Pollen analysis of a new core from Joe Lake indicates that the late Quaternary vegetation of northwestern Alaska was characterized by four tundra and two forest-tundra types. These vegetation types were differentiated by combining quantitative comparisons of fossil and modern pollen assemblages with traditional, qualitative approaches for inferring past vegetation, such as the use of indicator species. Although imprecisely dated, the core probably spans at least the past 40,000 yr. A graminoid-Salix tundra dominated during the later and early portions of the glacial record. The middle glacial interval and the transition from glacial to interglacial conditions are characterized by a graminoid-Betula-Salix tundra. A Populus forest-Betula shrub tundra existed during the middle potion of this transition, being replaced in the early Holocene by a Betula-Alnus shrub tundra. The modern Picea forest-shrub tundra was established by the middle Holocene. These results suggest that the composition of modem tundra communities in northwestern Alaska developed relatively recently and that throughout much of the late Quaternary, tundra communities were unlike the predominant types found today in northern North America. Although descriptions of vegetation variations within the tundra will always be restricted by the innate taxonomic limitations of their herb-dominated pollen spectra, the application of multiple interpretive approaches improves the ability to reconstruct the historical development of this vegetation type.

Pollen, plant macrofossil, and radiocarbon-dating studies of seven exposures of fluvial sediments in the Tunica Hills region of southeastern Louisiana and southwestern Mississippi provide new information on late Wisconsinan vegetation, flora, and environment of the region. The assemblages date between 25,250 and 17,530 yr B.P. Pollen and macrofossil assemblages are dominated by Picea, which comprises 40-70% of the pollen assemblages. Abies and Larix pollen and macrofossils are absent, in contrast to sites to the north in the central Mississippi Valley. Deciduous hardwoods (Quercus, Fagus, Fraxinus, Carya, Juglans nigra, Acer, Ulmus) are minor components of both pollen and macrofossil assemblages. Radiocarbon dates of Picea and Quercus wood indicate that these two genera grew contemporaneously in the region. Regional upland forests were dominated by Picea. Picea cones and cone fragments are not typical of any extant North American species, and probably represent either an extinct species or an extinct variety or subspecies of Picea glauca. Late Wisconsinan climate of the region was cooler than present, but not necessarily as cool as implied by P. glauca or other "boreal" taxa.

Major, trace, and rare earth element analyses of volcanic glass are used separately or in combination for correlating Quaternary tephras, often by graphical or simple comparative methods. We have taken a statistical approach using discriminant function analysis (DFA) to assess the relative discriminating power of the different elements in volcanic glasses from several tectonovolcanic provinces. We found that major oxides are powerful discriminating variables for widespread tephras from the Taupo Volcanic Zone in New Zealand and here they can be more discriminating than trace elements. A wide selection of tephras from the western United States can also be distinguished on major oxides alone, particularly those from Cascade Range volcanoes. For tephras from large intracontinental calderas, such as Long Valley or Yellowstone, REE and trace elements are more effective at discriminating than major oxides. However, tephras erupted from the Long Valley area can be distinguished on major oxide composition by DFA, despite their similar chemistry. The selection and relative significance of different elements for discriminating tephras depends on the total data set being compared, as well as the source volcano and the individual eruptive events. Caution must be exercised in the nonstatistical selection of compositional data for characterizing tephras: DFA is a more powerful and objective tool for the comparison of tephra chemistry.

A radiocarbon-dated pollen record near the Boylston Street Fishweir site in Boston; Massachusetts, provides a regional and local record of vegetation changes from the middle Holocene to present. The stratigraphy begins about 5630 ± 90 yr B.P. with a marine transgression and is continuous up to the historic backfilling of the Back Bay area about 100 yr B.P. When pollen began accumulating at the site, the immediate area resembled the swamp forests growing today in southern New England. Fresh- and brackish-water vegetation was present before the area near the site was submerged. While these vegetation changes occurred locally, oak forest grew in the region. Correlation of this stratigraphy with archaeological data collected from 500 Boylston Street indicates that between 4700 and 3700 yr B.P., a number of fence-like alignments ("weirs"), were probably placed within existing channels and/or along shorelines to capture fish and other marine animals as they moved with the tidal flow.

Stable carbon isotope ratios of total organic matter were measured in two cores collected from the Kashiru peatbog in Burundi, Equatorial Africa. The record, which spans at least the last 40,000 yr, documents the C3-C4 biomass balance in the organic sediment. Among the major modern peat formers, most plants are C3 species and are characterized by δ13C values of -25.5 ± 2.3% (vs PDB). The C4 plants, which are characterized by higher δ13C values (-11.3 ± 0.7%) belong to the Gramineae (Miscanthidium sp.) and Cyperaceae families (Cyperus latifolius, C. papyrus, Pycreus nigricans). In the fossil record, δ13C values of total organic matter vary between -28 and -15% in response to the relative fluxes of C3 and C4 plants. Before 30,000 yr B.P., low δ13C values (-23.5 ± 1.1%) match high arboreal pollen contents. From 30,000 to 15,000 yr B.P., higher δ13C values (-17.6 ± 1.1%) correspond to a significant increase in percentages of grass pollen. During this episode, a short and sharp shift toward lighter carbon isotopic compositions at 21,000 yr B.P. is synchronous with higher input of arboreal pollen. From 15,000 to 12,000 yr B.P., the 13C content decreases (δ13C = -22.9 ± 1.4%). This shift, which cannot be explained by an increase in the arboreal vegetation, could be explained by the spreading of C3 Gramineae or C3 Cyperaceae. The interval from 12,000 to 7000 yr B.P. is poorly documented in these cores due to much lower organic matter accumulation. Low δ13C values (δ13C = -25.2 ± 1.3%) are observed from 7000 to 5000 yr B.P., when the pollen data show development of C3 mountain forest. The Late Holocene is characterized by a mixed C3-C4 organic matter accumulation (δ13C = -20.9 ± 1.6%). This study depicts a change in the dominant photosynthetic pathway among the herbaceous components, notably at the glacial-interglacial transition, when C3 plants were favored by increased water supply and/or higher atmospheric CO2 concentration.

Using samples from a prehistoric quarry site in the Mojave Desert, we tested and compared the two principal methods of rock-varnish cation-ratio dating, analysis of rock-varnish scrapings (R. I. Dorn, 1983, Quaternary Research 20, 49-73 and 1989, Physical Geography 13, 559-596) and analysis of rock varnish in situ (C. D. Harrington and J. W. Whitney, 1987, Geology 15, 967-970). Because we found no consistent relationship between the varnish cation ratio (K + Ca)/Ti, and the relative age of the varnish, neither method could be used to "cation-ratio date" the underlying chert artifacts. Moreover, in situ analyses yielded systematically higher cation ratios and lower Ti abundances than bulk analyses of scraped varnish or microprobe analyses of varnish in cross section. Our results, when considered along with other recently published evidence, indicate that these two methods of varnish chemical analysis and cation-ratio dating may not be comparable, nor consistently reliable, chronometers.

The middle to late Holocene history and early Anglo-European settlement impacts on Santa Rosa Island, California, were studied through the analysis of sediments in a small estuarine marsh. A 5.4-m-long sediment core produced a stratigraphic and pollen record spanning the last 5200 yr. Three major zones are distinguishable in the core. The lowermost zone (5200 to 3250 yr B.P.) represents a time of arid climate with predominantly marine sediment input and high Chenopodiaceae and Ambrosia pollen values. The intermediate zone (3250 yr B.P. to 1800 A.D.) is characterized by greater fresh water input and high values for Asteraceae and Cyperaceae pollen and charcoal particles. The uppermost zone (1800 A.D. to present) documents the unprecedented erosion, sedimentation, and vegetation change that resulted from the introduction of large exotic herbivores and exotic plants to the island during Anglo-European settlement. The identification of pollen grains of Torrey Pine (Pinus torreyana) documents the persistence of this endemic species on the island throughout the middle to late Holocene.

The Banyoles lacustrine sequence shows that the vegetational history of the northeastern Iberian Peninsula during the last 30,000 yr follows the North Atlantic pattern of climatic oscillations. The Banyoles pollen diagram, supported by two calibrated 14C dates and nine U/Th dates, shows a clear interstadial event between 30,000 and 27,000 yr B.P., a Pleniglacial period with minor oscillations that ended abruptly ca. 14,420 ± 410 yr B.P., and a late-glacial sequence that records the classical stages described in Northern Europe: the Bølling-Allerød Interstade, the Younger Dryas event at 12,000 yr B.P. (U-series age), and a short warming phase between the Younger Dryas and the last cold event (dated at 11,000 yr B.P., U-series age).

The Miami site, excavated in 1937, is in a small "playa" basin on the High Plains surface. The site is one of the earliest documented co-occurrences of Clovis points and mammoth. Reinvestigation of the site and related collections was undertaken to better understand the stratigraphy, geochronology, and archaeology. The basin, 23 m diameter × 1.6 m deep, filled with (1) dark gray silty clay, and (2) near the top of the section, a lens of well-sorted silt or loess. The basin started to fill ca. 13,700 yr B.P., the loess dates to ca. 11,400 yr B.P., and the bone bed probably dates to ca. 11,400-10,500 yr B.P. The loess may be the local manifestation of a "Clovis drought." The partial remains of five mammoths (three adults and two juveniles) were recovered in 1937; no other animal remains are known. The bone is heavily weathered and there are no clear indications of human modification. Artifacts found at the site include three Clovis points and a scraper found among the bones and two flakes and a scraper found on the surface near the playa. The origins of the bone and stone assemblage are uncertain but four scenarios are offered: a successful mammoth kill, an unsuccessful kill with wounded animals dying at the watering hole, opportunistic scavenging following natural deaths, or a palimpsest of multiple deaths following both natural and human causes.

Fossil remains of beetles and oribatid mites from a peat deposit dated 18,870 ± 300 yr B.P. near Verona, northeastern Italy, represent the first insect fauna of its kind from the last glacial maximum to be described from Italy. The assemblage includes the ground beetle Amara alpina, whose distribution today in Europe is restricted to mountains in Scandinavia and Scotland. Ecological requirements and geographic distribution of recent populations of the identified species suggest mesic habitats with standing water and peat bogs during the glacial maximum. The paleoenvironment was comparable to present-day lowland moors in Scandinavia or mesic environments above 1000 m altitude on the southern slopes of the Alps. The climate is inferred to have been colder and wetter than today. Mean July temperature may have been 8-9°C lower than at present.

During the past 70,000 years, Israeli carnivore faunas have varied only slightly in richness (number of species), but widely in diversity (including the relative abundances of the different species). Human population growth, sedentism, and the beginning of agriculture are often considered to have caused habitat destruction toward the end of the Pleistocene that may have affected animal populations. However, patterns of changing carnivore diversities from three sites in the Mediterranean region of Israel correlate well with changes in precipitation: species diversity is higher during more humid phases and lower during drier phases. The environmental effect of sedentism may have accelerated the later drop in carnivore diversity but cannot account for fluctuations in diversity preceding it.

A total of 93 plant taxa were identified from 11 packrat (Neotoma sp.) midden samples from the Sierra Bacha on the coast of the Gulf of California near Puerto Libertad, Sonora, Mexico. Nine indurated samples have radiocarbon dates ranging from 9970 to 320 yr B.P. Sonoran desertscrub was present on rocky slopes throughout the Holocene. Early Holocene assemblages dominated by Fouquieria columnaris (boojum tree) reflect vegetation and climate more like modern Baja California with greater winter rainfall and cooler summers. Middle Holocene vegetation was essentially modern with modest indications of greater monsoonal rainfall even though cold-water upwelling locally inhibits summer precipitation. The results are similar to all previous midden reconstructions of early and middle Holocene climates in the Sonoran Desert, but contradict general atmospheric circulation model simulations.

Stable carbon isotope analysis of organic carbon in alluvial deposits and soils of three streams in central Texas reveals significant shifts in the ratio of C3 to C4 plant biomass production during the past 15,000 yr. These temporal changes in vegetation appear to be in response to changes in climate. During the late Pleistocene, C4 plants comprised only about 45 to 50% of the vegetative biomass in this area, suggesting that conditions were cooler and wetter than at any time during the past 15,000 yr. The time between 11,000 and 8000 yr B.P. is interpreted as transitional between late Pleistocene conditions and warmer and drier Holocene conditions based on a slight increase in the abundance of C4 plant biomass. During the middle Holocene, between approximately 6000 and 5000 yr B.P., mixed C3/C4 plant communities were replaced almost completely by C4-dominated communities, indicating prairie expansion and warmer and drier climatic conditions. By 4000 yr B.P., the abundance of C4 plant biomass decreased to levels similar to the early Holocene transitional period, suggesting a return to cooler and wetter climatic conditions. No significant shift in the ratio of C3 to C4 productivity has occurred during the last 4000 yr, except for a slight increase in the abundance of C4 plant biomass around 2000 yr B.P. The results of this investigation correlate well with other regional late Quaternary climatic interpretations for central and north Texas, the Southern Plains region, and with other portions of the Great Plains.

A fossil assemblage containing molluscs, mammals, insects, ostracodes, and plants has been recovered from a silt-filled depression near Lima, in west-central Illinois. The reversed remanent magnetic signature of the sediments and the temporal ranges of two mammals, Microtus paroperarius and Lasiopodomys deceitensis, constrain the age of the assemblage to between 730,000 and 830,000 yr B.P. The extent of isoleucine epimerization in the molluscan shell is consistent with this age interpretation. The fauna includes at least 43 taxa of beetles from 11 families, 35 nominal species of molluscs, and two genera of ostracodes. The mammals include two shrews, three rodents, and a rabbit. The plant macrofossils (no pollen recovered) include 25 species of seed plants and four kinds of terrestrial or wetland mosses. Most of the plant species identified still occur in the upper Midwest, although a few of the taxa are found mainly to the north of the site. The fauna is characterized by an almost total absence of true aquatic taxa. The association of both boreal and thermophilous faunal and floral elements suggest that summer temperatures were not greatly different from present ones, but cooler, moist areas must have been available to support the boreal elements. Local conditions were probably similar to those now found in northeastern Iowa, where rains blocks, fissures, and joints in carbonate bedrock serve as traps for debris accumulations, provide shade, and are kept cool and moist during the hot summer months by cold-air drainage and groundwater seepage. Summer mean temperature in these microhabitats was probably between 18 and 20°C, similar to temperatures that now occur near the northern hardwood spruce-fir transition in the eastern United States.

A 5.4-m sequence of peat and marl overlying a basal clay in a northern Belize wetland was studied to assess salinity changes over the past 7000 yr. The distribution of ostracods, gastropods, and foraminifers revealed initially freshwater conditions in a terrestrial wetland, changing to at least mesohaline conditions by about 5600 yr B.P. The mesohaline conditions corresponded to the formation of an open-water lagoon (and precipitation of a lacustrine marl) that was contemporaneous with rapidly rising sea level in the area. A mangrove peat filled the lagoon by 4800 yr B.P. probably as a result of increasingly shallow waters as sea level rise slowed and marl precipitation continued. A new lagoon began to form sometime after 3400 yr B.P. Freshwater ostracods and gastropods found in the marl of this lagoon suggest that it formed under near-limnetic conditions. Freshwater input likely resulted from massive deforestation by the Maya that began by 4400 yr B.P. Subsidence of the mangrove peat likely permitted the formation of a lagoon. A peat has filled the lagoon since at least 500 yr B.P.

Fossiliferous sediments from The Pit Locality in Porcupine Cave, Park County, Colorado, span multiple glacial and interglacial cycles between ca. 800,000 and 400,000 yr B.P. at 2900 m elevation. The uppermost cycle probably represents oxygenisotope glacial Stage 12 (stratigraphic Levels 4 and 3) and interglacial Stage 11 (Levels 3, 2, and 1) as determined by arvicoline biostratigraphic data that place the sequence between 487,000 and 365,000 yr B.P. A no-analog assemblage comprising the bog lemming Mictomys meltoni (of the M. borealis lineage) and the sagebrush vole Lemmiscus curtatus (=Lagurus curtatus) characterizes both the glacial and the interglacial deposits. Climatographs for these two species suggest that, compared to present conditions, the following local climatic limits can be imposed on oxygenisotope Stages 12 and 11: (1) January effective moisture was greater, resulting from average temperatures of at least 5°C cooler and/or at least 15 mm more precipitation; (2) January solar radiation was lower by at least 75 Langleys; (3) effective July solar radiation might have been lower by at least 50 Langleys. Seasonal temperature extremes were not necessarily different from those of the present, but seasonal precipitation was more equable. Relative abundance of mammal taxa and sedimentological data indicate that aridity increased at the onset of the interglacial.

An assemblage of micromammals, recovered from the Holocene levels of a rockshelter at 2400 m in the montane forest of the Mau Escarpment, were examined with the goal of testing and contributing to prior reconstructions of paleoenvironments in the Central Rift Valley of Kenya. Species representation in the assemblage is consistent with a drying of the Rift Valley lakes in the middle Holocene and suggests a decrease in forest accompanied by expanding grasslands near the site. Changes in the abundance of grassland species suggests an increase in the frequency of fires, probably the result of pastoral burning. The body size of the root rat (Tachyoryctes splendens) decreases from the early Holocene to the middle Holocene, and this may indicate increasing aridity or increasing temperature. We compare measures of species diversity (number of taxa, species richness, and the Shannon diversity index) for both micromammals and macromammals since species diversity may change with paleoenvironmental change. The macromammals show no changes in species diversity that are assignable to paleoenvironmental change, while the micromammals show a trend toward decreasing diversity from the early to middle Holocene, and then show an increase in diversity during the peak of the middle Holocene dry phase, though sample size effects may be confounding the patterning.