Radiocarbon dates of plant materials from packrat middens in caves below the elevation of the last high stand of Pleistocene Lake Lahontan, in conjunction with radiocarbon dates of ancient archaeological materials, provide evidence that the last high stand terminated before 12,070 yr B.P. This new information suggests that the last major fluctuation in the level of Lake Lahontan was approximately synchronous with those of Lakes Bonneville and Russell, and that these three lakes probably responded contemporaneously to the same regional paleoclimatic changes.

Reconstructions of Plio-Pleistocene habitats and lake levels from a brief interval (∼200,000 yr) of hominid-bearing sediments from Bed I and lower Bed II, Olduvai Gorge, Tanzania, are for the first time integrated with climatic data from the deep-sea in order to examine the effect of largescale climatic events in East Africa. The period of densely vegetated habitats and wet conditions at Olduval at ∼1.82 myr shows a best fit with a period of northern hemisphere warming as evidenced by oxygen isotopes from deep-sea core V28-239 from the western Pacific. This environmental response is similar to that witnessed in East Africa during the period of warming climates associated with deglaciation at ∼10,000–8,000 yr.

An estimate of the sliding velocity and basal meltwater discharge of the Puget lobe of the Cordilleran ice sheet can be calculated from its reconstructed extent, altitude, and mass balance. Lobe dimensions and surface altitudes are inferred from ice limits and flow-direction indicators. Net annual mass balance and total ablation are calculated from relations empirically derived from modern maritime glaciers. An equilibrium-line altitude between 1200 and 1250 m is calculated for the maximum glacial advance (ca. 15,000 yr B.P.) during the Vashon Stade of the Fraser Glaciation. This estimate is in accord with geologic data and is insensitive to plausible variability in the parameters used in the reconstruction. Resultant sliding velocities are as much as 650 m/a at the equilibrium line, decreasing both up- and downglacier. Such velocities for an ice sheet of this size are consistent with nonsurging behavior. Average meltwater discharge increases monotonically downglacier to 3000 m3/sec at the terminus and is of a comparable magnitude to ice discharge over much of the glacier's ablation area. Paleoclimatic inferences derived from this reconstruction are consistent with previous, independently derived studies of late Pleistocene temperature and precipitation in the Pacific Northwest.

Historical observations of fjord glaciers in southern Alaska suggest that their fluctuations may be asynchronous with one another and with climate. Possible causes for anomalous fjord glacier response to climatic changes involve the instabilities inherent in a calving terminus, the interaction of fjord geometry with the extent of the calving terminus, and the intersection of low-lying snow line by a thickening glacier. The Lituya glacier system in southeast Alaska is analyzed with reference to these phenomena to test the reliability of its Holocene glacial deposits as paleoclimatic records. This fjord glacier is shown to have been probably free of major, nonclimatic responses during the Holocene except for an insensitive period of up to 1000 yr that followed calving retreats. The insensitive period has no functional effect on how the relatively coarse geologic record is interpreted in terms of paleoclimate. However, it makes impossible any inference of climatic trends from this glacier's behavior in historical times. Fjord glacier deposits can be dependable records of paleoclimate but each glacier requires careful scrutiny. Anomalous glacier responses to climate are most likely to occur in fjords of complex geometry, in regions of relatively low snow line, and where glaciers terminate at fjord mouths bordering the open sea.

A new quantitative, reproducible method for determining relative ages of unconsolidated Quaternary deposits containing granitic clasts has been developed and tested. The technique makes use of a microseismic timer to determine the compressional wave velocity (clast-sound velocity) in each clast of a group chosen from a single deposit. From these data a group mean velocity is determined that is proportional to the age of the deposit: the youngest deposits having the highest velocities. The clast-sound velocity method was used to study Quaternary deposits in the San Gabriel Valley and San Gabriel Mountains, California, using a previously proposed four-part age classification for the deposits. The clast-sound velocity group means for the four age groups were found to be statistically separable at better than 99% confidence. A velocity/age correlation curve was determined for these deposits using two radiocarbon dates and one paleomagnetic determination. This curve suggests that the clast-sound velocity method may be used to determine ages of deposits up to one million years old when calibrated with sufficient radiometric dates and may also be used as a tool for correlating undated deposits.

Annual pollen deposition (pollen cm−2 yr−1) was sampled for 6 yr in the southern Sierra Nevada, the Owens Valley, and the Inyo Mountains of central California. Spatial pollen distributions were examined for anemophilous plants which are limited to growth sites (1) at high elevations on one or both mountain ranges, (2) on lower mountain slopes, and (3) on the Owens Valley floor, Pollen deposition values for high- and low-elevation plants changed by a factor of 5 to 10 between adjacent sample sites which straddle boundaries of the plant communities in which the source plants grow. Pollen deposition values were high and variable within communities in which the species grew. Pollen deposition beyond source populations was low and uniform despite great differences in distance of sites from pollen sources. The step function that seems to characterize pollen distribution patterns implies that pollen from distant populations may be of little value in interpreting vegetation stability and change reflected by fossil pollen stratigraphy. The physical isolation of pollen embedded in the local mountain-valley wind system from that in the prevailing westerlies may explain part of the spatial distribution of the pollen.

Stream channels in the Paria River basin were eroded and partially refilled between 1883 and 1980. Basin-wide erosion began in 1883; channels were fully entrenched and widened by 1890. This erosion occurred during the well-documented period of arroyo cutting in the Southwest. Photographs of the Paria River channel taken between 1918 and 1940 show that the channel did not have a floodplain and remained wide and deep until the early 1940s. A thin bar (<50 cm), now reworked and locally preserved, was deposited at that time. Basin-wide aggradation, which began in the early 1940s, developed floodplains by vertical accretion. The floodplain alluvium, 1.3–3 m thick. consists of two units recognizable throughout the studied area. An older unit was deposited during a time of low flow and sediment yield whereas the younger unit was deposited during times of high flow, sediment yield, and precipitation. Tree-ring dating suggests that the older unit was deposited between the early 1940s and 1956, and the younger between 1956 and 1980. The units are not time transgressive, suggesting that deposition by knickpoint recession was not an important process. High peak-flood discharges were associated with crosion and low flood discharges with aggradation. The erosional or aggradational mode of the streams was determined principally by peak-flood discharge, which in turn was controlled by precipitation.

The sequence of glacial deposits near McCall, Idaho, previously assigned to the Pinedale and Bull Lake glaciations, contains deposits of four different ages. These ages are defined by multiple relative-age criteria, including weathering rinds, soil development, surface-rock weathering, morainal morphology, and loess stratigraphy. The thickness of weathering rinds on basaltic clasts is statistically representative and reproducible and can be used to estimate numerical ages. Following in order of decreasing relation to age are soil development, surface-rock weathering, and moraine morphology. The glacial deposits near McCall appear to correspond to times of high worldwide ice volume indicated by the marine oxygen-isotope record. Pilgrim Cove and McCall deposits, both assigned to the Pinedale glaciation, are late Wisconsin in age, perhaps 14,000 and 20,000 years, respectively. They represent a rare case in which deposits of Pinedale age can be separated by relative-age data. Timber Ridge deposits, assigned to the Bull Lake glaciation, have subdued, but well-preserved morainal morphology; relative-age data indicate that they are pre-Wisconsin in age, probably about 140,000–150,000 years old, although we cannot exclude an older age. Williams Creek deposits are clearly distinct from, and intermediate in age between, McCall and Timber Ridge deposits. Weathering rinds and the inferred ages of the other deposits suggest an early Wisconsin age for Williams Creek deposits.

Latest Pleistocene-Holocene basaltic lava fields of the Snake River Plain, Idaho, have been dated by the radiocarbon method. Backhoe excavations beneath lava flows typically yielded carbon-bearing, charred eolian sediment. This material provided most of the samples for this study; the sediment typically contains less than 0.2% carbon. Charcoal fragments were obtained from tree molds but only from a few backhoe excavations. Contamination of the charred sediments and charcoal by younger carbon components is extensive; the effects of contamination were mitigated but appropriate pretreatment of samples using acid and alkali leaches. Twenty of the more than 60 lava flows of the Craters of the Moon lava field have been dated; their ages range from about 15,000 to about 2000 yr B.P. The ages permit assignment of the flows to eight distinct eruptive periods with an average recurrence interval of about 2000 yr. The seven other latest Pleistocene-Holocene lava fields were all emplaced in short eruptive bursts. Their 14C ages (yr B.P.) are: Kings Bowl (2222± 100), Wapi (2270 ± 50), Hells Half Acre (5200 ± 150), Shoshone (10,130 ± 350), North Robbers and South Robbers (11.980 ± 300), and Cerro Grande (13,380 ± 350).

Glacial Lake Devlin was an ice-margin lake that formed during Pinedale time when the North Boulder Creek valley glacier blocked the mouth of a tributary near the downvalley limit of glaciation. The lake was about 2 km long and as much as 60 m deep, and was fed by meltwater from small valley glaciers. It accumulated sediment, apparently without interruption, for about 10,000 yr. The inception of Lake Devlin between 23,500 and 21,000 yr B.P. provides a limiting date for the maximum expansion of Pinedale ice, because the lake could not form until the North Boulder Creek glacier had advanced to within 2 km of its downvalley limit. The maximum lowering of timberline, which occurred about 19,000 yr B.P., is assumed to coincide with the Pinedale glacial maximum. Apparent timberline elevations determined from Picea/Pinus pollen ratios suggest that ice-front positions during the time between 23,000 and 19,000 yr B.P. may have been only slightly less extensive than at the Pinedale maximum. Lake Devlin drained catastrophically about 13,000 yr B.P., which was after ice had receded from the outlet area. Pinedale glaciation in the Front Range began about or after 30,000 yr B.P. and final deglaciation occurred between 15,000 and 12,000 yr B.P.

During recent humid episodes, stromatolites were built along paleolake margins, some 60 m above the modern water level of Lakes Natron and Magadi (southern Gregory Rift Valley). Three generations of stromatolites are observed, the more recent ones frequently covering pebbles and boulders eroded from the older ones. The youngest one yielded 14C ages ranging from approximately 12,000 to 10,000 yr B.P. Their δ13C values (≥2.6%) suggest isotopic equilibrium between the paleolake total inorganic dissolved carbon and the atmospheric CO2, thereby lending credence to the reliability of the 14C. An initial 230Th/232Th ratio in the detrital component was determined by Th/U measurements on the 14C dated stromatolites. Using this value a 230Th/234U chronology for the older stromatolites was calculated. Ages of ≥240,000 and 135,000 ± 10,000 yr were obtained for the first and second generations, respectively. A humid episode apparently characterized eastern Africa during each glacial-interglacial transition. 18O and 13C measurements on stromatolites, when compared to values on modern waters and carbonates, provide paleohydrological information. Long residence time of the paleolake waters and less seasonally contrasted regimes are inferred.

The recent discovery of a post-Roman-aged marine calcarenite terrace, at an altitude of 7 to 8 m at the ancient harbor of Akhziv, supports the hypothesis of oscillatory tectonic movements along the coastline of Israel during the late Holocene. The marine to estuarine origin of this terrace is indicated by the presence within it of a biocoenosic, sea-marginal to estuarine, assemblage of well-preserved molluscs consisting of taxa tolerating brackish-water, together with a few fresh-water specimens. The transgressive marine sediments onlap and overlie a 2700-yr-old middle Iron Age (Phoenician) tomb, which was built on a dark clay layer, containing a middle Bronze HB (3750 yr B.P.) settlement. In post-Roman times the coastal zone at this site, both east and west of the present coastline, was first subjected to tectonic subsidence of a few meters, and was then tectonically uplifted to its present altitude.

The 104 available [230Th/234U] analyses of unrecrystallized corals from stable emerged terraces indicate that the last interglacial period occurred approximately 125,000 yr ago. An estimate is made of the exact duration of this period by accounting for the surprisingly small distribution width observed among the 80 most reliable analyses. This distribution width is compared with those obtained for model populations generated by assuming various characteristic analytical errors and various lengths of the last interglacial period. The results show that (1) if there was only a single rise in sea level, it probably lasted no more than 12,000 yr, in agreement with previous estimates; and (2) if there were two separate rises of sea level, the gap between them must have been less than 7500 yr and not 12,200 as proposed by some authors.

The dating of the 6 to 8-m shoreline in southern Africa as early Pleistocene prompted a reassessment of evidence from archaeological cave sites on the south coast of the Cape Province which had suggested that this shoreline is of last interglacial age. The successions in the caves at Klasies River Mouth and Die Kelders have been misinterpreted. In fact, they confirm a pre-last interglacial age for this feature. These sites, together with Herolds Bay Cave, indicate that in southern Africa the only last interglacial shoreline above present sea level is at about 4 m, and that it dates from isotope substage 5e.

Human occupation and utilization of plant resources have affected vegetation in the lower Little Tennessee River Valley of East Tennessee for 10,000 yr. Changes in Indian cultures and land use are documented by radiocarbon chronologies, lithic artifacts, ceramics, settlement patterns, and ethnobotanical remains from 25 stratified archaeological sites within the Holocene alluvial terrace. The ethnobotanical record consists of 31,500 fragments (13.7 kg) of wood charcoal identified to species and 7.7 kg of carbonized fruits, seeds, nutshells, and cultigens from 956 features. Pollen and plant macrofossils from small ponds both in the uplands and on lower stream terraces record local vegetational changes through the last 1500 to 3000 yr. Human impact increased after cultigens, including squash and gourd, were introduced ca. 4000 yr B.P. during the Archaic cultural period. Forest clearance and cultivation disturbed vegetation on both the floodplain and lower terraces after 2800 yr B.P., during the Woodland period. Permanent Indian settlements and maize and bean agriculture extended to higher terraces 1.5 km from the floodplain by the Mississippian period (1000 to 300 yr B.P.). After 300 yr B.P., extensive land clearance and cultivation by Historic Overhill Cherokee and Euro-Americans spread into the uplands beyond the river valley.

Field surveys of several sea-level indicators (exposed in situ reef framework, conglomerates, coral colonies and Tridacna shells in a growth position, sea-corrosion notches) carried out on six atolls from the NW Tuamotus (Mataiva, Rangiroa, Arutua, Kaukura, Apataki and Takapoto) and data from four subsurface boreholes drilled through Mataiva show that during the late Holocene mean sea level (MSL) reached a maximum elevation at approximately + 0.9 m. It remained above the present MSL from between 6000 and 5500 yr B.P. until at least 1200 yr B.P. Human settlements on the atolls were extremely unlikely and probably impossible throughout this time. The area investigated seems to have been tectonically quite stable during the late Holocene. A local curve of MSL variations may be representative of the regional eustatic pattern.

Glacial lake sedimentary records from Hector Lake in Banff National Park, Alberta, are examined with the aim of clarifying the relationship between glacial activity and outwash sedimentation. Rhythmic couplets in lake bottom sediments provide a method for accurate documentation and dating of recent sedimentation chronology, as analysis of 137Cs content indicates that the couplets are of annual periodicity (i.e., varves). A comparison of sedimentation rate records with recent regional glacial history indicates that high sedimentation rates of one to a few decades duration occur either during and immediately following periods of moraine deposition (ie, maximum ice stands) or during periods of rapid ice recession. Comparison of sedimentary records with somewhat longer term glacial and vegetation records indicates that sedimentation rate variations of several centuries duration closely parallel changes in upvalley ice extent. Over this time scale, high sedimentation rates occur during periods of relatively increased ice extent, low rates during periods of reduced ice extent. Within this general pattern, however, sedimentation rates may remain very high for nearly a century following maximum ice stands, due to the exposure of unstable glaciogenic deposits to fluvial reworking during ice recession.

The isotopic composition of paleosol carbonates from Olduvai Gorge can be used to give estimates of the paleoclimatic history of East Africa during the Plio-Pleistocene. Oxygen isotopes are indicative of major changes in the isotopic composition of meteoric water, while carbon isotopes are indicative of major changes in the fraction of biomass using the Hatch-Slack (C4) photosynthetic pathway. Three major periods of climatic change in this terrestrial setting occurred at about 1.67, 1.3, and 0.6 myr. These involved major changes in flora as well as major changes in the isotopic composition of meteoric water. Overall, the climate has become warmer and more arid over the past 2.2 myr.

U-series analyses have been made of travertine samples from the archaeological site at Ehringsdorf-Weimar. The lower Travertine, in which hominid remains were found, yields apparent ages from >350,000 to 200,000 yr. Correlation between isotope ratios yields an age of about 230,000 yr. One sample of the upper Travertine gives and age of 111,000 ± 47,000 yr. Both upper and lower travertines contain fauna and flora indicative of interglacial conditions. The lower Travertine was deposited during isotope stage 7, while the upper Travertine may have been formed during isotope stage 5. This contradicts most earlier biostratigraphic studies which assigned both lower and upper Travertines to stages of the Eem interglaciation (correlative with isotope stage 5). The fauna and flora of the lower Travertine are significantly different from those of the upper Travertine, however, and in some respects are transitional to those of the travertines of Bilzingsleben, which appear to be older yet.

Under the near-ideal hydrologic conditions provided by the Uitenhage artesian aquifer, reliable estimates of paleotemperatures were obtained by measurement of the dissolved N2 and Ar in groundwater, confidently dated in the range 28,000 yr B.P. to the present using 14C. The mean annual air temperature in this area during the last glaciation is calculated to have been, on average, 1/2°C lower than during the Holocene. The change in the 18O/16O ratio of the groundwater, corrected for changes in ocean water, amounted to 0.4‰ per °C.