Based on the threshold of 2% alloying element content, ancient Chinese bronzes fall into three main categories: tin bronze, lead bronze, and lead-tin bronze. The lead content in lead bronze and lead-tin bronze is higher than 2%, thus the lead isotope ratios mainly point to the source of lead material , whereas lead isotopes of pure copper and tin ware should indicate the source of copper ore and tin ore respectively . There are two points about the lead isotope information of tin bronze can be discussed in two cases: (1) In terms of alloy proportions, the content of copper in tin bronze is about 5~20 times that of tin, so the results of the lead isotope of this kind of tin bronze should reflect the source of the copper ore. (2) If the copper material is relatively pure, then lead mainly comes from tin material, and the results should indicate the source information of tin material . In terms of the slags produced during the smelting process of copper ores, the results of lead isotope ratios directly reflect the source information of copper ore .
4.1 Lead isotope analysis
4.1.1 Lead isotope ratios analysis of slags from the Tangjiadun site
The results of the lead isotope ratios analysis of the slags from the Tangjiadun site in Zongyang and other sites in Nanling and Tongling are shown in Table 2 (the data of every point for the samples shown in Table S3). Table 2 reveals that the data are distributed between 2.085 and 2.136 on 208Pb/206Pb, 0.844 and 0.871 on 207Pb/206Pb. All the data fall into the range of common lead, with no anomalous lead being detected. Fig.5 shows that the lead isotope ratios of all the slags exhibit a linear distribution as a whole. The lead isotope ratios of the Tangjiadun slags are relatively concentrated with those of the Jinshansheng, Muyushan, Wanyingshan, and Jiangmuchong sites in the Western Zhou Dynasty. The results indicate that the ore materials used for the Tangjiadun slags could be the same as those of the smelting sites in the Wannan region.
Until now, nearly 100 ancient copper mining and smelting sites have been found along the Yangtze River basin in Anhui Province, which are mainly distributed in the Wannan region (Tongling, Nanling, Fanchang, Qingyang, Guichi, and Jingxian, etc.), Zong-Lu (Zongyang-Lujiang) and Chu-Ma (Chuzhou-Ma'anshan) areas. The sites spread over thousands of square kilometers and are ranging from the Shang and Zhou Dynasties to the Ming-Qing Dynasties. With the excavation of the Shigudun site in Tongling in 2010, the date of the smelting and casting activities in the Wannan region were brought forward to the third and fourth stages of the Erlitou Culture (1750BC-1530BC ) [19-20]. Yet the smelting and casting activities occuring in Zongyang area in the late Shang dynasty, as represented by the Tangjiadun site, were later than those in the Wannan region. Geologically, the Tangjiadun site of Zongyang is located on the west bank of the Yangtze River and faces Wannan region across the river, having convenient water transport conditions. Tangjiadun is also closer to the mining area in the Wannan region than to the Zongyang mining area. Therefore, it is possible that the metal resources in Wannan region could have been transferred westward to the Zongyang region during the late Shang Dynasty.
4.1.2 Lead isotope ratios of bronzes unearthed in Zongyang area
The elemental composition of the Zongyang bronzes shows that they are all lead-tin bronzes (Table 3 and Table S4). All the samples from the Zongyang contain more than 2% lead. For data with lead content >2%, the lead isotope data in this paper are applied to discuss the sources of lead rather than of copper. The distribution of tin in the Zongyang bronzes is 2.96-13.19%. According to the research done by Pollard et al. [21, 22], the approximately normal distribution of tin centred suggests that it should have been added to the primary alloy deliberately rather than inadvertently in the recycling and remelting process.
We can see from Fig.6 and Table 4 that there are two categories of lead materials used for the Zongyang bronzes. Class Ⅰ lead materials, with the lead isotope ratios of 208Pb/206Pb ranging from 2.09 to 2.12, and the ratios of 207Pb/206Pb ranging from 0.84 to 0.86, were mainly used for the bronzes between the Western Zhou Dynasty and the Spring and Autumn period. Class Ⅱ lead materials were mainly used for the bronzes in the Warring States period, and the lead isotope ratios range from 2.12 to 2.16 for 208Pb/206Pb and from 0.86 to 0.88 for 207Pb/206Pb. The ratios of one single bronze in the Western Zhou Dynasty fall into the concentration area of Class Ⅱ, indicating that Class Ⅱ lead materials could have been used in a small amount in the Western Zhou Dynasty and this situation continued to the Warring States period.
To know the differences between the lead materials used for the bronzes from Zongyang and those from the surrounding area, the lead isotope ratios of the bronzes unearthed from Shou County , Lu'an , and Heying (Table 4) in the Jianghuai region were selected for a comparative study. It can be seen from Fig.7 that like Zongyang bronzes, those from Shou County, Lu'an, and Heying also used two kinds of lead materials: Class Ⅰ for the bronzes from the Western Zhou Dynasty to the Spring and Autumn period and Class Ⅱ for those in the Warring States period, which is consistent with the results of lead materials of the Zongyang bronzes. The shift in lead materials has aroused interest among some domestic scholars .
According to historical records, wars between the Zhou court and the Huaiyi tribes in the Jianghuai region had been frequent from the middle Western Zhou period onward. During the Spring and Autumn period and the Warring States period, the Jianghuai region became a battleground for the states fighting for supremacy. Hence, we conjecture that the change of the lead isotope ratios of bronzes might have been a by-product of the dynamic of the political arena at that time.
For exploring the sources of Class Ⅰ and Ⅱ lead materials, we selected the lead isotope ratios of the mining areas in Zongyang, Wannan, and Hubei for a comparative study. We can see from Fig.8 that although the lead isotope ratios of the Zongyang bronzes all fall into the ratio range of the mining areas along the Yangtze River in Anhui Province, Class Ⅰ and Class Ⅱ lead materials could come from different mines in this region. The lead isotope ratios of Class Ⅰ all fall into the range of the Wannan mining area, but far from the data of lead ores in Zongyang, which indicates that Class Ⅰ lead materials were probably from Wannan, rather than the Zongyang mining area. The isotope ratios of Class Ⅱ lead materials roughly coincide with those from the Zongyang and the Hubei mining areas.
Based on a scientific analysis of pottery samples and clay cores of the Pre-Qin period from the Tongling and the Feidong areas, it is proved that the bronzes unearthed in Anhui were not all cast in the Central Plain . Given that the bronzes unearthed in Zongyang bear the local characteristics  and that people in Zongyang involved in smelting and casting activities during the late Shang Dynasty, the bronze vessels of Zongyang should have been cast locally. With the exploitation of the local metal resources during the Eastern Zhou dynasty, it was more convenient to use the local mineral resources than those in Hubei which is far away. In addition, scholars have proved that "Jin Dao Xi Hang"(“金道锡行”) , a term recorded in history, meaning the routes along which copper, tin and other bronze raw materials were transported to the Central Plain, did exist in the Shang and Zhou dynasties [26-28], and that there were different routes in Jianghan region and Jianghuai region. Obviously, with the local resources at hand, it is unlikely that people sought long-distance mineral materials from Hubei. We believe that during the Warring States period the Class Ⅱ lead materials must have come from the local mines in Zongyang.
4.2 Trace element analysis
Trace elements in copper alloys can provide useful information for the sources of metals . The trace elements Co, Ni, As, Au, Ag, Se, Te, and Bi are indicative of the provenance of the raw material for the bronze vessels [30, 31]. The above indicative elements in the Zongyang bronzes were measured with LA-ICP-MS by Qin et al. in 2017 , the results shown in Table S5. To further explore the ore material sources of the Zongyang bronzes, we re-analyzed the trace element data with the Oxford Research System.
We used the Oxford Research System, carried out by the archaeometallurgical team at Oxford University, as a main data process method to summarize the chemical composition of Zongyang bronzes. The research system has been discussed in detail and made some achievements in a research of some Chinese bronzes [21-22, 32-41]. In this study, we applied the ‘Copper Groups’ method. Based on the presence/absence (Y/N) of the four most commonly reported trace elements--arsenic (As), antimony (Sb), silver (Ag), and nickel (Ni), the method could allocate the metal composition to one of 16 categories. For most datasets, 0.1% (after mathematical removal of any major alloying elements present and renormalization) is used as a cut-off for the division between presence and absence .
Table 5 reveals the results of the groups analysis. There are two groups present in the Western Zhou and the following Spring and Autumn period: CG9 (As-Ag) and CG12 (As-Sb-Ag). During the Warring States period, CG9 disappeared, whereas CG16 (As-Ag-Sb-Ni) and CG1 (clean metal) appeared, suggesting that the source of the ore material for the Zongyang bronzes changed during this period. However, CG12 continued from the Western Zhou Dynasty to the Warring States period, indicating that one of the ore material could have been used consistently throughout the Zhou Dynasty.
It is worthwhile to pay special attention to the element silver. It can be seen from Table 5 that all the samples contain silver and the silver content is 0.12%-0.81%. Scholars have long hypothesized that silver can offer an independent proxy to characterize lead minerals. It has also been proved that the silver concentration can help to identify the number of the lead sources of the Shang Dynasty bronzes . In the past, much of the world’s silver was obtained from silver-containing minerals, especially galena (PbS), with a small amount coming from copper ores. Galena was smelted to produce lead, which might contain up to 1% silver, while raw copper typically contained < 0.3% silver . Fig.9 shows that the correlations between Ag and Cu are negative, and those between Ag and Pb are positive, which suggests that the silver in the Zongyang bronzes mainly resulted from the addition of lead, with no connection to the base copper.
The results of the trace elements of the Zongyang bronzes are consistent with those of the lead isotope analysis. All the results imply that the provenance of the Zongyang bronzes was stable between the Western Zhou Dynasty and the Spring and Autumn period, whereas the changes of ore source happened during the Warring States period. In the Western Zhou Dynasty and the Spring and Autumn period, on a relatively stable political landscape under the rule of the Qunshu state, the Zongyang area was strongly influenced by the Dongyi and the Qunshu cultures . After 615 BC around the middle of the Spring and Autumn period, as the Qunshu state was annexed by the Chu state , Zongyang became part of the Chu state, where the dominant Chu culture coexisted with other cultures such as the Central Plain, Wu, and Yue . To compete for powerand hegemony, the states of Chu and Wu fought a 60-year war in the south of Yangtze River and the Wu-Yue wars lasted about two decades. The continuous wars inevitably had an impact on the output of copper materials in Wannan and to a large degree shaped the changes of the ore sources of the Zongyang bronzes during the Warring States period.