Terminal Pleistocene Early Holocene Research Articles

The Excavation of Gua Harimau's Western Gallery: A Contribution to the Terminal Pleistocene-Early Holocene archaeological records in Sumatra

Gua Harimau (gua=cave) is a crucial site for understand the prehistory of Sumatra and its relationship to the peopling of Insular Southeast Asia. A large number of archaeological records has been discovered at the western part of the cave, including human burials, lithic artefacts, potteries, and faunal remains that have been anthropogenically altered. The chronological stratigraphy of these records extends to the Late Pleistocene Period, enabling us to reconstruct the dynamics of prehistoric culture in Sumatra. At least three episodes of human occupation have been recognized in the cave: The Preneolithic, Neolithic, and Paleometalic (early metal age). However, the distributional pattern of various archaeological remains, such as mollusk shells, ground tools, unifacially-shaped pebble tools, and bone artefacts, reveals significant differences. These differences provide insight into the cultural development that occurred in Sumatra during the prehistoric occupation of Gua Harimau, and demonstrate changes in cultural affinity that may correspond to past environmental conditions and human adaptation.

Paleoindian land use at pluvial Lake Mojave: A temporally and seasonally dynamic wetlands resource patch in California’s central Mojave Desert

This study creates and preliminarily tests a biotic resource structure and optimal foraging theory inspired land use model for Silver Lake, one of two playa lakes that once formed pluvial Lake Mojave in California’s central Mojave Desert. The land use model predicts that when Silver Lake was a high-rank resource patch, Paleoindians optimally decreased inter-basin mobility as part of a wetland stable strategy (WSS); conversely, when Silver Lake was a low-rank resource patch, Paleoindians optimally increased inter-basin mobility using a wetland transient strategy (WTS). Specific land use expectations are developed in 500-year timesteps for the period 14.0–9.0 ka based on yearly and seasonal variations in the productivity of Silver Lake’s wetland habitat. Data limitations with the fine-grained version of the model led us to test a simplified version that supports the hypothesis that Paleoindians selected the optimal land use solution, a WTS, when Silver Lake was a low-rank wetlands resource patch during the terminal Pleistocene/Early Holocene (TP/EH) and more recent groups a WSS when patch rank improved after the Early Holocene. Though the simplified model’s conclusion that Paleoindians employed a WTS during the TP/EH fits the available archaeological evidence at Silver Lake and broader Great Basin/Mojave Desert, it is a static interpretation that implies Paleoindians used a single land use strategy during the several thousand-year TP/EH. This is unrealistic because pluvial lakes were dynamic water bodies that fluctuated through time and directly influenced the extent of and caloric returns from the available wetland habitat. Despite the current inability to test the fine-grained model, which we think would reveal temporal and seasonal variations in Paleoindian land use, not considering or integrating such factors into land use models misses opportunities to better understand the myriad decisions Paleoindians made under ever changing resource structure conditions.

New lithic evidence from Terminal Pleistocene-Early Holocene Zhongshan Rockshelter, Guangxi, southern China

The archaeological materials recovered during the 2018 excavation of Zhongshan Rockshelter (ZSRS) in Bubing Basin, Guangxi, southern China helps to fill a cultural gap during the terminal Paleolithic in the region. The evidence from the ZSRS, dated to ∼14,523–8,472 cal. BP, is comparable to other parts of southern China and mainland Southeast Asia where grinding techniques, complex settlement patterns and broad-spectrum subsistence strategies appeared. What distinguishes the ZSRS from other sites in the region however is the widespread use of flaked stone tools and ochre utilization, but with an absence of pottery. This is an excellent example of variation in foraging behavior in this spatio-temporal facies, where one size does not fit all. Regional differences suggest a more complex human foraging strategy rather than a single and fixed behavioral model across southern China during the Terminal Pleistocene to Early Holocene transition.

Allometry of unifacial flake tools from Mojave Desert terminal pleistocene/early holocene sites: Implications for landscape knowledge, tool design, and land use

This study evaluates the allometry of terminal Pleistocene-early Holocene age unifacial flake tools in the northcentral Mojave Desert of California to reveal linkages between landscape knowledge, tool design, and land use. Analyses of 438 unifacial flake tools (all types) show at the population scale that Mojave Desert Paleoindians, like other post-Clovis Paleoindians across North America, possessed good knowledge of the landscape and designed flake tools for short-term purposes. This design strategy repeats in three study areas but to varying degrees: flake tools in the Fort Irwin study area were designed for shorter-term use than those from the pluvial China Lake and Lake Mojave study areas. Chert flake tools were designed for somewhat shorter-term use than those made from fine-grained volcanic stones, with obsidian flake tools designed for longevity. The regional ubiquity of knappable lithic raw materials, especially cherts, partially explains the short-term design strategy given Paleoindians’ knowledge of the landscape, including where to find knappable stones. Given that the design strategy of unifacial flake tools varies by raw material type, it provides a reminder that Great Basin Paleoindian land use models derived solely or largely from obsidian projectile points and bifacial tools may result in biased interpretations.

Faunal assemblages and demography during the Late Pleistocene (MIS 2-1) to Early Holocene in Highland Pang Mapha, Northwest Thailand

This paper examines the changes in taxonomic diversity and demography of faunal assemblages hunted by the human inhabitants of the Tham Lod and Ban Rai rockshelters in the Pang Mapha highlands of northwestern Thailand during the Late Pleistocene (MIS 2-1) to Terminal Pleistocene-Early Holocene. Evidence from these excavated sites indicates that environmental changes influenced animal dispersal from high latitudes to the highland region of Pang Mapha. During the Late Pleistocene the taxonomic diversity was dominated by Cervidae and Testudines, with other faunal remains representing Himalayan goral, Suidae and Bovidae. A ternary analysis of the Suidae indicates a mixed pattern, mostly represented by Juveniles-Prime-Old (JPO), with the Prime dominating (Bovinae and Caprinae). By the Terminal Pleistocene-Early Holocene, Cercopithecidae had become dominant and the Himalayan goral absent, and presumably extinct. According to the changes in taxonomic diversity and the paleodemographic data obtained from the Pang Mapha faunal assemblages, we have increased our understanding of paleoenvironments and subsistence strategies during the Late-Terminal Pleistocene and Early Holocene.

Holocene savanna dynamics in the seasonal tropics of northern Australia

An environmental history is presented from Girraween Lagoon, Darwin region of the Northern Territory, Australia. Pollen and charcoal analysis of a 5-meter sediment core provides a record of vegetation change, fire history and climate spanning 12,700 cal BP to the present day. This study focusses on tree-grass vegetation dynamics, eucalypt to non-eucalypt plant interactions, and climate–fire–human relationships in an area where few long-term savanna records exist. The dataset suggests wetlands experienced alternating episodes of ephemeral waterlogging and seasonal inundation due to post-glacial monsoon variability up until permanent inundation from approximately 6000 cal BP. The surrounding catchment transformed from a terminal Pleistocene–early Holocene wooded-savanna to a later Holocene open forest. This increase in woody cover was a prominent site feature, primarily driven by climate–moisture availability. In turn, the extent of fire and fire impact, is a function of climate–vegetation feedbacks. Such interplay between fire history, climate change and vegetation pattern was also influenced by more intense human management of the area, in the last 4000 years of the record. It is proposed Girraween may have become a much-socialized and managed human landscape in this late Holocene phase. Results provide essential baseline data describing savanna dynamics linked to contemporary ecological observation, understanding and management goals, and serves as an important resource for the Quaternary sciences and archeology of northern Australia.

Do Uncharred Plants Preserve Original Carbon and Nitrogen Isotope Compositions?

The isotopic compositions of plants can provide significant insights into paleodiets, ancient agricultural activities, and past environments. Isotopic compositions of charred (aka carbonized) ancient plant remains are typically preferred over those of uncharred/uncarbonized plants, both because charred plants are more commonly preserved and because early research suggested they experience less post-depositional isotopic alteration. In this paper, we re-explore the question of whether uncharred plants experience large-magnitude post-depositional changes in carbon and nitrogen isotope compositions by analyzing Terminal Pleistocene–Early Holocene plant specimens from rockshelters in the Escalante River Basin (Colorado Plateau, southeastern Utah). Several lines of evidence, including C3-CAM differences, plant-part comparisons, and dietary estimates from ancient herbivore collagen, suggest that the original carbon isotope compositions of these plants have not been significantly altered. The preservation status of plant nitrogen isotope compositions is equivocal. The direction of temporal shifts in plant δ15N matches global trends and the magnitude of the shift may have been exacerbated by the extinction of megafauna in an arid environment. However, the Pleistocene plant δ15N values are higher than would be expected based on herbivore bone collagen δ15N. Nevertheless, in contrast to previous research, the ancient uncharred plants in this study did not have exceptionally high δ15N values (> + 25‰). Overall, our research suggests that uncharred plants could be useful substrates for isotopic paleodietary and/or paleoenvironmental studies.

Survival histories of marsupial carnivores on Australian continental shelf islands highlight climate change and Europeans as likely extirpation factors: implications for island predator restoration

Predators are critical components of ecosystems, but large marsupial carnivores have suffered major declines and extinctions in Australia. To inform predator restoration efforts on Kangaroo Island (South Australia) we examined the survival histories and potential extirpation factors of large marsupial carnivores that previously occurred on Kangaroo Island, King Island and Flinders Island, located off the southern coastline of the Australian mainland. Through a review of historical accounts and fossil evidence, we determined that the pattern of species persistence and extirpation on Kangaroo Island parallels that observed on King and Flinders Islands. Fossil data supports the terminal Pleistocene–early Holocene extinction of the thylacine (Thylacinus cynocephalus) from Kangaroo Island and the Tasmanian devil (Sarcophilus harrisii) from both Kangaroo and Flinders Islands. Though eastern quoll (Dasyurus viverrinus) fossils have been found on both Kangaroo and Flinders Islands, and western quoll (D. geoffroii) on Kangaroo Island, contemporary evidence for their post-European persistence is unclear. In contrast, fossil, museum and anecdotal data supports the presence of the spotted-tailed quoll (D. maculatus) on all three islands and, contrary to established knowledge, its post-European persistence on Kangaroo Island. The loss of T. cynocephalus, S. harrisii, D. geoffroii and D. viverrinus from these islands appears to be commensurate with late to terminal Pleistocene–early Holocene climate change and associated changes in vegetation communities. In contrast, anthropogenic persecution of D. maculatus appears to be the principal cause of its post-European extirpation. We recommend D. maculatus as a suitable candidate marsupial carnivore for reintroduction to Kangaroo Island.

A First Look at the Terminal Pleistocene/Early Holocene Record of Guano Valley, Oregon, USA

ABSTRACTSurface sites play a vital role in interpreting terminal Pleistocene/early Holocene (TP/EH) lifeways in the Great Basin. Two years of work in Oregon’s Guano Valley by crews from the University of Nevada, Reno have revealed a rich record of Western Stemmed Tradition occupations associated with an extensive delta system that brought freshwater into the valley from the adjacent tablelands. To date, we have recorded nearly 700 diagnostic TP/EH artifacts within the relatively small delta, making it one of the densest concentrations of Paleoindian artifacts in the region. This paper presents preliminary results from our archaeological and geomorphological investigations in Guano Valley as well as an extensive toolstone source provenance analysis.

Human maritime subsistence strategies in the Lesser Sunda Islands during the terminal Pleistocene–early Holocene: New evidence from Alor, Indonesia

The islands of Wallacea are remarkable on a world scale as settlement occurred by at least 43,000 cal BP and must have involved the use of watercraft. The majority of the islands are depauperate in terrestrial fauna and human subsistence must have focused on the marine environment. Although few islands have been archaeologically explored, some such as Timor have yielded abundant remains of pelagic and reef fishes, as well as the earliest evidence of fishhook manufacture, demonstrating that modern humans in this region were well equipped to undertake complex exploitation of the marine environment. However, a holistic understanding of human subsistence strategies on these islands from initial colonisation in the late Pleistocene through to the Recent remains elusive. Here, we present survey and excavation data from the site of Tron Bon Lei on the small island of Alor in Nusa Tenggara Timur, Indonesia. This study focuses on the terminal Pleistocene – early Holocene human fishing behaviour in the region, and specifically the zooarchaeological sequence from Tron Bon Lei dated to this period. The Holocene preference for reef fish relative to pelagic fish, observed in mid-to late Holocene occupation levels in East Timor, occurred earlier in Tron Bon Lei, suggesting a shift from the larger abundance of carnivore taxa observed in the late Pleistocene deposits. Comparisons with other archaeological deposits in Wallacea indicates that fishing was an important subsistence activity in Nusa Tenggara Timur, unlike more northern Wallacean islands where shellfish make up almost all zooarchaeological records. Tron Bon Lei confirms that the faunally limited nature of Alor spans the Holocene, and we find no evidence that terrestrial fauna made up a significant component of the subsistence activities undertaken at this site.

Biogeographic distribution and metric dental variation of fossil and living orangutans (Pongo spp.).

The genus Pongo has a relatively richer Quaternary fossil record than those of the African great apes. Fossil materials are patchy in terms of anatomical parts represented, limited almost exclusively to isolated teeth, jaw and bone fragments. Fossil evidence indicates that the genus Pongo had a broadly continuous distribution across the southern part of the Indomalayan biogeographic region, ranging in time from Early Pleistocene to Holocene: southern China (77 fossil sites), Vietnam (15), Laos (6), Cambodia (2), Thailand (4), Peninsular Malaysia (6), Sumatra (4), Borneo (6) and Java (4). Within this distribution range, there are major geographical gaps with no known orangutan fossils, notably central and southern Indochina, central and southern Thailand, eastern Peninsular Malaysia, northern and southern Sumatra, and Kalimantan. The geological time and place of origin of the genus remain unresolved. Fossil orangutan assemblages usually show greater extent of dental metrical variation than those of modern-day populations. Such variability shown in prehistoric populations has partially contributed to confusion regarding past taxonomic diversity and systematic relationships among extinct and living forms. To date, no fewer than 14 distinct taxa have been identified and named for Pleistocene orangutans. Clear cases suggestive of predation by prehistoric human are few in number, and limited to terminal Pleistocene-Early Holocene sites in Borneo and a Late Pleistocene site in Vietnam.

Epipalaeolithic occupation and palaeoenvironments of the southern Nefud desert, Saudi Arabia, during the Terminal Pleistocene and Early Holocene

The transition from the Terminal Pleistocene to the Early Holocene is poorly represented in the geological and archaeological records of northern Arabia, and the climatic conditions that prevailed in the region during that period are unclear. Here, we present a new record from the site of Al-Rabyah, in the Jubbah basin (southern Nefud desert, Saudi Arabia), where a sequence of fossiliferous lacustrine and palustrine deposits containing an archaeological assemblage is preserved. Sedimentological and palaeoenvironmental investigations, both at Al-Rabyah and elsewhere in the Jubbah area, indicate phases of humid conditions, during which shallow lakes developed in the basin, separated by drier periods. At Al-Rabyah, the end of a Terminal Pleistocene phase of lake expansion has been dated to ∼12.2 ka using optically stimulated luminescence (OSL), with a mid-Holocene humid phase dated to after ∼6.6 ka. Palaeoecological reconstructions based primarily on non-marine molluscs and ostracods from the younger lacustrine deposits indicate a relatively shallow body of freshwater surrounded by moist, well-vegetated environments. A lithic assemblage characterized by bladelets and geometric microliths was excavated from sediments attributed to a drier climatic phase dated to ∼10.1 ka. The lithic artefact types exhibit similarities to Epipalaeolithic industries of the Levant, and their occurrence well beyond the ‘core region’ of such assemblages (and at a significantly later date) has important implications for understanding interactions between Levantine and Arabian populations during the Terminal Pleistocene–Early Holocene. We suggest that the presence of foraging populations in the southern Nefud during periods of drier climate is due to the prolonged presence of a freshwater oasis in the Jubbah Basin during the Terminal Pleistocene–Early Holocene, which enabled them to subsist in the region when neighbouring areas of northern Arabia and the Levant were increasingly hostile.

Terminal Pleistocene–Early Holocene lithic technological organization around Lake Mojave, California

The surface lithic scatters at two areas around Soda Lake were intensively surveyed and 3133 artifacts were analyzed in the field using four main variables to infer how Terminal Pleistocene–Early Holocene foragers organized lithic technology around pluvial Lake Mojave, California. Results indicate that early stage bifaces and flake tool blanks were created at a fine-grained volcanic (felsite) quarry/workshop complex in the Soda Mountains survey area and transported elsewhere. In addition, fine-grained volcanic bifaces were reduced and bifaces and flake tools of cryptocrystalline silicates and obsidian were finished, used, and/or discarded at a habitation area on the ancient shorelines near Little Cowhole Mountain. Comparisons with nearby sites of similar ages (at Ft. Irwin and China Lake) reveal many similarities in lithic technological organization. Lake Mojave—an important locus of prior research—can now be integrated into recent Mojave Desert and Great Basin technological organization studies.

Paleoindian rock art: establishing the antiquity of Great Basin Carved Abstract petroglyphs in the northern Great Basin

One of the principal ways that researchers date archaeological sites is by using temporally diagnostic projectile points as index fossils; however, this practice has not been widely employed to date rock art sites. We use this approach here to test the hypothesis that the Great Basin Carved Abstract (GBCA) petroglyph style found in the northern Great Basin was produced by Paleoindians. Using frequencies of projectile points at 55 GBCA sites, we demonstrate that Paleoindian points are significantly overrepresented there relative to their occurrence on the general landscape, providing evidence that Great Basin populations produced rock art sometime during the Terminal Pleistocene/Early Holocene (TP/EH), ∼12,500–8000 radiocarbon years ago. Additionally, we examine several environmental variables at GBCA sites and propose a model of Paleoindian land-use in the northern Great Basin that highlights seasonal visits to uplands to procure geophytes (i.e., root crops).

The Bonneville Estates Rockshelter rodent fauna and changes in Late Pleistocene–Middle Holocene climates and biogeography in the Northern Bonneville Basin, USA

Excavations at Bonneville Estates Rockshelter, Nevada recovered rodent remains from stratified deposits spanning the past ca. 12,50014CyrBP (14,800calyrBP). Specimens from horizons dating to the late Pleistocene and early Holocene include species adapted to montane and moist and cool habitats, including yellow-bellied marmot (Marmota flaviventris) and bushy-tailed woodrat (Neotoma cinerea). Shortly after 900014CBP (10,200calyrBP) these mammals became locally extinct, or nearly so, taxonomic diversity declined, and the region became dominated by desert woodrats (Neotoma lepida) and other species well-adapted to xeric, low-elevation settings. The timing and nature of changes in the Bonneville Estates rodent fauna are similar to records reported from nearby Homestead and Camels Back caves and provide corroborative data on terminal Pleistocene–early Holocene environments and mammalian responses to middle Holocene desertification. Moreover, the presence of northern pocket gopher (Thomomys talpoides) at Bonneville Estates adds to a sparse regional record for that species and, similar to Homestead Cave, it appears that the ca. 950014CyrBP (10,800calyrBP) replacement of the northern pocket gopher by Botta's pocket gopher in the Great Salt Lake Desert vicinity was also in response to climate change.

Natural and human impacts in a 35 000-year vegetation history in central New Britain, Papua New Guinea

Phytoliths and micro-charcoal from the Yombon Airstrip archaeological site in central New Britain, Papua New Guinea, provide the longest vegetation history record yet available for the New Guinea islands. The record begins about 35 kya with the first evidence for human presence at the site and, with the exception of the Last Glacial Maximum period, is continuous to the present. Three other sites provide supplementary evidence, including plant macro-remains, from the early Holocene onwards. The record is punctuated by a series of volcanic events, which are reflected in the vegetation record by alternating frequencies of closed forest and regrowth elements. Micro-charcoal is present from the oldest levels and fluctuates in frequency throughout the sequence, increasing substantially from the terminal Pleistocene-early Holocene onwards. This coincides with the first appearance of panicoid grasses and a range of potential cultivars including bananas and Saccharum. Increased levels of burning coinciding with the appearance of potential plant cultivars may indicate shifts in plant food production leading to cultivation from the early Holocene onwards. This compares favourably with previously reported evidence from Garua Island off the north coast of New Britain. The combination of trends in burning, vegetation clearance and appearance of potential cultivars on New Britain appears to parallel changes in the Papua New Guinea highlands at a similar time, and suggests regional similarities in subsistence and vegetation management practices from before the LGM onwards. Further studies are needed to clarify the timing and extent of these shifts across the region, and to provide a vegetation picture for the period before human colonisation of New Britain.

Nacurrie 1: Mark of ancient Java, or a caring mother’s hands, in terminal Pleistocene Australia?

There has been a protracted debate over the evidence for intentional cranial modification in the terminal Pleistocene Australian crania from Kow Swamp and Coobool Creek. Resolution of this debate is crucial to interpretations of the significance of morphological variation within terminal Pleistocene–early Holocene Australian skeletal materials and claims of a regional evolutionary sequence linking Javan Homo erectus and Australian Homo sapiens. However, morphological comparisons of terminal Pleistocene and recent Australian crania are complicated by the significantly greater average body mass in the former. Raw and size-adjusted metric comparisons of the terminal Pleistocene skeleton from Nacurrie, south-eastern Australia, with modified and unmodified H. sapiens and H. erectus, identified a suite of traits in the frontal, parietal, and occipital bones associated with intentional modification of a neonate’s skull. These traits are also present in some of the crania from Kow Swamp and Coobool Creek, which are in close geographic proximity to Nacurrie, but not in unmodified H. sapiens or Javan H. erectus. Frontal bone morphology in H. erectus was distinct from all of the Australian H. sapiens samples. During the first six months of life, Nacurrie’s vault may have been shaped by his mother’s hands, rather than though the application of fixed bandages. Whether this behaviour persisted only for several generations, or hundreds of years, remains unknown. The reasons behind the shaping of Nacurrie’s head, aesthetics or otherwise, and why this cultural practice was adopted and subsequently discontinued, will always remain a matter of speculation.

Distinguishing Holocene Microblades from a Paleoindian Component at the Mesa Site, Alaska

The terminal Pleistocene-early Holocene Mesa complex (ca. 9750–10,250 B.P. in uncalibrated radiocarbon years) of arctic Alaska, with its distinctive and technologically diagnostic bifacial tools, is often thought to be part of the broader Paleoindian tradition found throughout Pleistocene North America. In addition to artifacts of Paleoindian type, however, the Mesa site (type site of the Mesa complex) has a wedge-shaped core microblade technology, which is decidedly non-Paleoindian in character. Unfortunately, the Mesa is a shallow site, with no appreciable vertical structure or stratigraphic development. In essence, artifacts must be treated as if they were sitting on a single surface. This situation permits conflicting interpretations of the microblades and their relationship to the Paleoindian component. This paper presents a basic description of the microblades and associated artifacts at the Mesa site and the results of an analysis designed to discriminate these materials from the rest of the site assemblage. Using data on technology, raw material, and spatial patterning, I argue that the microblades and associated artifacts constitute a component that is distinct and separate from the Mesa Paleoindian component. The paper concludes with a discussion of several possible interpretations of the Mesa complex in light of these findings.

The palaeolimnology of northern Lake Malawi over the last 25 ka based upon the elemental and stable isotopic composition of sedimentary organic matter

Sediments accumulating in offshore areas of northern Lake Malawi typically contain 1–4% organic carbon. This organic matter (OM) is of mainly phytoplankton origin with varying mixtures of terrestrial and degraded, reworked material. High-resolution, stable carbon and nitrogen isotopic, elemental and Rock-Eval pyrolysis analyses of the bulk OM contained in three cores have been used to characterise the OM and provide detailed insights into the palaeolimnology and palaeoclimatology of northern Lake Malawi over the last 25 ka. Preservation of the OM as reflected by its Rock-Eval hydrogen index (HI) is highly variable. Periods characterised by notably low HI (<250) indicate a lowered lake level when oxic conditions episodically reached the deep lake floor. Particularly well-defined lowstands occurred between ca 11.8 and 10 cal ka and around 700 years ago. Isotopic and other evidence also suggest lowstand conditions before and during the Last Glacial Maximum (LGM) lasting until ca 17 ka. However, the OM that accumulated at that time has relatively high HI. Generally high HI values persisted until ca 11.8 ka, when there was a rapid decline leading in to the terminal Pleistocene–early Holocene lowstand. This abrupt change divides the OM record into two distinct phases, one preceding and the other following the decrease in HI. Higher HI values in the earlier period, even during times of low lake level, reflect high rates of primary production stimulated by vigorous vertical mixing of the water column that was driven by a greater incidence of northerly winds. After ca 11.8 ka the wind regime changed to one similar to today's, with mixing and nutrient cycling largely controlled by the southerly trade wind system, and the locus of high primary production at the south end of the lake. Periods of low δ 15 N during the mid- and late Holocene suggest intervals of stable stratification due to significantly reduced wind-driven mixing in the northern part of the lake, when N-fixing cyanophytes formed a major component of the phytoplankton. The geochemical data support recent suggestions of lowstand conditions in Lake Malawi at the time of the LGM, and provide additional evidence of periods of intense but localised aridity in the terminal Pleistocene–earliest Holocene and at the time of the Little Ice Age. The change in mixing regime at ca 11.8 ka seems to have been the result of a radical reorganisation of the dominant surface wind flow over south-eastern Africa. The transition was, within the resolution of currently available chronologies, apparently contemporaneous with marked changes in the circulation patterns of adjacent oceans, and the climate of more distant parts of the Northern Hemisphere.

A Possible Ritual Cache of Great Basin Stemmed Bifaces from the Terminal Pleistocene—Early Holocene Occupation of Nw Nevada, USA

Recently, an assemblage of 19 obsidian tools has come to light following the death of their original collector. This collection apparently resultedfrom his excavations in an unknown cave in 1974 and may have been found buried together as a cache and ritual offering. These unusually large tools correspond to the Parman variety of the Great Basin Stemmed cluster, which is associated with the terminal Pleistocene-early Holocene occupation of the Great Basin region of North America. The assemblage contains six Parman projectile points, four stemmed bifacial preforms, five bipointed preforms, two ovoid bifacial blanks, one small triangular biface, and a thick scraper. Trace element studies of these materials suggest origins from five different obsidian sources near the High Rock country of NW Nevada. Comparison of this collection with Paleoindian biface caches from outside the Great Basin suggests some interesting parallels that may imply shared cultural behaviors between these groups.