Holocene desertification,traditional ecological knowledge,and human resilience in the eastern Gobi Desert,Mongolia

Abstract

Dryland regions are particularly challenging for human survival over the course of deep time. This is true for institutionally complex communities as well as small-scale societies that have existed in semi-arid regions throughout the Holocene. This paper examines some of the successful strategies employed by small-scale mobile communities which enhanced their ability to thrive in drylands over the course of thousands of years. Small-scale societies living in drylands must rely on the transmission of Traditional Ecological Knowledge across generations. Some of this knowledge relates to the availability and use of wetlands and other more ephemeral water sources, the exploitation of a diverse range of resources, and the potential for natural storage of food resources as a buffer against regularly occurring drought years in these regions. We compare this understanding with our environmental archeological findings at the Mid-Holocene site of Zaraa Uul in the eastern Gobi Desert of Mongolia. At the site of Zaraa Uul, we show how hunter-gatherer groups returned to a campsite near the edge of a wetland environment over the course of at least two phases during the Mid-Holocene. Here they took advantage of a greater diversity of animal species and plants, including small-grained-grasses and sedges, which could enhance their caloric intake and increase the potential for storable commodities which could be collected as needed from their natural habitat.

Keywords

dryland adaptations hunter-gatherers Middle-Holocene paleoenvironment Mongolia traditional ecological knowledge wetlands

Introduction

Dryland regions of the world present significant challenges to human communities living in those environments. Although institutionally complex societies have engineered intricate technologies to cope with uncertainties of rainfall regimes and water sources under those conditions, small-scale societies also can be remarkably resilient by relying on several “ground-up” mechanisms to thrive in semi-arid landscapes. The pursuit and acquisition of resources in these environments are challenged by plant and animal distributions that are highly sensitive to annual variations in rainfall and droughts, and therefore are less predictable than resources in more temperate zones. Small-scale societies can be especially resilient when dealing with the challenges of climate change, both secular and abruptly punctuated. In this paper we explore some of the parameters of their resilience to desertification and compare this to our findings at the Mid-Holocene forager campsite of Zaraa Uul in the eastern Gobi Desert of Mongolia.

Human communities in many regions of the world have adapted in innovative ways to environmental changes through time. In 1998 Sillitoe published a review paper in Current Anthropology, highlighting the importance of accounting for traditional indigenous knowledge in developmental planning (Sillitoe, 1998). He cited a number of studies demonstrating that traditions other than Western science have their own expert knowledge and effective “science” relating to resource use within their home environments. In the spirit of this understanding, many social and natural scientists have studied the ways in which small-scale societies living in present-day dryland environments use their deeply rooted systems of Traditional Ecological Knowledge (TEK) to thrive in these challenging settings.

These adaptations are far from trivial feats, and sometimes entail a great deal of technological, economic, and social flexibility. In recent years, a number of studies have focused on the critical role of wetland environments within semi-arid regions, as an important buffer zone for foragers, pastoralists, and incipient cultivators (Hammer, 2018; Holguín, 2019; Janetski and Madsen, 1990; Janz et al., 2017, 2021; Kelly, 1990; Lillie et al., 2007; Murphy and French, 1988; Nicholas, 1998; Ramsey and Rosen, 2016; Ramsey et al., 2015). Some of this research shows that small-scale mobile societies can expand across the landscape in moist years allowing them to exploit a wide range of available plants and animals across their home ranges, while always staying within reach of perennial wetlands which often provide a wide array of resources in lean seasons as well as overall drought years. This can be a resilient and sustainable strategy for long-term existence within semi-arid zones (Ramsey and Rosen, 2016). The TEK comes into play when mobile societies must decide whether to invest in travel to an ephemeral water source which may or may not hold water by the time they arrive there. This specialized knowledge is held in the collective memory of a community and passed on through generations.

Much of our understanding of how foragers and pastoralists thrive in semi-arid zones with unpredictable resource distributions comes from insightful studies of mobile foragers in the drylands of Australia and the Great Basin of the Western United States, and pastoralists in East and Southern Africa (Asmussen and McInnes, 2013; Bird and Bird, 2005; Janetski and Madsen, 1990; Kelly, 1990; Linstädter et al., 2016; Lourandos, 1980; Veth, 2005; Wright, 2019). The combination of ethnohistoric observations with data from archeological remains allows us to use deep-time perspectives to identify cycles of adaptive strategies used by small-scale societies to respond to ever-shifting climatic conditions leading to alternating phases of abundant and sparce resource distributions and availability over the course of thousands of years. We can best understand the long-term stability of these societies if we view them not as social/economic entities that are frozen in time, but rather those which acknowledge the inevitability of changing rainfall patterns, temperature distributions, resources and water availability, and are able to deftly shift social and economic strategies to maintain social stability (Nelson, 1996; Wright, 2019).

These societies are able to manage climatic and environmental shifts by using a “toolkit” of long-term memories and traditions of TEK that have been survival strategies passed down through generations via mechanisms such as generational memory, stories of the recent past, and finally myths and legends that embed this ecological knowledge within them (Butzer, 1982; Rosen and Rivera-Collazo, 2012). We can view these as “memory-messages” that convey multiple ways of exploiting plants and animals that may include preparation of poisonous plants and not-so-palatable animals, methods for finding sources of sub-surface water, and rituals that promote social bonding and resilient social institutions such as traditions of sharing. The archeological record, in the form of settlement distribution, tool kits, and plant and animal remains within the archeological sites, provides us with perspectives on this TEK, and how it might have been employed throughout deep time frames in dryland regions for the long-term sustainable use of unpredictable resources.

Some of the most important strategies that are identifiable in the archeological record include: (1) a knowledge of variability in availability of unmanaged water sources, and/or the ability to manage and control these sources. Archeologists can detect these by settlement patterns around extant and extinct water resources, and identification of artificially altered landscape features associated with local hydrological systems. This is accomplished through a program of geoarchaeological studies within the site and its vicinity, coupled with laboratory analyses of sediment depositional environments (Barker et al., 1997; Harrower, 2016; Kaptijn, 2018; Lourandos, 1980; Mithen, 2010; (2) another well-known resilient strategy in dry-land environments is the diversification of resources, diet, social institutions, and toolkits. Again, most of these elements are detectable in the archeological record through the systematic collection of archaeobotanical remains such as charred seeds, phytoliths, and starches, as well as zooarchaeological analyses of faunal remains. Social institutions are ascertained by settlement and architectural configurations, and the identification of “taskscapes” within the sites, and on the surrounding landscapes; and (3) a third resilient strategy is the use of some form of food storage. This can be a natural characteristic of a particular plant food which remains in its habitat over multiple seasons, or technologies for caching and/or preserving food items in times of abundance.

These three strategies are often reinforced by thousands of years of TEK involving trials, errors and successes accumulated as knowledge handed down through generations. For small-scale societies such as hunter-gatherers and mobile herders living in dryland regions, these three elements of resilient subsistence strategies are all facilitated by living in proximity to wetland environments or incorporating these microenvironments within their seasonal rounds. Wetlands provide essential resources for long-term sustainable access to water and food, and also offer increased diversity of plant and animal foods as well as natural storage of starchy, high caloric foods (Nicholas, 1998).

Water sources in arid and semi-arid environments are among the most critical resources for the support of human occupation, settlement, and subsistence in these challenging regions. Dryland regions may include lakes, streams, springs, marshy wetlands, and ephemeral playa lakes. The localities with these watery places typically become “landscape anchors” (Hammer, 2014) and “persistent places” (Maher, 2019; Olszewski and Al-Nahar, 2016; Schlanger, 1992) that may attract human groups for millennia. Small-scale societies such as foragers and mobile pastoralists are notable for their resilient strategies of mobility, social flexibility for altering group size, and knowledge of a broad range of environmental zones across the landscape. In dryland regions water sources, plant communities, and animal populations are in flux depending upon the amount and timing of rainfall events throughout the year. These mobile groups must have a detailed knowledge of how these resources shift depending upon these weather conditions, and the patterns of changing resource availability across their home ranges. This Traditional Ecological Knowledge is seamlessly integrated into many aspects of their society and culture (Charlier, 2016; Fernandez-Gimenez, 2000; Marin, 2010; Miller and Davidson-Hunt, 2013; Sepie, 2017).

For more than a decade, social and natural scientists have been gathering data on the Traditional Ecological Knowledge of mobile herders in Mongolia as an important guide to maintaining the health of herds and rangelands (Charlier, 2016; Fernandez-Gimenez, 2000; Fernánedez-Giménez, 1999; Marin, 2010). Although we have no direct information about the TEK of ancient mobile foragers and hunters in the Mongolian steppe lands, it’s possible we might find occasional clues to their perceived environments and ecological knowledge within the transmitted knowledge of modern herders, or at the very least, it may help us question the archeological record for evidence of ecological sustainable patterns of land-use which contribute to the resilience of these past mobile hunting societies.

One possible way to access traditional knowledge among ancient foragers is to track the availability and use patterns of water sources through time. However, water availability can be a moving target. This is a challenge at all temporal and spatial levels. Water sources vary seasonally with fluctuations in the amount and timing of rainfall events. Yearly averages rise and fall dramatically around a statistical mean, and total amounts can shift greatly over the course of decades and millennia in a secular or stochastic manner (Butzer, 1982; Rosen, 2022). In the short term of yearly availability, watery places such as ephemeral streams and playa lakes can be sources of abundant water or completely dry with seasonal or yearly variability. Therefore, these small-scale societies dependent upon such water sources must be able to read the rainfall patterns in any given year before investing in travel to these water sources. A critical factor in this process of decision-making then, is the varying levels of predictability. Predicting the availability of water sources would be greatly enhanced through memory and traditions of ecological knowledge transmitted across generations. Modern Mongolian herders still transmit such knowledge during seasonal festival gatherings (Sattler et al., 2021). Evidence for long-term ancient ecological knowledge related to water sources might be obtained through multi-disciplinary investigations including landscape geoarchaeology, settlement pattern survey, and analyses of faunal and botanical remains within archeological sites which are associated with wetlands, streams, and lakes.

The geography of water sources can be highly diverse in dryland areas. The configuration of these localities is one important factor influencing the patterns of movement across the landscape, which highlights the importance of archeological survey to track this movement through time. In the northern Gobi Desert/Steppe region of southeastern Mongolia, the most important sources for ancient mobile foragers and herders were the basin-centered lakes and wetlands which were prominent during the late Pleistocene and Mid-Holocene, as well as the system of linear valleys which sometimes supported spring-fed perennial streams. For pre-modern and recent pastoralists in Mongolia, researchers have documented that availability of water for herds dictates much of the settlement patterns, economic choices, and politics of mobile herders (Marin, 2010; Tugjamba et al., 2021). These patterns come from traditional knowledge handed down through generations. Some of this water-source understanding may have been transferred over the course of centuries and even millennia from previous generations of hunter-gatherers who depended on following wild herd animals.

Water and desert adaptations in the Mongolian Gobi Desert

In the Mongolian Gobi Desert, there were two main kinds of water resources available to the Middle and Later-Holocene foragers, hunter-pastoralists, and early pastoralists. These were (1) water sources in the broad basins typical of the geography of this region and (2) the drainages that formed within the systems of linear valleys.

Basin wetlands

The topographic basins were catchment areas for runoff from surrounding ridges and perennial or ephemeral streams, and many of them formed endorheic lakes with no outlets (Holguín and Sternberg, 2018). These sinks formed large saline lakes during the Late Pleistocene, which attracted a variety of plants, big-game, and hunters whose sites appear as lithic scatters on the highest beach terraces (Janz et al., 2017, 2021). These deep lakes were for the most part dry during the early Holocene (Felauer et al., 2012; Feng et al., 2005; Klinge and Sauer, 2019; Lehmkuhl et al., 2018; Mischke et al., 2020; Yu et al., 2019), but smaller freshwater ponds appeared with associated wetlands during the Mid-Holocene Climatic Optimum (Janz et al., 2021).

Many researchers point to the importance of wetlands in the seasonal rounds of small-scale mobile societies living in dryland environments (Janetski and Madsen, 1990; Kelly, 1990; Nicholas, 1998; Ramsey and Rosen, 2016; Ramsey et al., 2016). Most of these traditions of plant foraging are based in TEK that spans throughout deep time (Turner et al., 2011). Wetland microenvironments yield a variety of vital resources such as reeds and sedges that provide materials for shelters, matting, and basketry, reliable sources of animal protein from fish, birds and their eggs, and mammals of all sizes that are attracted to the water sources. Foragers in the Gobi may have abundant sources of protein in their diets, but many of these animal products are lean and yield few calories from fats and starches. There are few other sources of essential calories in the semi-arid environments of southern Mongolia, especially in the interfluves. The wetland zones have dense stands of starch-yielding plants which provide essential calories to supplement a protein-rich diet. These calorie-rich foods come from the roots, rhizomes, and shoots of cattails (Typha sp.) and sedges (Cyperaceae), for example Cyperus sp. and Scirpus sp. (Hillman et al., 1989; Liu et al., 2013, 2014; Wollstonecroft et al., 2008).

Given the wealth of resources available in the vicinity of wetland microenvironments, they can be a key focal point in the seasonal movements of foragers living in dry-land regions, especially during episodes of drought and unpredictable rainfall. The same is true for modern and historical Mongolian herders (Fernánedez-Giménez, 1999; Tugjamba et al., 2021). Research on archeological sites around basin wetlands in the Southern Levant highlights these ecozones as essential to forager resilience and adaptations to dryland environments in the Southern Levant (Garrard and Byrd, 2013; Maher, 2019; Maher et al., 2016; Ramsey and Rosen, 2016; Ramsey et al., 2015; Richter et al., 2013; Rosen, 2013), and also for herder-gatherer populations in ancient Arabia (Preston et al., 2012). Ramsey describes these small-scale mobile populations in the Southern Levant as being “tethered” to the wetlands, since they were often an essential component of their seasonal rounds (Ramsey and Rosen, 2016). The attraction of establishing a base camp near the basin wetlands during a seasonal migration route might result in a radial pattern of movement, allowing hunters-gatherers to follow game and collect plants in any direction, while maintaining a base camp around a secure source of supplies. In years of abundant rainfall, migrations could extend further from the wetland points. In drought years it is likely the migrations would be more restricted in distance since wetlands would offer an essential buffer zone for minimizing risk (Preston et al., 2012). Given that basin-centric wetlands existed for millennia in the Gobi Desert of southern Mongolia during the late Pleistocene through the Middle-Holocene, and the importance of wetlands for the resilience of small-scale mobile societies, we can expect them to be a key focal point for forager and hunter-pastoralists during the middle and Late-Holocene in the Mongolian Gobi. These traditions for movement and prediction of sources of water and pasture lands, might have been handed down through the generations as important elements of ecological knowledge embedded in the modern TEK of Mongolian herder communities (Fernandez-Gimenez, 2000).

Drainage systems in linear valleys

Other significant landscape features influencing the movement of foragers and pastoralists across the Gobi are the “hydrological corridors” containing belts of springs and seasonal flow within the narrow linear valleys. Many of these valleys trend north/south providing corridors of movement between the lush stipa-grass steppe zones to the north, and the paleolake basins and wetlands to the south. Holguin studied sites distributed along these types of valleys near the present-day lake of Ulaan Nuur (Holguín, 2019; Holguín and Sternberg, 2018). Rosen and others researched these types of sites at the Ikh Nart Nature Reserve (Rosen et al., 2019; Schneider et al., 2016). The foragers and subsequent mobile herders who followed these hydrological corridors were able to reduce their subsistence risk by exploiting a variety of micro-environments associated with these extensive linear palaeohydrological systems. Movement along these valleys would have allowed greater access to the microenvironments around the occasional springs that existed there, especially during seasons with lower rainfall. These passages also enabled mobility and movement throughout the Gobi in both east/west and north/south directions. Some of these traditions of movement through the landscape seeking springs and following stream flows have survived in strategies of modern herders (Kakinuma et al., 2014; Tugjamba et al., 2021).

Another critical factor for survival in drylands, and especially in situations of increasing desertification is the ecological knowledge which allows the identification, exploitation, and processing of a highly diverse suite of subsistence resource. This diversification of food stuffs compliments social resilience and flexibility (Turner et al., 2011).

Diversification of subsistence resources

The ability of humans to diversify diet, material culture, and social institutions is one of our most effective elements of survival. Our species can consume a wide variety of foods, withstand many types of climatic conditions and ecological zones, move easily from habitat to habitat, and adapt our material and non-material culture accordingly (Rosen, 2022). exploiting the natural resource diversity in dryland regions requires a sophisticated knowledge of edible plants and animals, methods for processing them, their availability, and how this availability shifts given the unpredictability of rainfall and temperature from year to year. This understanding can only come from traditional knowledge handed down through generations of experience. Most of our information on Mongolian TEK about wild plants relates to herding practices and pastureland plants (Fernandez-Gimenez, 2000; Kakinuma et al., 2014), however we do have abundant information relating to forager plant use in other dryland regions.

In periods of favorable climatic conditions, subsistence resources are denser, and patches of some items become more extensive across the landscape. This situation would allow mobile groups to target the foods that most appeal to them more intensively by narrowing down the range of different resource types and focus on high yield and large packets of foodstuffs, highly nutritious foods, or ones that contribute to social status (Bettinger, 2009; Kelly, 2013; Rosen and Rivera-Collazo, 2012; Stiner and Munro, 2002). Some examples of this may be big game hunting, or heavy reliance on collection of nuts from groves of nut-bearing trees (Abrams and Nowacki, 2008; Mason, 1995). These groups can afford the luxury of preferred foods in the absence of significant risk, and target highly profitable resources, since the food supplies and diversity of foods would be higher. Some societies may choose to increase their risk and select for more “costly” resources, while others may choose not to do so, but the option is there.

In periods of climatic degradation, environmental uncertainty, drought and decreased predictability, the most resilient strategy is to diversify subsistence resources. This is especially pronounced in situations when rainfall events become more stochastic and unpredictable over the long term. Diversification is a strategy of food procurement referred to by many archeologists as a “Broad Spectrum” approach (Janz, 2016; Stiner and Munro, 2002; Stutz et al., 2009; Weiss et al., 2004). It may take the form of exploiting more animal food sources that are small, fast, difficult to catch, and have less palatable meat with little fat content such as rabbits and other small mammals (Stiner and Munro, 2002). Foragers may exploit harder to collect plants such as small-seeded grasses (Rosen, 2010; Weiss et al., 2004), or those that need more intensive labor for collection, processing, and cooking. Diversification of plant resources may also include plants that require extensive processing to remove toxins which would otherwise be harmful to humans if consumed directly. The need to target diverse plant species highlights the essential importance of wetland environments. Many diverse plants grow in wetlands, which are environments with high biodiversity, and therefore attract mobile foragers for the diversity in food resources, lending a large measure of increased resilience to droughts and otherwise drying climatic conditions and/or unpredictable rainfall events.

Australian researchers have developed robust theoretical perspectives on the use of desert landscapes as foragers arrive in these regions and begin to understand the cycles of abundance and fluctuations of resource microenvironments. They suggest that the first foragers arrived in the Australian desert zones during times of moister climate giving these initial populations the opportunity to develop a sophisticated ecological knowledge of the behavior of the plant and animal resources there. With drying climate, these populations then used this ecological understanding to broaden the types of resources they were able to exploit (Hiscock and Wallis, 2005; Smith et al., 2005; Veth, 1989).

Storage and caching

Storage of foodstuffs is critical for societies living in environments with low predictability of rainfall. This is especially true for agricultural societies. Stored seeds may allow farmers to recover from at least 1 year of severe drought, and often 2 or even 3 years of rainfall shortages and crop failures. A number of researchers suggest that storage and food caching can be just as important to mobile foragers and early pastoralists living in dryland regions as well (Bettinger, 2009; Gerber et al., 2004; Morgan, 2012; Rowley-Conwy and Zvelebil, 1989; Tushingham and Bettinger, 2013). Tushingham and Bettinger (2013) adopted an Optimal Foraging perspective in their study of “Front-Loaded” and “Back-Loaded” storability of resources. They propose that the selection of food items is best understood from the view of energy output for collecting, storing, processing, and planning future return. “Front-loaded” resources are costly to acquire and store, but preparation for consumption is easy. “Backloaded” resources are easy to acquire and store, but energy intensive to prepare. Backloaded resources involve less risk because the energy expenditure needed to acquire and store them is lower, and therefore if the cache is lost or not needed, then there was little wasted effort. We can explore this concept from the perspective of wetland exploitation as well.

Wetland zones in dryland environments have great potential for natural storage of important high-value year-round resources with high starch and caloric values (Ramsey and Rosen, 2016). Some of the most important of these are the Underground Storage Organs (USOs) of sedges such as club rush (Bolboschoenus maritimus), yellow nutsedge (Cyperus esculentus), and purple nutsedge (Cyperus rotundus) (Hillman et al., 1989; Wollstonecroft et al., 2008). These resources are available for most of the year, thus exploitable at a level consistent with low-risk “back-loaded” stored resources. They are rich in caloric value, being the starch repository for the plants. As mentioned above, calories are hard to come by in some dryland regions such as the Gobi Desert where the animal foods are lean, and unlike more temperate zones, there is less access to other forms of calorie-rich wild plants such as nuts, or large-grained cereals grasses. Archaeobotanical evidence from dryland areas of northern China and Inner Mongolia shows that ancient foragers did indeed target USOs in their plant food selection strategies (Dubreuil et al., 2021; Liu et al., 2013, 2014). There are a number of Mid-Holocene sites in the eastern Gobi Desert of Mongolia that have caches of ground stone mortars (Dubreuil et al., 2021; Schneider et al., 2016). These localities might also have been associated with caches of stored edible plants which were cached in preparation for processing. This possibility can be tested through sampling for microbotanical remains in the sediments of pits at these sites.

Environmental archeological studies at Zaraa Uul

The Mid-Holocene was a pivotal period for human societies worldwide. This is the time frame in which many communities shifted from a predominantly forager lifestyle to a greater investment in the intensification of foodways. The study of human adaptations to environmental change in this episode in human civilization has focused primarily on the gradual shift from foragers to farming societies (c.f. Fuller et al., 2018; Liu et al., 2018; Smith, 2001). Many fewer studies have concentrated on the transition from primarily foraging societies to early pastoralists (Hermes et al., 2021; McCorriston and Martin, 2010; Rosen, 2021; Rosen et al., 2019). This is especially true of research areas in East Asia (Honeychurch, 2017; Janz et al., 2017, 2020; Wright, 2006; Wright et al., 2019). The eastern Gobi Desert of Mongolia provides a good laboratory for studying how transitions from foraging to mobile pastoralism occurred, the role of climate changes in this process, and the story of the unique and varied ways human societies reacted to these major climate changes using resilient strategies of food procurement, exploitation of the mosaic of wetland patches, and management of habitats in this region. These strategies likely began among forager groups who preceded the first pastoral societies.

Much research is currently underway to define the climatic and environmental changes that occurred in the western and central Gobi Desert of Mongolia during the Mid-Holocene. This includes investigations of lake cores using proxies such as sediment grain size, pollen, and geochemistry, as well as analyses of alluvial sediment sections and paleosols (Klinge and Sauer, 2019; Mischke et al., 2020; Yu et al., 2019). Far less information has come from the eastern localities of the Mongolian Gobi Desert. In this section, we report on Holocene sediment and phytolith evidence from alluvial and lacustrine deposits at the site of Zaraa Uul, in the Gobi Desert of southeastern Mongolia. These deposits provide us with a climatic and environmental record of the shift from the Mid-Holocene through the Late-Holocene periods. We are then able to match this environmental evidence for shifting climatic episodes with the data recovered from archeological sites in that locality which show evidence for the use of plants and animals associated with wetlands which once existed in this region. This may help us understand some of the risk-buffering that the mobile hunter-gatherers in this region used to enhance resilience to the climatic fluctuations in this semi-arid environment, and perhaps some of the strategies that were handed down to later generations as traditional knowledge.

Paleoenvironmental background

Much of the previous paleoclimatic research in the Mongolian Gobi Desert is summarized in several recent articles (Bazarova et al., 2019; Felauer et al., 2012; Holguín and Sternberg, 2018; Klinge and Sauer, 2019; Lehmkuhl et al., 2018; Li et al., 2018; Yu et al., 2019). There are also important parallels from northern China and Inner Mongolia (Zhang et al., 2011). However, very few studies have been conducted in the eastern Gobi Desert area of southern Mongolia. This is problematic because it leaves us with an incomplete understanding of the climatic history of a large region of the Gobi. Thus, there is a lack of information critical to reconstructing the human challenges and advantages in the process of adaptation to desertification in this region, as well as more detailed information about the movement of mobile foragers and pastoralists across this area. Our geoarchaeological and microbotanical findings at Zaraa Uul will provide a landscape and environmental context to the archeological data in the region and increase our understanding of the ways human inhabitants of the region found solutions to the effects of ever-increasing desertification there.

Klinge and Sauer (2019) cite records from Ulaan Nuur Lake as studied by Lehmkuhl et al. (2018), Lee et al. (2013), and Felauer et al. (2012), based on pollen, total organic carbon and C/N ratios. They indicate dry conditions during the Late Glacial period, with an increase in humidity from around 11,000 to 4000 BP, punctuated by several dry episodes around 8700, 7600, and 4700 BP. Aridity increased throughout the Late-Holocene. Klinge and Sauer (2019) conclude that the early Holocene was dominated by warm and humid conditions across most of southern and central Mongolia with diversifying climatic conditions beginning in the early Mid-Holocene. They note that warm and humid conditions dominated western Mongolia. In the latter part of the Mid-Holocene, the climatic signals became more diverse and complex in terms of temperature and humidity shifts. By the Late-Holocene period, most records show warm and dry conditions toward the eastern regions of the Gobi.

Generally, the Mid-Holocene increase in warmth and humidity initiated rising lake levels across much of the southern Gobi Desert. Klinge and Sauer (2019) point to a spatiotemporal trend of rising lake levels from the Late Glacial to the early Holocene which progressed in an eastward trajectory. They suggest this might have been due to a longitudinal warming trend resulting from the retreat of the glaciers and thawing of permafrost in the Altai and later Khangai Mountains. But in their concluding remarks, they reference the complete lack of data for south-eastern Mongolia. Our preliminary results at Zaraa Uul provide some initial data to address this gap.

Summary of geoarchaeological and microbotanical research

The focus of our study is the complex of prehistoric sites at Zaraa Uul (N 46°07′47.2″/E111°42′37.5″), in the desert-steppe zone of eastern Mongolia, located within the Tuvshinshiree Sum (county), Sükhbaatar Aimag (province) (Figure 1) (Janz et al., 2021). The precipitation in this region currently averages around 200 mm per annum (Venable et al., 2015). The sites investigated in this research are distributed along the western margin of a large longitudinal basin surrounded by low hills and shallow valleys. The region is considered a desert steppe zone with vegetation consisting of small desert steppe shrubs including Caragana sp. (Siberian peashrub – an edible legume), Salsola sp. (saltwort), and Artemisia sp. (sagebrush), as well as xerophytic tuft-forming short Stipa grasses (Stipa gobica, Stipa glareosa, etc.), and wild onions (Allium polyrhizum, Allium mongolicum). There are also sparse stands of Siberian elm (Ulmus pumila) along the ephemeral drainages (Undarmaa et al., 2015).

Figure 1.

Map of Mongolia showing the location of Zaraa Uul. (After open-source file: Mongolia location map.svg: NordNordWestderivative work Виктор_В, CC BY-SA 3.0 <https://creativecommons.org/licenses/by-sa/3.0>, via Wikimedia Commons), Insert from Google Earth.

The Mid-Holocene archeological site of Zaraa Uul is situated on the west side of the basin (Figure 2a). The western margin of this depression is bounded by uplifted terrestrial conglomerates and extrusive basalt flows. The floor of the basin is composed of incised marl, the remains of a vast Pleistocene saline lake that extended along the entire length of the valley. Today this marl forms hummock and terrace lines where it was incised by streams coming down from alluvial fans on the margins and bisected in length by a larger ephemeral stream. In many places these hummocks are covered with coarse-grained alluvium washing out from the surrounding hills after storms. Today these alluvial valleys are ephemeral and usually dry. There is at least one active fresh-water spring at the south end of the basin, forming a pool where livestock drink and bathe (Figure 2b). At the east side of the basin there is a residual salt marsh where a small amount of saline water is ponded (Figure 2c).

Figure 2.

Photographs of the vicinity of Zaraa Uul: (a) the Zaraa Uul site and vicinity, (b) freshwater spring at southern end of the basin, and (c) the remaining extant salt marsh in the Zaraa Uul basin.

During the 2016 and 2017 field seasons of the Zaraa Uul project Rosen conducted a geomorphological and geoarchaeological investigation of the landscape and environmental history of the Zaraa Uul site and surroundings. The primary goal of this study was to reconstruct Late Pleistocene – Holocene paleoenvironments in the vicinity of Zaraa Uul in order to understand how ancient foragers and early herders used the landscape and resources in the eastern Gobi throughout the Holocene. The work was conducted within the framework of the archeological survey and excavation project directed by L. Janz and O. Davaakhuu. The initial geoarchaeological research consisted of a pedestrian survey of the landscape along the western perimeter of the basin, and an examination of the tributary valleys searching for Quaternary alluvial terraces and beach ridges as well as exposed geological sections through these geomorphological features. We also used geoarchaeological techniques to excavate, record, and sample sections along a transect from the archeological excavations on the slope at the edge of the basin, down into the depression itself. Here we report on several key geological sections and terraces which span much of the late Pleistocene through Late-Holocene. Geo-sections were cleaned, described, and sampled for both sediment and phytolith analyses.

GeoSections

Pleistocene outcrops

Zuun Shovkh (N46°08′23.4″/E111°43′00.7″) (1044 m asl) (OSL date 32,300 ± 1650 BP) and Otson Tsokhoi (1040 m asl) (both initially discovered by Davaakhuu) are sandy Late Pleistocene beach terraces associated with the remains of numerous flint artifacts consistent with early Upper Paleolithic assemblages in this region. These terraces mark the latest high stand of the Zaraa Uul Pleistocene Paleolake which is indicated by numerous exposures of thick white saline marls across the basin and exposed as the base level in all of our geo-trenches within the basin (Figure 3).

Figure 3.

The upper paleolithic site of Otson Tsokhoi. Photograph shows the sandy high beach terrace above the white marl of the former Pleistocene lake.

Middlep-Holocene exposures and deposits

Fluvial facies

GeoSection ZU-16-1 (Figure 4a) (N46°08′06.2″, E111°42′28.9″): This section represents the sediment remains of an active Mid-Holocene stream which drains from the hilly catchment due north of the basin directly down into the basin depression. It is capped by about 10 cm of slope wash from the basalt hill immediately behind the section to the west. The deposits of Unit 1 (10–109 cm below surface) consist of sets of (a) well-sorted small (10⁻¹ mm diameter) gravels and well-sorted sands from channel facies, and (b) fine-grained slack-water floodplain deposits with high organic matter content indicating a partially perennial meandering stream system which drained the upland hills, and supplied the ponds along the valley floor. Unit 2 (109–159 cm bs) was radiocarbon dated to 7500–6300 cal BP (¹⁴C 6900 ± 300 UW3685). The deposits are very dark and loamy with CaCO₃ suggesting marshy facies along the paleo-stream. The basal Unit 3 (159–210 cm bs) is a silty loam with signs of illuviation and CaCO₃ flecks suggesting a truncated B-Horizon in the upper portion of the unit, formed on the remnants of a sandy Pleistocene beach or alluvium. Today, these Mid-Holocene stream deposits are deeply incised down through the bedrock. The current drainage channel accommodates a mostly dry ephemeral streambed with a seasonal flash-flood regime.

Figure 4.

Geological sections of two Mid-Holocene landscape facies: (a) geo-section ZU-16-1 showing the alluvial sediments of a Mid-Holocene perennial stream feeding into the Zaraa Uul basin and (b) geo-section ZU-16-3, showing the Mid-Holocene wetland phases of organic-rich marshy deposits overlaying the Pleistocene marls.

Wetland facies

GeoSection ZU-16-3 (Figure 4b) (N46°07′41.2″, E111°42′46.0″): These deposits are located on the valley floor of the basin and represent wetland and ponded deposits dating to the Mid-Holocene (ca. 6300 cal BP) (See Table 1). They overlay truncated marls that are remnants of the Late Pleistocene saline lake deposits. The upper 15 cm of this section (Unit 1) is blanketed in coarse sandy colluvial sheet wash, this overlies Unit 2, a deposit of over a meter of fine-grained, organic rich compacted clayey silts containing in situ CaCO₃ nodules. These deposits represent a thick layer of still-water sediments indicating marshy ponded conditions. Unit 3, the basal layer of this section was exposed between 120 and 130 cm below surface. Here we exposed the whitish marls of the Late Pleistocene saline lake. There was an unconformable boundary between Units 2 and 3 indicating a long episode of truncation and erosion of the Pleistocene marls under dry conditions before the Mid-Holocene marshy wetlands were established.

Table 1.

Radiocarbon dates for Mid-Holocene geological sections ZU-16-1 (alluvial facies) and ZU-16-3 (wetland facies).

UOC-4480	Geological	GEO-16-ZU3_2b	Bulk sediment	5534 ± 31	6399–6288 (95.4%)
UOC-4481	Geological	GEO-16-ZU3_2b	Bulk sediment (humic)	3651 ± 29	4013–3890 (70.8%)
UOC-4482	Geological	GEO-16-ZU3_2b	Bulk sediment (humin)	5547 ± 27	6399–6296 (95.4%)
UW3684	Geological	GEO-16-ZU3_2a	Sediment (IRSL)	6700 ± 300	7300–6100 (95.4%)
UW3685	Geological	GEO-16-ZU1	Sediment (IRSL)	6900 ± 300	7500–6300 (95.4%)

Calibration was performed using OxCal v4.4.2, and the IntCal20 calibration curve (Janz et al., 2021).

Interpretation

A summary of geomorphological changes in the Zaraa Uul Basin from the late Pleistocene through the present is illustrated in Figure 5. Our preliminary geoarchaeological study found that there are remnants of several high beach and stream terraces surrounding the former Pleistocene lake basin. The highest of these beach terraces that we observed was at an approximate elevation of 1043 m (Garmin GPS reading) (N46°08′21.7″/E111°43′02.7″) and is dated by OSL to the early Upper Paleolithic Period (ca. 28,000–34,000 cal BP) (Janz et al., 2021). Davaakhuu found and identified lithic scatters in several of these high beach ridges at Otson Tsokhio (see Figure 3) and Zuun Shovkh (Janz et al., 2021). The artifacts there are eroding out of the whitish/yellow silty beach sand deposits through deflation. It is likely that this represents a late Pleistocene maximum extent of the lake level, which would place this high stand of the Pleistocene lake near the end of the Oxygen Isotope Stage 3 Interglacial. At this elevation, the lake would have filled most of the entire basin (Figure 5a).

Figure 5.

Schematic geological profile of the Zaraa Uul Basin showing geomorphological changes from (a) late pleistocene high saline lake stand, (b) terminal pleistocene drying of the saline lake, (c) Mid-Holocene marshes and ponds forming in the topographic swales within the basin, and (d) modern eroded landscape after the drying out of the basin.

A very severe drop in lake level – quite possibly a drying up of most of the basin – ensued as the next phase of geomorphological activity. It is likely that this episode coincided with the Terminal Pleistocene Younger Dryas (YD) cold/dry episode (GISP 2) recorded across Europe, Central and East Asia, dating to ca. 13,500–ca. 11,700 BP. A similar low lake level is recorded at Gun Nuur Lake in northern Mongolia within the Terminal Pleistocene/early Holocene time frame (Zhang et al., 2012). In the side valleys leading to the lake basin, the former Late Pleistocene beach deposits were deeply incised. The exposed marl of the former Pleistocene Lake basin was also incised, leaving behind hummocks of marl and deeper swales between them (Figure 5b).

The return to humid/warm conditions is well known across East Asia during the Mid-Holocene, ca 8000–ca. 4000 BP. Although the eastern Gobi Desert is quite far north vis-à-vis the East Asian Monsoon’s northward migration, there is ample paleoenvironmental data suggesting that the Mongolian Plateau also enjoyed moister climatic intervals at this time (Feng et al., 2005), although some lake records suggest localized dry phases (Chen et al., 2003). At Zaraa Uul, our chrono-stratigraphic model suggests that the Mid-Holocene Climatic Optimum manifested itself with the resumption of stronger stream activity under a regime of increased rainfall events, as demonstrated by the alluvial and marsh sediment deposits recorded in our geomorphological sections (ZU-16-1 and ZU-16-3). These deposits show renewed stream activity, floodplain buildup, and back-swamp formation from overbank flooding in the stream valley draining into the main basin (Janz et al., 2021). On the basin plain, increased rainfall and outwash from side valleys filled the swales which had been sculpted into the marl from the previous erosional episodes. These swales became fresh-water seasonal marshes as indicated by the heavy clay build-up with very high dark organic matter content as seen from the ca. one-meter build-up of these clays in geomorphological sections we excavated into these marsh deposits (Figures 4b and 5c). These deposits had radiocarbon dates on humin of 6399–6296 cal BP, with a later intrusion of humic acids dated to 4013–3890 cal BP (Janz et al., 2021) suggesting that the marshy environment dried and became a soil by that later date. The black clay unit in these sections also contain large well-developed CaCO₃ nodules, indicating that these ponds were seasonally dry. In the Late-Holocene, the ponds disappeared, and there was another major erosional phase which likely continued into the present. Former hillslope sediments and soils have been eroded away and redeposited onto modern exposed surfaces (Figure 5d).

The geoarchaeological field observations show that during most of the Late Pleistocene period, the Zaraa Uul Basin held a large saline lake which was visited by big game hunters. During the terminal Pleistocene cold/dry Younger Dryas period, the lake disappeared, leaving behind thick marl deposits that were subsequently incised and eroded into an uneven hummocky surface. The renewed moisture of the Middle-Holocene Optimum was fueled by strong East Asian Monsoonal systems. At this time, perennial streams flowed into the basin, filling the swales in the old lakebed, creating marshy wetlands and ponds. These wetlands facilitated the multiple occupations of the forager campsite at the Zaraa Uul site.

The site of Zaraa Uul contained abundant remains of hunter-gatherers who inhabited the basin valley from around 8500 BP. Another occupation period occurred 6276–6003 cal BP. These two occupation periods span the period in which small freshwater ponds and marshes existed on the valley floor, within the hummocks created by the erosion of the former Pleistocene marl deposits (Janz et al., 2021). It is also notable, that the locality of the Zaraa Uul site appears to be the only location around the basin with archeological remains dating to this Mid-Holocene moist period. It indicates that Neolithic Period foragers recurringly returned to this particular spot. This is in contrast to the Late Pleistocene Paleolithic sites which appear in at least two localities associated with outcrops of Pleistocene beach terraces. Although the evidence is sparse, it might suggest different patterns of movement, and placement of basecamps between the two periods.

Phytoliths

We were able to reconstruct some of the vegetation at the site from the phytolith samples taken from the Mid-Holocene period archeological excavation units (Table 2 and Figure 6) (also see Janz et al. (2021)). The vegetation picture we obtained from the phytoliths suggests a grassland steppe that is lusher than the groundcover in the basin today. Figure 6 is a graph showing the phytolith types divided into categories of the microenvironments they represent. It indicates that temperate steppe and wetland forms dominate the phytoliths from all of the samples. Stipa grasses are well-represented across most of the samples for the Zaraa Uul Site excavation units. “Dendriform” long-cell morphotypes occur in the samples from both the earlier occupation phases identified from Locus 2 and the later Locus 1, in the form of phytoliths with multiple cells joined together. Dendriforms occur only in the florets and seed casings (glume, palea, lemma) of grasses, and are good proxies for the collection and consumption of small-grained grasses and/or the season of site occupation. Since grasses in this region flower from late June through August, and form seeds from late July through September (Jigjidsuren and Johnson, 2003), we can posit a possible site occupation during the summer months for both Mid-Holocene occupation phases at the site. This is supported by faunal remains at the site including waterfowl and hibernating mammals including marmot (Marmota sibirica) and badger (Meles leucurus) in Locus 2 (Janz et al., 2021). This is a time period in which the summer monsoons would have brought abundant rainfall to this area, filling the streams and ponds with runoff water, and contributing to an array of lush grasses attracting herds of wild grazing animals, and proliferating Underground Storage Organs from an abundance of sedges (Cyperaceae).

Table 2.

Phytolith results from Mid-Holocene archeological sediments (test pits 1 and 2, levels 2 and 3 from Zaraa Uul), arranged according to microenvironmental zones they represent.

Sample	GSN-16-6	GSN-16-7	GSN-16-8	GSN-16-9	GSN-16-10	GSN-16-11	GSN-16-12	GSN-16-16	GSN-16-17	GSN-16-18
Context	L2	L2	L2	L2	L2	L2	L2	L1	L1	L1
Single-cell	n/gm	n/gm	n/gm	n/gm	n/gm	n/gm	n/gm	n/gm	n/gm	n/gm
Wetland forms
Sedge long cells (rods)	1152	1662	1320	1613	836	30	1920	6807	4431	7200
Bulliform (Panicoid)	1498	1329	2160	1152	2230	90	3264	5236	4431	8064
Fan-shaped Bulliform	115	443	840	461	1579	90	192	1833	1329	2304
Phragmites Bulliform	0	0	0	230	0	0	0	0	0	0
Sedge cones	3802	997	2160	1037	1394	0	960	4713	3545	7776
Total wetland	6567	4431	6480	4493	6039	210	6336	18,589	13,736	25,344
Temperate grassland forms (Pooid)
Rondels	1036	7200	9240	8525	6875	30	19,968	20,422	20,382	27,360
Stipa-type Rondel	346	6757	1296	9331	5203	0	17,664	11,520	11,742	7776
Total temperate grassland	1382	13,957	10,536	17,856	12,078	30	37,632	31,942	32,124	35,136
Semi-arid grassland forms (chlorodoid)
Saddle-shape forms	1498	4985	1440	806	372	0	2880	2095	443	2016
Total semi-arid grassland	1498	4985	1440	806	372	0	2880	2095	443	2016
Phytoliths from florets
Dendriform long cells	346	332	840	115	0	0	576	0	0	0
Papillae	0	0	0	115	0	0	960	0	0	0
Shrubs/trees
Platelets	691	332	600	230	1301	2340	576	524	665	864
Total shrubs/tree	691	332	600	230	1301	2340	576	524	665	864
Multi-cell forms	n/gm	n/gm	n/gm	n/gm	n/gm	n/gm	n/gm	n/gm	n/gm	n/gm
Leaf/stem: Psilate	0	240	540	300	360	210	384	0	420	480
Leaf/stem: Echinate	0	0	0	0	60	0	192	0	0	0
Leaf/stem: Sinuate	60	0	0	0	0	0	0	0	0	0
Wild grass husk	0	0	0	0	0	0	0	180	60	60
Panicoid	0	60	60	60	0	0	0	0	60	0
Cyperaceae	120	240	0	0	0	0	0	360	60	60
Stipa leaf/stem	0	0	240	0	180	30	192	0	60	0
Polyhedron	120	0	0	300	0	840	2496	240	180	300

Figure 6.

Graph showing numbers of phytolith forms per sample according to the microenvironments they represent.

All the samples contain “bulliform” morphotypes from reed-grasses, but these are more prevalent in samples from the later phase of occupation seen in Locus 1. This trend continues with somewhat better representation of the “cone” phytolith morphotype representing sedges, along with multi-celled forms from sedges (Cyperaceae) which grow in moist soils.

These are plants that prefer growing in moist and marshy environments. This may indicate that the freshwater ponds of the earlier Mid-Holocene period were giving way to more marshy wetlands, and their use suggest a possible increase in wetland exploitation as the moister conditions begin their trend to dryer environments which later culminated in the modern dry landscapes of the Late-Holocene period. The presence of sedges in the archeological deposits also suggests the USOs were consumed as a source of high-calorie starches.

Human adaptations to desertification in the Gobi Desert

Water sources, diversification of resources and natural storage of food sources

Our work at Zaraa Uul, demonstrates the use of wetlands is a key component of the resilience of small-scale mobile societies who had transitioned from megafauna hunting near the shores of a large saline lake during the late Pleistocene, to more intensive use of a broader spectrum of plants and animals in the Mid-Holocene. All of this took place in a volatile climatic milieu of increasingly moist conditions which eventually turned to persistent and profound desertification. In previous research, Janz defined three phases of terminal Pleistocene through Late-Holocene occupation, which are distinctly focused on the intensive use of wetlands and their resources in the semi-arid desert-steppe region (Janz, 2012). Janz’s terminology is outlined as:

(1) Oasis 1/Mesolithic (13,500–8000 cal BP), defined by sites showing the earliest use of the wetland environments and their adjacent ecozones in the vicinity of rivers or lakes.

(2) Oasis 2/Neolithic (8000–5000 cal BP) with sites indicating the intensive exploitation of wetland oases, characterized by camp sites within dune-fields and marshlands, and notably a common occurrence of grinding stones, suggesting the exploitation of small-seeded grasses and other steppe plants, along with a wider range of microlithic tool types, also suggesting the possibility of composite tools for hunting small-fast prey, harpooning fish, and possible reaping activities associated with these wetlands.

(3) Oasis 3/Bronze Age or Eneolithic (5000–3000 cal BP), overlapping with the introduction of herd animals, provides continuing evidence for the intensive use of wetland habitats and even intensified use in western regions, including larger numbers and types of ceramics, and bifacial flaking of projectile points, knives, and other small tools (Janz, 2012).

In the desert-steppe zones of the eastern Mongolian Gobi, our research has shown that Mid-Holocene foragers began intensive exploitation of a broader range of animal species than at the late Pleistocene sites near the basin, as well as small-fast “r-selected species” with high reproductive rates, in addition to the larger mammals such as camel, deer, horse and cattle, as the megafauna populations diminished at the end of the Pleistocene. Janz attributes this to an intensification of land-use potential (Janz et al., 2017). These foragers also exploited different plant resources from the Paleolithic hunters such as sedges, perhaps for the starchy “Underground Storage Organs” (UGS), and wild small-grained grass seeds. This went together with increasing warmth and humidity, and the early development of wetlands, as they began to appear within swales of eroded Pleistocene lake beds, and other localities in the Gobi dryland regions (Feng et al., 2005; Janz et al., 2021).

The subsistence activities of the mobile desert-steppe inhabitants during the Mid-Holocene are distinctly different from the megafauna hunting specialists which preceded the terminal Pleistocene, and the later Bronze Age pastoral specialists which succeeded these populations in the later Holocene (Janz et al., 2017). It is useful to think of the adaptations of the Mid-Holocene forager strategies to more diverse resource selection in terms of “Push” and “Pull” factors. The decline of megafauna populations pushed the Mid-Holocene foragers to use the landscape more intensively and exploit a greater diversity of resources, while the increased rainfall from the strengthening of the East Asian Monsoons, raising ground water tables and forming ponds and small lakes across the landscape of the eastern Gobi acted as a “Pull” factor expanding the diversity of starchy plants and smaller fauna within the home ranges. In southeastern Mongolia, the region would have been at the very northernmost edge of the East Asian Monsoonal systems, and thus still a semi-arid zone. Therefore, these newly formed wetlands would have been an attractive draw to the hunters and foragers living in this region for the reasons outlined above, including (a) the attraction of a water source for human consumption, (b) the presence of diverse species of plants and animals living within the mosaic of wetland microenvironments, and c) the potential for low-risk “back-loaded” storage of sedge rhizomes and other perennially available foodstuffs in a marshland environment.

These Mid-Holocene foragers would have employed adaptive strategies that were not available to the later Holocene inhabitants who encountered an ever increasingly dry environment. In the later Holocene the Gobi Desert experienced weakened East Asian Monsoons, which led to the disappearance of the lakes, marshlands, and small perennial streams in this region. It is also a period in which the inhabitants of the eastern Gobi adopted animal husbandry and a nomadic pastoral lifestyle (Honeychurch, 2017; Honeychurch and Makarewicz, 2016; Wright et al., 2019). At this time, the Basin wetlands had disappeared, along with a potentially decreased potential for subsistence diversity, although at Zaraa Uul, there is an extant fresh-water spring at the southern end of the basin. Presumably, this had been available throughout the Late-Holocene period of desertification and might have been an important source of water for flocks of the later incipient herders. The linear valley drainages in the Mongolian Gobi still host seasonal stream activity, and some still contain small springs and wet patches, which would have been a draw to the foragers and early herders as attractive corridors of movement as they migrated across the Gobi (Holguín, 2019; Holguín and Sternberg, 2018; Rosen et al., 2019). These Mid- to Late-Holocene patterns of movement that focused on a familiarity with water sources within home ranges might later have linked up over a much broader geographic range with the intensification of trade and the advent of the Silk Road in later Holocene times (Frachetti et al., 2017).

The Traditional Ecological Knowledge of survival in these environments, possibly adjusted to changing environments as wetlands formed in Mid-Holocene times. Traditions of plant and animal use were likely handed down throughout the generations. Even as the larger Mid-Holocene wetlands were shrinking down to tiny wet patches on the landscape, the knowledge of how to exploit them more efficiently was honed through the experience of generations.

Clearly, it is difficult to draw a direct link from the Traditional Ecological Knowledge of historic period herders back through thousands of years to the ancient Mid-Holocene foragers, but there is room for informed speculations which could lead to testable hypotheses. There are a growing number of studies about modern Mongolian herders, and their traditional knowledge about rangeland management and responses to recent climate changes (cf. Fernandez-Gimenez, 2000). One vehicle for accessing this information in the past is through analysis of settlement patterns and systematic collection of geobotanical remains at archeological sites to find converging strategies of land use and resource exploitation between ancestral and descendant communities. For Mongolia, one of these points of convergence might be the patterns of movement throughout the Gobi, the way water sources were located, and the array of settlement types around the water sources themselves.

One study of pre-modern Mongolian herder TEK indicates that the primary driver of migration patterns is the seasonal availability of water sources, and good pastureland is a secondary concern. Tugjamba et al. (2021: 9) write that the main concern with desertification is the reduction of water resources and this determines herders’ seasonal migration patterns and livelihoods. They quote one herder as saying: “If there is no water, it does not matter how beautiful is a grassland, it is not a pastureland.” They make the point that in drought years, herders will reduce their mobility to be located for longer periods around viable water sources. This allows us to formulate hypotheses for deep-time traditions of migration patterns of hunter/foragers, hunter/herders, and later full-time herders – not as ethnographic analogy, but as patterns that existed and transmitted through generations of ancestral/descendant knowledge.

If we compare this to information gathered from the archeological site of Zaraa Uul, it frames the site location and its longevity in new ways. Our geoarcheological research has shown that there were freshwater ponds within the basin near the site during the Mid-Holocene. Archeological survey and excavations showed there were recurring occupations at that one location at different time periods in the Mid-Holocene, yet little or no evidence for contemporary occupations in other localities around the basin. This might indicate that mobile hunter/foragers were tethered to that “persistent place” where tradition indicated there would be a reliable source of water for the hunters as well as the game animals, and thus a low-risk locality to visit and revisit over generations until the water source dried up permanently (Maher, 2019; Olszewski and Al-Nahar, 2016; Schlanger, 1992). The same might be hypothesized for the linear patterns of movement north and south through the valleys of the Ikh Nart Nature Reserve (Dornogovi Province) (Rosen et al., 2019; Schneider et al., 2016) and the southern Gobi region around Ulaan Nuur (Holguín and Sternberg, 2018) where the wild herds also would have followed the line of springs.

Another possible link between the TEK of the present and that of the distant past might be in the knowledge and timing of the ripening of grassy pasture lands, and best times to harvest wild plants (Tugjamba et al., 2021: 7). Grassy pastureland provides important forage for domestic herds, and also would have attracted wild herds that were prey for ancient hunters. Archeologically we could approach this information from the past through archaeobotanical remains, and particularly phytoliths. Our findings of grass floral parts at Zaraa Uul, suggest that the Mid-Holocene hunter-gatherers were visiting the site during a period of abundant grass growth, something that would attract ancient wild herds as well as modern domestic ones.

Finally, sites such as Zaraa Uul which were visited multiple times perhaps functioned as base camps. The importance of such gathering spots extends far beyond their economic role as places to accumulate an abundance of resources. They also served as localities to gather yearly to exchange information and transmit ecological knowledge to a large number of related groups. Sattler et al. (2021) note that throughout the year, mobile herders do not have many opportunities to meet and exchange information relating to obtaining food, water, and shelter. However, this kind of knowledge exchange is essential for survival in the challenging desert and mountain environments of Mongolia – especially in times of rapid climate change. The authors point to the importance of a long tradition of celebrating the summer festival of Naadam. This annual event provides an opportunity for nomadic herders to come together, participate in traditional games, and also to exchange information about pasture and water sources that is critical for sustaining lifeways and herds.

Final remarks

Geographers have pointed out the diversity of landforms, vegetation, and hydrological systems in desert landscapes. Humans have greatly benefited from this diversity by the long-term transmission of experience relating to the kinds of resources available in each of the diverse microenvironments. In this way mobile populations can gage the availability of water, plants, and animals in a climatic regime of unpredictable rainfall timing and amounts. These factors of water source predictability, diversification of subsistence resources, storability, and the transmission of expertise through thousands of years of accumulated Traditional Ecological Knowledge, are the core and foundation of resilience among societies living in dryland regions. Although it is difficult to access information about TEK for many past societies, perhaps an understanding of the archeological evidence for recurring patterns of resource exploitation under varying climatic conditions will allow us to gain a better understanding of the importance of long-term memory-messages through deep time and TEK in past societies (Rosen and Rivera-Collazo, 2012).

Footnotes

Acknowledgements

We gratefully acknowledge Chunag Amartuvshin for facilitating the archeological research at Zaraa Uul. The phytolith analyses were conducted at the Environmental Archeology Laboratory, Department of Anthropology, University of Texas at Austin. Steve Rosen provided helpful editorial comments. We also thank the two anonymous reviewers for their insightful comments.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Funding for the Zaraa Uul fieldwork was provided to LJ by the American Philosophical Society Franklin Grant (https://www.amphilsoc.org/), the Wenner-Gren Foundation (Grant No. 9050) (http://www.wennergren.org/), and the Social Sciences and Humanities Research Council of Canada’s Post-Doctoral Fellowship (Grant No. 756-2015-0019) and Insight Development Grant (Grant No. 430-2016-00173) ().

References

Abrams

Nowacki

(2008) Native Americans as active and passive promoters of mast and fruit trees in the eastern USA. The Holocene 18(7): 1123–1137.

Asmussen

McInnes

(2013) Assessing the impact of mid-to-late-holocene ENSO-driven climate change on toxic Macrozamia seed use: A 5000 year record from eastern Australia. Journal of Archaeological Science 40(1): 471–480.

Barker

Creighton

Gilbertson

, et al. (1997) The Wadi Faynan project, southern Jordan: A preliminary report on geomorphology and landscape archaeology. Levant 29: 19–40.

Bazarova

Tsydenova

Lyaschevskaya

, et al. (2019) Reconstruction of paleoenvironmental conditions of ancient people habitation in the Togootyn Gol River valley (Eastern Mongolia). Quaternary International 503: 105–114.

Bettinger

(2009) Hunter–Gatherer Foraging: Five Simple Models. Clinton Corners, NY: Eliot Werner Publications Inc.

Bird

(2005) Evolutionary and ecological understandings of the economics of desert societies comparing the Great Basin USA and the Australian deserts. In: Veth

Smith

Hiscock

(eds) Desert Peoples: Archaeological Perspectives. Oxford: Blackwell, pp.81–99.

Butzer

(1982) Archaeology as Human Ecology: Method and Theory for a Contextual Approach. Cambridge: Cambridge University Press.

Charlier

(2016) Actions rituelles, mobilité et attachement au «pays natal» parmi des éleveurs nomades de mongolie. Études mongoles et sibériennes, centrasiatiques et tibétaines. Epub ahead of print 2016. DOI: 10.4000/emscat.2779.

Chen

C-TA

Lan

H-C

Lou

J-Y

, et al. (2003) The dry holocene megathermal in inner Mongolia. Palaeogeography Palaeoclimatology Palaeoecology 193(2): 181–200.

10.

Dubreuil

Evoy

Janz

(2021) The new oasis: Potential of use-wear for studying plant exploitation in the Gobi desert neolithic. In: Pedersen

Jörgensen-Lindahl

Sørensen

, et al. (eds) Proceedings of the 3rd Meeting of the Association of Ground Stone Tools Research. Oxford: Archaeopress Publishing Ltd, pp.116–138.

11.

Felauer

Schlütz

Murad

, et al. (2012) Late quaternary climate and landscape evolution in arid Central Asia: A multiproxy study of lake archive Bayan Tohomin Nuur¢, Gobi desert, southern Mongolia. Journal of Asian Earth Sciences 48: 125–135.

12.

Feng

Wang

Guo

, et al. (2005) Lacustrine and eolian records of holocene climate changes in the Mongolian plateau: Preliminary results. Quaternary International 136(1): 25–32.

13.

Fernandez-Gimenez

(2000) The role of Mongolian nomadic pastoralists’ ecological knowledge in rangeland management. Ecological Applications 10(5): 1318–1326.

14.

Fernánedez-Giménez

(1999) Sustaining the steppes: A geographical history of pastoral land use in Mongolia. Geography Review 89(3): 315–342.

15.

Frachetti

Smith

Traub

, et al. (2017) Nomadic ecology shaped the highland geography of Asia’s silk roads. Nature 543(7644): 193–198.

16.

Fuller

Lucas

González Carretero

, et al. (2018) From intermediate economies to agriculture: Trends in wild food use, domestication and cultivation among early villages in Southwest Asia. Paléorient 44(2): 59–74.

17.

Garrard

Byrd

(2013) Beyond the fertile crescent: Late palaeolithic and neolithic communities of the Jordanian Steppe. The Azraq Basin project volume 1: Project background and the late palaeolithic (geological context and technology). Oxford: Oxbox Books.

18.

Gerber

Reichman

Roughgarden

(2004) Food hoarding: Future value in optimal foraging decisions. Ecological Modelling 175(1): 77–85.

19.

Hammer

(2014) Local landscape organization of mobile pastoralists in southeastern Turkey. Journal of Anthropological Archaeology 35: 269–288.

20.

Hammer

(2018) Water management by mobile pastoralists in the middle east. In: Holt

(ed.) Water and Power in Past Societies. New York: State University of New York Press, pp.63–88.

21.

Harrower

(2016) Water histories of ancient Yemen in global comparative perspective. In: Harrower

(ed.) Water Histories and Spatial Archaeology: Ancient Yemen and the American Wes. Cambridge: Cambridge University Press, pp.51–83.

22.

Hermes

Shnaider

Semibratov

, et al. (2021) Revisiting the emergence of pastoralism in the Altai Mountains through interactions between local hunter-gatherer and Afanasievo communities. Archaeological Research in Asia 26: 100281.

23.

Hillman

Madeyska

Hather

(1989) Wild plant foods and diet at late Palaeolithic Wadi Kubbaniya (Upper Egypt): Evidence from charred remains. In: Wendorf

Schild

(eds) The Prehistory of Wadi Kubbaniyavol: Stratigraphy, Subsistence and Environment. Dallas: Southern Methodist University, pp.162–242.

24.

Hiscock

Wallis

(2005) Pleistocene settlement of deserts from an Australian perspective. In: Veth

Smith

Hiscock

(eds) Desert Peoples: Archaeological Perspectives. Oxford: Blackwell, pp.34–57.

25.

Holguín

(2019) The changing neolithic landscape of Ulaan Nuur: Modelling hydro-social dynamics in the Mongolian Gobi desert. PhD Dissertation. University of Southampton.

26.

Holguín

Sternberg

(2018) A GIS based approach to holocene hydrology and social connectivity in the Gobi Desert, Mongolia. Archaeological Research in Asia 15: 137–145.

27.

Honeychurch

(2017) The development of cultural and social complexity in Mongolia. In: Habu

Lape

Olsen

(eds) Handbook of East and Southeast Asian Archaeology. New York: Springer, pp.513–532.

28.

Honeychurch

Makarewicz

(2016) The archaeology of pastoral nomadism. Annual Review of Anthropology 45(1): 341–359.

29.

Janetski

Madsen

(1990) Wetland adaptations in the Great Basin. In: Brigham Young University Museum of Peoples and Cultures Occasional Papers 1. Salt Lake City: University of Utah Press.

30.

Janz

(2012) Chronology of post-glacial hunter-gatherers in the Gobi Desert and the neolithization of arid Mongolia and China. Unpublished PhD Dissertation. University of Arizona.

31.

Janz

(2016) Fragmented landscapes and economies of abundance: The broad-spectrum revolution in arid East Asia. Current Anthropology 57(5): 537–564.

32.

Janz

Cameron

Bukhchuluun

, et al. (2020) Expanding frontier and building the sphere in arid East Asia. Quaternary International 559: 150–164.

33.

Janz

Odsuren

Bukhchuluun

(2017) Transitions in palaeoecology and technology: Hunter-gatherers and early herders in the Gobi Desert. Journal of World Prehistory 30(1): 1–80.

34.

Janz

Rosen

Bukhchuluun

, et al. (2021) Zaraa Uul: An archaeological record of pleistocene-Holocene palaeoecology in the Gobi Desert. PLoS One 16(4): e0249848.

35.

Jigjidsuren

Johnson

(2003) Forage Plants in Mongolia. Ulaanbaatar: Admon.

36.

Kakinuma

Okayasu

Jamsran

, et al. (2014) Herding strategies during a drought vary at multiple scales in Mongolian rangeland. Journal of Arid Environments 109: 88–91.

37.

Kaptijn

(2018) Irrigation and social organization: A Longue duree perspective from the Jordan Valley. In: Holt

(ed.) Water and Power in Past Societies. Albany: State University of New York Press, pp.41–61.

38.

Kelly

(1990) Marshes and mobility in the western Great Basin. In: Janetski

Madsen

(eds) Wetland Adaptations in the Great Basin. Salt Lake City: University of Utah Press, pp.259–276.

39.

Kelly

(2013) The Lifeways of Hunter-Gatherers: The Foraging Spectrum. New York: Cambridge University Press.

40.

Klinge

Sauer

(2019) Spatial pattern of late glacial and holocene climatic and environmental development in western Mongolia - A critical review and synthesis. Quaternary Science Reviews 210: 26–50.

41.

Lee

Lim

, et al. (2013) Late pleistocene-Holocene records from Lake Ulaan, southern Mongolia: Implications for east Asian palaeomonsoonal climate changes. Journal of Quaternary Science 28(4): 370–378.

42.

Lehmkuhl

Grunert

Hülle

, et al. (2018) Paleolakes in the Gobi region of southern Mongolia. Quaternary Science Reviews 179: 1–23.

43.

Madsen

Jin

, et al. (2018) Orbital scale lake evolution in the Ejina Basin, central Gobi Desert, China revealed by K-feldspar luminescence dating of paleolake shoreline features. Quaternary International 482: 109–121.

44.

Lillie

Ellis

Fenwick

(2007) Wetland Archaeology & Environments: Regional Issues, Global Perspectives. Oxford [England] Oakville, CT: Oxbow Books, D. Brown Book Co.

45.

Linstädter

Kuhn

Naumann

, et al. (2016) Assessing the resilience of a real-world social-ecological system: Lessons from a multidisciplinary evaluation of a South African pastoral system. Ecology and Society 21(3): 35–50.

46.

Liu

Bestel

Shi

, et al. (2013) Paleolithic human exploitation of plant foods during the last glacial maximum in North China. Proc Natl Acad Sci USA 110(14): 5380–5385.

47.

Liu

Kealhofer

Chen

, et al. (2014) A broad-spectrum subsistence economy in neolithic inner Mongolia, China: Evidence from grinding stones. The Holocene 24(6): 726–742.

48.

Liu

Levin

Bonomo

, et al. (2018) Harvesting and processing wild cereals in the upper palaeolithic yellow river valley, China. Antiquity 92(363): 603–619.

49.

Lourandos

(1980) Change or stability?: Hydraulics, hunter-gatherers and population in temperate Australia. World Archaeology 11(3): 245–264.

50.

Maher

(2019) Persistent place-making in prehistory: The creation, maintenance, and transformation of an epipalaeolithic landscape. Journal of Archaeological Method and Theory 26(3): 998–1083.

51.

Maher

Macdonald

Allentuck

, et al. (2016) Occupying wide open spaces? Late pleistocene hunter–gatherer activities in the eastern levant. Quaternary International 396: 79–94.

52.

Marin

(2010) Riders under storms: Contributions of nomadic herders’ observations to analysing climate change in Mongolia. Global Environmental Change 20(1): 162–176.

53.

Mason

SLR

(1995) Acornutopia? Determining the role of acrons in past human subsistence. In: Wilkins

Harvey

Dobson

(eds) Food in Antiquity. Exeter: Exeter University Press, pp.12–14.

54.

McCorriston

Martin

(2010) Southern Arabia’s early pastoral population history: Some recent evidence. In: Petraglia

Rose

(eds) The Evolution of Human Populations in Arabia: Paleoenvironments, Prehistory and Genetics. Dordrecht: Springer Netherlands, pp.237–250.

55.

Miller

Davidson-Hunt

(2013) Agency and resilience: Teachings of Pikangikum first nation elders, northwestern Ontario. Ecology and Society 18(3): 9–24.

56.

Mischke

Lee

(2020) Climate history of southern Mongolia since 17 ka: The Ostracod, gastropod and Charophyte record from Lake Ulaan. Earth Science Frontiers 8: 221.

57.

Mithen

(2010) The domestication of water: Water management in the ancient world and its prehistoric origins in the Jordan valley. Philosophical Transactions Series A, Mathematical, Physical, and Engineering Sciences 368(1931): 5249–5274.

58.

Morgan

(2012) Modeling modes of hunter-gatherer food storage. American Antiquity 77(4): 714–736.

59.

Murphy

French

CAI

(1988) The Exploitation of Wetlands. Oxford: B.A.R.

60.

Nelson

(1996) Technological strategies responsive to subsistence stress. In: Tainter

Tainter

(eds) Evolving Complexity and Environmental Risk in the Prehistoric Southwest. Reading: Addison-Wesley, pp.107–143.

61.

Nicholas

(1998) Wetlands and hunter-gatherers: A global perspective. Current Anthropology 39(5): 720–731.

62.

Olszewski

Al-Nahar

(2016) Persistent and ephemeral places in the early epipaleolithic in the Wadi al-Hasa region of the western highlands of Jordan. Quaternary International 396: 20–30.

63.

Preston

Parker

Walkington

, et al. (2012) From nomadic herder-hunters to sedentary farmers: The relationship between climate change and ancient subsistence strategies in south-eastern Arabia. Journal of Arid Environments 86: 122–130.

64.

Ramsey

Jones

Richter

, et al. (2015) Modifying the marsh: Evaluating early epipaleolithic hunter-gatherer impacts in the Azraq wetland, Jordan. The Holocene 25(10): 1553–1564.

65.

Ramsey

Maher

Macdonald

, et al. (2016) Risk, reliability and resilience: Phytolith evidence for alternative ‘Neolithization’ pathways at Kharaneh IV in the Azraq Basin, Jordan. PLoS One 11(10): e0164081.

66.

Ramsey

Rosen

(2016) Wedded to wetlands: Exploring late pleistocene plant-use in the eastern levant. Quaternary International 396(7): 5–19.

67.

Richter

Maher

Garrard

, et al. (2013) Epipalaeolithic settlement dynamics in southwest Asia: New radiocarbon evidence from the Azraq Basin. Journal of Quaternary Science 28(5): 467–479.

68.

Rosen

(2010) Natufian plant exploitation: Managing risk and stability in an environment of change. Eurasian Prehistory 7(1): 117–131.

69.

Rosen

(2013) Natufian foragers and the ‘Monocot Revolution’: A phytolith perspective. In: Bar-Yosef

Valla

(eds) Natufian Foragers in the Levant: Terminal Pleistocene Social Changes in Western Asia. Ann Arbor, MI: International Monographs in Prehistory, pp.638–648.

70.

Rosen

(2022) Resilience at the edge: Strategies of small-scale societies for long-term sustainable living in dryland environments. In: Izdebski

Haldon

Filipkowski

(eds) Perspectives on Public Policy in Societal-Environmental Crises: What the Future Needs from History. New York: Springer Nature, pp. 161–176.

71.

Rosen

Hart

Farquhar

, et al. (2019) Holocene vegetation cycles, land-use, and human adaptations to desertification in the Gobi Desert of Mongolia. Vegetation History and Archaeobotany 28(3): 295–309.

72.

Rosen

Rivera-Collazo

(2012) Climate change, adaptive cycles, and the persistence of foraging economies during the late pleistocene/Holocene transition in the levant. PNAS 109(10): 3640–3645.

73.

Rosen

(2021) The herding revolution in the desert: Adoption, adaptation, and social evolution in the Negev and levantine deserts. In: Rowland

Lucarini

Tassie

(eds) Revolutions – The Neolithisation of the Mediterranean Basin: The Transition to Food Producing Economies in North Africa, Southern Europe and the Levant. Berlin: Excellence Cluster Topoi, pp.231–244.

74.

Rowley-Conwy

Zvelebil

(1989) Saving it for later: Storage by prehistoric hunter–gatherers in Europe. In: O’Shea

Halstead

(eds) Bad Year Economics: Cultural Responses to Risk and Uncertainty. Cambridge: Cambridge University Press, pp.40–56.

75.

Sattler

Bishkhorloo

Graham

(2021) Climate change threatens nomadic herding in Mongolia: A model of climate change risk perception and behavioral adaptation. Journal of Environmental Psychology 75: 101620.

76.

Schlanger

(1992) Recognizing persistent places in Anasazi settlement systems. In: Rossignol

Wandsnider

(eds) Space, Time, and Archaeological Landscapes. Boston, MA: Springer US, pp.91–112.

77.

Schneider

Yadmaa

Hart

, et al. (2016) Mongolian “neolithic” and early bronze age ground stone tools from the northern edge of the Gobi Desert. Journal of Lithic Studies 3(3): 479–497.

78.

Sepie

(2017) More than stories, more than myths: Animal/human/nature(s) in traditional ecological worldviews. Humanities 6(4): 78.

79.

Sillitoe

(1998) The development of indigenous knowledge: A new applied anthropology. Current Anthropology 39(2): 223–252.

80.

Smith

(2001) Low-level food production. Journal of Archaeological Research 9(1): 1–43.

81.

Smith

Veth

Hiscock

, et al. (2005) Global deserts in perspective. In: Veth

Smith

Hiscock

(eds) Desert Peoples: Archaeological Perspectives. Oxford: Blackwell, pp.1–13.

82.

Stiner

Munro

(2002) Approaches to prehistoric diet breadth, demography, and prey ranking systems in time and space. Journal of Archaeological Method and Theory 9(2): 181–214.

83.

Stutz

Munro

Bar-Oz

(2009) Increasing the resolution of the broad spectrum revolution in the southern levantine epipaleolithic (19–12 ka). Journal of Human Evolution 56(3): 294–306.

84.

Tugjamba

Walkerden

Miller

(2021) Climate change impacts on nomadic herders’ livelihoods and pastureland ecosystems: A case study from Northeast Mongolia. Regional Environmental Change 21(4): 1–12.

85.

Turner

Łuczaj

ŁJ

Migliorini

, et al. (2011) Edible and tended wild plants, traditional ecological knowledge and agroecology. Critical Reviews in Plant Sciences 30(1–2): 198–225.

86.

Tushingham

Bettinger

(2013) Why foragers choose acorns before salmon: Storage, mobility, and risk in aboriginal California. Journal of Anthropological Archaeology 32(4): 527–537.

87.

Undarmaa

Okuro

Manibazar

, et al. (2015) Rangeland Plants of Mongolia: Desert Steppe and Desert Zones, Extra-Zone. Ulaanbaatar, Mongolia: Munkhiin Useg.

88.

Venable

NBH

Fassnacht

Hendrick

(2015) Spatial changes in climate across Mongolia. In: Férnández-Giménez

Baival

Fassnacht

, et al. (eds) Proceedings of Building Resilience of Mongolian Rangelands: A Trans-disciplinary Research Conference. Ulaanbataar, Mongolia: Nutag Action and Research Institute, pp.73–79.

89.

Veth

(1989) Islands in the interior: A model for the colonization of Australia’s arid zone. Archaeology in Oceania 24(3): 81–92.

90.

Veth

(2005) Cycles of mobility and human mobility: Risk minimization among late pleistocene foragers of the Western Desert, Australia. In: Veth

Smith

Hiscock

(eds) Desert Peoples: Archaeological Perspectives. Oxford: Blackwell, pp.100–115.

91.

Weiss

Wetterstrom

Nadel

, et al. (2004) The broad spectrum revisited: Evidence from plant remains. Proc Natl Acad Sci USA 101(26): 9551–9555.

92.

Wollstonecroft

Ellis

Hillman

, et al. (2008) Advances in plant food processing in the near Eastern Epipalaeolithic and implications for improved edibility and nutrient bioaccessibility: An experimental assessment of Bolboschoenus maritimus (L.) Palla (sea club-rush). Vegetation History and Archaeobotany 17: 19–27.

93.

Wright

(2019) Long-term dynamics of pastoral ecology in northern Kenya: An old model for new resilience. Journal of Anthropological Archaeology 55: 101068.

94.

Wright

(2006) The adoption of pastoralism in northeast Asia: Monumental transformation in the Egiin Gol valley, Mongolia. Unpublished PhD Dissertation. Harvard University, Cambridge.

95.

Wright

Ganbaatar

Honeychurch

, et al. (2019) The earliest Bronze Age culture of the south-eastern Gobi Desert, Mongolia. Antiquity 93: 393–411.

96.

Lehmkuhl

Schlütz

, et al. (2019) Late quaternary environments in the Gobi Desert of Mongolia: Vegetation, hydrological, and palaeoclimate evolution. Palaeogeography Palaeoclimatology Palaeoecology 514: 77–91.

97.

Zhang

Feng

, et al. (2012) Holocene hydrological and climatic change on the northern Mongolian Plateau based on multi-proxy records from Lake Gun Nuur. Palaeogeography Palaeoclimatology Palaeoecology 323–325: 75–86.

98.

Zhang

Chen

Holmes

, et al. (2011) Holocene monsoon climate documented by oxygen and carbon isotopes from lake sediments and peat bogs in China: A review and synthesis. Quaternary Science Reviews 30(15–16): 1973–1987.