Abstract
Secure environmental contexts are crucial for hominin interpretation and comparison. The discovery of a Denisovan individual and associated fauna at Tam Ngu Hao 2 (Cobra) Cave, Laos, dating back to 164–131 ka, allows for environmental comparisons between this (sub)tropical site and the Palearctic Denisovan sites of Denisova Cave (Russia) and Baishiya Karst Cave (China). Denisovans from northern latitudes foraged in a mix of forested and open landscapes, including tundra and steppe. Using stable isotope values from the Cobra Cave assemblage, we demonstrate that, despite the presence of nearby canopy forests, the Denisovan individual from Cobra Cave primarily consumed plants and/or animals from open forests and savannah. Using faunal evidence and proxy indicators of climates, results herein highlight a local expansion of rainforest at ~ 130 ka, raising questions about how Denisovans responded to this local climate change. Comparing the diet and habitat of the archaic hominin from Cobra Cave with those of early Homo sapiens from Tam Pà Ling Cave (46–43 ka), Laos, it appears that only our species was able to exploit rainforest resources.
Similar content being viewed by others
Introduction
The Denisovans were initially identified through their genome, which was extracted from a handful of finger bones, teeth, and sedimentary DNA from Denisova Cave in southern Siberia, Russia1,2,3,4,5,6,7. Additional evidence has since emerged from the analysis of ancient proteins and the morphology of a partial mandible8, as well as sedimentary DNA from the Baishiya Karst Cave in Xiahe, China9 (Fig. 1). Other populations, such as the Xujiayao hominins from northern China (identified through a set of teeth10), the Penghu 1 individual from Taiwan (identified through a mandible11) and the cranium from Harbin, China, known as Homo longi (proposed as a putative new species12) have been suggested as potential Denisovans. Recently, a molar of a young female Denisovan was discovered in the Tam Ngu Hao 2 Cave (Cobra Cave), Laos, with an age range of 164–131 ka13, with morphological similarity to that of molars of the Baishiya Karst Cave mandible (Supplementary Fig. S1).
[Source: (a) Base map from https://capcarto.fr, (b) Authors].
(a) Location of key sites of Denisovans in Russia (Denisova Cave), China (Baishiya Karst Cave), and Laos (Tam Ngu Hao (Cobra) Cave). The yellow line represents the biogeographic limit between the Palearctic and Oriental realms. (b) Location of the southeast Asian mammalian assemblages used in the study: Tam Pà Ling, Tam Hang South, Nam Lot, Tam Hay Marklot, and Tam Ngu Hao (Cobra) in northeastern Laos; Coc Muoi and Duoi U’Oi in Vietnam. Only three sites produced hominin remains: Tam Ngu Hao (Cobra) Cave (Denisovan, 164–131 ka), Tam Pà Ling (H. sapiens, 86–43 ka), and Duoi U’Oi (Homo sp., 70–60 ka)
Palaeogenetic evidence suggests that the Denisovans have the physiological capacity to live in high-altitude hypoxic environments14. This feature likely resulted from their adaptation to the extreme conditions of the Tibetan plateau from around 160,000 years ago (160 ka)8 up to approximately 60 ka9. Baishiya Karst Cave, where Denisovan remains were discovered, is located at an altitude of 3280 m above sea level (asl), which is much higher than Denisova Cave in the foothills of the Altai Mountains (700 m asl) or that of Cobra Cave in the karstic mountains of northeastern Laos (1116 m asl). Furthermore, the high rate of introgression of Denisovan DNA in the genome of modern populations from New Guinea, east Indonesia, the Philippines (Mananwa population), and Australia strongly suggests that Denisovans were present in southern and/or southeast Asia15. The locations and age estimates of the sites where Denisovans were unearthed therefore indicate that from around 200 ka to 50 ka, they adapted to a variety of environments ranging from temperate habitats in the Altai5 to tropical habitats in southeast Asia13.
Ancient DNA analyses indicate that Neandertals, Denisovans, and Homo sapiens interbred several times in the Middle to Late Pleistocene, throughout their evolution in Eurasia6,16,17,18,19,20,21,22. A major gene flow event between Neandertals and early H. sapiens23 likely occurred in the Levant > 170 ka24, whereas the Neandertal contribution to modern-day humans is constrained by the timing of the dispersal of our species outside Africa after 60–50 ka25. Similarly, Denisovans contributed up to ~ 4–6% to the genomes of ancestors of present-day Melanesian and Australasian populations15 and ~ 0.2% to the genomes of ancestors of mainland Asians and Native Americans17. The timing of interbreeding events between Denisovans and H. sapiens in Asia remains unclear. Palaeogenomic evidence indicates that interbreeding occurred over 50,000 years ago in the northern areas of the Denisovan distribution7. However, another study suggests that interbreeding may have occurred much more recently in the southern regions22.
In tropical latitudes, the scarcity of hominin fossils, as well as hot and humid conditions implying difficulties in retrieving DNA sequences from both fossils and sediments, poses a challenge in addressing the population history of Denisovans and H. sapiens26. Our discovery at Cobra Cave13 provides a new opportunity to explore the interaction between Denisovans and low-latitude tropical environments. While the dearth of archaeological material limits direct assessment into potential adaptations to tropical rainforests, geochemical proxies such as stable isotopes can offer a valuable source of data. Based on the principle that animal tissues metabolise and incorporate or reflect the isotopic composition of their diet, carbon isotope analysis from tooth enamel can provide crucial information on palaeodiets and, therefore, on palaeoenvironments27,28. Furthermore, because of the broad ecological range of ruminant ungulate taxa (i.e., browser, mixed-feeder, or grazer) and their unpredictable responses to climate changes in southeast Asia29, only such proxy records can help reveal the structure of past ecosystems and, therefore, their level of heterogeneity. This holds particular significance because, during the Pleistocene, the environments of southeast Asia consisted of a diverse range of biomes that underwent continuous fluctuations ranging from closed-canopy forests to grasslands30.
Here, we present the first analysis of the carbon (δ13Capatite) and oxygen (δ18O) isotope composition of a broad spectrum of mammalian taxa (Artiodactyla, Perissodactyla, Proboscidea, Carnivora, Primates, and Rodentia), as well as the Denisovan individual from Cobra Cave (164–131 ka13), to describe its diet and habitat. The Denisovan tooth (TNH2-1) is a developing first or, more likely, second lower molar of a juvenile female individual who died between 3.5 and 8.5 years13 (Supplementary Fig. S2). The age at which Denisovans were weaned is not known, but evidence from their closest relatives, the Neandertals, suggests an early weaning process similar to that of extant humans31,32. Therefore, considering the fact that the isotope values of TNH2-1 were obtained from a sample at the bottom of the crown, the young girl from Cobra Cave likely consumed the same food as that of adults of the group.
The δ13C values of bioapatite are used to investigate palaeodiets based on values associated with C3-plants (trees, bushes, shrubs, and grasses) versus C4-plants (grasses, sedges), and their respective environments. The δ13Ccarbon source values in the diet of animals were then calculated from δ13Capatite (“Material and methods”33) to more accurately explore the proportions of these isotopically-distinct carbon sources over the period studied, including sub-partitioning biomes such as closed-canopy forests34. The δ18O values are used to contribute palaeoecological information related to variation in abiotic conditions (latitude, climate, temperature, moisture content, amount, and isotopic composition of precipitation35, and references therein). Thus, these directly complement δ13C values and provide additional insights into past conditions.
To investigate the Denisovans’ environments in temperate versus tropical regions in Marine Isotopic Stage [MIS] 6 (191–130 ka36), we compared habitats inferred from fauna and isotopic records from Cobra Cave with those inferred from fauna and pollen evidence from Denisova Cave over the same period (Main Chamber, layers 19–17, 151 ± 17–128 ± 13 ka5). We evaluated the habitats for the other Asian hominin Homo erectus on Java until ~ 120 ka37 and questioned to what extent the ecological niches of Denisovans and H. erectus were comparable. Some works over the last two decades refined the contours of the ecological niche of Indonesian H. erectus, which is clearly that of open habitats in lowland areas38,39,40,41,42,43. Furthermore, in an attempt to compare habitats and diets between the Denisovans from Cobra Cave and the earliest H. sapiens in the area, we used available data from Tam Pà Ling (TPL) Cave, the two sites being located ~ 300 m apart (Fig. 1). Isotopic data from TPL include the H. sapiens individual TPL-1 (the upper left molar of the partial skull of a young mature adult, dated to 46–43 ka44,45) (Supplementary Fig. S2) and a handful of herbivores’ teeth (Artiodactyla and Perissodactyla) recovered in the sedimentary section46 prior to 33 ka, i.e., before the settlement of the Last Glacial Maximum conditions47. Previous research has documented that it foraged in a highly forested habitat46,47.
In a second step, using carbon and oxygen faunal records from Cobra Cave along with a series of five Middle to Late Pleistocene faunas of comparable composition (Artiodactyla, Perissodactyla, Proboscidea, Carnivora, Primates, and Rodentia) from northern Vietnam and Laos29,48,49 (Fig. 1), we identified large-scale climatic shifts that have transformed the palaeoenvironments locally. Thus, despite a discontinuous and patchy record (Supplementary Table S1), the mammalian faunas from Cobra Cave (164–131 ka), Coc Muoi (148–117 ka), Tam Hang South (94–60 ka), Nam Lot I (86–72 ka), Duoi U’Oi (70–60 ka) and Tam Hay Marklot (38.4–13.5 ka) may nevertheless provide key insights into major changes of ecosystems—both functional (species diversity and abundance) and structural (distribution of ecological niches)—over the period and, therefore, into the adaptive capacity of hominins to novel environments. Overall, at the scale of continental and insular southeast Asia, the Middle Pleistocene is seen as a period of open habitats that favoured the settlement and expansion of archaic hominins30, whereas the Late Pleistocene is marked by the expansion of rainforests at the time of H. sapiens’ dispersal events, thus revealing potentially two different adaptive strategies. But one may question what environment prevailed in northern Indochina’s latitudes. Given this background, the present study aims to describe the environmental contexts into which Denisovans and H. sapiens inhabited successively in the studied area, considering the new extended TPL chronology with evidence of earliest H. sapiens at least 68 ka ago50.
Results
The Cobra Cave Denisovan and associated fauna: The δ13Ccarbon source and δ18Oapatite values of every specimen are compiled in Supplementary Annex S1. As illustrated in Fig. 2, the δ13Ccarbon source values for Cobra Cave range from − 31.3 to − 11.9‰ (average δ13Ccarbon source = − 25.18 ± 4.6‰ (1 σ), n = 54). The δ18Oapatite values for the site range from − 10.5 to − 2.6‰ (average δ18Oapatite = − 6.7 ± 2.0‰ (1 σ), n = 54). The δ13Ccarbon source and δ18Oapatite values of the Denisovan individual TNH2-1 are − 16.3‰ and − 7.0‰, respectively.
(Source: Authors).
(a) Distribution of δ13Ccarbon source values of animal and Denisovan specimens from Cobra Cave by taxon (see “Material and methods” for range of values associated with biomes). (b) Histogram distribution of the relative frequency (%) in δ13Ccarbon source values for all taxa. Each bin represents a spacing of 1%. The dashed red line refers to the mean value
Post-hoc Dunn’s test pair-wise comparisons between the δ13Ccarbon source values of Cobra Cave (164–131 ka) and those of the other sites (Coc Muoi, Tam Hang South, Nam Lot, Duoi U’Oi and Tam Hay Marklot) demonstrate significant differences only with Coc Muoi (148–117 ka) and Duoi U’Oi (70–60 ka) (Supplementary Tables S6, S7 and Annexes S3, S4). The δ18Oapatite values of Cobra Cave show no significant differences from those of the other faunas.
Discussion
High latitude ecosystems, like those of Denisova Cave (151–128 ka; Main Chamber, Layers 19–175, and Baishiya Karst Cave (~ 160 ka8; ~ 100 ka9), and the medium latitude ecosystem of Cobra Cave (164–131 ka13) harboured diverse herbivore communities. They encompass megafaunas adapted to very different environmental conditions, Palearctic versus Oriental51. This biogeographic division is reflected in the little commonality in taxonomic composition at the genus level between the cold-adapted Mammuthus-Coelodonta and the warm-adapted Stegodon-Ailuropoda faunal units (Fig. 1a and Supplementary Tables S8, S9).
What do we know about the biodiversity of Denisovans’ ecosystems? Figure 2 shows that at the latitude of Cobra Cave, the majority of the mammalian specimens (62%) exhibited δ13Ccarbon source values not associated with closed-canopy forest (i.e., > − 27.2‰), thus rather reflecting intermediate and open woodland to savannah environments (“Material and methods”, Supplementary Table S5). Large ruminants (i.e., large bovines (Bos) and sambar deer (Rusa)) are those that predominantly foraged in this open landscape. We also note a gain in biodiversity among medium-sized ruminants due to the increased number of ecological niches29. Caprines such as gorals (Naemorhedus) and other medium-sized deer grazed on grasses in these open areas. In this ecosystem, the C3 canopy forests contained most of the other ground-dwelling herbivores (specimens having δ13Ccarbon source values < − 27.2‰), including megaherbivores > 1000 kg52, tapirs (Tapirus), rhinoceroses (Rhinoceros and Dicerorhinus), and stegodon (Stegodon), while more open forests supported primates, macaques (Macaca) and orangutans (Pongo), wild boars (Sus), panda (Ailuropoda) and porcupines (Hystrix). At northern Indochina's latitudes, the Leizhou Peninsula pollen record53 reveals two major phases during MIS 6, with the latest half characterized by a relatively high percentage of Poaceae, which matches with the presence of savannah at Cobra Cave at the same period.
At the latitude of the Altai Mountains, environmental indicators also show a mosaic of biomes. The palynological evidence from the lower part of Layer 19 at Denisova Cave (starting ~ 168 ka5) suggests an association between meadows and steppe environments with forests composed of temperate elements (birch, pine, with a mixture of alder, linden, and elm) under relatively warm climatic conditions in the context of the Palearctic zone5. In this environment type, open landscapes such as tundra and steppe contained most of the megaherbivore biomass54 (contra tropical environment). Sedimentary DNA at Denisova indicates that the ‘mammoth’ steppe was occupied by non-ruminant grazers preferentially eating grasses and sedges (e.g., woolly rhinoceros (Coelodonta), woolly mammoths (Mammuthus)) with steppe bisons (Bison), and a large spectrum of gazelles (Procapra, Saiga), ibex (Capra) and argali (Ovis) adapted to grassy steppe and particularly abundant at that time (Layer 1924) (Supplementary Tables S9, S10). The occurrence of red deer (Cervus elaphus) and horse (Equus sp.) at the site also supports the presence of shrubs and trees, as indicated by the isotopic investigation conducted on these species from the Palearctic zone55,56,57.
The faunas from both the Tibetan Plateau and Altai Mountain share a common Palearctic origin (Fig. 1a). Today, the community of large herbivores adapted to live in high altitudes in excess of 3500 m consist primarily of medium-sized cervids, red deer (Cervus) and Siberian roe (Capreolus), medium-sized bovids (e.g., gazelle (Procapra), argali (Ovis), goral (Naemorhedus) and serow (Capricornis), and only one large bovid, the yak (Poephagus/Bos) (Supplementary Tables S8, S9). At the precise location of the Baishiya Karst Cave, the foothills of the mountain are dominated by alpine meadows composed of a variety of grasses, sedges, and herbs, whereas some wooded areas are present in riparian environments and along the mountain slopes9. The analysis of mtDNA from Late Pleistocene sediments of the Baishiya Karst Cave revealed that ~ 100 ka Denisovans lived within a richer herbivore community than today, dominated by rhinocerotids and equids that are now absent at high altitudes and with large bovids and cervids9 (Supplementary Tables S9, S10, S11).
The known Denisovan populations occupying either temperate or tropical environments could therefore predate on a wide choice of herbivores. At the Denisova and Baishiya sites, herbivore remains have been found associated with abundant Palaeolithic stone artefacts8,9, and with direct evidence of human activities suggested by animal bones with cut-marks8. In the absence of comparable archaeological evidence, the δ13Ccarbon source value of the Denisovan individual from Cobra Cave can be used to assess its diet. It reflects the consumption of plants and/or animals from open landscapes (δ13Ccarbon source = − 16.3‰; Fig. 3). Around Cobra Cave, open landscapes favoured the range expansion of mixed-feeders and grazers (bovines, caprines, deer), which exposed a diversified large game to hominin predation. This might have resulted in Denisovans foraging preferentially in open areas at the fringes of nearby forests, although a dense canopy forest was present in the environment.
(Source: Authors).
Distribution of δ13Ccarbon source values of ungulate and hominin specimens from two sites (see “Material and methods” for range of values associated with biomes): Tam Ngu Hao (Cobra) Cave (Denisovan TNH2-1, 164–131 ka) (Supplementary Annex S1) and Tam Pà Ling (Homo sapiens TPL-1, 46–43 ka)46 (Supplementary Annex S2)
When comparing the carbon isotope value of the Denisovan individual from Cobra Cave (164–131 ka) with that already published of the H. sapiens individual (46–43 ka46) from nearby TPL site44, the results display notable differences. The TPL individual data (δ13Ccarbon source = − 26.4‰; Fig. 3) reflect a food procurement strategy that preferentially selects a C3 forest biome, possibly from a dense canopy forest. Various caprines (goral and serow), rhinoceroses, and large bovids46 are associated to this biome (Fig. 3, “Material and methods”, Supplementary Annex S2). We also infer patches of open vegetation with C4 plants based on two caprine teeth, further supported by the isotopic composition of Camaena massiei shells, a terrestrial gastropod, over the period 70–33 ka47 (Supplementary Text and Fig. S3). Therefore, conversely to the Denisovan individual from Cobra Cave, the TPL H. sapiens consumed food from a more forested area.
Despite the scarcity of archaeological material and poor organic matter preservation in tropical latitudes, we now have evidence of rainforests occupation by H. sapiens in Asia by ~ 70 ka45,50,58,59 and increasing evidence underlying the reliance on diverse settings ~ 45 ka46,60,61,62. δ13C values of H. sapiens from all these sites highlight the capacity of our species to adopt various behaviours in similar environments where both C3 forest and C4 open biomes are present in the vicinity. The data suggest specialization, such as hunting of arboreal species63, use of coastal resources64, opportunistic use of resources from mosaic and/or open forest edge environments65. In relation to Homo sapiens from TPL ~ 46 ka, reliance on deep forest resources would suggest the exploitation and processing of plants60,66,67,68 and the use of diverse hunting strategies such as traps, microlithics, and other tools made of organic material60,61,63,69,70. In contrast, the Denisovan from Cobra Cave, as well as the other archaic hominin Homo erectus from Java41,43, exhibit δ13C values that are solely indicative of a dietary reliance on open environments. Furthermore, while Denisovans adapted to diverse climates and habitats (i.e., from high latitudes at Denisova Cave and Baishiya Cave to medium latitudes at Cobra Cave)6,8,9,13,22,71, their reliance on grassland and woodland resources seemingly persisted.
The evolutionary path of H. sapiens since ~ 300 ka is marked by both a structural and genomic reorganization of the brain and a moderate increase in its size, in comparison with other contemporaneous large-brained hominins such as Neandertals and Denisovans72. Meyer et al.16 identified derived genomic features in H. sapiens that are not present in Denisovans and showed that some substitutions on human genes resulted in critical changes in brain function or nervous system development, notably greater synaptic plasticity in our species. That seems in accordance with southeast Asian palaeoenvironmental data, that suggests that H. sapiens expansions involved reliance on biome-specific specializations (versus Denisovans or H. erectus), thanks to a unique ecological plasticity within the hominin clade73.
In a previous analysis29, we revealed that late Middle to Late Pleistocene ecosystems were locally dynamic and diverse, based on notable changes in the distribution of δ13Ccarbon source values (vegetation cover) in a series of faunas geographically close (Fig. 1b). This is confirmed here with the new data of Cobra Cave (164–131 ka) as illustrated in Fig. 4 by using violin plots, and further supported by statistical data with significant differences with Coc Muoi (148–117 ka) and Duoi U’Oi (70–60 ka) (Supplementary Table S6). Overall, δ18Oapatite values (rainfall regime) show a general trend towards higher values from Cobra Cave (164–131 ka) to Nam Lot (86–72 ka), before a change likely related to increased aridity over the Last Glacial period from ~ 70 ka, as suggested by the Duoi U’Oi (70–60 ka) and Tam Hay Marklot records (38.4–13.5 ka) (but not statistically significant) (Supplementary Table S7).
(Source: Authors).
Distribution of δ13Ccarbon source and δ18O values for all taxa in the southeast Asian faunas (Supplementary Annexes S1, S3, S4) from Tam Hay Marklot48, Duoi U'Oi29, Nam Lot49, Tam Hang South29, Coc Muoi29, and Tam Ngu Hao (Cobra) Cave (this paper). All sites were found within a relatively narrow latitudinal belt between 23° and 20° running through the northern regions of Laos and Vietnam (Fig. 1). The sites follow a chronological order from left to right highlighting environmental changes through vegetation cover variation (δ13Ccarbon source) and likely rainfall regimes (δ18O). The outline of the violin plots represents kernel probability density, where the width shows the proportion of the data found there. The boxes from the box and whisker plots inside the violin represent the 25th–75th percentiles, with the median as a bold horizontal line
By enabling the reconstruction of past environments, faunal isotopic data can, therefore, be useful tools to identify external drivers of hominin evolution, even if correlating chronologies between faunas constrained by luminescence dating and better chronologically constrained palaeoclimatic signals from the speleothem records reveals challenging74 (“Material and methods”). With regards to the region studied, we used the δ18O curves from speleothems of the nearest Chinese reference sites as indicators of the intensity of East Asian summer monsoon75 (Fig. 5a) and histograms of the distribution of δ13Ccarbon source values associated with the different biomes inferred from each fauna (Supplementary Table S5, Fig. 5b).
[Source: (a) Modified from Wang et al.75; (b–d) Authors].
(a) Chinese caves δ18O records (left ordinate axis, ‰ VPDB International Standard Vienna Pee Dee Belemnite) showing millennial-scale climate shifts related to changes in East Asian summer monsoon intensity for the last 224 ka (black line) published in 2008 by Wang et al.75. Numbers refer to the marine isotope stages and substages. The right ordinate axis corresponds to the Northern Hemisphere summer insolation (red line) (65°N, W m−2). (b) Histograms of faunas showing the distribution of δ13Ccarbon source values of specimens associated with each biome (Supplementary Table S5). (c) Age ranges of faunas: Tam Ngu Hao (Cobra) Cave (MIS 6.3, 164–131 ka) analyzed herein, Coc Muoi (MIS 6.2, 148–117 ka), Tam Hang South (MIS 5, 94–60 ka), Nam Lot (MIS 5, 86–72 ka), Duoi U’Oi (MIS 4, 70–60 ka) and Tam Hay Marklot (MIS 3–2, 38.4–13.5 ka). (d) Occurrence of hominins in the region, Denisovan (Cobra Cave, 164–131 ka), Homo sp. (Duoi U’Oi, 70–60 ka) and H. sapiens (Tam Pà ling, 86–43 ka)
Our results highlight two repeated episodes of rainforest expansion as climates fluctuated (Fig. 5b). Each episode was a singular event that led to novel plant communities and structures (i.e., the density of canopy, shrub, and floor strata). The first one, which occurred at the time of the dispersal of H. sapiens has been documented previously by Bacon et al.29, based on changes in the distribution of biomes between Nam Lot (MIS 5, 86–72 ka) and the Homo sp.-bearing site of Duoi U’Oi (MIS 4, 70–60 ka) (Fig. 5b,c). Overall, MIS 5 was a period of relatively strong monsoons and high precipitation69, associated with a mosaic of biomes. It was followed by a rapid decrease in monsoon strength at the onset of MIS 4 (Fig. 5a). At that time, a rapid forest transformation resulted in an increase of temperate forest elements, most notably conifers, in a relatively cooler climate53. These changes were also accompanied by a novel type of shrub, fern, and herb strata53 that likely rendered the forests easier for hunter-gatherers to navigate and forage29. The presence of H. sapiens in the region has been confirmed recently by the extended chronology of Tam Pà Ling at least 68 ka50. Further evidence for the ability of H. sapiens to occupy a broad spectrum of rainforests comes from Lida Ajer, Sumatra (71–68 ka)59. At this latitude, humans exploited a landscape dominated by closed-canopy forests (based on the isotopic faunal record, Supplementary Fig. S4 and Annex S7) that were not so different from the equatorial forests of Sumatra today.
Furthermore, the comparison in the distribution of biomes between Duoi U’Oi (MIS 4, 70–60 ka) and Tam Hay Marklot (MIS 3–2, 38.4–13.5 ka) (Fig. 6) shows how dramatic were the environmental changes over MIS 4–247, a period that accompanied the dispersal of hunter-gatherers through the savannah corridor65. Duoi U’Oi witnesses low biodiversity and this ecosystem dominated by canopy forests favored the abundance of the sambar deer (61.3%) and muntjacs (31.8%) (Supplementary Table S12). Tam Hay Marklot shows that the diversity of herbivores increased as landscapes opened and biomes diversified, owing to the increased number of ecological niches. This gain of biodiversity, most likely through dispersal events, concerns various deer (17.2%) known to live in large herds in open areas and caprines (5.3%). Their relative abundance also points to the extent of grassland and its important carrying capacity, as also observed at the same latitude in Tham Lod Rockshelter (34–12 ka)65, a unique condition not observed in our MIS 6–5 faunal records (Supplementary Fig. S5).
(Source: Silhouette image from public domain https://www.phylopic.org; charts, Authors).
Compared biodiversity between Duoi U’Oi (MIS 4, 70–60 ka) and Tam Hay Marklot (MIS 3–2, 38.4–13.5 ka) faunas. We combined in a same scheme the 3 descriptors of biodiversity, species diversity among identified ruminant taxa (i.e., cervids and bovids being able to browse and/or graze, Supplementary Tables S2, S3), species abundance (by using percentages of the minimum number of individuals (MNI), Supplementary Tables S12), and distribution of ecological niches (biome associated with each taxon)
Another episode is highlighted here for the first time by the changes in the distribution of biomes between Cobra (164–131 ka) and Coc Muoi (148–117 ka) (Fig. 5b), a period that likely impacted archaic hominins locally (Fig. 5d). Cobra Cave clearly provides evidence for the presence of savannah and woodland savannah biomes along with fragmented rainforests, and its age range coincides with the MIS 6.3 period of relatively strong monsoons (Fig. 5a). This environment was established ~ 160 ka when grassy areas, notably composed of Poaceae and Cyperaceae, replaced Artemisia steppe76. The shift in vegetation cover documented by the expansion of the canopy forest biome at Coc Muoi (148–117 ka), seems to coincide with the weaker monsoon interval 135.5–129 ka (MIS 6.2 in the Hulu/Sanbao records75). This shift can be associated with the abrupt reappearance of forest montane elements in lowland zones with cool and relatively wet conditions53.
Considering the subsistence strategy of the Denisovan individual from Cobra Cave, could this cold event have been a driving factor in archaic hominin evolution? If data from a single individual cannot fully reflect the diversity of the subsistence of the entire group locally, the fact that it relied on mixed to open landscapes for food resources supports the idea that these biomes might have played a significant role in the mobility and settlement of this archaic hominin in a tropical ecosystem77. This raises the question of how they adapted to climate changes ~ 130 ka that resulted in the expansion of rainforests. Foragers faced different challenges in forest biomes based on their behavioural flexibility, and archaic hominins may have experienced a population contraction in response to the emergence of dense rainforests78.
It is tempting to draw a parallel with the history of Homo erectus. On Java, the pollen record from Sangiran around 800 ka documents the settlement of a landscape dominated by grasslands while rainforests underwent a severe fragmentation in high altitude areas, river streams, and swamps39. Although H. erectus occupied a mosaic of habitats with grassland on Java from ~ 1.2 Ma79, this vegetational change ~ 800 ka is associated with a greater abundance of hominin remains along with the spread of Acheulean-like industries40, a settlement of populations likely favoured by this landscape. For subsequent periods, the carbon isotope analysis of faunas from H. erectus-bearing sites Trinil H.K. (540–430 ka80) and Ngandong (117–108 ka37) suggests a mixed woodland-savannah environment41,42 (Supplementary Fig. S4 and Annexes S5, S6). At these lower latitudes, a major biogeographical event led to the dispersal of a rainforest-adapted fauna resulting from a lowstand sea level of up to – 120 m81. This is documented by the replacement of the Ngandong archaic fauna (117–108 ka) by the fully modern Punung fauna (128–118 ka82,83), but the synchronicity with the climatic episode that led to rainforest expansion ~ 130 ka farther north remains to be demonstrated. Ngandong also witnessed the last occurrence of H. erectus37, raising the question of a population range contraction in more suitable areas on Sundaland shortly before its extinction.
Conclusion
The present study highlights that ecosystems occupied by Denisovans, whether temperate or tropical, shared mixed vegetation covers with a significant part of open landscapes. At Cobra Cave, the presence of open woodlands and savannahs promoted a high diversity of herbivores, with a notable expansion in the range of cervids and bovids through increased niche partitioning between taxa. Despite the presence of closed forested areas, Denisovans likely preferentially targeted large game visible in open areas or at forest edges. In contrast, our results suggest that early Homo sapiens of the same region had a different ecological niche that relied on rainforests at least ~ 70 ka, most likely due to the development of new behavioral skills.
Hence, our findings are relevant to the debate regarding the potential role of rainforests as a primary driver of hominin evolution in southeast Asia and raise the question of whether the expansion of rainforests acted as a regional barrier to Denisovans. Recent genomic analyses reveal multiple Denisovan groups that were geographically isolated from each other during the Pleistocene in southeast Asia. The repeated episodes of rainforest expansion might have played a key role in this population contraction range that shaped the hominin evolution.
Material and methods
Geographical and chronological context of Tam Ngu Hao 2 (Cobra) Cave
The karstic cave is located in northeastern Laos (Supplementary Fig. S1). The dates for breccia deposits in which the human tooth and faunal remains were found range between 164 and 131 ka13 based on Bayesian modelling of luminescence dating of sediments, uranium-series dating of flowstones, and coupled U-series and electron spin resonance dating of three bovid teeth. This time interval corresponds to the second half of Marine Isotopic Stage [MIS] 6 (191–130 ka36).
Composition of the Cobra Cave assemblage
It is dominated by isolated teeth of large mammals (N = 186), including the Denisovan tooth (Supplementary Tables S2, S3). Due to the deposition in a context of high energy13, the selective conservation and poor preservation of most specimens constrained their identification at the genus or family level. Artiodactyla: Sus scrofa (n = 38), Bos sp. (n = 35), Naemorhedus sp. (n = 10), Muntiacus sp. (n = 24), medium-sized Cervidae (n = 6), large-sized Cervidae (n = 16); Perissodactyla: Rhinocerotina indet. (n = 12), Rhinoceros sp. (n = 3), Rhinoceros sondaicus (n = 2), Dicerorhinus sp. (n = 1), Tapirus sp. (n = 1); Proboscidea: Stegodon sp. (fragments of enamel); Carnivora: small-sized carnivora (n = 2), small-sized Felidae (n = 1), Paradoxurus sp. (n = 1), Ursus thibetanus (n = 3), Ailuropoda sp. (n = 2); Primates: Macaca sp. (M. cf. nemestrina) (n = 10), Pongo sp. (n = 1), Homininae (Denisovan) (n = 1); Rodentia: Hystrix sp. (n = 18).
New sub-sample for isotopic analyses
A sub-sample of 54 specimens was selected within the Cobra Cave assemblage: Sus scrofa (n = 8), Large-sized Bovidae (n = 8), medium-sized Caprinae (Naemorhedus sp.) (n = 10), Muntiacus sp. (n = 3), medium-sized Cervidae (n = 2), large-sized Cervidae (n = 7), Rhinocerotidae (n = 3), Rhinoceros sp. (n = 1), Tapirus sp. (n = 1), Stegodon sp. (n = 1), small-sized carnivora (n = 2), small-sized Felidae (n = 1), Ailuropoda sp. (n = 2), Macaca sp. (n = 4), Pongo sp. (n = 1), Homininae (Denisovan) (n = 1), Hystrix sp. (n = 5) (Supplementary Tables S4, S5).
Stable carbon and oxygen isotope data
In terrestrial food webs, analyses of stable carbon isotopes of bioapatite (δ13Capatite) are an effective way to assess the relative proportion in a consumer's diet of ingested carbon derived from food webs’ primary sources (plants) using either C3 or C4 photosynthetic pathways84. In tropical and subtropical regions specifically, C4 plants (grasses, sedges) are found in open environments and exhibit high δ13C values, whereas forests and woodland habitats are associated with C3 plants (trees, bushes, shrubs, and grasses) and low δ13C values85. Densely forested conditions induce even lower δ13C values in plants due to a "canopy effect"86, allowing the identification of additional ecological partitioning in C3 forested environments. Using δ13Capatite values and diet-enamel spacing, the average initial δ13C values of the carbon source in the animal's diet (herein labeled as "δ13Ccarbon source") were estimated to allow more accurate environmental reconstructions. To account for the atmospheric CO2 shift due to fossil fuel burning, the δ13C ranges of values for C3 and C4 plants were adjusted accordingly (~ 1.3‰87). In Figs. 2, 3 and 5b, the δ13Ccarbon source values associated with closed-canopy forests are < − 27.2‰34, intermediate rainforests and woodland biomes > − 27.2‰ and < − 21.3‰88, and savannah-like environments > − 15.3‰84. Values between > − 21.3‰ and < − 15.3‰ are associated with the consumption of both C3 and C4 resources and do not correspond to any specific ecological environment.
Stable oxygen isotopes of bioapatite (δ18O values) vary according to the oxygen isotopic composition of drinking water and chemically-bound water in diet (i.e., water found in plants) and are controlled by various environmental and geographic conditions such as latitude, climate, temperature, moisture content, amount and isotopic composition of precipitation35. At low latitudes, such as in the studied area, variations of δ18O rainfalls are primarily indicative of the amount of precipitation. Just as with δ13C values, a "canopy effect" can also be characterized by low δ18O values on the forest floor86.
Fossil teeth of the TNH2-1 individual and sympatric mammal specimens (n = 54) from Cobra Cave were sampled and analyzed for the present study (Supplementary Annexes S1–S3). Using a handheld dental drill equipped with a diamond-tipped burr, the enamel surface of each specimen was cleaned mechanically, and powder samples were subsequently taken along the full height of the crown. Powdered enamel teeth samples were subsequently pretreated to remove exogenous carbonate. Samples were thus soaked in 1 ml of CH3COOH (0.1 M) for 4 h at room temperature, rinsed several times in distilled water, and then dried overnight at 65 °C. Measurements of stable carbon and oxygen isotopic ratios of the carbonate phase of enamel were performed at the Service de Spectrométrie de Masse Isotopique du Muséum (SSMIM) in Paris using a Thermo Scientific Delta V Advantage isotopic mass spectrometer along with a Thermo Scientific Kiel IV Carbonate Device chemical preparer. Isotopic abundances are presented in delta (δ) notation and expressed as deviation per mil (‰), as follow: δ13C = (13C/12Csample/13C/12Cstandard − 1) × 1000 and δ18O = (18O/16Osample/18O/16Ostandard − 1) × 1000.
During every mass spectrometer run, we analyzed an internal laboratory standard (Marble LM, accepted δ13C = + 2.13‰ and δ18O = − 1.83‰) normalized to the International Atomic Energy Agency reference material NBS 18 and NBS 19. LM was used for tooth sample correction (one point-correction) and for controlling the precision (1σ) of the mass spectrometer (δ13C = 0.035‰ and δ18O = 0.051‰; n = 24). We usually analyzed each tooth sample one or two times. Four samples were analyzed three times to test for intra-individual heterogeneity and analytical reproducibility of enamel analysis. Maximum standard deviations were 0.527‰ and 0.342‰ for δ13Cenamel and δ18O analysis, respectively.
Palaeoenvironmental reconstruction
For the purpose of this study—the comparison between the landscapes inhabited by Denisovans and H. sapiens locally—we used published isotopic data from Tam Pà Ling, including the TPL-1 individual (46–43 ka) and a handful of herbivore teeth (Artiodactyla and Perissodactyla) recovered in the sedimentary section between 70 and 33 ka46 (Supplementary Annex S2). This range has been selected to document the environment of early H. sapiens locally before the major changes that began ~ 33 ka and led to the settlement of the Last Glacial Maximum conditions47 (Supplementary Text and Fig. S3).
In this study, original carbon (δ13Ccarbon source) and oxygen (δ18O) isotope values from tooth enamel of mammals from Cobra Cave (164–131 ka) are compared to our published data (Bacon et al., 2021, Sci. Rep.)29 of faunas of comparable composition (Artiodactyla, Perissodactyla, Proboscidea, Carnivora, Primates, and Rodentia) from the following sites: Coc Muoi (148–117 ka), Tam Hang South (94–60 ka), Nam Lot I (86–72 ka), Duoi U’Oi (70–60 ka), and Tam Hay Marklot (38.4–13.5)29,48 (Supplementary Tables S1–S4 and Annexes S3, S4).
Results of δ13Ccarbon source values (‰ VPDB) from these six subsamples are used herein to build the histograms presented in Fig. 5b. They show the distribution (expressed as a percentage) of δ13Ccarbon source values of specimens within each biome: closed-canopy forests (< − 27.2‰); intermediate rainforests and woodland biomes (> − 27.2‰ and < − 21.3‰); no specific ecological environment (> − 21.3‰ and < − 15.3‰); and savannah-like environments (> − 15.3‰) (Supplementary Table S5). Our goal here is to correlate the faunal isotopic data with available proxy indicators of climates. The palaeoclimatic data used from the last ~ 200,000 years are speleothem δ18O values recorded in the Sanbao/Hulu Chinese caves (Fig. 275), which allow for tracking the fluctuation of the summer monsoon intensity for the period as shown in Fig. 5a. In this record, low speleothem δ18O values (‰, VPDB, left ordinate axis) correspond to low rainfall δ18O values and therefore to increases in precipitation, i.e., the amount effect89.
Correlating the chronologies between faunas constrained by luminescence dating and better chronologically constrained palaeoclimatic signals is challenging, particularly for assemblages with a wide age range like that of Tam Hang South (94–60 ka). However, the δ13C values of teeth selected from this assemblage could not result from a mixture with the Duoi U’Oi period (70–60 ka) as both subsamples show different distributions (Fig. 4), furthermore supported by statistics29. From our previous published analysis29, overall, results of the MIS 5 Tam Hang South and Nam Lot faunas are consistent and indicate a high level of environmental heterogeneity, showing a mosaic of habitats from closed-canopy forests to more open biomes during relatively strong monsoonal periods and high precipitation75. Supplementary Table S5 shows the percentages of teeth associated to the different biomes with 27.4% (C3), 69.2% (C3–C4), 3.2% (C4) for Tam Hang South and 42.1% (C3), 54.3% (C3–C4), 3.5% (C4) for Nam Lot I. During this period, C3–C4 forested to open landscapes were occupied by ruminant taxa thanks to their great dietary flexibility and capacity to shift from browsing to grazing (essentially Rusa unicolor, medium-sized cervids, large bovines Bos and Bubalus, and caprines Capricornis and Naemorhedus). The range of δ13C values of these taxa, the highest among herbivores, has been associated with either the variability of diet within a species or dietary flexibility according to seasons65,90,91. In these types of environments, the C3 canopy forests contained most of the megaherbivore biomass (Tapirus, Megatapirus, Rhinoceros, Dicerorhinus, Elephas, and Stegodon).
Our previous results also highlighted the potential of canopy forests for contraction during some periods. In our sample, two faunas illustrate these changes, Duoi U’Oi (70–60 ka) and Coc Muoi (148–117 ka), which match with significant drops of monsoon intensity at the onset of MIS 4 and the end of MIS 6, respectively. Data of the Supplementary Table S5 show a predominant proportion of C3 plants (up to ~ 73% at Duoi U’Oi), which corresponds to phases of expansion of the canopy rainforests: 73.3% (C3), 25% (C3–C4), 1.6% (C4) for Duoi U’Oi and 65.4% (C3), 34.4% (C3–C4), 0% (C4) for Coc Muoi. In these ecosystems, closed rainforests contained all of the mammalian biomass. Isotopic results29 and palynological proxies53,76 indicate major abiotic (rates of insolation, seasonality, amount of rainfall, etc.) and environmental changes with different types of forests (plant communities, structure of the vegetation, etc.) during these events of weak monsoons. Furthermore, an estimate of the abundance of herbivore taxa (Supplementary Fig. S6) also supports functional changes of ecosystems: that of Coc Muoi favoured rhinoceroses and large bovids, whereas that of Duoi U’Oi favoured large cervids.
Statistical analyses
Kruskal–Wallis one-way analysis of variance and post-hoc Games-Howell pairwise comparisons (Supplementary Tables S6, S7) were performed across the dataset to identify statistical differences in δ13Ccarbon source and δ18Oapatite values between sites (and thus also periods). Preliminary tests and visual inspection were carried out to check for normally distributed data and equal variance, which revealed that non-parametric testing (i.e., Kruskal–Wallis) was preferred over parametric methods (i.e., ANOVA). A total of 6 different sites and 390 samples were used for the analyses: Cobra Cave (n = 55), Coc Muoi (n = 84)29, Tam Hang South (n = 62)29, Duoi U’Oi (n = 60)29, Nam Lot (n = 57)49 and Tam Hay Marklot (n = 72)48. All statistical analyses were conducted using the free program R software (version 4.2.292) and packages “car” (version 3.193), stats (version 3.6.2; R Core Team, 2022), tidyverse (version 1.3.294), and ggplot2 (version 3.495).
Data availability
All data generated and analysed during this study are included in this article and its supplementary file.
References
Kraus, J. et al. The complete mitochondrial DNA genome of an unknown hominin from southern Siberia. Nature 464, 894–897 (2010).
Reich, D. et al. Genetic history of an archaic hominin group from Denisova Cave in Siberia. Nature 468, 1053–1060 (2010).
Sawyer, S. et al. Nuclear and mitochondrial DNA sequences from two Denisovan individuals. Proc. Natl Acad. Sci. USA 112, 15696–15700 (2015).
Slon, V. et al. A fourth Denisovan individual. Sci. Adv. 3, e1700186 (2017).
Jacobs, Z. et al. Timing of archaic hominin occupation of Denisova Cave in southern Siberia. Nature 565, 594–599 (2019).
Douka, K. et al. Age estimates for hominin fossils and the onset of the Upper Palaeolithic at Denisova Cave. Nature 565, 640–644 (2019).
Massilani, D. et al. Denisovan ancestry and population history of early East Asians. Science 370, 579–583 (2020).
Chen, F. et al. A late Middle Pleistocene Denisovan mandible from the Tibetan Plateau. Nature 569, 409–412 (2019).
Zhang, D. et al. Denisovan DNA in Late Pleistocene sediments from Baishiya Karst Cave on the Tibetan Plateau. Science 370, 584–587 (2020).
Xing, S., Martinon-Torres, M., Bermudez de Castro, J. M., Wu, X. & Liu, W. Hominin teeth from the Early Late Pleistocene Site of Xujiayao, Northern China. Am. J. Phys. Anthropol. 156, 224–240 (2015).
Chang, C.-H. et al. The first archaic Homo from Taiwan. Nat. Commun. 6, 6037 (2015).
Ni, J. et al. Massive cranium from Harbin in northeastern China establishes a new Middle Pleistocene human lineage. Innovation 2, 100130 (2021).
Demeter, F. et al. A middle pleistocene Denisovan from the annamite chain of northern Laos. Nat. Commun. 13, 2557 (2022).
Huerta-Sanchez, E. et al. Altitude adaptation in Tibetans caused by introgression of Denisovan-like DNA. Nature 512, 194–197 (2014).
Reich, D. et al. Denisova admixture and the first modern human dispersals into Southeast Asia and Oceania. Am. J. Hum. Genet. 89, 516–528 (2011).
Meyer, M. et al. A high-coverage genome sequence from an archaic Denisovan individual. Science 338, 222–226 (2012).
Prüfer, K. et al. The complete genome sequence of a Neanderthal from the Altai Mountains. Nature 505, 43–49 (2014).
Kuhlwilm, M. et al. Ancient gene flow from early modern humans into Eastern Neanderthals. Nature 530, 429–533 (2016).
Vernot, B. et al. Excavating Neandertal and Denisovan DNA from the genomes of Melanesian individuals. Science 352, 235–239 (2016).
Browning, S. R., Browning, B. L., Zhou, Y., Tucci, S. & Akey, J. M. Analysis of human sequence data reveals two pulses of archaic Denisovan admixture. Cell 173, 53–61 (2018).
Slon, V. et al. The genome of the offspring of a Neanderthal mother and a Denisovan father. Nature 561, 113–116 (2018).
Jacobs, G. S. et al. Multiple deeply divergent Denisovan ancestries in Papuans. Cell 177, 1010–1021 (2019).
Bergström, A. et al. Insights into human genetic variation and population history from 929 diverse genomes. Science 367, eaay5012 (2020).
Zavala, E. I. et al. Pleistocene sediment DNA reveals hominin and faunal turnovers at Denisova Cave. Nature 595, 399–403 (2021).
Bergström, A., Stringer, C., Hajdinjak, M., Scerri, E. M. L. & Skoglund, P. Origins of modern human ancestry. Nature 590, 229–237 (2021).
Morley, M. W. The geoarchaeology of hominin dispersals to and from tropical Southeast Asia: A review and prognosis. J. Archaeol. Sci. 77, 78–93 (2017).
DeNiro, M. & Epstein, S. Influence of the diet on the distribution of carbon isotopes in animals. Geochim. Cosmochim. Acta 42, 495–506 (1978).
Lee-Thorp, J. A., Sealy, J. C. & van der Merwe, N. J. Stable carbon isotope ratio differences between bone collagen and bone apatite, and their relationship to diet. J. Archaeol. Sci. 16, 585–599 (1989).
Bacon, A.-M. et al. A multi-proxy approach to exploring Homo sapiens’ arrival, environments and adaptations in Southeast Asia. Sci. Rep. 11, 21080 (2021).
Louys, J. & Roberts, P. Environmental drivers of megafauna and hominin extinction in Southeast Asia. Nature 586, 402–406 (2020).
Guatelli-Steinberg, D. et al. Chronology of linear enamel hypoplasia formation in the Krapina Neanderthals. PaleoAnthropology 431−445 (2014).
Nava, A. et al. Early life of Neanderthals. Proc. Natl Acad. Sci. USA 117, 28719–28726 (2023).
Tejada-Lara, J. V. et al. Body mass predicts isotope enrichment in herbivorous mammals. Philos. Trans. R. Soc. B 285, 20181020 (2018).
Tejada, J. V. et al. Comparative isotope ecology of western Amazonian rainforest mammals. Proc. Natl Acad. Sci. USA 117, 26263–26272 (2021).
Pederzani, S. et al. Subarctic climate for the earliest Homo sapiens in Europe. Sci. Adv. 7, eabi4642 (2021).
Lisiecki, L. E. & Raymo, M. E. A Pliocene-Pleistocene stack of 57 globally distributed benthic δ18O records. Paleoceanography 20, PA1003 (2005).
Rizal, Y. et al. Last appearance of Homo erectus at Ngandong, Java, 117,000–108,000 years ago. Nature 577, 381–385 (2020).
Bettis, E. A. III. Way out of Africa: Early Pleistocene paleoenvironments inhabited by Homo erectus in Sangiran, Java. J. Hum. Evol. 56, 11–24 (2009).
Sémah, A.-M., Sémah, F., Djubiantono, T. & Brasseur, B. Landscapes and Hominids’ environments: Changes between the Lower and the Early Middle Pleistocene in Java (Indonesia). Quat. Int. 223–224, 451–454 (2010).
Sémah, A.-M. et al. The palaeoenvironmental context of the Palaeolithic of Java: A brief review. Quat. Int. 416, 38–45 (2016).
Janssen, R. et al. Tooth enamel stable isotopes of Holocene and Pleistocene fossil fauna reveal glacial and interglacial paleoenvironments of hominins in Indonesia. Quat. Sci. Rev. 144, 145–154 (2016).
Puspaningrum, M. R., van den Bergh, G. D., Chivas, A. R., Setiabudi, E. & Kurniawan, I. Isotopic reconstruction of Proboscidean habitats and diets on Java since the Early Pleistocene: Implications for adaptations and extinction. Quat. Sci. Rev. 228, 106007 (2020).
Kubat, J. et al. Dietary strategies of Pleistocene Pongo sp. and Homo erectus on Java (Indonesia). Nat. Ecol. Evol. 7, 279–289 (2023).
Demeter, F. et al. Anatomically modern human in Southeast Asia (Laos) by 46 ka. Proc. Natl Acad. Sci. USA 109, 14375–14380 (2012).
Shackelford, L. L. et al. Additional evidence for early modern human morphological diversity in Southeast Asia at Tam Pa Ling, Laos. Quat. Int. 466, 93–106 (2018).
Bourgon, N. et al. Trophic ecology of a Late Pleistocene early modern human from tropical Southeast Asia inferred from zinc isotopes. J. Hum. Evol. 161, 103075 (2021).
Milano, S. et al. Environmental conditions framing the first evidence of modern humans at Tam Pà Ling, Laos: A stable isotope record from terrestrial gastropod carbonates. Palaeogeogr. Palaeoclimatol. Palaeoecol. 511, 352–363 (2018).
Bourgon, N. et al. Zinc isotopes in Late Pleistocene fossil teeth from a Southeast Asian cave setting preserve paleodietary information. Proc. Natl Acad. Sci. USA 117, 4675–4681 (2020).
Bacon, A.-M. et al. Nam Lot (MIS 5) and Duoi U’Oi (MIS 4) Southeast Asian sites revisited: Zooarchaeological and isotopic evidences. Palaeogeogr. Palaeoclimatol. Palaeoecol. 512, 132–144 (2018).
Freidline, S. et al. Early presence of Homo sapiens in Southeast Asia by 86–68 kyr at Tam Pà Ling, Northern Laos. Nat. Commun. 14, 3193 (2023).
Norton, C. J., Jin, C., Wang, Y. & Zhang, Y. Rethinking the Palearctic-Oriental biogeographic boundary in Quaternary China. In Asian Paleoanthropology: From Africa to China and Beyond (eds. Norton, C.J. & Braun, D.R.) 81–100 (Vertebrate Paleobiology and Paleoanthropology, 2010).
Owen-Smith, R. N. Megaherbivores. The Influence of Very Large Body Size on Ecology (Cambridge University Press, 1988).
Zheng, Z. & Lei, Z.-Q. A 400,000 years record of vegetational and climatic changes from a volcanic basin, Leizhou Peninsula, southern China. Palaeogeogr. Palaeoclimatol. Palaeoecol. 145, 339–362 (1999).
Schwartz-Narbonne, R. et al. Reframing the mammoth steppe: Insights from the analysis of isotopic niches. Quat. Sci. Rev. 215, 1–21 (2019).
Ma, J. et al. The Mammuthus-Coelodonta Faunal Complex at its southeastern limit: A biogeochemical paleoecology investigation in Northeast Asia. Quat. Int. 591, 93–106 (2021).
Kelly, A. et al. Dietary paleoeocology of bison and horses on the mammoth steppe of eastern Beringia based on dental microwear and mesowear analyses. Palaeogeogr. Palaeoclimatol. Palaeoecol. 572, 110394 (2021).
Drucker, D. G. The isotopic ecology of the mammoth steppe. Annu. Rev. Earth Planet. Sci. 50, 395–418 (2022).
Westaway, K. E. et al. An early modern human presence in Sumatra 73000–63000 years ago. Nature 548, 322–325 (2017).
Louys, J. et al. Speleological and environmental history of Lida Ajer cave, western Sumatra. Philos. Trans. R. Soc. B. 377, 20200494 (2022).
Wedage, O. et al. Specialized rainforest hunting by Homo sapiens ~45,000 years ago. Nat. Commun. 10, 1–8 (2019).
Langley, M. C. et al. Bows and arrows and complex symbolic displays 48,000 years ago in the South Asian tropics. Sci. Adv. 6, eaba3831 (2020).
Scerri, E. M., Roberts, P., Maezumi, S. Y. & Malhi, Y. Tropical forests in the deep past. Philos. Trans. R. Soc. B. 377, 20200500 (2022).
Roberts, P. et al. Fruits of the forest: Human stable isotope ecology and rainforest adaptations in Late Pleistocene and Holocene (~36 to 3 ka) Sri Lanka. J. Hum. Evol. 106, 102–118 (2017).
Roberts, P. et al. Isotopic evidence for initial coastal colonization and subsequent diversification in the human occupation of Wallacea. Nat. Commun. 11, 2068 (2020).
Suraprasit, K., Shoocongdej, R., Chintakanon, K. & Bocherens, H. Late Pleistocene human paleoecology in the highland savanna ecosystem of mainland Southeast Asia. Sci. Rep. 11, 16756 (2021).
Barton, H. The case for rainforest foragers: The starch record at Niah Cave, Sarawak. Asian Perspect. 44, 56–72 (2005).
Barker, G. et al. The ‘human revolution’ in lowland tropical Southeast Asia: The antiquity and behavior of anatomically modern humans at Niah Cave (Sarawak, Borneo). J. Hum. Evol. 52, 243–261 (2007).
Summerhayes, G. R. et al. Human adaptation and plant use in highland New Guinea 49,000 to 44,000 years ago. Science 330, 78–81 (2010).
Barton, H., Piper, P. J., Rabett, R. & Reeds, I. Composite hunting technologies from the Terminal Pleistocene and Early Holocene, Niah Cave, Borneo. J. Archaeol. Sci. 36, 1708–1714 (2009).
Bacon, A.-M. et al. Late Pleistocene mammalian assemblages of Southeast Asia: New dating, mortality profiles and evolution of the predator-prey relationships in an environmental context. Palaeogeogr. Palaeoclimatol. Palaeoecol. 422, 101–127 (2015).
Kuzmin, Y. V., Bondarev, A. A., Kosintsev, P. A. & Zazovskaya, E. P. The Paleolithic diet of Siberia and Eastern Europe: Evidence based on stable isotopes (δ13C and δ15N) in hominin and animal bone collagen. Archaeol. Anthropol. Sci. 13, 179 (2021).
Hublin, J.-J. & Changeux, J.-P. Paleoanthropology of cognition: An overview on Hominins brain evolution. C. R. Biol. 345, 1–19 (2022).
Roberts, P. & Stewart, B. A. Defining the ‘generalist specialist’ niche for Pleistocene Homo sapiens. Nat. Hum. Behav. 2, 542–550 (2018).
Bobe, R. & Wood, B. Estimating origination times from the early hominin fossil record. Evol. Anthropol. 31, 92–102 (2022).
Wang, Y. et al. Millennial- and orbital-scale changes in the East Asian monsoon over the past 224,000 years. Nature 451, 1090–1093 (2008).
Sun, X., Luo, Y., Huang, F., Tian, J. & Wang, P. Deep-sea pollen from the South China Sea: Pleistocene indicators of East Asian monsoon. Mar. Geol. 201, 97–118 (2003).
Roberts, P., Boivin, N., Lee-Throp, J., Petraglia, M. & Stock, J. Tropical forests and the genus Homo. Evol. Anthropol. 25, 306–317 (2016).
Stewart, J. R. & Stringer, C. B. Human evolution out of Africa: The role of refugia and climate change. Science 335, 1317–1321 (2012).
Hilgen, S. L., Hilgen, F. J., Adhityatama, S., Kuiper, K. F. & Joordens, J. C. A. Towards an astronomical age model for the Lower to Middle Pleistocene hominin-bearing succession of the Sangiran Dome area on Java, Indonesia. Quat. Sci. Rev. 297, 107788 (2022).
Joordens, J. C. A. et al. Homo erectus at Trinil on Java used shells for tool production and engraving. Nature 518, 228–231 (2015).
Spratt, R. M. & Lisiecki, L. E. A Late Pleistocene sea level stack. Clim. Past Discuss. 11, 3699–3728 (2015).
Westaway, K. E. et al. Age and biostratigraphic significance of the Punung Rainforest Fauna, East Java, Indonesia, and implications for Pongo and Homo. J. Hum. Evol. 53, 709–717 (2007).
Storm, P. & de Vos, J. Rediscovery of the Late Pleistocene Punung hominin sites and the discovery of a new site Gunung Dawung in East Java. Senckenb. Lethaea 86, 121–131 (2006).
Farquhar, G. D., Ehleringer, J. R. & Hubick, K. T. Carbon isotope discrimination and photosynthesis. Annu. Rev. Plant Biol. 40, 503–537 (1989).
Cerling, T. E. et al. Dietary changes of large herbivores in the Turkana Basin, Kenya from 4 to 1 Ma. Proc. Natl Acad. Sci. USA 112, 11467–11472 (2015).
van der Merwe, N. J. & Medina, E. The canopy effect, carbon isotope ratios and foodwebs in Amazonia. J. Archaeol. Sci. 18, 249–259 (1991).
Friedli, H., Lötscher, H., Oeschger, H., Siegenthaler, U. & Stauffer, B. Ice core record of the 13C/12C ratio of atmospheric CO2 in the past two centuries. Nature 324, 237–238 (1986).
Kohn, M. J. Carbon isotope compositions of terrestrial C3 plants as indicators of (paleo)ecology and (paleo)climate. Proc. Natl Acad. Sci. USA 107, 19691–19695 (2010).
Dansgaard, W. Stable isotopes in precipitation. Tellus 16, 436–468 (1964).
Stacklyn, S. et al. Carbon and oxygen isotopic evidence for diets, environments and niche differentiation of early Pleistocene pandas and associated mammals in South China. Palaeogeogr. Palaeoclimatol. Palaeoecol. 468, 351–361 (2017).
Sun, F. et al. Paleoecology of Pleistocene mammals and paleoclimatic change in South China: Evidence from stable carbon and oxygen isotopes. Palaeogeogr. Palaeoclimatol. Palaeoecol. 524, 1–12 (2019).
R Core Team. R: A Language and Environment for Statistical Computing (R Foundation for Statistical Computing, 2022).
Fox, J. & Weisberg, S. An R Companion to Applied Regression 3rd edn. (Sage, 2019).
Wickham, H. et al. Welcome to the tidyverse. J. Open Source Softw. 4, 1686 (2019).
Wickham, H. ggplot2: Elegant Graphics for Data Analysis (Springer, 2016).
Acknowledgements
We thank the Ministry of Information, Culture and Tourism of Laos PDR for encouraging and supporting our work. We thank also the authorities of Xon district, Huà Pan Province, and the villagers of Long Gua Pa village for their continuous support of our fieldworks since 2003. Funding for isotopic analyses of the Cobra Cave specimens came from the research laboratory BABEL UMR 8045 CNRS/Université Paris Cité, France, to A.M.B. and the Department of Human Evolution, Max Planck Institute for Evolutionary Anthropology, Leipzig, Germany, to N.B.
Author information
Authors and Affiliations
Contributions
A.M.B. and N.B. designed and performed research; N.B., E.D., O.T. and D.F. performed sample preparation; A.M.B., N.B., F.D., C.Z., K.E.W., R.J.B., P.D., J.L.P., M.W.M., E.S., S.F., Q.B., P.O.A., S.B., P.S., D.S., E.P.E., A.Z., T.L., V.S., T.E.D., L.S., J.J.H. are part of the LAOS project; N.T.M.H., N.A.T., A.M.B., P.O.A., P.D., J.L.P., K.E.W., R.J.B., F.D. are part of the VIETNAM project; A.M.B., N.B., J.J.H. wrote the paper with contributions of all other co-authors All authors reviewed the manuscript.
Corresponding author
Ethics declarations
Competing interests
The authors declare no competing interests.
Additional information
Publisher's note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary Information
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Bacon, AM., Bourgon, N., Dufour, E. et al. Palaeoenvironments and hominin evolutionary dynamics in southeast Asia. Sci Rep 13, 16165 (2023). https://doi.org/10.1038/s41598-023-43011-2
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/s41598-023-43011-2