 years should be added to the floating series for segments 3 and 4, so that the first calendar year of RS03 was AD 591, while adding 590 was suggested as choice #6 for the second segment (Table 2). For the first segment, however, COFECHA suggested adding 589 years to the sample years or the first sample year was assigned as AD 590 instead of 591. This was because a missing year should be added in AD 644, which would make the entire series match to the master chronology. A total of 5 samples series contained missing rings (RS03, 04, 06, 12, and 18). The sample series RS13 posed the greatest challenge in cross-dating for us. Most of the COFECHA’s top suggestions (Table 2) were far away from the timeframe indicated by previous studies (7th– 8th century) [4, 7, 8, 26]. According to these previous studies, we picked AD 591 as the calendar year prior to the first year of the sample series and the following visual dating process proved that it was the correct year, even though it appeared only twice in the top choices provided by COFECHA (Table 2).
Please refer to the text for details.
The years underlined are the final choices.
Table 3 summarizes the results of cross-dating for all sample series, including the suggested years prior to the first sample years and the corresponding correlation coefficients for the floating series. It can be seen that COFECHA provided good suggestions to the ages for most of the floating sample series. The absolute ages of the sixteen series (Table 3) range from AD 564 to AD 715. There are eight missing rings in the sixteen specimens and AD 681 was seen in three sample series as a year of missing rings (RS04, RS06 and RS18), while RS18 had the most missing rings (AD 681, 691, 700, 711) (Table 3). The overall percentage of missing ring is only 0.471% for our samples. Previous studies have found relatively high percentages of missing rings in moisture sensitive ring-width series in this region [36–37, 40]. The low frequency of occurrence of missing rings in our samples is probably mostly due to the young ages of these trees [50].
Table 4 shows the final results from the COFECHA correlation analysis between AD 550 and AD 724 for the dated sample series. It contains the correlation coefficients for each 50-year segments with 25-year lags for each sample series and the mean correlation coefficients with the COFECHA master series, which were used to evaluate the quality of cross-dating. All sixteen series had statistically significant correlations for all 50-year segments tested and the correlations ranged from 0.41 to 0.88. Inter-series correlations (the correlation of one series against a composite created from the other fifteen series) ranged from 0.498 to 0.830. The average inter-series correlation was 0.696 over the full periods (Table 4), suggesting strong consistency among the 16 dated sample series.
The quality of cross-dating can also be illustrated by Fig 6 that contains the time-series plots of the 16 dated sample series. It can be seen that the variation patterns of the 16 ring-width series exhibit an excellent match among themselves, especially for the years with narrow rings. Also plotted in Fig 6 is the Qaidam Chronology (QC) together with two additional chronologies adjusted based on QC [46] from Dulan, one constructed by Zhang et al. [36] (as ZDC) and the other by Sheppard et al. [40] (as SDC). Again, the match between the 16 series to these reference chronologies was also very good (Fig 6). Fig 7 shows the correlation coefficients between the sample series and the three reference chronologies. Sample series RS16 has the highest correlations with the three chronologies (r = 0.745 to QC, 0.611 to ZDC, and 0.734 to SDC), while the lowest correlations were found for sample series RS13 (r = 0.447 to QC, 0.366 to ZDC and 0.382 to SDC). The mean correlation coefficients to the reference chronologies are 0.595 to QC, 0.485 to ZDC, and 0.544 to SDC, respectively.
RS01-06, 08, 09, 11–18: The 16 dated series from archaeological wood sampled; QC: The Qaidam Chronology constructed by Shao et al. [46]; ZDC: The Dulan chronology constructed by Zhang et al. [36]; SDC: The Dulan chronology constructed by Sheppard et al. [40]. The black dots show the locations of the missing rings.
As stated above, the latest year for most of the 16 dated series was AD 715 (Table 3). In fact, seven discs and four out the nine increment cores all ended in this year (Fig 6). For the remaining five increment cores (RS06, 08, 12, 14, and 16), the outermost rings were decayed due to weathering and the latest years that had been dated are AD 707, 707, 664, 712, and 679, respectively. To make sure that the missing of the outermost rings in these samples was not due to wood processing, we carefully examined the morphology of the discs and increment core samples. Fig 3 shows the conditions of the 7 wood discs and it can be seen that there are no signs of wood processing, which means that the outermost rings of these discs should represent the years when the trees were cut. In the meantime, we observed in the field that barks were still attached to some of the beams sampled by increment borers (Fig 4), so that the outermost rings of these samples should also represent the years when the trees were felled. Since the ages of the outermost rings from eleven out of the sixteen trees are AD 715 (7 discs and 4 increment cores), it can be concluded that all these trees sampled were not processed before they were used to build the tomb, which is consistent with the earlier observations of wood used in other ancient tombs in the area [11]. In addition, the latewood of the outermost rings of the RS11 and 13 was visible under the microscopes. Therefore, these trees must have been cut down in late AD 715 or before the growing season started in AD 716.
Discussions
The dates of ancient events and ages of artifacts have important meanings in archaeology. For example, determining the ages and construction histories of monuments is important to relate religious symbolism, scientific knowledge, and cultural continuity to groups within a region [53]. It is well known that tree-ring data have the advantages of accurate dating, annual resolution, wide availability, and good replication [50]. Prehistoric sites and ancient tombs that contained wood specimens can be dated using regional reference tree ring chronologies, but it can be difficult at times due to the lack of well-preserved rings and collaborations with field archaeologists are often necessary to achieve accurate dating results [18, 22]. The dry climate in the eastern Qaidam Basin provides the optimal condition for preservation of wood [26, 46, 54]. As shown in Fig 4, the barks are still attached to most of the beams we sampled using increment borers, showing low degrees of decay and weathering after more than 1000 years of tree harvesting.
COFECHA was very effective in providing fairly accurate suggested ages for most sample series in our study (Table 3). When comparing with other studies that employed COFECHA, our cross-dating results apparently have a similar level of quality. For example, the correlations of the sample series segments with the COFECHA master series from our study ranged from 0.41 to 0.88 (Table 4), while those in Grissino-Mayer et al. [12] are from 0.19 to 0.76. For the mean correlation with the COFECHA master series, our result is from 0.498 to 0.830, while those in Therrell and Stahle [55] is from 0.190 to 0.732.
In the light of the local tradition, the beams should have been used to build the tomb after the trees were cut down for about ten days [7]. Therefore, we conclude that the tomb was built using fresh wood soon after the trees were harvested. As to the 5 increment cores dated prior to AD 715 even with barks still remaining on the beams, the wood was weathered over time and very fragile, so the rings immediately beneath the bark fell off at the time of sampling by the increment borer. Therefore, it is reasonable to conclude that the Reshui-1 Tomb was completed in late AD 715 or early AD 716, since the samples were taken from the roofing beams whose installation should be the last step of tomb construction. That is to say, the Reshui-1 Tomb was finished in the Tubo Kingdom Period. This date was consistent with those estimated ages (ranging from the second half of the 7th century to the first half of the 8th century) based on the unearthed artifacts from the Reshui-1 Tomb [4, 6], which lead to the speculations that the tomb owner was of Tubo ethnicity.
Wang et al. [8] used dendrochronology to date 7 tombs of the Reshui Tomb Cluster. They determined that one tomb (numbered 00DRXM10) was built in AD 611 belonging to the Tuyuhun Kingdom Period, while the other six tombs were built in the Tubo age after AD 663. The Dulan area was the territory of the Tuyuhun Kingdom during AD 329–663 and Dulan was the capital of the Tuyuhun Vassal State of the Tubo Kingdom after AD 663 [5]. Therefore, the Reshui Tomb Cluster was probably the relics of the people of Tuyuhun ethnicity [1, 4, 6]. Others argued that the Reshui Tomb Cluster was started by the Tuyuhun people but later was continued by the Tubo people after AD 663, while the ancient Tombs in Delingha during the same period exclusively belonged to the Tuyuhun people [56].
Our analysis results suggested that the trees were harvested in late AD 715 or early AD 716 before the growing season started. According to the ancient Tibetan inscribed wooden slips and coins of the Tang Dynasty (Kaiyuan Tongbao or Currency, first minted in AD 621) unearthed from the Reshui-1 Tomb excavation, it was determined that the tomb was built during AD 713–741 [6], which would imply the tomb owner as a member of the Tubo Kingdom. On the other hand, Tong [4] dated the Reshui-1 Tomb to the time from the end of the 7th century to the beginning of the 8th century based on silk pieces, color-painted wooden planks, and gold and silver wares unearthed from the tomb, and speculated that the owner of the tomb was a Tuyuhun King (Khan) established by the Tubo Kingdom after AD 663. However, Tong [4] also pointed out that this Tuyuhun Khan Bendayanchisong (dBon da rgyal khri zung) died in AD 694. So it is unlikely that his tomb was finished 21 years after his death, as the deceased should be buried soon after the death according to the local funeral tradition during the Tubo and Tuyuhun age [57]. Even if we consider the uncertainties in the exact year of his death (AD 689–706) [4], it was still at least approximately 10 years before the completion of the Reshui-1 Tomb. The exact date of the tomb construction plus the knowledge of the tomb structure and unearthed artifacts from the Reshui-1 Tomb indicated that the tomb owner was someone of high importance, either as a Tuyuhun under Tubo’s ruling or a member of the Tubo Kingdom. Hence, we still cannot determine the exact date of tomb usage or the tomb owner’s time of death at this point. If we want to know more about the exact ethnicity and identify of the Reshui-1 Tomb, we need to comprehend the social and political characteristics of the Tuyuhun-Tubo age and have a more complete survey of the ages of the ancient tombs in this area with the help from archaeologists. We also need a thorough investigation on historical documents from both Tibetan and Han Chinese sources, which may provide the times of death of important people in this region around AD 715 or 716.
Conclusions
In this study, we sampled 18 roofing beams from the Reshui-1 Tomb, the largest tomb of the Tuyuhun-Tubo ages in Qinghai Province, and 16 of them could be accurately dated. Based on cross-dating to an existing master dating chronology, the Qaidam Chronology, we dated the roofing beams of the Reshui-1 Tomb as being harvested in late AD 715 or before the growing season started in AD 716, and the Tomb itself was probably finished soon after and belonged to the Tubo age. This date was supported by high-quality cross-dating among the sample series themselves and between these series and three reference chronologies from the study area. Our results corroborated with previous archaeological studies and settled the question whether the Tomb belonged to the Tuyuhun or Tubo age. As to the tomb owner’s identify, the ethnicity of the tomb owner still could not be determined at this moment. The same is true for the exact date of tomb usage or the time of death of the tomb owner. Additional archaeological analysis and research of historical documents need to be performed to further our understanding of the history of this unique cultural relic.
Acknowledgments
We are grateful to two anonymous reviewers who provided constructive suggestions. We also thank Yang Wang and Feng Chen from the Institute of Geographic Sciences and Natural Resources Research, Chinese Academy of Sciences, for collecting samples.
Author Contributions
Conceived and designed the experiments: XMS MQL ZYY XGX. Performed the experiments: XMS MQL. Analyzed the data: MQL XMS ZYY. Contributed reagents/materials/analysis tools: XMS MQL XGX. Wrote the paper: MQL XMS ZYY.
References
- 1. Xu XG (2001) Searching for the lost kingdom—Discovery and excavation of the ancient tombs in Dulan. Chaidamu Kaifa Yanjiu (2): 66–70. (in Chinese)
- 2. Cheng QJ (2004) The historical records of Tubo’s vassal state—Tuyuhun Kingdom. China’s Tu nationality (1): 38–41. (in Chinese)
- 3. Huo W (2009) Tubo archaeology and Tubo civilization. Journal of Tibet university 24: 57–74. (in Chinese)
- 4. Tong T (2012) The structure, date and occupant of the tomb Reshui-1 in Dulan, Qinghai. Kaoguxuebao (4): 467–488. (in Chinese)
- 5.
Zhou WZ (2006) The histories of Tuyuhun. Guangxi Normal University Press, Guilin. pp1–244.
- 6. Cheng QJ, Mao WB (2006) The story of thousands of tombs—re-discussion about the tomb group’s ethnicity from Dulan with Xinguo Xu. China’s Tu nationality (3): 9–12. (in Chinese)
- 7. Wang S, Zhao X (2013) Re-evaluating the Silk Road's Qinghai Route using dendrochronology. Dendrochronologia 31: 34–40.
- 8. Wang SZ (2004) Tree-ring chronology since 515 BC in Dulan area, Qinghai and its applications. Kaogu and Wenwu (6): 45–50. (in Chinese)
- 9. Fletcher JM (1982) Tree rings and archaeology: dating Roman oaks. Nature 298: 227.
- 10. Kuniholm PI, Kromer B, Manning SW, Newton M, Latini CE, Bruce MJ (1996) Anatolian tree rings and the absolute chronology of the eastern Mediterranean, 2220–718 BC. Nature 381: 780–783.
- 11. Wang SZ, Shao XM, Xu XG, Xiao YM (2008) Establishment of a 2332-year Dendro-archaeological Chronology and Dating of Xiatatu Tombs. Archaeology (Kaogu) (2): 80–86. (in Chinese)
- 12. Grissino-Mayer HD, Maxwell JT, Harley GL, Garland NA, Holt DH, Absher C, et al. (2013) Dendrochronology reveals the construction history of an early 19th century farm settlement, southwestern Virginia, USA. Journal of Archaeological Science 40: 481–489.
- 13. Lorentzen B, Manning SW, Cvikel D, Kahanov Y (2014) High-precision dating the Akko 1 shipwreck, Israel: wiggle-matching the life and death of a ship into the historical record. Journal of Archaeological Science 41: 772–783.
- 14. Douglass AE (1929) The secret of the Southwest soved by talkative tree rings. National Geographic Magazine 56(6): 736–770.
- 15.
Douglass AE (1935) Dating Pueblo Bonito and Other Ruins of the Southwest, National Geographic Society, p. 76.
- 16. Kuniholm PI, Kromer B, Manning SW, Newton M, Latini CE, Bruce MJ, (1996) Anatolian tree rings and the absolute chronology of the eastern Mediterranean, 2220–718 BC, Nature 381: 780–783.
- 17.
Nash S (1999) Time, Trees, and Prehistory: Tree-Ring Dating and the Development of North American Archaeology, 1914–1950. University of Utah Press, Salt Lake City.
- 18. Koerner SD, Grissino-Mayer HD, Sullivan LP, Deweese GG (2009) A Dendroarchaeological Approach to Mississippian Culture Occupational History in Eastern Tennessee, USA, Tree-Ring Research 65: 81–90.
- 19.
Stahle DW (1978) Tree-Ring Dating of Selected Arkansas Log Buildings. MA thesis, University of Arkansas, Fayetteville.
- 20. Stahle DW (1979) Tree-Ring Dating of Historic Buildings in Arkansas. Tree-Ring Bulletin 39: 1–28.
- 21.
Cook ER, Callahan W (1992) The Development of a Standard Tree-Ring Chronology for Dating Historical Structures in the Greater Philadelphia Region. Tree-Ring Laboratory, Lamont-Doherty Geological Observatory, Columbia University, New York.
- 22. Grissino-Mayer HD, and Van de Gevel SL (2007) Tell-tale trees: The historical dendroarchaeology of log structures at Rocky Mount, Piney Flats. Tennessee. Historical Archaeology 41:30–47.
- 23. Grissino-Mayer HD (2009) An introduction to dendroarchaeology in the southeastern United States. Tree-Ring Research 65: 5–10.
- 24. Schneider EA, Stachowiak LA, Grissino-Mayer HD (2015) The historical dendroarchaeology of two log structures at the Wynnewood State Historic Site, Castalian Springs, Tennessee, USA. Dendrochronologia 33: 34–41.
- 25. Baatarbileg N, Park WK, Jacoby GC, Davi NK (2008) Building history of Mandal Monastery in Mongolia based on tree-ring dating. Dendrochronologia, 26(2): 63–69.
- 26.
Hollstein E (1980) Mitteleuropäische Eichenchronologie: Trierer dendrochronologische Forschungen zur Archäologie und Kunstgeschichte, P. von Zabern, Mainz am Rhein. pp 1–273.
- 27. Wang SZ, Shao XM, Xu XG, Xiao YM (2008) Dating of tombs in Delingha, Qinghai province, on the basis of a 2332-year long tree-ring juniper chronology (Sabina przewalskii Kom) (1575 BC–756 AD). Dendrochronologia 26: 35–41.
- 28. Fraiture P (2009) Contribution of dendrochronology to understanding of wood procurement sources for panel paintings in the former Southern Netherlands from 1450 AD to 1650 AD, Dendrochronologia 27: 95–111.
- 29. Pichler T, Nicolussi K, Goldenberg G (2009) Dendrochronological analysis and dating of wooden artefacts from the prehistoric copper mine Kelchalm/Kitzbuhel (Austria), Dendrochronologia 27: 87–94.
- 30. Grissino-Mayer HD, Sheppard PR, Cleaveland MK, Cherubini P, Ratcliff P, Topham J (2010) Adverse implications of misdating in dendrochronology: Addressing the re-dating of the "Messiah" violin, Dendrochronologia 28: 149–159.
- 31. Klein A, Nemestothy S, Kadnar J, Grabner M (2014) Dating furniture and coopered vessels without waney edge—Reconstructing historical wood-working in Austria with the help of dendrochronology. Dendrochronologia 32: 90–96.
- 32. Tegel W, Elburg R, Hakelberg D, Stäuble H, Büntgen U (2012) Early Neolithic Water Wells Reveal the World’s Oldest Wood Architecture. PLOS ONE 12: e51374.
- 33. Bauch J, Eckstein D, Meier-Siem M (1972) Dating the wood panels by a dendrochronological analysis of tree-rings, nederl. kunsthist. Jaarboek 23: 485–496.
- 34. Baillie MGL (1999) Seven thousand years of alternative history: the tree-ring story, irish forestry 56: 29–38.
- 35. Zhang QB, Qiu HY (2007) A millennium-long tree-ring chronology of Sabina przewalskii on northeastern Qinghai-Tibetan plateau. Dendrochronologia 24: 91–95.
- 36. Zhang QB, Cheng GD, Yao TD, Kang XC, Huang JG (2003) A 2,326-year tree-ring record of climate variability on the northeastern Qinghai-Tibetan Plateau. Geophysical Research Letters 30:
- 37. Shao XM, Xu Y, Yin ZY, Liang EY, Zhu HF, Wang SZ (2010) Climatic implications of a 3585-year tree-ring width chronology from the northeastern Qinghai-Tibetan Plateau. Quaternary Science Reviews 29: 2111–2122.
- 38. Zhang Y, Gou XH, Chen FH, Tian QH, Yang ML, Peng JF, et al. (2009) A 1,232-year tree-ring record of climate variability in the Qilian Mountains, northwestern China. IAWA Journal 30: 407–420.
- 39. Liu XH, Qin DH, Shao XM, Chen T, Ren JW (2005) Temperature variations recovered from tree-rings in the middle Qilian Mountain over the last millennium. Science in China Series D-Earth Sciences 48: 521–529.
- 40. Sheppard PR, Tarasov PE, Graumlich LJ, Heussner KU, Wagner M, Österle H, et al. (2004) Annual precipitation since 515 BC reconstructed from living and fossil juniper growth of northeastern Qinghai Province, China, Climate Dynamics 23: 869–881.
- 41. Shao XM, Huang L, Liu HB, Liang EY, Fang XQ, Wang LL (2005) Reconstruction of precipitation variation from tree rings in recent 1000 years in Delingha, Qinghai. Science in China Series D-Earth Sciences 48: 939–949.
- 42. Liu Y, An ZS, Ma HZ, Cai QF, Liu ZY, Kutzbach JK, et al. (2006) Precipitation variation in the northeastern Tibetan Plateau recorded by the tree rings since 850 AD and its relevance to the Northern Hemisphere temperature. Science in China Series D-Earth Sciences 49: 408–420.
- 43. Gou XH, Yang MX, Peng JF, Zhang Y, Chen T, Hou ZD (2006) Maximum temperature reconstruction for Anmaqing Mountains over the past 830 years based on tree-ring records. Quaternary Science 26: 991–998.
- 44. Zhu HF, Zheng YH, Shao XM, Liu XH, Xu Y, Liang EY (2008) Millennial temperature reconstruction based on tree-ring widths of Qilian juniper from Wulan, Qinghai Province, China. Chinese Science Bulletin 53: 3914–3920.
- 45. Yang B, Qin C, Wang J, He M, Melvin TM, Osborn TJ, et al. (2014) A 3,500-year tree-ring record of annual precipitation on the northeastern Tibetan Plateau. Proceedings of the National Academy of Sciences of the United States of America 111: 2903–2908. pmid:24516152
- 46. Shao XM, Wang SZ, Zhu HF, Xu Y, Liang EY, Yin Z-Y, et al. (2009) A 3585-year ring-width dating chronology of Qilian Juniper from the northeastern Qinghai-Tibetan Plateau. IAWA Journal 30: 379–394.
- 47. Xu X (2009) Initial research into the Tuyuhun Mausoleum of Chaka. Journal of Qinghai Nationalities Institute (Social Sciences) 35: 132–135. (in Chinese)
- 48.
Huo W (1995) The history of ancient mausoleum tradition in Tibet. Sichuan People’s Publishing House. pp 1–424.
- 49. San MC, Liu SJ (2001) The review of funeral and interment folk-custom about the Dulan-1 tomb. Chaidamu Kaifa Yanjiu (6): 51–58. (in Chinese)
- 50.
Fritts HC (1976) Tree rings and climate. Academic Press, London.
- 51.
Stokes MA, Smiley TL (1968) An introduction to tree-ring dating. University of Chicago Press, Chicago.
- 52. Holmes RL (1983) Computer-assisted quality control in tree-ring dating and measurement. Tree-ring bulletin 43: 69–78.
- 53. Herrmann EW, Monaghan GW, Romain WF, Schilling TM, Burks J, et al. (2014) A new multistage construction chronology for the Great Serpent Mound, USA, Journal of Archaeological Science 50:, 117–125.
- 54. Tarasov P, Heussner KU, Wagner M, Osterle H, Wang SZ (2003) Precipitation changes in Dulan 515 BC– 800 AD inferred from tree-ring data related to the human occupation of NW China, Eurasia Antiqua (9): 302–321.
- 55. Therrell MD and Stahle DW (2012) Tree-ring dating of an Arkansas Antebellum Plantation House. Tree-ring research, 68(1): 59–67.
- 56. Xiao YM (2008) The application of dendrochronology in the research of Tuyuhun and Tubo tombs in western Qinghai area. Nationalities research in Qinghai 19: 57–63. (in Chinese)
- 57.
Cai R (2007) The Tubo’s history. Gansu People’s Publishing House, pp1–313.