Two millennia of anthropogenic landscape modification and nutrient loading at Dian Lake,Yunnan Province,China

Abstract

Landscapes have been shaped by human activities for millennia and there is a pressing need to characterize pre-industrial impacts in order to mitigate present-day effects. We present the analysis of two sediment cores from Dian Lake in Yunnan, China, which span 4000 years. We compare cores from the northern and southern ends of the lake to investigate spatial variability in natural and anthropogenic environmental changes in this large (300 km²) lake. To document the initiation of human impacts on the landscape and characterize the attendant changes in the lake water and sediment quality, we rely on organic and inorganic geochemical measurements as well as sedimentology and stratigraphy. The character and magnitude of proxy changes are coherent between the two core sites with slight differences in the timing of events. At both core sites, we find definitive evidence for substantial anthropogenic change beginning AD 100 (1850 yr BP), coincident with the introduction of terraced agriculture. Sedimentological shifts are distinctive and characterized by an increase in magnetic susceptibility values and a visible change to red, fine-grained clay. The geochemistry of this sediment suggests that it was sourced from the eastern catchment of the lake and delivered into the basin following intensive agriculture and soil erosion. Anthropogenic impacts intensify after AD 900 through hydrologic modification and cultural eutrophication resulting from increased nutrient loading. This study presents evidence that human-affected landscapes have been present in this region of China for longer than previously believed and that ‘small-scale’ land use change can have measureable impacts on lakes.

Keywords

anthropogenic change archaeology carbon isotopes China geochemistry landscape erosion nitrogen isotopes paleolimnology

Introduction

The proliferation of the term ‘Anthropocene’ and the debate regarding its characterization as a geologic epoch (Waters et al., 2016) in recent years have arisen out of the need to characterize the long-lasting and substantial impacts humans have on planetary systems. One question that remains regarding the initiation of the Anthropocene is the global nature of human-induced environmental change (Lewis and Maslin, 2015). Paleoenvironmental records in conjunction with archaeological data are a crucial tool used to document the extent of these human activities. The study of lake sediments allows for the clarification of the relationship among humans, climate change, and freshwater ecosystems, both providing records of regional environmental change and constraining temporal questions in archaeological investigations and interpretations.

The Yunnan Province of southwestern China is an ethnically diverse region that has historically existed on the margins of the Chinese empire. Compared with the heart of the Chinese empire, this has created challenges in documenting how human activities have impacted the landscape and what role the landscape has had on influencing human behavior. The need to characterize pre-industrial human impacts has arisen as a pressing concern for Yunnan, which is plagued by severe erosion and metal pollution caused by interactions among rapid population growth (Xu and Wilkes, 2004) and land use change (Diallo et al., 2009). Erosion and contamination problems from new and old industries are likely to persist or intensify as the Chinese population continues to expand and urbanize.

Here, we present a set of sediment cores from Lake Dian in central Yunnan that illustrate the profound role of human activity in modifying the landscape of Southwestern China beginning around AD 100. The purpose of this study is to characterize the timing, magnitude, and duration of anthropogenic impacts to Dian at a location proximal to ongoing archaeological excavations and in doing so investigate the potential couplings between humans and environmental change. We use a variety of proxies from lake sediment cores, including trace metal geochemistry, organic carbon and nitrogen isotopes, ratios of carbon and nitrogen, and magnetic susceptibility (MS), and interpret these changes within the context of archaeological and historical records. We present proxy records from both the southern and northern ends of the lake to investigate the temporal and spatial patterns of human impacts to the lake.

Study site

Dian is a large (surface area of 300 km²) and shallow tectonic lake with a maximum depth of 5 m and an average depth of 4.4 m (Li et al., 2007). The catchment area of Dian is 3000 km², and in the north, it drains the urban watershed of Kunming, a city of nearly 6 million people (Figure 1a). In the south, the watershed is more rural with extensive areas of flat land used for agricultural activities. Most of the surficial catchment geology is composed of Quaternary alluvium with portions of limestone and slate to the west (Bureau of Geology and Mineral Resources of Yunnan Province, 1990) (Figure 1b). Numerous small streams from the eastern side of the catchment drain into Dian, but the only outflow is the artificial Tanglang River to the west (Whitmore et al., 1997). Soils in the east of the catchment, where most of the hydrologic input to Dian originates from, are eutric gleysols, whereas on the western side the soils are a mixture of chromic luvisols and chromic cambisols (Fengrang, 1990) (Figure 1a). Chromic luvisols have a clay-rich brown/red B horizon and chromic cambisols have a weakly developed brown/red B horizon (FAO/UNESCO, 1974). Gleysols have no diagnostic horizons but display characteristics of prolonged water saturation such as grayish-blue coloring and/or reddish/orange mottling of iron minerals (FAO/UNESCO, 1974).

Figure 1.

(a) Central Yunnan lakes and soil map adapted from Fengrang (1990). The hatched shaded area is the Dian catchment. (b) Geologic map of Dian catchment from Bureau of Geology and Mineral Resources of Yunnan Province (1990) and faults with bathymetry (1 m contours) adapted from Zhang et al. (1996).

Temperatures in Kunming are generally mild, ranging between 9°C and 20°C (IAEA/WMO, 2017). Precipitation is highly seasonal due to the Indian Summer Monsoon (ISM) and 80% of annual precipitation falls during the months June–September (IAEA/WMO, 2017). Due to the effects of rain-out, distillation, and monsoonal transport processes (Vuille et al., 2005), the oxygen isotopic composition of precipitation associated with the monsoon season has more negative values, ranging between −8‰ and −13‰ VSMOW (IAEA/WMO, 2017). Calculations of the amount of precipitation arriving during the summer months as well as its isotopic composition show that the average weighted isotopic composition of precipitation at Kunming is −9.9‰ VSMOW. Water samples collected from Dian between the summers of 2012–2015 range between −1.81‰ and −4.66‰ VSMOW, 5‰ to 8‰ higher than average weighted precipitation. Although Dian undoubtedly loses water through outflow to the Tanglang River to the west as well as an unknown proportion through groundwater throughflow, the lake water oxygen isotopes strongly suggest that Dian loses the majority of water through evaporation. Previous research into the water quality of Yunnan lakes observed a lake water isotopic composition of −6.4‰ VSMOW for Dian (Whitmore et al., 1997), demonstrating that in the last 20 years, lake water has become increasingly enriched in heavy isotopes, possibly due to these evaporative processes. Water and sediment quality problems have been noted at Dian over the past few decades, and it is currently classified as Grade V, making it unfit for agricultural or industrial uses (Kahn and Yardle, 2007). It is one of the three most eutrophic lakes in all of China due to its high fertilizer and sewage inputs (Le et al., 2010).

The southern end of the lake presently has a lower population density than the northern end on which Kunming is located. However, prior to establishment of Kunming as the seat of the provincial capital in AD 765 under Nanzhao Kingdom, the southern end was the focus of political and social activity. Beginning with the establishment of Bronze Age occupation mounds in the early to middle first millennium BC under the ‘Dian’ kingdom, the southern end experienced successive population expansions, in particular following Han imperial incorporation and the founding of the provincial capital in the southern end (Fang, 1987; Yao and Zhilong, 2012). Therefore, this study utilizes sediment cores from both northern and southern ends of the lake basin to document spatial variability in anthropogenic impacts associated with differences in settlement patterns.

In regions such as Yunnan, which were typically on the margin of the Chinese dynasties, historical records are sparse (Allard, 1998) and it is difficult to find documentation regarding human impacts to the environment. Paleolimnological studies have been used in other locations such as Central America (Brenner, 1983), North America (Pompeani et al., 2013), and Europe (Brännvall et al., 2001) to supplement existing historical and archaeological data. These types of studies are limited in that human impacts must be of a large enough magnitude to be recorded in the sediment record, but have the advantage of allowing spatial and temporal differences to be explored. Paleolimnological study of the human impacts on central Yunnan lakes has been extensive on lakes such as Xingyun (Chen et al., 2014; Gao et al., 2018; Hillman et al., 2014; Wu et al., 2015, among many others), Qilu (Brenner et al., 1991), and Yangzong (Zhang et al., 2012). At Xingyun and Qilu, indications of human impact are visible by ~AD 500 through the deposition of red, fine-grained clay (Brenner et al., 2000; Hillman et al., 2014; Hodell et al., 1999). Further intensification of human impacts in the last few hundred years can be seen at all central Yunnan lakes through deforestation and increased heavy metal concentrations (Brenner et al., 1999; Liu et al., 2013; Zeng and Wu, 2009; Zhang et al., 2012).

While the recent (last 50 years) deterioration of water and sediment quality has been characterized for Lake Dian (e.g. Le et al., 2010; Yang et al., 2010; Zeng and Wu, 2009; Zhan et al., 2017), paleolimnological work has been relatively limited. Previous study of pollen in Dian sediments focused on analysis of two sediment cores from the northernmost part of the lake over the last 16,000 years (Sun et al., 1986). This study primarily focused on vegetation dynamics driven by climate variability, but did note evidence of human disturbance around AD 500 marked by an increased proportion of clay sediments, increased sedimentation rates, and pollen changes indicating deforestation.

Theoretical framework

This study relies on a multi-proxy approach, including the carbon and nitrogen isotopic composition of organic matter, trace element geochemistry, and MS. Combined, these proxies can reveal natural climate variability, past human–environment interactions, and limnological response to landscape alterations.

The ratio of organic carbon to nitrogen (C/N) and the stable isotopic composition of both carbon (δ¹³C_org) and nitrogen (δ¹⁵N_org) in organic matter are used to assess lake trophic status and infer nutrient loading arising from human activities. Carbon isotope dynamics in lakes are influenced by many factors, including changes in organic and inorganic carbon flux and source, CO₂ exchange with the atmosphere, primary productivity, and organic matter burial and respiration (Diefendorf et al., 2007). Since a full exploration of these factors in the Dian lake system is beyond the present study, δ¹³C_org values are interpreted within the context of C/N ratios, which can indicate the source of the organic matter being preserved in the lake (i.e. terrestrial or aquatic) (Myers, 1994). Terrestrial organic matter typically has higher values of C/N (>14), while aquatic organic matter has lower values of C/N (<10) (Meyers and Laillier-verges, 1999). When changes in organic matter source and changes in terrestrial vegetation are well constrained, organic carbon isotopes have the potential to provide insight into primary productivity because lighter isotopes of carbon tend to be preferentially incorporated into organic matter. With increased primary productivity, ¹²C gradually becomes depleted and ¹³C is increasingly utilized in organic matter production, driving up δ¹³C_org values (Hodell and Schelske, 1998).

While a complete review of nitrogen isotopes is also beyond the scope of this paper, in general δ¹⁵N_org is a function of naturally occurring reactive nitrogen sourced from biological fixation, lightning, and forest fires. Controls on δ¹⁵N_org include primary productivity, nitrification, denitrification, ammonification, and volatilization. Lighter isotopes of nitrogen are preferentially used in biochemical and inorganic processes, leaving the remaining pool of nitrogen relatively enriched in ¹⁵N. Determining the process responsible for shifts in δ¹⁵N_org values in lakes is further complicated by post-depositional loss and inorganic nitrogen adsorbed by clay surfaces (Talbot, 2001). In addition, modern processes such as NO_x deposition from fossil fuel combustion and fertilizer and sewage runoff have perturbed the natural nitrogen cycle and influenced isotopic signature of the sediments (Kendall et al., 2007). Nonetheless, δ¹⁵N_org values when used in conjunction with other proxies such as δ¹³C_org and C/N can yield insight into trophic level shifts (Talbot and Johannessen, 1992), primary productivity (Brenner et al., 1999), cultural eutrophication (Hollander et al., 1992), and sewage inputs (Elliott and Brush, 2006). Consequently, δ¹⁵N_org values of aquatic macrophytes have been found to increase in proportion to the amount of wastewater input (Cole et al., 2004). To more definitively interpret sewage and wastewater inputs, we pair these data with lipid biomarkers, specifically 5β-stanols, which are produced in the intestines of mammals (Bull et al., 2002) and have been used as indicators of fecal contamination in modern (Leeming et al., 1996) and past environments (D’Anjou et al., 2012).

MS and elemental geochemistry are used to assess land use change and sediment inputs arising from soil erosion. MS primarily reflects the concentration of magnetic minerals with secondary influences from mineral composition and crystal size and shape (Dearing, 1994). Since ~50% of the Dian catchment is composed of gleyed soil (Figure 1a), MS is used here to indicate increased inputs of gleyed soil arising from erosional processes occurring within the watershed. Gleyed soils do not have any diagnostic horizons but rather are characterized by poor drainage and anoxic conditions, leading to the build-up of ferromagnetic elements such as Fe, Co, Mn, Mo, and V and a deficiency of Cu and Zn (Kabata-Pendias, 2011). Therefore, the concentrations of metals weakly sorbed to lake sediments are additionally used to indicate erosional inputs arising from both natural and human activities (Boyle, 2002) and supplement the MS data.

Methods

Core collection

In 2012, two sets of cores (A-12 and B-12) were collected in the same location from the northern end of the lake (Figure 1b). From site B-12, a piston corer was used to collect a 56-cm surface core that preserved the water–sediment interface. The upper 20 cm was sliced in the field at 0.5 cm intervals and used for geochemical analysis and ²¹⁰Pb dating. From site A-12, the upper 30 cm of sediment was not preserved due to the high water content and extremely flocculent sediment at the water–sediment interface. A Livingston corer was used to collect deeper samples (Wright et al., 1984) in 1 m sections. Five drives were collected, each overlapping by approximately 20 cm. Each of the A-12 drives and the B-12 surface core were correlated on the basis of field measurements and confirmed with stratigraphic correlation of MS profiles and sediment composition (such as organic matter) to form a composite record of 395 cm.

In 2014, the F-14 core was collected from the southern end of the basin (Figure 1b). A piston corer was used to collect surface sediments that preserved the water–sediment interface, and the upper 20 cm was sliced in the field at 0.5 cm and used for geochemical analysis and ²¹⁰Pb dating. Deeper samples were collected using a Livingston corer (Wright et al., 1984) in 1 m sections. Five drives were collected, each overlapping by approximately 20–30 cm. These drives were correlated on the basis of field measurements and confirmed with stratigraphic correlation of MS profiles and sediment composition (such as organic matter and MS). The composite record from the F-14 site is 383 cm.

Soil samples were collected from approximately 24°42′25′N, 102°41′22″E at an archaeological excavation pit in order to characterize potential sources of erosive materials to the lake (Supplemental Figure S1, available online). Sample 1 was taken 211 cm below the surface in the B horizon which was comprised of reddish/brown organic-rich clay (~50% clay, 30% silt, 20% sand estimated by feel in the field) matrix intermixed with a shell midden layer and archaeological artifacts. Sample 2 was taken at 286 cm below the surface and had similar soil properties as sample 1 but did not have any archaeological artifacts in it. Sample 3 was taken 288 cm below the surface at a transition into mottled grey/blue gleyed clay (~90% clay, <10% silt and sand, estimated by feel in the field) with high cohesiveness. Sample 4 was taken 316 cm below the surface and had similar soil properties as sample 3 but was fully in the gleyed unit.

Age control

Radiocarbon ages were measured on identifiable terrestrial macrofossils (e.g. charcoal and leaves). Seven dates were measured from the A-12 cores and 10 from the F-14 cores (Table 1). Samples were analyzed at Keck Center for Accelerator Mass Spectrometry at the University of California Irvine. Prior to analysis, samples were pretreated using a standard acid–base–acid procedure (Olsson, 1986). The resulting ages were calibrated using CALIB 7.0 and the INTCAL13 calibration curve (Reimer et al., 2013). The upper 20 cm of the B-12 and F-14 cores was lyophilized and analyzed for ²¹⁰Pb, ²¹⁴Pb, and ¹³⁷Cs activities by direct gamma (γ) counting in a broad energy germanium detector (Canberra BE-3825) at the University of Pittsburgh.

Table 1.

AMS radiocarbon dates from Dian cores.

UCI number	Composite core depth (cm)	Material	¹⁴C age (BP)	Error ±	Median probability calibrated age (yr BP)	2σ calibrated age range (yr BP)
A-12
116862	36.5	Plants	795	35	713	672–810
116863	97.5	Plants and charcoal	1140	20	1027	974–1172
201166	139.0	Charcoal	2745	15	2826	2768–2868
116864	147.5	Plants and charcoal	2940	35	3098	2976–3208
122328	272.0	Charcoal	10,285	50	12,064	11,827–12,380
116865	351.5	Charcoal	13,110	60	15,736	15,457–15,985
116866	354.5	Charcoal	12,670	40	15,093	14,855–15,310
F-14
201167	78.5	Seeds	815	20	719	686–733
152051	92.0	Charcoal and wood	585	50	598	526–657
152052	152.0	Wood	2260	15	2312	2183–2341
201168	198.0	Charcoal	6575	20	7467	7431–7508
172625*	214.5	Charcoal	2235	15	2220	2159–2328
180963	276.5	Charcoal	9395	15	10,631	10,572–10,690
152053	285.0	Charcoal and leaf	9580	40	10,934	10,742–11,122
152054	320.5	Wood	10,810	140	12,726	12,428–13,038
180964	330	Wood	12,390	20	14,444	14,193–14,695
152055	349.5	Wood	15,275	50	18,550	18,395–18,695

date excluded from age model.

Geochemistry

All sediment cores not sliced in the field were split lengthwise and immediately described following the methods of Schnurrenberger et al. (2003). Sedimentary structures, grain size by feel, and Munsell color were characterized for each core. Water content and bulk density were measured at 2 cm intervals using 1 cm³ samples from all sets of cores. Weight percent organic matter and carbonate content within these same samples was determined by loss-on-ignition (LOI) analysis at 550°C and 1000°C, respectively (Dean, 1974). Sediment core MS was measured on all split cores at room temperature using a Bartington^® Instruments Ltd. ME2EI surface-scanning sensor equipped with a TAMISCAN-TSI automatic logging conveyer. Since the upper 20 cm of sediment from both coring locations was sliced in the field, MS measurements on these intervals are lacking.

Weight percent nitrogen, weight percent organic carbon, δ¹⁵N_org, δ¹³C_org, and atomic C/N ratio were measured at 2 cm intervals from the A-12 cores and at 4 cm intervals from the F-14 cores. Samples were covered in 1 M HCl for 24 h to dissolve carbonate minerals, rinsed to neutral, and lyophilized. Samples from the A-12 cores were analyzed at Idaho State University using an ECS 4010 (Elemental Combustion System 4010) interfaced to a Delta V mass spectrometer through the ConFlo IV system. Samples from the F-14 cores were analyzed at the University of Arizona using a continuous-flow gas-ratio mass spectrometer (Finnigan Delta Plus XL) coupled to an elemental analyzer (Costech). Organic carbon isotopes are expressed in conventional delta (δ) notation as the per mil (‰) deviation from the Vienna Peedee Belemnite standard (VPDB), whereas nitrogen isotopes are reported relative to atmospheric N₂.

Weakly sorbed metal concentrations were measured on the soil samples and at 3–5 cm intervals on the A-12 and F-14 cores. All samples were lyophilized and homogenized prior to analysis. Elements were extracted by reacting ~0.1 g of sample with 10 mL of 1 M HNO₃ for ~24 h (Graney et al., 1995). The supernatant was diluted before being measured on an inductively coupled plasma mass spectrometer (ICP-MS) at the University of Pittsburgh. Every 10th sample, the first sample was measured again as a duplicate, resulting in a total of 15 replicate measurements. For the metals presented here, samples were on average within 5% of each other for V, 4% for Cr, 2% for Fe, 3% for Co, 3% for Ni, 1% for Zn, and 1% for Pb. Blanks were measured every 10 samples to check for bleed-through.

Five samples were analyzed for organic biomarker compounds at the Lamont-Doherty Earth Observatory Organic Geochemistry Lab from the A-12 cores following the methods outlined in Polissar and Freeman (2010). Freeze-dried sediments were extracted with a Dionex Accelerated Solvent Extraction (ASE) system. Two to five grams of dried sediment was placed in stainless steel sample holders and extracted at 100°C and 1000 psi with 10% (v/v) methanol/dichloromethane with a total volume of 50 mL. These total lipid extracts (TLEs) were evaporated to near dryness with N₂ in a Turbovap solvent evaporator and transferred to 4 mL borosilicate vials with Teflonlined caps. All remaining solvent was then evaporated and the TLE stored in a few drops of hexane at 4°C. TLEs were separated into aliphatic (F1), ketone/alcohol/acid (F2), and polar (F3) fractions with silica gel column chromatography. Silica gel was transferred in hexane to the SPE column and the column then rinsed twice with 6 mL hexane. The TLE was loaded on the column in 100 μL hexane and the F1, F2, and F3 fractions eluted with a sequence of 5 mL of 10% dichloromethane in hexane, 8 mL ethyl acetate, and 5 mL methanol. Compounds were characterized by gas chromatography–mass spectrometry (GC-MS) using an Agilent 6890 GC and an Agilent 5973 quadrapole mass spectrometer. Compounds were identified by elution time, comparison with published spectra, and standards.

Results

Geochronology

We attempted to date the upper 20 cm of both the B-12 and F-14 cores using the constant rate of supply (CRS) ²¹⁰Pb age model method (Appleby and Oldfield, 1983); however, excess ²¹⁰Pb never reached background values in either core. ¹³⁷Cs peaks from AD 1963 are commonly found in lake sediments (Ritchie and McHenry, 1990), but no ¹³⁷Cs activity was detected in either B-12 or F-14 cores. This may be because Dian is a shallow lake, making it more susceptible to wind-driven mixing, which can cause homogenization of radioisotopes (Binford et al., 1993). Alternatively, Dian may have been dredged, removing the most recent sediment from our coring locations.

While this paper is only focused on the last 4000 years, the entirety of the radiocarbon age results is presented here since it serves to anchor the age models for each core (Table 1, Figure 2). The calibrated radiocarbon dates were used in the BACON code (Blaauw and Christen, 2011) in the statistical software package ‘R’ (R Core Team, 2013). The BACON age–depth modeling code uses Markov chain Monte Carlo statistics to create age–depth models and uses posterior probabilities to determine radiocarbon outliers. In the A-12 core, the BACON age–depth modeling software suggests that the base of the core is ~3000 years older than the radiocarbon dates at 351.5 and 354.5 cm. BACON likely identified these dates as outliers on the basis of prior sediment accumulation rates. It is possible that this approach is incorrect and that the base of the core should span 14,000 yr BP instead of 17,000 yr BP; however, this would not substantially change the timing of any of the events focused on in this study. Of particular note for the F-14 core, the radiocarbon date from 214.5 cm is much younger than would otherwise be expected given the ages both above (152 and 198 cm) and below (276.5 cm). We therefore conclude that this date may have been contaminated by modern carbon during processing and pre-treatment, subsequently making it appear younger than its true age. On the basis of these age models, the last 4000 years span 163 cm in the A-12 core and 168 cm in the F-14 core.

Figure 2.

Sedimentology, BACON age–depth models with 95% confidence intervals and dates (blue circles) with 2σ error bars, and sedimentation rate.

Sedimentology

The sedimentology of the A-12 and F-14 cores is broadly similar, albeit units occur at slightly different depths and corresponding ages (Table 2). Until unit 3, diatoms are present in both cores and comprise <10% of sediment on the basis of estimation from smear slides. In unit 3, diatoms are absent and there are abundant charcoal fragments (~10%). The color of unit 3 was black (5 YR 1/1) when opened but changed color to bright reddish brown (2.5 YR 4/8) within 1–2 h.

Table 2.

Sedimentology of the A-12 and F-14 cores.

Unit	A-12 depth (cm)	A-12 age (year AD/BC)	F-14 depth (cm)	F-14 age (year AD/BC)
1. Massive, very fine-grained oxidized clay	104–0	AD 600–present	68–0	AD 1300–present
Gradational contact	109–104	AD 360–600	78–68	AD 1200–1300
2. Massive brownish gray (5YR 5/1) calcareous fine-grained clay	150–109	1220 BC–AD 360	160–78	950 BC–AD 1200
Gradational contact	160–150	1840–1220 BC	170–160	2300–950 BC
3. Massive reddish gray (2.5 YR 4/1) fine-grained clay	395–160	15,400–1840 BC	383–170	18,200–2300 BC

Approximately 30–50 cm prior to the visible change to the red clay layer in both cores, MS values begin to increase in both cores (Figure 3). Throughout the rest of this paper, this horizon characterized by high MS values and a change in color and grain size will be referred to as the ‘red clay layer’.

Figure 3.

A comparison of proxies from A-12 (blue solid line) and F-14 (black dotted line) over the last 4000 years: (a) Weight percent carbonate, (b) nitrogen isotope values of organic matter, (c) carbon isotope values of organic matter, (d) weight percent carbon, (e) ratio of organic carbon to nitrogen, and (f) magnetic susceptibility (MS) values. Red bars indicate the gradual increasingly intensive human impact on the lake. The red dotted lines indicate notable archaeological/historical events around the lake. Black rectangles are Yunnan cultural periods and white rectangles are Chinese Dynasties.

Geochemistry of the A-12 core

From 2000 to 700 BC, sediments gradually increase to 15% weight percent carbonate (Figure 3). All other proxies (δ¹⁵N_org, δ¹³C_org, weight percent carbon, C/N, and MS) are stable at an average of 4.1‰, −28.43‰, 2.5%, 10.86, and 63 SI units, respectively. From 700 BC to AD 90, sediment composition fluctuates between 10% and 15% weight percent carbonate. At the same time, δ¹³C_org values begin to gradually increase to −27‰ and C/N values gradually decline to 10. Weight percent carbon, δ¹⁵N_org, and MS values continue to remain stable.

Around AD 90, coincident with the beginning of an increase in MS values, carbonate content decreases from 15% to <1%. Concentrations of weakly bound Co, Ni, and Pb roughly double, while Fe increases by 250% (Figure 4). In addition, sedimentation rates increase from a stable average of 0.02 cm/yr to 0.14 cm/yr (Figure 2). The sediments visibly transition into the red clay layer around AD 600. At around AD 800, weight percent carbon and C/N ratio decline rapidly to 0.8% and 4, respectively, while most other proxies remain stable (Figure 3). The upper 28 cm of these sediment cores (~AD 1400) have higher weight percent carbonate (~30%) and organic carbon (9%), as well as an increase in δ¹³C_org and δ¹⁵N_org values by 3.4‰ and 7.0‰, respectively.

Figure 4.

Concentrations of weakly bound metals from A-12 (blue solid line) and F-14 (black dotted line): (a) iron (Fe), (b) nickel (Ni), (c) lead (Pb), and (d) cobalt (Co).

Geochemistry of the F-14 core

From 2000 to 900 BC, sediments are composed of <5 weight percent carbonate. Sediment composition gradually increases to 10–12 weight percent carbonate and remains relatively stable until ~AD 100 from 900 to 400 BC (Figure 3). From 2000 to 900 BC, δ¹³C_org, δ¹⁵N_org, weight percent carbon, and C/N values are stable at an average of −28.5‰, 4.9‰, 2.8%, and 11.9, respectively. MS values are relatively low and stable, ranging between 40 and 60 SI units. From 900 BC to AD 40, δ¹³C_org increases by 2.3‰, δ¹⁵N_org remains stable, weight percent carbon slightly declines to ~2.0%, and C/N ratio gradually declines to 10.

Around AD 100, coincident with the beginning of an increase in MS values, carbonate content decreases from 12% to <1%. Concentrations of weakly bound Fe, Co, and Pb roughly double, while Ni increases by ~50% (Figure 4). In addition, δ¹³C_org values gradually increase to −25.7‰ and C/N ratio gradually declines to 9. From AD 1000 to 1200, weight percent carbonate, weight percent carbon, and δ¹³C_org values briefly increase to a maximum of 16%, 3.2%, and −22.4‰, respectively, and MS values dip slightly. At around AD 1200, the sediments visibly transition into the red clay unit and sedimentation rates increase from a stable average of 0.04 cm/yr to 0.10 cm/yr. Coincident with this visible change in sedimentology, δ¹³C_org values decrease to around −26.3‰, weight percent carbonate declines to <5%, weight percent carbon decreases rapidly to ~1.2%, and MS values further increase. The upper 22 cm of these sediment cores (~AD 1800) has higher δ¹³C_org values of −20.0‰, δ¹⁵N_org values of 12.3‰, weight percent carbon of 11.7%, and weight percent carbonate of 30%.

Geochemistry of the soil samples

Weak acid digestions indicate that soil samples 1 and 2 are relatively high in elements such as Co, Ni, P, and Zn (Supplemental Figure S2, available online). Sample 3 has relatively lower concentrations of many elements such as Al, Ca, K, Mg, and Sr, but is noticeably enriched in Ti and V. Sample 4 has low concentrations of P, Ti, and Zn but high concentrations of Ca, Fe, K, Mn, Se, and V.

Discussion

Pre-anthropogenic disturbance (<AD 100)

Prior to AD 100, C/N ratios between 11 and 12 and δ¹³C values around −28.5‰ in both the A-12 and F-14 cores (Figure 3) suggest the Dian sediments were composed of a mix of aquatic and terrestrial organic matter, with the majority of the terrestrial vegetation being produced by C3 plants (Myers, 1994). The low and stable MS values, low concentrations of ferromagnetic metals, and low concentrations of lead (Pb) in both A-12 and F-14 (Figure 4) also suggest moderate inorganic inputs arising from either aeolian or fluvial processes.

Small changes in organic matter proxies occur in both cores around 700 BC (A-12) and 900 BC (F-14) and continue until AD 90 (A-12) and AD 100 (F-14). The decrease in C/N ratio is primarily driven by a decline in weight percent organic carbon. Lower C/N ratios suggest more aquatic sources of organic matter, which combined with increased δ¹³C values suggest an increase in primary productivity as lighter isotopes of carbon are consumed.

These lower C/N values are accompanied by an increase in weight percent carbonate in both cores. The increased carbonate content could be explained multiple ways. The authigenic precipitation of carbonate minerals can be caused by the supersaturation of CO₂ within the water column, often due to a decrease in dissolved CO₂. An increase in algal productivity, as suggested by an increase in δ¹³C values and decrease in C/N ratios, may be responsible for a decrease in dissolved CO₂ and the precipitation of carbonate (Teranes et al., 1999). Other possible mechanisms for increased authigenic carbonate deposition include (1) increased evaporation and loss of CO₂ (Kelts and Hsu, 1978), which may commonly be caused by a drop in lake level, or (2) an increase in water temperature resulting in reduced CO₂ solubility (Talbot, 1990).

Alternatively, the increased carbonate may not be authigenic material and may be material washing in from the catchment since there are areas of limestone (Figure 1b). If so, this may indicate increased weathering and erosion within the catchment. However, chemical weathering and erosion usually increase with increased rainfall, and by 2000 BC, there are indications of substantial aridity on the Tibetan Plateau due to a weakening of the ISM (Bird et al., 2014; Cai et al., 2012; Morrill et al., 2006). Palynological records from nearby Yunnan Lakes such as Tiancai (Xiao et al., 2014), Xing Yun (Chen et al., 2014), and Erhai (Shen et al., 2006) show no indications of pronounced shifts in temperature during this time and confirm a slight and gradual increase in aridity.

With the available proxy data at Dian, it is difficult to attribute this increase in primary productivity and carbonate content broadly from 900 BC to AD 100 to a specific cause. Regional records would suggest that hydroclimate change probably did not result in an increase in weathering and erosion and that lake levels may have declined due to aridity, although this would need to be confirmed with additional proxy data. Alternatively, we cannot discount the possibility of human activities resulting in increased weathering and erosion causing increased amounts of carbonate and nutrients being washed into the lake, driving this primary productivity and carbonate deposition.

Many archaeological sites including shell mounds, settlement structures, and rice remains on the southern end of the lake date to ~750 BC (Yao et al., 2015), very close in timing to the observed geochemical changes in the sediment core within the limits of the age models (Figure 3). The presence of slags associated with smelting and melting of copper ores at many of these sites indicates intensification of metal working in the Dian lake basin, an economic activity that has also been detected at neighboring Fuxian, Xingyun, and Qilu lakeside settlements of this period (Yao and Zhilong, 2012; Yunnan Provincial Institute of Cultural Relics and Archaeology et al., 2015; Zou et al., 2017). Permanent human settlements and nutrient loading associated with agriculture, fertilization, metal working, and/or runoff may have driven primary productivity changes at Dian; however, without additional proxy data, this human influence cannot be definitively demonstrated. Therefore, we suggest that anthropogenic impacts on the lake were relatively limited prior to ~AD 100. That is not to say human impacts on the landscape were absent or non-existent; rather, our current data suggest that anthropogenic impacts were not large enough or persistent enough to be unambiguously recorded in the Dian sediments.

Initiation of intensive human impact (beginning at AD 100)

Though gradual, the initiation of human impacts on Lake Dian is distinctive and characterized by (1) an increase in sedimentation rate (Figure 2), (2) an increase in MS values (Figure 3) and a concomitant increase in the concentrations of ferromagnetic metals (Figure 4), (3) a decline in weight percent carbonate (Figure 3), and (4) a visible change in color and grain size. In both sediment cores, the change in color and grain size does not take place until MS values reach a threshold of around 100 SI units. Particularly for the F-14 core, there is an 1100-year lag between the decline in carbonate content and the visible appearance of the red sediment color. There are several possibilities to explain this lag. These could be two distinct units, one characterized by high MS only and one characterized by high MS and a distinct red color from high iron content. Each of these units may represent erosion of different types of soils. Alternatively, there may have been a lag in the transport of iron. As iron was mobilized from soil erosion, it may have first formed iron coatings on the fine-grained sediments in fluvial systems before being transported to the lake (Whitney, 1975). Similarly, iron coatings could have been progressively stripped from sediments following years of disturbance (Carpenter and Hayes, 1980).

Previous study of Dian sediment cores noted similar changes and interpreted them as evidence of human impacts on the landscape (Sun et al., 1986). The timing of this transition was thought to be around AD 500, relatively close to the results suggested by our own age models, despite bulk sediment radiocarbon measurements being used in the previous study. Sediments in the red clay layer exhibit many of the same characteristics noted in previous studies of Yunnan lake sediment cores (Xingyun (Hillman et al., 2014), Qilu (Brenner et al., 1991), and Chenghai (Hillman et al., 2016)) and indicate a similar time period of initiation (~AD 500 in Xingyun and Qilu). Previous work attributed these changes to the onset of intensive land use and noted that the low carbonate content, high residual mineral matter, and red, fine-grained sedimentology are similar to what was found in Guatemalan lakes known to have been impacted by human deforestation and erosion (Binford et al., 1987; Brenner, 1983).

The transition into this distinctive unit at Dian coincides closely, within the errors of the age models for both the A-12 and F-14 cores, with documentation from AD 19 that the Han Governor built water tanks, opened irrigation channels, and opened up nearly 121 km² of new tracts of farm land (Sun and Xiong 1983) (Figure 3). Therefore, we conclude that the deposition of this layer throughout Dian, as well as a number of other Yunnan lakes, is the result of soil erosion associated with whole-scale landscape change, agriculture, and/or intense working of existing sediments within the lake catchment. The earlier occurrence of these changes at Dian relative to other Yunnan lakes may be because it is located proximal to Kunming, a larger center of population than the regional towns near lakes such as Xingyun and Qilu. Population density estimates are limited for this period of time; rough calculations suggest 25 people/km² for the entirety of the Yunnan province (Ban, 1938), but regional information is not given. It is therefore difficult to use historical records to explore the spatial distribution of human impacts as related to population density during this time. The similar timing of the transition into this unit in both the A-12 and F-14 cores suggests comparable patterns of land use between the northern and southern catchment areas of the lake or extensive mixing and circulation within the lake. The apparently less rapid increase in MS values at the F-14 site may be due to the relatively fewer inflowing rivers carrying sediment from portions of the catchment composed of eutric gleysols compared with the northern end of the lake near core site A-12 (Figure 1a).

Eutric gleysols compose a large portion of the Dian catchment, especially from the eastern side of the basin where the majority of inflowing water originates from (Fengrang, 1990) (Figure 1a). Gleysols are common in water-logged areas where the oxidation of organic matter causes the accumulation of Fe²⁺ as well as other redox-sensitive elements (Blume, 1988). The presence of ferromagnetic minerals results in high MS values; therefore, we would expect the increased delivery of material derived from eutric gleysols to drive MS values up. The geochemistry of the soil sample that was within the gleyed horizon (#4) was depleted in Zn but rich in Ca, Fe, K, Mg, Mn, and V relative to the B horizon samples (such as #1) (Supplemental Figure S2, available online), thereby displaying many of the geochemical characteristics expected of gleyed horizons (Kabata-Pendias, 2011). To distinguish the soil profile samples from each other and characterize them as potential sources of sediment to the lake, we examined patterns in elemental ratios, a method that reveals source signatures by normalizing large variations in individual elemental concentrations. The mixing space of lead to vanadium (Pb/V) and chromium to zinc (Cr/Zn) was found to suggest mixing of a small number (2-3) of distinct sources of metal fluxes (i.e. endmembers) (Figure 5).

Figure 5.

Molar elemental ratios of Pb/V and Cr/Zn for sediment cores (A-12 = blue circles, F-14 = black squares), soil samples (1 and 2 = ‘B horizon’ in red, 3 and 4 = ‘gleyed horizon’ in green), and the average upper continental crust (UCC) of the Yangtze craton (purple hexagon) (Gao et al., 1998). Shifts toward the left of the mixing space can be seen in the sediment cores beginning around AD 60 as a result of soil erosion. Further shifts toward the top begin around AD 300 as a result of the intensification of erosional processes.

Molar elemental ratios of the soil samples indicate that the B horizon samples (#1 and 2) have relatively high Pb/V but low Cr/Zn ratios, whereas the samples within the gleyed horizon (#3 and 4) have lower Pb/V and higher Cr/Zn ratios (Figure 5). Prior to AD 10, the molar elemental ratios of weakly bound metals from both A-12 and F-14 sediment cores were characterized by high Pb/V and low Cr/Zn ratios. Around AD 60 and just prior to the increase in MS values, samples trend toward lower Pb/V and higher Cr/Zn ratios, suggesting that this layer may be partially sourced from the gleyed horizon. The eastern shores of the lake are where the soil profile is dominated by gleyed horizons (Figure 1b), suggesting that the eastern catchment may be the primary source of erosive material and thus where land use change was most intense. The change in these molar elemental ratios through time suggests an evolution of erosion, possibly as channels of erosion cut through the B horizon into the gleyed units. A soil map of Yunnan indicates that other lakes with the red clay layer (e.g. Xingyun, Qilu) have watersheds where gleyed horizons are commonly found (Figure 1b).

While additional soil pits would be needed to more definitively attribute the source of this sediment, our interpretation fits well with other sedimentological indicators – the delivery of iron-rich soil to the lake, the reduction of iron under anoxic lake conditions, and the subsequent re-oxidation of the material upon exposure to oxygen can account for many of the properties of this layer in the lake. In addition, it can explain the increase in MS values in this unit.

Further shifts in the molar elemental ratios occur around AD 300 closer toward values that are representative of the upper continental crust (UCC) of the Yangtze craton, of which northeast Yunnan is a part of (Gao et al., 1998). In the A-12 cores, this occurs just after the visible transition into the red clay sediments. This may suggest that there was a progression of erosion from the upper soil layers (such as the B horizon) down to lower soil layers (such as the gleyed horizon) and further down to an increasingly bedrock-like source.

Intensification of human-driven environmental change (after AD 900)

Following the deposition of the red clay layer in the A-12 cores and further increases in sedimentation rate in the F-14 cores, C/N ratios drop to their lowest values from AD 900 to 1100, which is driven primarily by a decrease in weight percent organic carbon (Figure 3). This is not accompanied by major changes in other organic proxies, and the only other proxy that shows variability during this time is MS values, which continue to increase. Organic carbon loss may be driven by remineralization of particulate organic matter, where the most labile parts are broken down and recycled during suspension and settling. This process can occur without concomitant changes in δ¹³C_org values (Meyers and Eadie, 1993). However, normally when remineralization occurs, nitrogen is more labile than carbon and nitrogen losses are greater than carbon losses, resulting in an increase in C/N ratio (Galman et al., 2008) and a decrease in δ¹⁵N_org values (Galman et al., 2009).

Alternatively, these low C/N ratios may suggest a change in organic matter source, with a relatively lower proportion of carbon and a relatively higher proportion of nitrogen indicating more aquatic organic material. However, a C/N ratio of 4 in the A-12 cores is less than that of the Redfield ratio (6.6) (Muller, 1977), which suggests that there may be a source of inorganic nitrogen at the northern end of the lake. A plot of %C against %N shows that in the A-12 cores there is an intercept of 0.0317 %N at 0 %C (Supplemental Figure S3, available online), showing the influence of inorganic nitrogen (Hu et al., 2006). Therefore, an increased influx of sediment, primarily composed of clay, may have increased nitrogen through adsorption. The δ¹⁵N values of inorganic nitrogen tend to be more negative than organic nitrogen (Schubert and Calvert, 2001), potentially influencing the isotopic signal of the A-12 cores through this interval.

The changes in the F-14 core from AD 1000 to 1200, namely, the increase in weight percent carbonate, organic carbon, and δ¹³C_org values, are close in timing to when dredging of the Tanglang River, on the southwestern side of the lake (Figure 1b), took place due to flooding problems in AD 1273 and 1301 (Yunnan Publishing Workbook, 1989). Population densities for the city of Kunming are unknown at this time, but the city of Dali in western Yunnan had densities of roughly 100 people/km² by AD 1379. Lower lake levels in the southern basin may have resulted in degassing of CO₂, causing the precipitation of carbonate. Alternatively, or in conjunction with these processes, lower lake levels may have increased water temperatures which promoted algal growth and could explain the increased weight percent organic carbon as well as higher δ¹³C_org values. Shallower lakes are also more susceptible to wind-driven mixing which can overturn nutrient-rich bottom waters and stimulate primary productivity. However, mixing is usually accompanied by a decrease in δ¹³C_org values rather than an increase (Hodell and Schelske, 1998).

The increased δ¹⁵N_org and δ¹³C_org values and the increased weight percent carbonate and organic carbon in both A-12 and F-14 over the last ~600 years are suggestive of cultural eutrophication. Lakes impacted by cultural eutrophication are characterized by high δ¹⁵N_org values (Brenner et al., 1999) and high sediment δ¹³C_org values (Schindler et al., 2008), typically due to high fertilizer and sewage inputs. Less dissolved CO₂ due to intense photosynthesis from algal productivity is a likely explanation for the increased carbonate precipitation. These changes are noticeable in the A-12 core beginning around AD 1400 but do not take place in the F-14 core until AD 1800. This difference in timing may be due to poorly constrained age control at the top of the cores from a lack of a ²¹⁰Pb profile or ¹³⁷Cs peak. Alternatively, this difference may reflect real heterogeneity in land use change and population change. The A-12 cores from the northern basin are in close proximity to the city of Kunming, whose increasing population at this time (Lee, 1982) is a likely source of nutrient loading. Rapid increases in population density likely stressed agricultural systems and may have led to intense fertilization (Lee, 1982).

An analysis of fecal 5β-stanols from the A-12 core, including coprostinol and epi-coprostanol, shows an increase in concentration at ~AD 1500 (Supplemental Figure S4, available online). Fecal 5β-stanols can be used as indicators of fecal contamination in modern aquatic environments (Leeming and Nichols, 1996) and have been used as evidence of human occupation of a watershed (D’Anjou et al., 2012). Increases in these compounds suggest that inputs of sewage to the lake are driving the cultural eutrophication, which is further reflected in the δ¹⁵N_org values reaching as high as 11.5‰ (A-12) and 12.25‰ (F-14) in the uppermost sediments (Figure 3). These findings fit well into the historical context. By AD 1550, population densities in Kunming were as high as 456 people/km² due to the influx of nearly 1 million settlers into Yunnan from the 13th–16th centuries (Lee, 1982).

Implications

The results of this study show evidence that pre-industrial human impacts are measurable in Dian Lake by at least AD 100, if not earlier. While ideally we would have more definitive evidence either confirming or denying potentially earlier human impacts around 900 BC, we can conclude that humans measurably impacted the landscape in this region of China ~2000 years ago. The close correspondence between archaeological data and the changes in lake sedimentology and carbon cycle suggest that human activities have been the dominant driver of environmental change in the late Holocene; this adds an additional constraint to both archaeologists and climatologists wishing to characterize natural and anthropogenic change. While previously anthropogenic impacts were thought to be relatively limited in the Yunnan region prior to ~AD 1700 (Elvin et al., 2002; Lee, 1982), we have shown that earlier events should be taken into account when considering landscape change in this region.

While land use change and cultural eutrophication are often perceived to be a modern-day problem that has plagued lakes in only the last ~50–100 years, this study presents evidence that it can be a historically severe environmental issue. This study, as well as several other previous studies (Brenner et al., 1991; Hillman et al., 2014, 2016; Hodell et al., 1999), has shown that Yunnan lake catchments have been subject to a great deal of soil erosion and subsequent metal contamination. A pervasive feature of the lake sediment records within central Yunnan is an increase in MS values followed by the visible appearance of a red clay layer that dates to between ~AD 100 and 500 (Brenner et al., 1991; Hillman et al., 2014; Hodell et al., 1999). Signs of cultural eutrophication occur several centuries following the deposition of this red clay layer at Dian (this study) and Xingyun (Hillman et al., 2014). Notably, all the lakes that display the red clay layer are dominated by gleyed soil horizons in the catchment (Figure 1a). As soil erosion associated with urban development and the expansion of agriculture continues to increase with population, this study suggests that mitigation of water and sediment quality issues must consider the several millennia of legacy environmental degradation.

Conclusion

In this study we compared sediment cores from the northern and southern basin of Lake Dian and found general coherence in the character and magnitude of proxy changes over the last 4000 years, albeit with some slight differences in the timing of events, which is likely due to chronological error associated with age–depth models. At both sites, there is tentative evidence for anthropogenic impacts at Dian Lake starting around 900 BC, close in timing to the initiation of settlements and metalworking around the lake (750 BC), although more definitive proxy evidence will be needed to suggest a link with confidence. After AD 100, changes in the Dian sediment record are clearly driven by human activities and closely correspond with recorded cultural and historical events. The initiation of anthropogenic disturbance within the lake catchment occurs nearly concurrently to the introduction of terraced agriculture as well as other regional records of intensive settlement and land use (Fang, 1987; Lee, 1982). The beginning of a trophic status shift in the lake is recorded ~AD 900 and is likely the result of cultural eutrophication which has intensified with time. This study adds to a growing number of records that show early human impacts on the Yunnan Plateau that are equal in magnitude to post-industrial impacts.

This study has suggested that the deposition of the red, iron-rich, very fine-grained clay layer is the result of human-induced erosion of gleyed soil horizons. As deforestation (Diallo et al., 2009) and soil loss (Barton et al., 2004) become a growing concern for China, and particularly Yunnan, it is critical that there is a recognition that these processes have been taking place over millennia. Consequently, mitigation of these detrimental processes must contend not just with the impacts of the last few hundred years associated with modern industrial-scale agriculture but the aggregated ‘small-scale’ agricultural practices of several thousand years of history that had a measurable impact on the landscape and functioning of lakes.

Supplemental Material

Supplemental_Material – Supplemental material for Two millennia of anthropogenic landscape modification and nutrient loading at Dian Lake, Yunnan Province, China

Supplemental material, Supplemental_Material for Two millennia of anthropogenic landscape modification and nutrient loading at Dian Lake, Yunnan Province, China by Aubrey L Hillman, Alice Yao, Mark B Abbott and Daniel J Bain in The Holocene

Footnotes

Acknowledgements

We thank David Pompeani for useful feedback and discussion and Matt Finkenbinder for assistance in the measurement of radiocarbon samples. We thank JunQing Yu, ChunLiang Gao, Jordan Abbott, Lauren Ledin, and Duo Wu for assistance with fieldwork. We thank two anonymous reviewers for their helpful comments and suggestions.

Funding

This work was financially supported by the US National Science Foundation (BCS-1357347). Aubrey Hillman recognizes the support of the Byrd Post-Doctoral Research Fellowship during the preparation of this manuscript.

Supplemental material

Supplemental material for this article is available online.

References

Allard

(1998) Stirrings in the periphery: History, archaeology, and the study of Dian. International Journal of Historical Archaeology 2: 321–341.

Appleby

Oldfield

(1983) The assessment of 210Pb data from sites with varying sediment accumulation rates. Hydrobiologia 103: 29–35.

Ban

(1938) The History of the Former Han Dynasty (trans. Dubs

), 3 vols. Baltimore, MD: Waverly.

Barton

Fullen

Mitchell

et al . (2004) Effects of soil conservation measures on erosion rates and crop productivity on subtropical Ultisols in Yunnan Province, China. Agriculture, Ecosystems & Environment 104: 343–357.

Binford

Brenner

Whitmore

et al . (1987) Ecosystem, paleoecology and human disturbance in subtropical and tropical America. Quaternary Science Reviews 6: 115–128.

Binford

Kahl

Norton

(1993) Interpretation of 210Pb profiles and verification of the CRS dating model in PIRLA project lake sediment cores. Journal of Paleolimnology 9: 275–296.

Bird

Polisar

Lei

et al . (2014) A Tibetan lake sediment record of Holocene Indian summer monsoon variability. Earth and Planetary Science Letters 399: 92–102.

Blaauw

Christen

(2011) Flexible paleoclimate age-depth models using an autoregressive gamma process. Bayesian Analysis 6: 457–474.

Blume

H-P

(1988) The fate of iron during soil formation in humid-temperate environments. In: Stucki

Goodman

Schwertmann

(eds) Iron in Soils and Clay Minerals. Bad Windsheim: D. Reidel Publishing Company, pp. 749–778.

10.

Boyle

(2002) Inorganic geochemical methods in palaeolimnology. In: Last

Smol

(eds) Tracking Environmental Change Using Lake Sediments. Dordrecht: Kluwer Academic Publishers, pp. 83–141.

11.

Brännvall

Bindler

Emteryd

et al . (2001) Four thousand years of atmospheric lead pollution in northern Europe: A summary from Swedish lake sediments. Journal of Paleolimnology 25: 421–435.

12.

Brenner

(1983) Paleolimnology of the Petén Lake district, Guatemala. II. Mayan population density and sediment and nutrient loading of Lake Quexil. Hydrobiologia 103: 205–210.

13.

Brenner

Dorsey

Xueliang

et al . (1991) Paleolimnology of Qilu Hu, Yunnan Province, China. Hydrobiologia 214: 333–340.

14.

Brenner

Whitmore

Curtis

et al . (1999) Stable isotope (δ13C and δ15Norg) signatures of sedimented organic matter as indicators of historic lake trophic state. Journal of Paleolimnology 22: 205–221.

15.

Brenner

Whitmore

Song

et al . (2000) Sediment records from Qilu Hu and Xingyun Yunnan Provine, China: Late Pleistocene to present. In: Gierlowski-Kordesch

Kelts

(eds) Lake Basins through Space and Time (AAPG Studies in Geology 46). Tulsa, OK: AAPG, pp. 619–624.

16.

Bull

Lockheart

Elhmmali

et al . (2002) The origin of faeces by means of biomarker detection. Environment International 27: 647–654.

17.

Bureau of Geology and Mineral Resources of Yunnan Province (1990) Regional Geology of Yunnan Province. Beijing, China: Geological Publishing House.

18.

Cai

Zhang

Cheng

et al . (2012) The Holocene Indian monsoon variability over the southern Tibetan Plateau and its teleconnections. Earth and Planetary Science Letters 335–336: 135–144.

19.

Carpenter

Hayes

(1980) Annual accretion of Fe-Mn oxides and certain associated minerals in a stream environment. Chemical Geology 29: 249–259.

20.

Chen

et al . (2014) Holocene vegetation history, precipitation changes and Indian Summer Monsoon evolution documented from sediments of Xingyun Lake, south-west China. Journal of Quaternary Science 29: 661–674.

21.

Cole

Valiela

Kroeger

et al . (2004) Assessment of a δ15N isotopic method to indicate anthropogenic eutrophication in aquatic ecosystems. Journal of Environment Quality 33: 124–132.

22.

D’Anjou

Bradley

Balascio

et al . (2012) Climate impacts on human settlement and agricultural activities in northern Norway revealed through sediment biogeochemistry. Proceedings of the National Academy of Sciences 109: 20332–20337.

23.

Dean

(1974) Determination of carbonate and organic matter in calcareous sediments and sedimentary rocks by loss on ignition: Comparison with other methods. Journal of Sedimentary Petrology 44: 242–248.

24.

Dearing

(1994) Environmental Magnetic Susceptibility - Using the Bartington MS2 System. Kenilworth: Chi Publishing.

25.

Diallo

Wen

(2009) Applications of remote sensing in land use/land cover change detection in Puer and Simao Counties, Yunnan Province. Journal of American Science 5: 157–166.

26.

Diefendorf

Patterson

Holmden

et al . (2007) Carbon isotopes of marl and lake sediment organic matter reflect terrestrial landscape change during the late Glacial and early Holocene (16,800 to 5,540 cal yr B.P.): A multiproxy study of lacustrine sediments at Lough Inchiquin, western Ireland. Journal of Paleolimnology 39: 101–115.

27.

Elliott

Brush

(2006) Sedimented organic nitrogen isotopes in freshwater wetlands record long-term changes in watershed nitrogen source and land use. Environmental Science and Technology 40: 2910–2916.

28.

Elvin

Crook

et al . (2002) The impact of clearance and irrigation on the environment in the Lake Erhai catchment from the Ninth to the Nineteenth Century. East Asian Studies 23: 1–60.

29.

Fang

(1987) Zhongguo xi nan li shi di li kao shi (Zhonghua shu ju: Xin hua shu dian Beijing fa xing suo fa xing). Beijing, China: Zhonghua shu ju (China Press).

30.

FAO/UNESCO (1974) Soil Map of the World. Paris: FAO/UNESCO.

31.

Fengrang

(1990) Soil Map of the People’s Republic of China (ed. World FUSMot). Beijing, China: State Land Administration, Beijing Agricultural University.

32.

Galman

Rydberg

Bigler

(2009) Decadal diagenetic effects on d13C and d15N studied in varved lake sediment. Limnology and Oceanography 54: 917–924.

33.

Galman

Rydberg

de-Luna

et al . (2008) Carbon and nitrogen loss rates during aging of lake sediment: Changes over 27 years studied in varved lake sediment. Limnology and Oceanography 53: 1076–1082.

34.

Gao

Min

et al . (2018) Heavy metal concentrations in sediments from Xingyun lake, southwestern China: Implications for environmental changes and human activities. Environmental Earth Sciences 77: 666.

35.

Gao

Luo

T-C

Zhang

B-R

et al . (1998) Chemical composition of the continental crust as revealed by studies in East China. Geochimica et Cosmochimica Acta 62: 1959–1975.

36.

Graney

Halliday

Keeler

et al . (1995) Isotopic record of lead pollution in lake sediments from the northeastern United States. Geochimica et Cosmochimica Acta 59: 1715–1728.

37.

Hillman

Abbott

et al . (2016) The isotopic response of Lake Chenghai, SW China, to hydrologic modification from human activity. The Holocene 26: 906–916.

38.

Hillman

Abbott

et al . (2014) Rapid environmental change during dynastic transitions in Yunnan Province, China. Quaternary Science Reviews 98: 24–32.

39.

Hodell

Schelske

(1998) Production, sedimentation, and isotopic composition of organic matter in Lake Ontario. Limnology and Oceanography 43: 200–214.

40.

Hodell

Brenner

Kanfoush

et al . (1999) Paleoclimate of Southwestern China for the past 50,000 yr inferred from lake sediment records. Quaternary Research 52: 369–380.

41.

Hollander

McKenzie

Haven

(1992) A 200 year sedimentary record of progressive eutrophication in Lake Greifen (Switzerland): implications for the origin of organic-carbon-rich sediments. Geology 20: 825–828.

42.

Peng

Jia

et al . (2006) Distribution and sources of organic carbon, nitrogen and their isotopes in sediments of the subtropical Pearl River estuary and adjacent shelf, Southern China. Marine Chemistry 98: 274–285.

43.

IAEA/WMO (2017) Global network of isotopes in precipitation: The GNIP database. Available at: https://nucleus.iaea.org/Pages/GNIPR.aspx

44.

Kabata-Pendias

(2011) Trace Elements in Soils and Plants. Boca Raton, FL: Taylor & Francis.

45.

Kahn

Yardle

(2007) As China roars, pollution reaches deadly extremes. The New York Times. Available at: https://www.nytimes.com/2007/08/26/world/asia/26china.html.

46.

Kelts

Hsu

(1978) Freshwater carbonate sedimentation. In: Lerman

(ed.) Lakes: Chemistry, Geology, Physics. New York: Springer, pp. 295–323.

47.

Kendall

Elliott

Wankel

(2007) Tracing anthropogenic inputs of nitrogen to ecosystems. In: Michener

Lajtha

(eds) Stable Isotopes in Ecology and Environmental Science. Malden, MA: Blackwell Publishing, pp. 375–449.

48.

Zha

et al . (2010) Eutrophication of lake waters in China: Cost, causes, and control. Environmental Management 45: 662–668.

49.

Lee

(1982) Food supply and population growth in Southwest China, 1250–1850. Journal of Asian Studies 414: 709–746.

50.

Leeming

Lathan

Rayner

et al . (1996) Concentrations of coprostanol that correspond to existing bacterial indicator guideline limits. Water Research 30: 2997–3006.

51.

Lewis

Maslin

(2015) Defining the anthropocene. Nature 519: 171–180.

52.

Dong

Zhao

et al . (2007) Assessment of water quality and identification of pollution sources of plateau lakes in Yunnan (China). Journal of Environmental Quality 36: 291–297.

53.

Liu

Zhang

et al . (2013) Historical reconstruction of atmospheric lead pollution in central Yunnan province, southwest China: An analysis based on lacustrine sedimentary records. Environmental Science Pollution Research 20: 8739–8750.

54.

Meyers

Eadie

(1993) Sources, degradation and recycling of organic matter associated with sinking particles in Lake Michigan. Organic Geochemistry 20: 47–56.

55.

Morrill

Overpeck

Cole

et al . (2006) Holocene variations in the Asian monsoon inferred from the geochemistry of lake sediments in central Tibet. Quaternary Research 65: 232–243.

56.

Muller

(1977) C/N ratios in Pacific deep-sea sediments: Effect of inorganic ammonium and organic nitrogen compounds sorbed by clays. Geochimica et Cosmochimica Acta 41: 765–776.

57.

Meyers

(1994) Preservation of elemental and isotopic source identification of sedimentary organic matter. Chemical Geology 114: 289–302.

58.

Meyers

Lallier-Verges

(1999) Lacustrine sedimentary organic matter records of Late Quaternary paleoclimates. Journal of Paleolimnology 21: 345–372.

59.

Olsson

(1986) Radiometric methods. In: Berglund

(ed.) Handbook of Holocene Palaeoecology and Palaeohydrology. Chichester: John Wiley & Sons, pp. 273–312.

60.

Polissar

Freeman

(2010) Effects of aridity and vegetation on plant-wax dD in modern lake sediments. Geochimica et Cosmochimica Acta 74: 5785–5797.

61.

Pompeani

Abbott

Steinman

et al . (2013) Lake sediments record prehistoric lead pollution related to early copper production in North America. Environmental Science & Technology 47: 5545–5552.

62.

R Core Team (2013) R: A Language and Environment for Statistical Computing 3.2.2. Vienna: R Foundation for Statistical Computing.

63.

Reimer

Bard

Bayliss

et al . (2013) INTCAL13 and Marine 13 radiocarbon age calibration curves 0–50,000 years CAL BP. Radiocarbon 55: 1869–1887.

64.

Ritchie

McHenry

(1990) Application of radioactive fallout cesium-137 for measuring soil erosion and sediment accumulation rates and patterns: A review. Journal of Environment Quality 19: 215–233.

65.

Schindler

Wolfe

Vinebrooke

et al . (2008) The cultural eutrophication of Lac la Biche, Alberta, Canada: A paleoecological study. Canadian Journal of Fisheries and Aquatic Sciences 65: 2211–2223.

66.

Schnurrenberger

Russell

Kelts

(2003) Classification of lacustrine sediments based on sedimentary components. Journal of Paleolimnology 29: 141–154.

67.

Schubert

Calvert

(2001) Nitrogen and carbon isotopic composition of marine and terrestrial organic matter in Arctic Ocean sediments: Implications for nutrient utilization and organic matter composition. Deep-Sea Research 48: 789–810.

68.

Shen

Jones

Yang

et al . (2006) The Holocene vegetation history of Lake Erhai, Yunnan province southwestern China: The role of climate and human forcings. The Holocene 16: 265–276.

69.

Sun

Xiong

(1983) Chronology. In: Rawson

(ed.) The Chinese Bronzes of Yunnan. London: Sidgwick and Jackson and Cultural Relics Publishing House, pp. 243–252.

70.

Sun

Qiao

et al . (1986) Late Pleistocene and Holocene vegetation history at Kunming, Yunnan Province, Southwest China. Journal of Biogeography 13: 441–476.

71.

Talbot

(1990) A review of the palaeohydrological interpretation of carbon and oxygen isotopic ratios in primary lacustrine carbonates. Chemical Geology 80: 261–279.

72.

Talbot

(2001) Nitrogen isotopes in Paleolimnology. In: Last

Smol

(eds) Tracking Environmental Change Using Lake Sediments: Volume 2 Physical and Geochemical Methods. Dordrecht: Kluwer Academic Publishers, pp. 401–439.

73.

Talbot

Johannessen

(1992) A high resolution palaeoclimatic record for the last 27,500 years in tropical West Africa from the carbon and nitrogen isotopic composition of lacustrine organic matter. Earth and Planetary Science Letters 110: 23–37.

74.

Teranes

McKenzie

Lotter

et al . (1999) Stable isotope response to lake eutrophication: Calibration of a high-resolution lacustrine sequence from Baldeggersee, Switzerland. Limnology and Oceanography 44: 320–333.

75.

Vuille

Werner

Bradley

et al . (2005) Stable isotopes in precipitation in the Asian monsoon region. Journal of Geophysical Research 110: D23108.

76.

Waters

Zalasiewicz

Summerhayes

et al . (2016) The anthropocene is functionally and stratigraphically distinct from the Holocene. Science 351: aad2622.

77.

Whitmore

Brenner

Jiang

et al . (1997) Water quality and sediment geochemistry in lakes of Yunnan Province, southern China. Environmental Geology 32: 45–55.

78.

Whitney

(1975) Relationship of manganese-iron oxides and associated heavy metals to grain size in stream sediments. Journal of Geochemical Exploration 4: 251–263.

79.

Wright

Mann

Glaser

(1984) Piston corers for peat and lake sediments. Ecology 65: 657–659.

80.

Zhou

Liu

et al . (2015) Changing intensity of human activity over the last 2,000 years recorded by the magnetic characteristics of sediments from Xingyun Lake, Yunnan, China. Journal of Paleolimnology 53: 47–60.

81.

Xiao

Haberle

Shen

et al . (2014) Latest Pleistocene and Holocene vegetation and climate history inferred from an alpine lacustrine record, northwestern Yunnan Province, southwestern China. Quaternary Science Reviews 86: 35–48.

82.

Wilkes

(2004) Biodiversity impact analysis in northwest Yunnan, southwest China. Biodiversity and Conservation 13: 959–983.

83.

Yang

Zhou

Guo

et al . (2010) Analysis of spatial and temporal water pollution patterns in Lake Dianchi using multivariate statistical methods. Environmental Monitoring and Assessment 170: 407–416.

84.

Yao

Zhilong

(2012) Rediscovering the settlement system of the ‘Dian’ kingdom, in Bronze Age southern China. Antiquity 86: 353–367.

85.

Yao

Jiang

Chen

et al . (2015) Bronze Age wetland/scapes: Complex political formations in the humid subtropics of southwest China, 900–100 BC. Journal of Anthropological Archaeology 40: 213–229.

86.

Yunnan Provincial Institute of Cultural Relics and Archaeology, University of Toronto and University of Michigan Museum of Anthropology (2015) Preliminary findings from the 2010 archaeological survey in Lake Dian Basin, Yunnan. Chinese Archaeology 15(1): 152–160.

87.

Yunnan Publishing Workbook (1989) Yunnan sheng di fang zhi bian zuan wei yuan hui (Yunnan Sheng Zhi 38: Shui Li Zhi (Hydrological History)). Kunming, China: Yunnan renmin chuban she.

88.

Zeng

H-A

J-L

(2009) Sedimentary records of heavy metal pollution in Fuxian Lake, Yunnan Province, China: Intensity, history, and sources. Pedosphere 19: 562–569.

89.

Zhan

Zhou

et al . (2017) Evidence for the importance of atmospheric nitrogen deposition to Eutrophic Lake Dianchi, China. Environmental Science & Technology 51: 6699–6708.

90.

Zhang

Liu

Shen

et al . (2012) One century sedimentary record of lead and zinc pollution in Yangzong Lake, a highland lake in southwestern China. Journal of Environmental Sciences 24: 1189–1196.

91.

Zhang

Wang

Lei

(1996) Authigenic mineralogy, depositional environments and evolution of fault-bounded lakes of the Yunnan Plateau, south-western China. Sedimentology 43: 367–380.

92.

Zou

Cui

Liu

et al . (2017) Investigation of early Bronze Age civilizations in Yunnan: A scientific analysis of metallurgical relics found at the Guangfentou ruins in Jiangchuan. Archaeological and Anthropological Sciences 10: 1–17.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.47 MB