Early Pastoral Economies and Herding Transitions in Eastern Eurasia

Abstract

While classic models for the emergence of pastoral groups in Inner Asia describe mounted, horse-borne herders sweeping across the Eurasian Steppes during the Early or Middle Bronze Age (ca. 3000-1500 BCE), the actual economic basis of many early pastoral societies in the region is poorly characterized. In this paper, we use collagen mass fingerprinting and ancient DNA analysis of some of the first stratified and directly dated archaeofaunal assemblages from Mongolia's early pastoral cultures to undertake species identifications of this rare and highly fragmented material. Our results provide evidence for livestock-based, herding subsistence in Mongolia during the late 3rd and early 2nd millennia BCE. We observe no evidence for dietary exploitation of horses prior to the late Bronze Age, ca. 1200 BCE - at which point horses come to dominate ritual assemblages, play a key role in pastoral diets, and greatly influence pastoral mobility. In combination with the broader archaeofaunal record of Inner Asia, our analysis supports models for widespread changes in herding ecology linked to the innovation of horseback riding in Central Asia in the final 2nd millennium BCE. Such a framework can explain key broad-scale patterns in the movement of people, ideas, and material culture in Eurasian prehistory.

Figure 1

Selected archaeological sites discussed in the text.

Figure 2

(A) Habitation structure and occupation surface at Bagsagiin Bulan, in northern Mongolia. (B) Structure shown as 3D photogrammetric model. Central hearth feature is eroding from the bank, while the upper third of the structure remains along with upright support stones. (C) Artist’s reconstruction of habitation structure at Bagsagiin Bulan (Drawing by V. Pham). (D) Aerial photograph of riverbank and structure, with excavation area highlighted. View to the north, facing the Hugiin Gol. Image by W.Taylor and N. Case.

Figure 3

Zooarchaeology by Mass Spectrometry (ZooMS) taxonomic identifications overlain on a schematic of a cladogram for bone remains from Bagsagiin Bulan. Specimens with conclusive markers are shown in color (blue, red, green, purple), while those with missing markers are shown in tan according to the greatest possible level of taxonomic specificity. Circles have been scaled to size according to quantity of identified specimens in each category and inset numbers refer to the number of specimens identified. Dotted circle in top left shows specimens with insufficient collagen for analysis.

Figure 4

Species composition of calcined bone fragments from Biluut, Peat Valley 1 (non-faunal data originally reported in Fitzhugh and Kortum 2012), based on new morphological identifications by T. Tuvshinjargal. The assemblage was unanalyzed in the initial report, and ZooMS/aDNA efforts yielded no collagen or DNA for biomolecular identifications.

Figure 5

(A) Aerial view of structure at Tsagaan Asga with excavated area highlighted, and (B) planview of excavated structure showing position of artifacts. Image by W.Taylor and N. Case.

Figure 6

Zooarchaeology by Mass Spectrometry (ZooMS) taxonomic identifications overlain on a schematic of a cladogram for bone remains from Tsagaan Asga. Specimens with conclusive markers are shown in color (blue, red, and black), while those with missing markers are shown in tan according to the greatest possible level of taxonomic specificity. Circles have been scaled to size according to quantity of identified specimens in each category. The unknown rodent was identified based on morphology to the subfamily of Arvicolinae (voles, lemmings, and muskrats). Dotted circle in top left shows specimens with insufficient collagen for analysis.

Figure 7

Frequency (%NISP) of horse remains at Bronze Age and Early Iron Age Central Asian archaeological assemblages, as compared to the earliest direct evidence for chariot use at Sintashta (ca. 2000 BCE) and horse riding in the DSK culture (ca. 1200 BCE). Green color indicates those with equivocal domestication status^–.

Figure 8

(A) Early and Middle Bronze Age archaeological sites (Afanasievo Chemurchek, and Munkhkhairkhan, ca. 3000–1500 BCE) and (B) Post-1200 BCE Late Bronze and Early Iron Age archaeological sites (DSKC and Slab Burial cultures) plotted against digital elevation model of Mongolia, with low elevation of 500 m shown in red, and high elevation (2500 + m) shown in blue. Data from Eregzen (2016). Icons in upper right show hypothesized horse transport technology in use during each period.

Figure 9

Idealized model for the impact of horse transport on pastoral lifeways in Mongolia during the late Bronze Age. Blue denotes hunting and gathering subsistence, red denotes low-mobility pastoral subsistence, and green denotes horse-based, high mobility pastoral subsistence. The top figure shows the clustering of pastoral lifeways in montane regions of western and central Mongolia during early and middle Bronze Age, and the bottom figure shows the widespread geographic distribution of Late Bronze Age and Early Iron Age pastoral sites, concurrent with both early evidence for horseback riding as well as increased economic exploitation of horses.

Figure 10

Timeline for horse domestication and key events related to early pastoralism in Mongolia and surrounding regions.