Temperature profoundly influences the physiology and life history characteristics of organisms, particularly in terms of body size. Because so many critical parameters scale with body mass, long-term temperature fluctuations can have dramatic impacts. We examined the response of a small mammalian herbivore, the bushy-tailed woodrat ( Neotoma cinerea), to temperature change from 20,000 yr B.P. to present, at five sites within the Colorado Plateau. Our investigations focused on the relationship between temperature, plant composition and abundance, and woodrat size. Body size was estimated by measuring fossil fecal pellets, a technique validated in earlier work. We found significant and highly covariable patterns in body mass over the five locations, suggesting that responses to temperature fluctuations during the late Quaternary have been very similar. Although woodrat mass and the occurrence of several plant species in the fossil record weresignificantly correlated, in virtually all instances changes in woodrat size preceded changes in vegetational composition. These results may be due to the greater sensitivity of woodrats to temperature, or to the shorter generation times of woodrats as compared to most plants. An alternative hypothesis is that winter temperatures increased before summer ones. Woodrats are highly sensitive to warmer winters, whereas little response would be expected from forest/woodland plants growing at their lower limits. Our work suggests that woodrat size is a precise paleothermometer, yielding information about temperature variation over relatively short-term temporal and regional scales.

The beginning of the Bølling-Allerød warm period was marked in Greenland ice by an abrupt rise in δ18O, an abrupt drop in dust rain, and an abrupt increase in atmospheric methane content. The surface waters in the Norwegian Sea underwent a simultaneous abrupt warming. At about this time, a major change in the pattern of global rainfall occurred. Lake Victoria (latitude 0°), which prior to this time was dry, was rejuvenated. The Red Sea, which prior to this time was hypersaline, freshened. Lake Lahontan, which prior to this time had achieved its largest size, desiccated. Whereas the chronologic support for the abruptness of the hydrologic changes is firm only for the Red Sea, in keeping with evidence obtained well away from the northern Atlantic in the Santa Barbara Basin and the Cariaco Trench, the onset and end of the millennial-duration climate events were globally abrupt. If so, the proposed linkage between the size of African closed basin lakes and insolation cycles must be reexamined.

The S5 is the most prominent paleosol unit in the Chinese Loess Plateau, and is distinguished by its great thickness, dark color, and well-developed clay coatings. Previous studies led to the proposal that the S5 developed under much warmer and wetter conditions than at present, implying a “climatic optimum.” However, the S5 is a paleosol complex corresponding in time to a period of weak glacial–interglacial oscillations, as recorded by marine oxygen-isotope stages 15 through 13. To understand the character of this climatic interval, two profiles of the paleosol S5 in the Guanzhong basin were studied, with most of the emphasis being placed on the uppermost and best-developed pedon, S5-I. Grain-size, mineralogical, and chemical data indicate that the major pedogenic processes experienced by the S5-I included complete decalcification and mechanical translocation of fine particles. Chemical alteration of silicate minerals has been weak. Stable carbon-isotope analyses of organic matter showed that C4 plant remains are abundant throughout the profiles. These results suggest that grasslands dominated the paleovegetation during S5-I development. The prominent development of the S5 paleosol complex may reflect a prolonged period of pedogenesis. In contrast to previous reports, paleoclimatic conditions perhaps even drier than at present are implied by the data for the S5 profiles. The relative strength of the glacial–interglacial oscillation in this region (related to the East Asian monsoon) was similar to that indicated for global climate during S5 time.

Pollen analysis of two cores from the Lake Sinnda, located in one of the driest areas of the southern Congo, reveals a history of vegetation and climate in this region during the past 5000 yr. A major change centered around 3000–2500 yr B.P. is indicated by an abrupt decrease in forest pollen and by a corresponding increase in grassland pollen. Concurrent drying up of the lake shows that climate, in particular aridity, was the major cause of this change. This paleoclimatic reconstruction conforms with evidence for drier conditions in other parts of western equatorial Africa, such as the development of isolated enclosed savannas and of heliophilous forested formations. The aridity is recorded more fully at Lake Sinnda than at the previous studied ones. It probably lasted longer, from 4200 to 1300 yr B.P., and was more progressive than previously inferred. The aridity predates agriculture marked by pollen of the oil palm at Lake Sinnda.

The paleoclimatic record from bottom sediments of Lake Baikal (eastern Siberia) reveals new evidence for an abrupt and intense glaciation during the initial part of the last interglacial period (isotope substage 5d). This glaciation lasted about 12,000 yr from 117,000 to 105,000 yr B.P. according to correlation with the SPECMAP isotope chronology. Lithological and biogeochemical evidence of glaciation from Lake Baikal agrees with evidence for the advance of ice sheet in northwestern Siberia during this time period and also with cryogenic features within the strata of Kazantzevo soils in Southern Siberia. The severe 5d glaciation in Siberia was caused by dramatic cooling due to the decrease in solar insolation (as predicted by the model of insolation changes for northern Asia according to Milankovich theory) coupled with western atmospheric transport of moisture from the open areas of Northern Atlantic and Arctic seas (which became ice-free due to the intense warming during preceeding isotope substage 5e). Other marine and continental records show evidence for cooling during 5d, but not for intense glaciation. Late Pleistocene glaciations in the Northern Hemisphere may have begun in northwestern Siberia.

Archaeological sites at Abeurador and Font-Juvénal have produced extensive charcoal-rich horizons. The results of the charcoal analysis, based on the identification of woody species from anatomical features, contribute to the elaboration of plain phases within the development of vegetation in southern France. They are presented here in a new diagram and are interpreted as four major vegetation phases from 11,00014C yr B.P. to the recent past. In order to interpret the palaeoecological significance of the charcoal assemblages we have used multivariate analysis. The data base consists of 42 taxa and 48 archaeological levels. The division into four vegetation phases based on analytical interpretation is complemented by the more synthetic interpretation based on correspondence factor analysis (CFA). The results reveal the patterns of the vegetation history over the last 11 millennia, each being characterized by a key plant species.

Diatoms, magnetic susceptibility, organic content, and14C ages of sediments from a 26-m core suggest that Lake Chalco, in the southern part of the basin of Mexico, went through a series of major fluctuations during the late Pleistocene and the Holocene. Before ca. 39,00014C yr B.P. the lake was very deep (about 8–10 m), alkaline, and saline. It then became shallow (<2 m) for most of the time between ca. 39,000 and 22,500 yr B.P. Chalco deepened to about 4–5 m about the time of a major eruption of nearby Popocatepetl volcano ca. 22,000 yr B.P. The lake remained relatively deep and fresh until ca. 18,500 yr B.P., when lower levels and alternating acidic to freshwater conditions were established. After 14,500 yr B.P. lake level rose slightly, but by ca. 10,000 yr B.P. Chalco became very shallow (<2 m), remaining as a low, alkaline saline marsh until ca. 6000 yr B.P. This period corresponds with the Playa cultural phase, during which the earliest human settlements in the basin were established. After ca. 6000 yr B.P. Chalco became a fresh to slightly alkaline shallow lake a few meters deep.

The megafaunal rodent Amblyrhiza inundatafrom Anguilla and St. Martin is often cited in lists of late Quaternary human-induced extinctions, but its date of disappearance has never been established. Here, we present a suite of uranium-series disequilibrium dates from three independent Amblyrhizasites in Anguilla, all of which cluster in marine isotope Stage 5. Thus, there is no indication that Amblyrhizasurvived into the late Holocene, when islands of the northern Lesser Antilles were first invaded by humans. We argue that the most probable cause of the extinction of Amblyrhizawas a failure of island populations to adjust to catastrophic reductions in available range which accompanied last interglacial sea-level maxima. We support this argument with quantitative extinction probability estimates drawn from persistence time models. Amblyrhizaexhibits body-size hypervariability, a common but underemphasized feature of island megafaunal species. We argue that hypervariability is a record of morphological response to oscillating natural selection, which in turn is driven by asymmetries in the relationship of population size, body mass, and persistence time. The fate of Amblyrhizastands in marked contrast to that of most other West Indian land mammals, whose losses increasingly appear to have been anthropogenically mediated.

In order to aid palaeoenvironmental research of Late Pleistocene and Holocene deposits of central Mexico, tephra layers collected from the sediments of the Texcoco and Chalco sub-basins, in the southern part of the Basin of Mexico, are geochemically characterized and used as stratigraphic markers. The tephra layers range in composition from basaltic andesites to rhyolites and are calc-alkaline. The tephras range in age from >34,000 to ca. 260014C yr B.P. New names are used informally to designate correlated tephras. The Tlahuac tephra is present in Chalco, at a depth of 18 m; in the southeastern part of Texcoco, at a depth of around 10 m; and at the Tlapacoya archaeological site, where it had been mistakenly described as basaltic. This basalt–andesite tephra is dated to at least 34,00014C yr B.P. The Tlapacoya 1 tephra is dated to between 15,020 ± 450 and 14,430 ± 190 yr B.P. and is present in all Chalco sections. The Tlapacoya 2 tephra corresponds to the previously described “pomez gruesa con fragmentos de andesita” (ca. 14,400 yr B.P.) and is present in all Chalco and Texcoco sections. The likely source of these three tephras is the volcano Popocatepetl. Tephra II at Chalco dates to 12,520 ± 135 yr B.P. and correlates with the Upper Toluca Pumice from Nevado de Toluca volcano. These represent the first geochemical glass-shard analysis of tephras from the Basin of Mexico, and so further research is necessary before a reliable tephrochronology can be established.

A controversy currently exists regarding the number of Toba eruptive events represented in the tephra occurrences across peninsular India. Some claim the presence of a single bed, the 75,000-yr-old Toba tephra; others argue that dating and archaeological evidence suggest the presence of earlier Toba tephra. Resolution of this issue was sought through detailed geochemical analyses of a comprehensive suite of samples, allowing comparison of the Indian samples to those from the Toba caldera in northern Sumatra, Malaysia, and importantly, the sedimentary core at ODP Site 758 in the Indian Ocean—a core that contains several of the earlier Toba tephra beds. In addition, two samples of Toba tephra from western India were dated by the fission-track method. The results unequivocally demonstrate that all the presently known Toba tephra occurrences in peninsular India belong to the 75,000 yr B.P. Toba eruption. Hence, this tephra bed can be used as an effective tool in the correlation and dating of late Quaternary sedimentary sequences across India and it can no longer be used in support of a middle Pleistocene age for associated Acheulian artifacts.