Very little is known about the prehistory of Siberia prior to the Sartan (“Main Würm” = “Main Wisconsin”) Stadial of the Last Glacial. It is not yet clear whether this is result of the inadequacy of investigations so far or of the fact that human occupation was relatively limited in pre-Sartan times. The Sartan occupation of Siberia has been well documented by the discovery of a relatively large number of open-air sites in the major river valleys of the region. The large numbers of broken-up animal bones found at many of these sites, in combination with rich inventories of bone and stone artifacts and remains of structures, hearths, ect., show that the Sartan inhabitants of Siberia were comparable in level of cultural development to their European (Upper Paleolithic) contemporaries. More particularly, they seem to have been effective big-game hunters, fully capable of exploiting the comparatively rich game resources of their open country environment. As in Europe, evidence from Siberia suggests that the environmental change (especially reforestation) which took place at the end of the Last Glacial led to decreased reliance on big-game hunting and increased emphasis on other modes of subsistence, especially fishing.

Dendrochronology is the science of dating annual growth layers (rings) in woody plants. Two related subdisciplines are dendroclimatology and dendroecology. The former uses the information in dated rings to study problems of present and past climates, while the latter deals with changes in the local environment rather than regional climate.

Successful applications of dendroclimatology and dendroecology depend upon careful stratification. Ring-width samples are selected from trees on limiting sites, where widths of growth layers vary greatly from one year to the next (sensitivity) and autocorrelation of the widths is not high. Rings also must be cross-dated and sufficiently replicated to provide precise dating. This selection and dating assures that the climatic information common to all trees, which is analogous to the “signal”, is large and properly placed in time. The random error or nonclimatic variations in growth, among trees, is analogous to “noise” and is reduced when ring-width indices are averaged for many trees.

Some basic facts about the growth are presented along with a discussion of important physiological processes operating throughout the roots, stems, and leaves. Certain gradients associated with tree height, cambial age, and physiological activity control the size of the growth layers as they vary throughout the tree. These biological gradients interact with environmental variables and complicate the task of modeling the relationships linking growth with environment.

Biological models are described for the relationships between variations in ring widths from conifers on arid sites, and variations in temperature and precpitation. These climatic factors may influence the tree at any time in the year. Conditions preceding the growing season sometimes have a greater influence on ring width than conditions during the growing season, and the relative effects of these factors on growth vary with latitude, altitude, and differences in factors of the site. The effects of some climatic factors on growth are negligible during certain times of the year, but important at other times. Climatic factors are sometimes directly related to growth and at other times are inversely related to growth. Statistical methods are described for ascertaining these differences in the climatic response of trees from different sites.

A practical example is given of a tree-ring study and the mechanics are described for stratification and selection of tree-ring materials, for laboratory preparation, for cross-dating, and for computer processing. Several methods for calibration of the ring-width data with climatic variation are described. The most recent is multivariate analysis, which allows simultaneous calibration of a variety of tree-ring data representing different sites with a number of variables of climate.

Several examples of applications of tree-ring analysis to problems of environment and climate are described. One is a specification from tree rings of anomalies in atmosphere circulation for a portion of the Northern Hemisphere since 1700 A.D. Another example treats and specifies past conditions in terms of conditional probabilities. Other methods of comparing present climate with past climate are described along with new developments in reconstructing past hydrologic conditions from tree rings.

Tree-ring studies will be applied in the future to problems of temperate and mesic environments, and to problems of physiological, genetic, and anatomical variations within and among trees. New developments in the use of X-ray techniques will facilitate the measurement and study of cell size and cell density. Tree rings are an important source of information on productivity and dry-matter accumulation at various sites. Some tree-ring studies will deal with environmental pollution. Statistical developments will improve estimation of certain past anomalies in weather factors and the reconstructtion of atmosphere circulation associated with climate variability and change. Such information should improve chances for measuring and assessing the possibility of inadvertent modification of climate by man.

Pollen in surface sediments from five lakes in southern Michigan shows evidence of differential deposition. In most lakes, pollen from ragweed occurs in a higher ratio to tree pollen in shallow-water sediment than in deep water. Pine and certain herbs with small pollen grains follow the same pattern. One very deep lake is an exception, with the ratio of ragweed to tree pollen highest in deep-water sediment. Pollen from local aquatics and from willow, which grows along the lake shore, is also unevenly distributed, occurring in highest frequencies near the parent plants. Pollen from deciduous trees, however, occurs in similar ratios at all sampling stations within each lake. Deciduous pollen occurs in uniform ratios, also in older sediment, deposited in the early 19th century, when the landscape was still forested.

Percentages of deciduous tree pollen (as percent tree pollen) were compared among lakes. Single samples were taken for this purpose from the deepest part of each lake basin. Oak pollen percentages are higher in three lakes in western Washtenaw County than in three lakes in eastern Washtenaw County. This difference reflects a similar difference in present-day vegetation: second-growth oak forests grow near the lakes in the western half of the county, while all but 5% of the area in the eastern part of the county is farmland. (The difference in the ratio of farmland to forestland in the two parts of the county is not reflected clearly in the ratio of herb pollen to tree pollen, because there is so much variation within each lake.) In 140-year-old sediment, on the other hand, tree pollen percentages in the six samples are homogeneous as shown by a chi-square test. The homogeneity in sediment deposited before the forest was cleared is surprising, because witness-tree data from presettlement time show that the frequencies of tree species in the two areas were quite different. Pollen dispersal at that time must have been effective enough, to counteract differences over distances of a few tens of kilometers in the amounts and kinds of pollen produced by the vegetation.

Les mesures du rapport isotopique 18O/16O le long de l'axe de stalagmites soigneusement choisies donnent u̇n enregistrement climatique continental qui peut être comparé aux données de la paléoclimatologie océanique. Les âges des différentes couches de la stalagmite sont établis soit par la méthode du carbone 14 pour la période de 0 à 35.000 ans b.p., soit par le rapport 230Th/234U jusqu'à 300.000 ans b.p.

La validité de cette méthode est discutée et une application en est donée pour la période allant de 130.000 à 90.000 ans b.p.

Measurements of the isotopic ratio 18O/16O along the axis of carefully chosen stalagmites give a continental climatic record which can be compared to oceanic paleoclimatological data. Ages of stalagmite layers are established either by carbon 14 for the period 0 to 35,000 years b.p. or by 230Th/234U ratio until 300,000 years b.p.

The validity of this method is discussed and an application is given for the period 130,000 to 90,000 years b.p.

Comparison of the glacial and periglacial deposits and soils of the Alps and Rocky Mountains suggests the following correlations. The Donau, Günz, and Mindel are correlated with the Washakie Point (Nebraskan), Cedar Ridge (Kansan), and Sacagawea Ridge (Illinoian). These glaciations are separated and followed by interglacials represented by thick deeply weathered soils, the last being the Mindel/Riss, or Sacagawea Ridge/Bull Lake (Sangamon). The Riss I (Paar) and Riss II (type Riss) Glaciations are correlated with the early and late advances of the Bull Lake Glaciation (early Altonian), and are believed to be different glaciations separated and followed by short interglacials represented by soils of intermediate development at these latitudes. The Alt-Würm period of restricted ice in the Alps is represented only by local unnamed deposits in the Rocky Mountains. In the Alps, at least three interstadial soils occur between the Riss/Würm soil and the oldest Main Würm end moraine. These coincide in time respectively with the beginning and end of the Port Talbot II interstadial and with Plum Point interstadial. The three major end moraines of the Main Würm are correlated with three or locally more end moraines of the early and middle stades of the Pinedale (Woodfordian). Moraines of the late stade of the Pinedale are equivalent to the late glacial and early postglacial moraines of post-Würm but pre-Atlantic age in the Alps.

K-Ar dating and tephrochronology in the Rockies suggest that the Washakie Point ended about 1.2 m.y. B.P., the Cedar Ridge about 700,000 B.P., the Sacagawea Ridge about 180,000 B.P., the last great interglacial about 130,000 B.P., and the interglacial separating early and late Bull Lake about 80,000 B.P. Radiocarbon dates from the Alps suggest that the Riss/Würm was about 70,000–60,000 B.P., and that Alt-Würm interstadials occurred at >50,000, 45,000–40,000, 34,000–32,000, and 28,000 B.P. Radiocarbon dates from rock shelter deposits in France indicate minor interstadials in the Main Würm at 21,000–20,000 B.P. and about 17,000 B.P. Both the Main Würm and Pinedale ended about 11,800 B.P.

The evolution of Pleistocene mammals is characterized by giantism and in the ungulates by the appearance of huge and often bizarre horns, antlers, ossicones, and tusks. An earlier study of sheep (Ovis) led to a theory explaining these phenomena. This theory, termed the “dispersal theory,” is described briefly and then applied to the evolution of Old World deer (Cervinea, Baird, 1857). It is shown that the zoogeography and physical characteristics of Old World deer follow the predictions of this theory. Attention is drawn to the parallels in zoogeography and convergent evolutionary changes between Old World deer and other ungulates such as New World deer, the goats and sheep and the genus Bison. The application of the “dispersal theory” to the evolution of the genus Bison shows that it is compatible with known facts and that it explains aspects of Bison morphology not usually considered. It also suggested a new and testable hypothesis of the origin of present-day Bison. In general the “dispersal theory” applied to Pleistocene conditions predicts that a specialization of social behavior and social organs leading to giantism, altered body proportions and bizarre hornlike organs occurs during postglacial dispersal into new habitat. It also describes the characteristics of these early, pioneering populations. In so doing it links the disciplines of animal behavior, ecology, animal science, population dynamics, and zoogeography into a functional whole. In particular it shows the relationship between ecology and social behavior. Finally, the “dispersal theory” is shown to be a general theory, applicable to Pleistocene and pre-Pleistocene conditions. In the tropics or pre-Pleistocene conditions social evolution follows ecological specialization, while in the periglacial zones social evolution only reflects the colonization history of the species. It is shown that the “dispersal theory” unites the apparently contradictory views on evolution and zoogeography of Matthews and Darlington. Matthews' views apply largely to Pleistocene conditions and Darlington's to pre-Pleistocene conditions. A preliminary application of the “dispersal theory” to pre-Pleistocene mammals suggests that it predicts correctly.

The prevailing surface temperatures in summer and winter at several different stages of the last ice age, indicated at various points scattered over the Northern Hemisphere, by botanical, glaciological, marine biological, oceanographic, etc. evidence, are used to derive probable distributions of 1000−500 mbar thickness, roughly equivalent to mean temperature of the lowest 5 km of the atmosphere and indicating the general flow pattern of the atmosphere in depth. From these thermal wind patterns computation of the tendency to cyclonic and anticyclonic development is possible. Maps of this development field, taken together with the indicated steering of surface cyclones and anticyclones by the thermal winds, make it possible to sketch probable distributions of surface pressure (and, by implication, surface winds) prevailing during each of the glacial stages studied. New light is thrown on the onset of glaciation and on the regimes associated with the maximum extent of glaciation, with the Alleröd warm epoch and the Post-Alleröd cold stage when there was some readvance of the ice.

The thermodynamic model of J. Adem has been applied to the determination of Arctic and hemispheric surface temperatures with both ice-covered and ice-free states of the Arctic Ocean. The effect of glaciated and nonglaciated continents is included in the investigation. With an ice cover over the Arctic, as at present, computed temperatures for the polar sea and the Northern Hemisphere correspond closely with present observations. Over a broad range of the critical parameters, removal of the ice cover yields computed temperatures that remain well above freezing level throughout the year. With glaciated continents computed Arctic temperatures are depressed.

The intricate pattern of moraines of the Laurentide ice sheet in the Great Lakes region reflects the marked lobation of the ice margin in late Wisconsin time, and this in turn reflects the distribution of steam-cut lowlands etched in preglacial times in the weak-rock belts of gentle Paleozoic fold structures. It is difficult to trace and correlate moraines from lobe to lobe and to evaluate the magnitude of recession before readvance, but three breaks stand out in the sequence, with readvances at about 14,500, 13,000, and 11,500 years ago. The first, corresponding to the Cary advance of the Lake Michigan lobe, is represented to the west by distant advance of the Des Moines lobe in Iowa, and to the east by the overriding of lake beds by the Erie lobe. The 13,000-year advance is best represented by the Port Huron moraine of the Lake Michigan and Huron lobes, but by relatively little action to west and east. The 11,500-year advance is based on the Valders till of the Lake Michigan lobe, but presumed correlations to east and west prove to be generally older, and the question is raised that these and some other ice advances in the Great Lakes region may represent surges of the ice rather than regional climatic change. Surging may involve the buildup of subglacial meltwater, which can provide the basal sliding necessary for rapid forward movement. It would be most favored by the conditions in the western Lake Superior basin, where the Superior lobe had a suitable form and thermal regime, as estimated from geomorphic and paleoclimatic criteria. The Valders advance of the Lake Michigan and Green Bay lobes may also have resulted from a surge: the eastern part of the Lake Superior basin, whence the ice advanced, has a pattern of deep gorges that resemble subglacial tunnel valleys, which imply great quantities of subglacial water that may have produced glacial surges before the water became channeled.

The morphology of paleosols and radiocarbon-dated charcoal from buried surface horizons of soils provide evidence to suggest that between periods of northward forest encroachment tundra climate has dominated areas at least 50 km south of the present forest/tundra border in southwest Keewatin. The present forest/tundra border climate is nearly as severe as any climate that has prevailed in the area since deglaciation.

Pleistocene exposures located near Cape Deceit on the south shore of Kotzebue Sound (Alaska) contain a complex sequence of organic sediments which have yielded fossil vertebrates of taxonomic, evolutionary, and zoogeographic significance. Cape Deceit Local Fauna from the Cape Deceit Formation, lowermost of three major stratigraphic units recognized at the site, contains representatives of the genera Canis, Rangifer, Cervus, Ochotona, Lemmus, Microtus, Pliomys, and a newly described genus and species, Predicrostonyx hopkinsi-predecessor of the extant genus Dicrostonyx. In addition to P. hopkinsi, new species of Ochotona, Pliomys, and Microtus are described from the Cape Deceit Local Fauna. the stage of evolution of Microtus deceitensis sp. n. and P. hopkinsi sp. n. indicate that the Cape Deceit Formation is at least of pre-Cromerian age. This assumption is substantiated by the stratigraphy of the overlying Inmachuk and Deering formations, fossils from these units (including Dicrostonyx torquatus and D. henseli), and C-14 dates. The Cape Deceit Local Fauna, therefore, contains the earliest North American record of several of the included genera in addition to the only North American record of Pliomys.

Stratigraphic features at the exposure indicate that the Cape Deceit Local Fauna mammals lived in a treeless (tundra) environment showing that the contemporary tundra mammalian fauna has a lengthy arctic-adapted evolutionary history. We postulate the existence in northeastern Eurasia and Alaska of a Pleistocene-Recent “Beringian” mammalian realm which has at times contributed to the fauna of more southern regions of the Palaerctic and Nearctic. The fact that this southern movement of arctic-adapted or arctic-derived Beringian mammals occurred at different times in Europe and North America (interpreted by some as indicating the existence of a “Beringian filter-bridge”) is, we think, due largely to differences of continental physiography and Pleistocene geography south of the Beringian area.

The thermodynamic ocean of the Sillen school offers little incentive to those who search the sedimentary record for evidence of changes in ocean chemistry during Cenozoic time. Their models predict a uniform chemical composition. However as the sediments presently accumulating in the ocean show little evidence of equilibration with the overlying water, the possibility that kinetic factors play an important role must be seriously explored. Such a model is presented in this paper. Material balance restrictions are substituted for some of the usual chemical equilibria. The role of organisms is shown to be dominant for at least some of the important components of sea salt (i.e., C, N, P, Si, …). If, as proposed here, the chemistry of sea water is dependent on rates of supply of individual components, the rate of vertical mixing in the sea, and the type of material formed by organisms, then substantial changes in the chemical composition have almost certainly taken place. Several means by which such changes might be reconstructed from chemical and isotopic measurements on marine sediments are discussed.

Earlier studies in Alaska and northwest Canada have shown inconsistent evidence for the expected northward extension of the Arctic tree line during the Hypsithermal Interval. Only megafossil evidence has supported this suggestion; the palynological findings have been inconclusive. The Tuktoyaktuk Peninsula, in the Northwest Territories of Canada, offers critical sites for studies of late-Pleistocene ecology, because of its geological, biotic, and climatological features. Palynological and megafossil evidence is presented from sites on the Tuktoyaktuk Peninsula, indicating northward advance of the Arctic tree line during the period 8500-5500 B.P. Relative pollen frequencies of a core of lake sediment suggest a late-Pleistocene sequence as follows: 12,900-11,600 dwarf birch tundra; 11600-8500 forest tundra; 8500-5500 closed-crown spruce-birch forest; 5500-4000 tall shrub tundra; 4000-present dwarf birch heath tundra. These results suggest that during the Hypsithermal Interval the Arctic Front (July position) was further north, over the Beaufort Sea, a displacement from its present position of about 350 km. The Tuktoyaktuk Peninsula, presently occupied by tundra, and dominated by the Arctic airstream in July, was apparently under forest, with warm, moist Pacific air during the Hypsithermal Interval.

It is becoming increasingly apparent that the pattern of early human occupation of the Southwestern United States was strongly influenced by the major paleoclimatic events of the period 9500 B.C. to A.D. 700. The size of human populations and the distribution of human settlement at both the regional-topographic and large-scale areal level, known from archaeological research, are directly correlated to climatic change documented by the evidence of geology and palynology.

The effect of climatic change is felt through the actions and reactions of the economic subsystem and its linkages with other subsystems. These reactions reflect not only the character of the climatic stimulus but also the existing state of the cultural system. Alternate reactions include direct systemic readaptation to the changed environment (through changed technologies, methods of population control, etc.); or small scale or large scale relocation of populations in different local niches, regions, or areas whose character most closely approximates the conditions to which the cultural system was initially adapted.

Rock-magnetic, paleomagnetic and petrologic properties of samples from the Laschamp and Olby basalt formations in France were studied to aid in determining the validity of the Laschamp geomagnetic field reversal reported by Bonhommet and Babkine. The Laschamp flow contains ilmenomagnetite, with partial alteration of the magnetite to hematite. Ilmenomagnetite in the Olby flow has largely recrystallized at high temperatures to a composite mozaic intergrowth of pseudobrookite, titanohematite and magnesioferrite, with rare residual magnetite and lamellae of ilmenite. The remanent magnetization is stable and resides primarily in single-domain magnetite particles. Our results indicate that the magnetizations of the Laschamp and Olby flows faithfully record the direction of the ambient magnetic field in which they cooled.

The most commonly used criteria in marine sediments for detecting climatic changes are the remains of organisms and the position of the shorelines, for these two types of criteria can have a relatively quick response to climatic change. The inorganic components of marine sediments, however, also provide useful criteria. On the inner continental shelf where the best correlation should be found between modern terrigenous marine sediments and modern climates, sediment texture is the main criterion. Where land ice reaches the sea, gravel may be deposited, but much of the inner shelf in polar climates receives abundant mud, containing a small amount of clay minerals. From tropical humid climates abundant mud is delivered composed mainly of clay minerals, but knowledge of their composition is required, because the largest rivers do not have a dominance of tropical sediment products. In arid climates and midlatitude moderate rainfall climates, inner shelf sand is indicative, although it also possibly reflects the common entrapment of mud in estuaries and the presence of the middle latitude cyclone belt in which storms remove the fine material present on the inner shelf. Climate also controls extensive carbonate deposits. In deep-sea sediments composition contains more important criteria than texture. Some criteria appear to be reliable for various aspects of modern climates and therefore should be useful in detecting climatic changes. These criteria include the size, surface texture, and mineralogical and chemical composition of eolian transported material downwind of arid lands; global dust in latitudinal bands of atmospheric circulation; volcanic ash downwind of geologically instantaneous events; surface texture of quartz grains and the abundance of terrigenous material in pelagic sediments as indication of glaciation; chlorite from a polar climate; kaolinite from a tropical climate, and inorganically precipitated calcium carbonate in enclosed seas. Less definitive criteria are possibly the rate of turbidity current activity, iron-rich layers in the sediment, sedimentation from the nepheloid zone, construction of features by bottom currents, organic matter content, and sedimentation rate. Speculations include the intensity of benthic faunal reworking of sediment. Using these criteria it is possible to identify the sediment products of the extreme climates: polar, tropical rainy, and dry (desert), and thereby to infer the existence of these climates. The moderate climates apparently are not so easily detected. The criteria also indicate the nature of the water, wind, and ice processes delivering the sediment products to the sea. Extreme values in the frequency or magnitude of the climate-associated processes have great significance in the supplying of terrigenous material, and changes in these extreme values could produce salient changes in the sedimentary sequence. The criteria of climatic change might well be considered criteria of change in extreme values of the processes.

Opal phytoliths, inorganic biogenetic plant particles of microscopic size, were investigated to determine the level at which their varied morphology is taxonomically significant. Extraction methods for the isolation of these particles from living plants and from soils were developed to provide maximum preservation of morphological features while remaining simple, rapid, and inexpensive. Extracts were made and microscopically viewed from 30 live specimens, of which 16 were systematically typed in order to develop taxonomic guidelines for their identification in soils. Clear taxonomic differentiation was noted between and within major plant groups with considerable indication that more detailed investigation will yield finer subdivisions. Since opal phytoliths are reported to be a particularly durable fossil found in deposits as far back as Tertiary Age, it appears that opal phytolith analysis can provide paleobotanical information comparable to palynological data in many areas where pollen is absent. When used in conjunction with pollen where present, they should serve to confirm pollen data as well as provide additional botanical data not available in pollen studies.

Deposits of Quaternary age from the Rocky Mountains to the Pacific Coast in the western conterminous United States represent a great variety of environments. The deposits include those of continental and alpine glaciers, glacial meltwater streams, nonglacial streams, pluvial lakes, marine environments, eolian environments, and masswasting environments. On two charts we have attempted to correlate representative sequences of deposits of many of these environments, based on published sources and recent unpublished investigations. Evidence for correlation is based mainly on stratigraphic sequence, soil characteristics, the amount of subsequent erosion and interlayered volcanic ash beds identifiable as to source. Chronologic control is based on numerous radiocarbon dates, U-series dates on marine fossils, and K-Ar dates on volcanic rocks. The Bishop volcanic ash bed and one of the Pearlette-like volcanic ash beds appear to represent significant regional key horizons, respectively about 700,000 and 600,000 years old. Rock magnetism is shown to suggest the paleomagnetic polarity at the time of rock deposition. Assigned land-mammal ages of included fossils help to put limits on the age of some units.