More than 10,000 nivation landforms occur in the higher mountain ranges of the western Great Basin. They range from small, subtle hollows with head scarps a few meters high and a few tens of meters long to broad, clearly defined terraces as much as 220 m wide bounded by bold, steeply sloping head scarps as much as 30 m high and 1600 m long. Distribution of these nivation hollows is strongly influenced by elevation, slope orientation, local relief, and substrate lithology. About 95% occur between 2200 and 3000 m elevation, and nearly 80% are situated on north-northwest-to east-northeast-facing slopes. They occur mainly in areas of moderately sloping terrain and moderate local relief, and they are preferentially developed on relatively incompetent substrates including terrigenous sedimentary deposits, volcanic and metavolcanic rocks of intermediate composition, and deeply weathered granitoid rocks. Nearly all of these nivation hollows are relict. They are most abundant near areas of late Pleistocene glaciation but rarely occur within such areas. Most are veneered with colluvium and are well vegetated, and many hollows in the Mono Basin area are veneered with volcanic ash at least 700 yr old. Distribution of nivation hollows suggests that (1) the full-glacial nivation threshold altitude (NTA) rose from north to south at 190 m per degree of latitude, subparallel to, and approximately 740 m lower than, the full-glacial equilibrium-line altitude (ELA) and about 1370 m lower than the estimated modern ELA; (2) the difference between the full-glacial and modern ELAs indicates an approximate 7°C full-glacial mean-annual-temperature depression throughout the Great Basin; and (3) the full-glacial mean annual temperature at the NTA is estimated to have been approximately 0° to 1°C, assuming little change in accumulation-season precipitation.

Nine distinct paleohydrologic regimes in the southwestern Great Basin over the last 3.2 my are recorded by the lacustrine deposits in KM-3, a 930-m core from Searles Lake, California. These are characterized as being “wet,” “intermediate,” or “dry” (like today). Excepting the present incomplete regime, each lasted 0.12 to 0.76 my. Major regime changes 0.01, 0.13, 0.6, and 2.5 my ago appear to coincide with recognized changes in global ice-sheet histories as represented by 18O and other records from marine sediments, but comparable changes 0.3, 1.0, 1.3, and 2.0 my ago do not appear to coincide closely with comparable perturbations in ice-sheet histories. However, all regime boundaries (during the last 1.75 my) coincide closely in time with changes in sea-surface temperatures in the tropical Atlantic, and many coincide with other deep-sea and continental paleoclimatic boundaries.The average duration of these paleohydrologic regimes was about 0.4 my (standard deviation, 0.2 my or less, depending on assumptions), and it is suggested that the regime boundaries reflect times of change in global(?) sea-surface temperatures, possibly controlled in part by the Earth's 413,000-yr orbital eccentricity cycle. During the wettest and driest regimes in the Searles Lake area, lake levels were not sufficiently affected by the 23,000-, to 42,000-, or 100,000-yr climate cycles related to high-latitude ice-sheet fluctuations to produce changes in the lacustrine sediment character. During intermediate regimes, however, when lacustrine sedimentation in this area was more sensitive to climate, the sediments, in KM-3, record lake fluctuations with average frequencies near those of the ice sheets. This seems to indicate that the high-latitude ice-sheet fluctuations caused local climatic perturbations but did not dominate the hydrologic component of climate in this area. Other lacustrine deposits in the southwestern Great Basin of California and Nevada have ages comparable in part to those of the wet to intermediate regimes indicated by KM-3, and they may all be products of finite periods when lake expansion, alluvial fan growth, increased spring discharge, and fluvial deposition were promoted in this area by widespread wet climates. Glacier expansion in the Sierra Nevada may also have been primarily an expression of, and in phase with, these wet regimes.

Cutting, fracturing, flaking, and polishing of bones and other osseous materials may in some instances be interpreted as evidence of former human activity. Such interpretations must avoid confusion with the wide variety of natural processes that alter bones. Reliable criteria are prerequisite to sound inferences based on bone, regardless of whether they have been redeposited or recovered from primary assemblages. Criteria must be defined by means of actualistic studies (neotaphonomy) and experiments that demonstrate causal relationships between patterns of alteration and the processes that produce them. The criteria can be empoyed in the interpretation of fossil bones (paleotaphonomy) on the basis of uniformitarian principles. At the present time, such relationships are relatively well understood in the case of cut marks on bones, and modest progress has been made in investigating fractured and flaked bones. Polished bones are little understood and difficult to interpret.

Late Wisconsinan age glacial landforms and deposits indicate that an ice shelf of at least 60,000 km2 flowed northwestward into Viscount Melville Sound, probably from the M'Clintock Dome of the Laurentide Ice Sheet. The ice shelf overlapped coastal areas and laid Winter Harbour Till up to 125 m above present sea level on the southern coast of Melville Island, to 135 m on Byam Martin Island, to possibly 90 m on the northeast tip of Banks Island, and to 150 m on the north coast of Victoria Island. The contemporary sea level was 50 to 100 m higher than present (it now rises eastward). A maximum age of 10,340 ± 150 yr B.P. for the till, and thus the ice-shelf advance, is provided by shells in marine sediments which underlie it, whereas a minimum age of 9880 ± 150 yr B.P. is provided by overlying shells that postdate the ice advance. The major advance of shelf ice into Viscount Melville Sound may be the result of the rapid disintegration of the M'Clintock Dome while the climate ameliorated in the western Arctic.

The Pleistocene outburst floods from glacial Lake Missoula, known as the “Spokane Floods”, released as much as 2184 km3 of water and produced the greatest known floods of the geologic past. A computer simulation model for these floods that is based on physical equations governing the enlargement by water flow of the tunnel penetrating the ice dam is described. The predicted maximum flood discharge lies in the range 2.74 × 106−13.7 × 106 m3 sec−1, lending independent glaciological support to paleohydrologic estimates of maximum discharge.

In 1982 we carried out a shallow-penetration, high-resolution seismic profiling study on the shelf of the southern Argolid, Peloponnesos, Greece, to identify and map the shores of the late Pleistocene and Holocene, and to establish their coastal environments. Portable equipment and a local boat were used. The lowest glacial shore occurs at −115 to −118 m, within the range of global values. The subsequent rise across the distinct old land surface left behind many shore features (scaps, beaches and beach ridges, channels and lagoons) now buried under a few meters of post-transgressive deposits. These features cluster at a small number of depths below present sea level, suggesting that the rise of the sea, usually too fast to leave an imprint, was episodically interrupted by brief stillstands or even temporary reversals of climatic or tectonic origin. The clusters can be roughly dated with reference to a global sea-level-rise curve; after 6000 yr B.P. sparse archaeological data establish a local curve. The seismic profiling technique, convenient and not costly, holds promise for the identification of postglacial shores elsewhere prior to sampling for dating. It has wide application for environmental reconstructions of vanished coastal zones as a basis for prehistoric resource assessments. These applications are illustrated with examples from this study.

Ages of fault scarps, as well as those other types of transport-limited slopes, can be estimated by comparing their morphology with the morphology of scarps of known age. Age estimates are derived by fitting the scarp profiles to synthetic profiles generated using a diffusion equation or, alternatively, by classification using a linear discriminant function. The usefulness of morphology-derived age estimates depends on the relative importance of non-age-related morphologic variation. Data from more than 200 scarp profiles demonstrate that morphologic variation not related to scarp age can introduce significant uncertainties into morphology-derived age estimates.

The annual freezeup and violent breakup of temperate and high-latitude rivers produces unique geomorphic features that can be recognized in ancient sediments. The forces and materials involved in the breakup process could modify entrained bones in ways similar to those attributed to human activity. An intuitive taphonomic model is developed to explain how river ice may affect bones, and to predict the modifications expected. Partial testing of the model by experiments designed to simulate river breakup produced bone modifications such as fractures, flakes, facets, striations, isolated cuts, and polish. The taphonomic effects of river ice must be discounted prior to the interpretation of such features as human in origin.

An unusually severe drought in 1982 led to a temporary die-off of elephants at a natural water source in Hwange National Park, Zimbabwe. Compared to living populations, the age structure of the animals killed by drought is strongly biased toward 2-to 8-year-old animals. However, the fresh carcasses of these young elephants were commingled with weathered remains of adults that had died earlier, creating a mixed skeletal sample whose age structure was much closer to that of living populations. Observations of elephant bones that have accumulated due to natural mortality at water holes might provide analogs for paleoecological interpretations of fossil proboscidean assemblages.

Newly examined exposures in northern Idaho and Washington show that catastrophic floods from glacial Lake Missoula during late Wisconsin time were repeated, brief jökulhlaups separated by decades of quiet glaciolacustrine and subaerial conditions. Glacial Priest Lake, dammed in the Priest River valley by a tongue of the Purcell trench lobe of the Cordilleran ice sheet, generally accumulated varved mud; the varved mud is sharply interrupted by 14 sand beds deposited by upvalley-running currents. The sand beds are texturally and structurally similar to slackwater sediment in valleys in southern Washington that were backflooded by outbursts from glacial Lake Missoula. Beds of varved mud also accumulated in glacial Lake Spokane (or Columbia?) in Latah Creek valley and elsewhere in northeastern Washington; the mud beds were disrupted, in places violently, during emplacement of each of 16 or more thick flood-gravel beds. This history corroborates evidence from southern Washington that only one graded bed is deposited per flood, refuting a conventional idea that many beds accumulated per flood. The total number of such floodlaid beds in stratigraphic succession near Spokane is at least 28. The mud beds between most of the floodlaid beds in these valleys each consist of between 20 and 55 silt-to-clay varves. Lacustrine environments in northern Idaho and Washington therefore persisted for two to six decades between regularly recurring, colossal floods from glacial Lake Missoula.

Glacial landforms in the drumlin drift belt of Ireland and the Alaska Range can be identified and mapped from Seasat synthetic-aperture radar (SAR) images. Drumlins cover 60% of the Ireland scene. The width/length ratio of individual drumlins can be measured on the SAR images, allowing regional differences in drumlin shape to be mapped. This cannot be done with corresponding Landsat multispectral scanner (MSS) images because of lower spatial resolution and because of shadowing effects that vary seasonally. The Alaska scene shows the extent and nature of morphological features such as medial and lateral moraines, stagnant ice, and fluted ground moraine in glaciated valleys. Perception of these features on corresponding Landsat MSS images is limited by seasonal differences in solar illumination. Because SAR is not affected by such differences or by cloud cover, it is particularly well suited for monitoring glacial movement. The disadvantage of distorted high-relief features on Seasat SAR images can be reduced in future SAR systems by modifying the radar illumination geometry.

Field investigations of caves along Alaska's Porcupine River document three major mechanisms which modify bone in patterns similar to alterations produced by man: (1) carnivore fracture; (2) rodent gnawing; and (3) rock fall and rubble scarring. A late Wisconsin faunal assemblage composed of Equus sp., Rangifer tarandus, Ovis dalli, Bison sp., proboscidean, numerous small mammal species, birds, and fish is well documented. This faunal assemblage suggests a mosaic environment of grassland-tundra-forest in the immediate vicinity of these caves and implies that the late Wisconsin environment in north-central Alaska may have been characterized by a number of microenvironments and colder, dryer, steppe conditions. Taphonomic data which have historically been interpreted to support human occupation of eastern Beringia during the Pleistocene are critically examined and the context of these discoveries (not the specimens themselves) provides the test essential to document the antiquity of man in North America prior to 12,000 yr ago.

The chemistry of water in the Cambrian-Ordovician aquifer in six midwestern states has been studied as part of the Northern Midwest Regional Aquifer-System Analysis of the U.S. Geological Survey. Dissolved-solids concentrations generally increase perpendicular to the direction of regional groundwater flow, from less than 400 mg/liter in southeast Minnesota, southwest Wisconsin, and northeast Iowa to more than 10,000 mg/liter in northwest Missouri. Isotopic ratios of hydrogen and oxygen are significantly depleted from north to south, with an areal distribution approximately parallel to the distribution of dissolved solids. For example, δ18O in southern Iowa and northern Missouri is about 6 parts per thousand lighter than δ18O of modern recharge water in Minnesota and Wisconsin. Covariance between δ18O and δD of the groundwater, similar to that of modern precipitation, suggests that the differences in isotopic ratios between groundwater and modern recharge water reflect meteoric signatures of water during past recharge events rather than geochemical processes such as isotopic exchange with aquifer materials. The pronounced parallelism between the distribution of isotopes and dissolved solids over large areas probably reflects largescale recharge of Pleistocene glacial meltwater into the aquifer system, which probably had a paleoflow system with a gradient from northeast to southwest rather than from northwest to southeast.

Radiocarbon-dated marine sediments from five coastal sites along the Strait of Magellan and Beagle Channel in southernmost Chile permit construction of a curve of relative sea-level fluctuations during the Holocene. Morphologic and stratigraphic data point to coastal submergence during the early Holocene as the sea rose to a maximum level at least 3.5 m higher than present about 5000 yr ago. Progressive emergence then followed during the late Holocene. Data from widely separated localities define a smooth curve, the form of which is explainable in terms of isostatic and hydroisostatic deformation of the crust resulting from changing ice and water loads. Apparently anomalous data from one site located more than 100 km behind the outer limit of the last glaciation may reflect isostatic response to deglaciation. The sea-level curve resembles one derived by Clark and Bloom (1979, In “Proceedings of the 1978 International Symposium on Coastal Evolution in the Quaternary, Sao Paulo, Brasil,” pp. 41–60. Sao Paulo) using a spherical Earth model, both in amplitude and in the timing of the maximum submergence.

Paleomagnetic analysis of samples from the original 600-m-long core from the San Agustin Plains, New Mexico, showed an unquestionable reversal stratigraphy dating the record back to at least 1.6 my. Analysis of pollen, ostracodes, and algae of a duplicate sample section in the vicinity of the original coring site and spanning the last 18,000 yr B.P. shows a much higher resolution than the earlier results.

Camelops was a major faunal element in late Wisconsin biotic communities over much of North America. Interpretations of possible human association with Camelops are often based on poorly evaluated evidence. Ideal standards for acceptable evidence are compared here to the actual evidence that has been advanced. Of 25 fossil assemblages examined, 2 might be examples only of geological contemporaneity of humans and Camelops; 2 might indicate behavioral association of humans and Camelops bones; and 2 might indicate actual human utilization of Camelops (killing and/or butchering). Camelops bones interpreted as artifacts are similar to modern specimens affected by noncultural processes.

Meltwater from a 299-m-long ice core was filtered and analyzed for fossil pollen and spores. Pollen concentration was higher in the late Holocene and interglacial intervals (ca. 7 liter−1) than in the early Holocene and Wisconsinan (ca. 1–2 liter−1) ones. The late Holocene and interglacial assemblages were dominated by Alnus (alder), whereas the early Holocene and Wisconsinan ones were dominated by Betula (birch) and Artemisia (sage). During the Holocene and probably the last interglaciation, most of the pollen and spores were blown a minimum of 1000 km from low arctic shrub tundra and adjacent subarctic Picea (spruce) forest; these areas were dominated by the arctic air mass during the summer pollinating season. During the Wisconsinan-early Holocene, glacier ice and arctic air were more widespread and pollen sources were more distant; thus, at this time relatively little pollen was incorporated into the ice.

The Devon ice-core data suggest that there should have been pollen in the continental ice sheet of Wisconsin time. When the ice sheet retreated this pollen would be carried by meltwater and redeposited with silt and clay together with contemporary pollen, producing an ecologically anomalous assemblage.

Pollen evidence from a 350-cm section of a fen in a moraine belt at Rucañancu (39°33′S, 72°18′W) bears on the controversy regarding interpretation of late-glacial and Holocene climate in midlatitude Chile. Earlier pollen studies, indicating a cooling trend between approximately 11,000 and 10,000 yr B.P., disagreed with observations of glacier fluctuations which show continuous glacier wastage and, by inference, warming after 12,500 yr B.P. and possibly earlier, up until Neoglaciation, beginning after 6850 yr B.P. Fossil beetle assemblage data in this time range support the interpretation of climate made from the observed glacier behavior. At Rucañancu, a pollen assemblage containing upper montane podocarp (Podocarpus andinus) in quantities reaching 34% and dating between 10,440 and 10,000 yr B.P. implies a cold climate with summer temperatures possibly 5–8°C lower than today's. Holocene warming began afterward, later than the glacier and beetle records indicate, and continued until at least 8350 yr B.P., as suggested by the sequence of assemblages dominated by Myrtaceae, by Aextoxicon punctatum, and by Gramineae. A subsequent assemblage of Nothofagus obliqua type implies an increase of moisture until 6960 yr B.P., following which N. dombeyi type, under a cool and humid Neoglacial climate, became dominant.

Thirteen packrat (Neotoma spp.) and two porcupine (Erethizon dorsatum) middens from 1555 to 1690 m elevation from the Sacramento Mountains, New Mexico, provide an 18,000-yr vegetation record in the northern Chiuahuan Desert. The vegetation sequence is a mesic, Wisconsin fullglacial (18,000–16,000 yr B.P.) pinyon-juniper-oak woodland; a xeric, early Holocene (ca. 11,000–8000 yr B.P.) juniper-oak woodland; a middle Holocene (ca. 8000-4000 yr B.P.) desert-grassland; and a late Holocene (ca. 4000 yr B.P. to present) Chihuahuan desertscrub. The frequency of spring freezes and summer droughts in the late Wisconsin probably set the northern limits of Pinus edulis and Juniperus monosperma at about 34°N, or 6° south of today's limit. Rising summer tempratures in the early Holocene eliminated pinyon and other mesic woodland plants from the desert lowlands and allowed the woodland to move upslope. At this time pinyon-juniper woodland and pine forest dominated by Pinus ponderosa probably began their spectacular Holocene expansions to the north. Continued warming in the middle Holocene led to very warm summers with strong monsoons, relatively dry, cold winters, and widespread desert-grasslands. Desertscrub communities in the northern Chihuahuan Desert did not develop until the late Holocene when the biseasonal rainfall shifted slightly back toward the winter, catastrophic winter freezes decreased, and droughts in all seasons increased. The creosote bush desertscrub corridor across the Continental Divide between the Chihuahuan and Sonoran deserts was probably connected for the first time since the last interglaciation.