Paleopedological data testify that during the Holocene the territory of Northwest Caucasus experienced contrasting environmental changes. Thus, in the Early and Middle Holocene steppe landscapes propagated up to 700-m altitude. The increasing humidity caused the expansion of broad-leaf forests from the mountains to the foot slopes (50-100-m height) which took place 3500-2500 years ago. Following this climatic and ecosystem change the evolution of Chernozems formed on loess towards Luvisols occurred in the foothills. However, the direction of soil evolution differed being determined by the composition of the parent material.

The study of soil profiles formed on artificial landsurfaces of kurgans (burial mounds) and ramparts in the second half of the Holocene allows to trace this diversity. Within the last 3000 years mature Luvisols have developed on the burial mounds composed of loess and material of Ah horizon of Chernozem. The leaching of calcite took place at the beginning of the forest stage, within the first hundreds of years. Afterwards the clay illuviation started. Rendzinas have developed on the mounds composed of calcareous material, because the release of bases from the weathering of limestone fragments prevents the destruction of clay-humus complexes and substantially retards the development of clay illuviation. Thus the formation of Rendzina on calcareous parent material shows evolutionary stability whereas in case of development of Luvisol on loess and Chernozem-loess substrates relatively fast profile transformation occurred, followed by long-term quasi-stability in a steady state of ecosystem (Luvisols on loess and Chernozem-loess substrates).

Specific trends of soil development along with the related quasi-stability states at the intermediate stages of their development were observed on the parent materials depending upon the content and localization of calcareous material. Leached Rendzinas and Humic Luvisol (Dark gray forest soils) with carbonate horizon are formed on kurgans constructed of loess and Chernozem materials with calcareous pavement on top, the latter provides the CaCO3 supply in the upper part of the profile. In cases when similar pavement covers natural land surface near kurgans the evolution of the middle Holocene Chernozem ceased at the stage of Leached Chernozems. The most probable way of further development of these soils is a quick evolution towards Luvisols after the components of pavement are destroyed by leaching.

Recent small climatic fluctuations on a 10²-10³ year time scale can be correlated worldwide for example by distinct moraines dated to about 1850 AD in similar positions above the present day glaciers in the southeastern Canadian Rockies, in the European Alps and in the Tian-Shan near Urumqi, China. These moraines result from glacier advances caused by a decline of mean annual temperature of only about 1 C. This suggests that major climatic changes on at least a 10⁵ year scale (glacial-interglacial cycles) and probably a 10⁴ year scale (the approximate length of an interglacial) must be of similar ages throughout the temperate climatic belt of the Northern Hemisphere. This concept is important for continental pedostratigraphical correlations; especially when loess-paleosol sequences correspond with time equivalent parts of the deep-sea oxygen isotope record in the Brunhes epoch we are in all probability close towards a continental or even a global chronostratigraphical correlation between loess-paleosol sequences in different continents. This allows the reconstruction of a quaternary climatic history especially of the Brunhes epoch in the temperate climatic belt of the northern hemisphere.

Detailed knowledge of the genesis of paleosols is needed to establish loess-paleosol stratigraphies that can be used for paleoclimatic reconstruction. Most paleosols, however, are truncated and largely recalcified by carbonate derived from overlying loess. Micromorphological studies allow primary and secondary carbonates to be distinguished and provide unequivocal evidence of clay illuviation. This enables the separation of typical loess, weakly weathered loess and the recognition of different genetic soil horizons as CB, BC, Ah, Bw, B and Bt horizons. A comparison of the loess-paleosol sequence of the Karamaydan section/Tadjikistan with the upper part of the Luochuan section/China clearly indicates that for the Brunhes chron the sequence at Karamaydan, Tadjikistan, is more detailed than the corresponding section in Luochuan, China and even more than in Central Europe except for the last glaciation. Single paleosols at Luochuan correspond to pedocomplexes at Karamaydan such as S3 soil to PK III and the S4 soil to PK IV; both PKs developed in a time period much longer than one interglacial (about 10⁴ years). Some pedocomplexes at Luochuan have more detailed successions at Karamaydan; e.g. the S5-complex in Luochuan, which corresponds with PKs VI and V. The loess-paleosol sequence in Karamaydan, therefore, should be regarded as a key sequence for reconstructing the climatic history of the Brunhes epoch. The good correlation with the deep-sea oxygen isotope record of Bassinot et al (1994), which includes the development of an accurate astronomical time scale, allows a detailed chronostratigraphical subdivision of the loess-paleosol sequence in Karamaydan for the Brunhes epoch. Also, for most of the Matuyama epoch the central and lower parts of the sequence at Chashmanigar, Tadjikistan, show more pronounced paleosols than the equivalent parts of the sequences at Luochuan and much more than in Central Europe, although some Ck or Ckm horizons indicate the presence of hiatuses or lost parts of the Chashmanigar section. Nevertheless it provides more paleoclimatic information regarding cold arid stages represented by loesses and warm humid stages (interglacials) represented by paleosols, than even the deep-sea cores known so far.

To reconstruct paleoclimates the primary and secondary minerals of the silt and clay fractions must be determined separately to evaluate the type and intensity of mineral weathering and clay mineral formation. This shows that there is little difference in the type and amount of pedogenic clay mineral formation between the Holocene soils and the paleosols of the Brunhes epoch at Karamaydan and of most of the paleosols in the Matuyama epoch at Chashmanigar. This suggests that the interglacial climates represented by the B or Bt horizons of the buried paleosols of young, middle and old Pleistocene age were similar to that of the Holocene.

The Qaidam Basin is one of the largest intermountain sedimentary basins in western China. It lies to north of Qing-Zang (Qinghai-Tibet) Plateau and is surrounded by three mountain chains including Kunlun, Qilian mountains and Argin mountains. The basin is about 2800 m above the present sea level and it covers an area of approximately 120,000km². The thickness of the materials accumulated during the Quaternary period in the Qaidam Basin is in the range of 20000~3000m, with the maximum being found at the last subsiding center. Such a thick sedimentary sequence may provide sufficiently detailed information for understanding the entire history of the basin during the Quaternary period, since such a continuously and rapidly deposited thick sequence is likely to have recorded small-scale geological events lasting for only very short periods, such as paleoclimatic cycles.

The material accumulated in the center came from the surrounding mountains, and their particle size is usually very small due to the long distance of transportation to the paleolake. On the basic of the index biostratigraphic units established according to the characteristics of the Ostracoda fossils, attempts were made to determine the depth of many important boundaries by comparing the stratigraphic sequences classified using the results obtained from paleomagnetic dating. For instances, in Seshen-1 BH the well-known M/G boundary (2.48Ma B.P.) was found from such analyses at the depth of 1460 m; and the Olduvai event corresponds to the interval of 710~880m in Sezhong-6 BH. By analyzing statistically the thickness of typical couplets found in the varve deposits and though correlating such results to the pattern of the geophysical logging curves, the yearly rate of deposition is obtained for the following typical facies: semi-deep lake facies, 0.45mm/a; shallow lake facies, 1.03mm/a; lakeshore facies, 1.28mm/a; lake basin bog facies, 0.94mm/a; and delta facies, 0.90mm/a. These rates were then used to cross-chick the paleomagnetically determined geological age of fossil Unit 1 in five different boreholes and the maximum difference was less than 1.5%.

It is the timescale that occupies an important position in the Quaternary paleoclimate research. Combined the biostratigraphy and magnetostratigraphy, the marker band of electrical measurement, as well as the sedimentology chronological calculation, the primary timescale of the Sezhong-6 BH (2.15~1.30Ma B.P), Tuozhong-2 BH (2.20~1.35Ma B.P), and Dacan-1 BH (2.85~0 Ma B.P) were obtained respectively. Based on the astronomical forcing theory, the turning astronomical timescale was obtained successfully during 2.85Ma B.P in Quaternary of Dacan 1 Core after carefully turning.

Based on the climate proxies development of the sediment faces and the assemblage characteristics of Gamma logging curve, the relation was discussed fully between the assemblage characteristics in the Gamma logging curve of all kinds of sediment faces and the fluctution in the water of the Caidam paleolake, consequently the paleoclimate-inversive model based on Gamma logging curve was constructed, which showed that the Gamma logging curve could be regarded as one of the favorable climate proxies. To extend this model into the Dacan-1 BH with a length of 2600m, which had no example to be analysis, the climate change during 2.85Ma B.P was reconstructed successfully.

To regarded the laminar layer of carbonaceous shale as the remarkable marker of the sharp retracts of the paleolake, considering the Gamma logging curve of Dacan-1 BH, it could be deduced that there were 5 paleoclimate abrupt events just like Heinrich event with accurate chronological data, which show in according with the ODP659 in the paleoclimate recovered through the dust flux. The study on the dacan 1 core since 1.00Ma B.P showed that there have been more than 30 climate abrupt events just like Heinrich event since the Last Glacial which indicated that the Gamma logging curve to be regarded as climate proxy had brought the great potential into researching the paleoclimate changes in the period of 1000 year, combined with the fractal research, it could provide a favorable clue for the project of the climate.

The following can be seen from the division of Quaternary series, calculating of geochronology, vertical relations of facies and reconstruction of facies distribution and paleolake shorelines. There is many factors controlled development of Quaternary deposits, of which tectonics and paleoclimatic cycles at different grades pre-primary, and landform and sedimentary supply are subordinate. In the whole quaternary interval, the depositional evolution can be deviled into three stages as following: a) in the interval between 3.05Ma B.P. and 1.95 Ma B.P., transgressions were developed gradually and the facies are dominated by the shore facies, near shore shallow water facies and shallow lacustrines; b) the transgression was remarkable, lake enlarged and deepen in water depth, especially after 1.35 Ma B.P., the hemi-deep lacustrines environment expanded; c) beginning with 1.00 Ma B.P., the lake level contracted remarkable and facies of the area studied are mainly swamps, shore and nearshore shallow lacustrines, even salt lake facies deposited widely in late periods.

The development of the paleolake in the Caidam Basin such as the wandering and the denivellation of the paleolake were allied with the stages uplifting of the Qinghai-Tibet plateau. During the periods of several enlarging of the sediment rate of the Dacan 1 core after 2.85Ma B.P, each period can find a corresponding event of the uplifting of Qinghai-Tibet plateau.

The eastern desert region in China refers to the vast expanse of sandy land and desert areas to the north of the Qinlin Mts. and to the east of the Helan Mts. and the Qinghai Lake (Dong Guangrong et al., 1990), situated in the mid-temperate monsoon climatic zone and alternately affected by modern East Asian Winter monsoon (it can be reflected by the cold wave invasion route) and Summer monsoon (it can be reflected by the mid-summer polar front location). There are many buried soils as a result of palaeomonsoon vicissitudes in the eastern desert region since the Holocene period, their development and evolution able to provide climate information. The stratigraphic assemblage forms in the eastern desert region are as follows: there exists a layer of black sandy soil or Heilu soil at outer fringe of sandy land and many places of the loess Plateau, their development ages generally range from 10 ka B.P. to present. However, there also exist 2~4 interbedding layers of aeolian sand or loess and black sandy soil, black soil or peat in the interior of the sandy land (Chen Weinan et al., 1993), their development ages also range from 10 ka B.P. to present. In addition, there exist a relative thickness and continue lager of black paleosils or lacustrine deposits in the valley and lower surface of sandy dune.

According to the sedimentary facies, age determination and magnetic susceptibility of the Holocene in the eastern desert region, its climatic variation can be further divided into three stages:

(1) Early Holocene (10~7.5 ka B.P.). During the period there formed two layers of sand intercalated a layer of weakly developed gray-black sandy soil (or thin humus layer in Horqin), the magnetic susceptibility of palaeosol was relatively lower. its vegetation was dominated by dry desert steppe species, such as Chenopodiaceae, Artemisia, etc.. While in Shouliao Plane, Ordos Plateau, Gonghe of Qinghai, Loess Plateau there are several periglacial phenomena developed. This shows that the climate retained some characteristics of the Last Glacial Period and was cold and dry this stage.

(2) Mid-Holocene (7.5-4 ka B.P.), this stage recorded in 2~4 layers of well developed black soils in the eastern desert region. For example, the Sandaogou Profile of the southeastern Mu Us Sandy Land has recorded in three layers of well developed black soils during the middle Holocene, its peak values of magnetic susceptibility occur in the ages of 5.5, 4.8 and 4.4 ka B. P., its vegetation belongs to sparse forest steppe species dominated by Pinus, Chenopodiaceae, Atemisia, etc., suggesting a strongest period of warm and humid climate. On the other hand, two layers of aeolian sand formed between 5.4-4.8-4.2 ka B.P. and their relatively lower magnetic susceptibility shows that even in the prime of warm and humid climate circulation there still existed two dry and cold stages.(3) Late Holocene (4 ka B.P. up to now), This stage has one main characteristic: the deposits in the eastern desert region are dominated by aeolian sand, however, there still existed a layer of weakly developed sandy-soil. For example the Sandaogou Profile of Yulin has two layers of aeolian sand intercalated a layer of weakly developed gray-black sandy soil 2.8 ka B.P. in age, their relatively lower magnetic susceptibilities are relatively lower and vegetation belongs to semiarid steppe species dominated by Chenopodiacease, Artemisia and other shrubs, while in Gonghe of Qinghai, Otinday and Horqin Sandy Land there are several layers of weakly developed thin black soils, and their ages vary between 1.3-1.5 or 1.7 and 0.5-0.6 or 0.11 ka B.P.. Afterwards, modern shifting sand covers the last black sandy soil layer, showing that from 4 ka B.P. onwards the dry and cold climate in the eastern desert region gradually became stronger.

Key words: Holocene period, eastern desert region, paleosoi, climatic change

Organic C-14, TL, IRSL and paleomagnetic dating of pedostratigraphy in the western Loess Plateau and the eastern Tibetan Plateau has show that paleosol sequences in these areas are not older than 0.8-1.1 Ma BP, suggesting that there must have had a big circualtion change occurred at that time. Uplift of the Tibetan Plateau to a critical height is thought the major factor driving this circulation shift. Spatial observation of paleosol sequences over the western Loess Plateau and the eastern Tibetan Plateau shows that paleosol sequence in the southern part of the eastern Tibetan Plateau is quite diferent from its northern part and the western Loess Plateau (the later two are very similar). It seems that in low frequency paleosol sequences in the western Loess Plateau and the northern part of the eastern Tibetan Plateau are chiefly controlled by eccentricity of 100-ka cycle overlapped with obliquity of 41-ka cycle and precession of 21-ka cycle; while in the southern part of the eastern Tibetan Plateau paleosol sequences are principally affected by the obliquity and precession, as number of paleosols in the southern part of the eastern Tibetan Plateau is much higher than that in its northern part and the western Loess Plateau.

In high frequency, multi-climatic proxies (rock magnetic, carbonate, soil color and grain size) from four late Pleistocene century scale loess-paleosol records in the Lanzhou-Linxia region on the Chinese western Loess Plateau have shown that in the last glacial cycle, Asian summer monsoon experienced 27 episodic pulse enhancements spanning only ca.1-2 ka in high frequency domain and having at least 13 sub-Milankovitch cycles of progressive weakening in low frequency domain, and winter monsoon had at least 13 episodic cold surges. Soil formation is in surprisingly fast response to these warm pulses. In low resolution (10⁴-year scale), both winter and summer monsoon changes match global climatic changes driven by global ice volume. But in high resolution (10² - 10³-year scale), winter and summer monsoon seem to vary broadly in pace (for winter monsoon) or by 1-2 ka lag (for summer monsoon) with North Atlantic rapid climatic changes, with the most identified cold surges and warm enhancements correlated to Heinrich events, major warm (Dansgaard-Oeschger) episodes and long-term cooling (Bond) cycles of North Atlantic climatic records, respectively. But in detail, both monsoons show different variations from North Atlantic climatic changes. This is particularly evident in the Holocene when both winter and summer monsoons had large fluctuations, but the North Atlantic region shows a fair stable Holocene in climate. These suggest that in one hand there have strong links between Asian monsoon system and North Atlantic climatic system in much of the last glacial cycle, and in other hand Asian monsoon worked in its own way in some periods. However, present dating precision has limited our clarification of which system is the driving force.

Paleomagnetic, rock magnetic, and grain size studies of a thick loess section in the West Qin Ling (mountain range), which forms the southern barrier of the Chinese Loess Plateau, show that loess in West Qin Ling started to be deposited about 800 ka and that it reveals a progressively increasing coarse grain size fraction upwards into the Holocene (> 32 ?m as well as > 63 ?m, i.e., coarser silt and sand size, respectively). The averages of these coarse size fractions are higher than in the central Loess Plateau, which was apparently farther from the source area, and slightly lower than those of the western Loess Plateau and the eastern Tibetan Plateau, which were therefore closer to the source area. The coarsening suggests (1) that Asian air circulation may have changed and intensified since 800 ka to allow dust to expand into West Qin Ling, (2) that dust-carrying winds were driven not only by the Asian winter monsoon caused by the Siberian-Mongolian High, but included also the westerlies and a winter monsoon caused by the Tibetan Plateau High, and (3) that intensification of all these air systems continues to the present. Increased elevation of the Tibetan Plateau so that it reached into the cryosphere by about 800 ka and a subsequent persistent uplift of the Plateau may have been the mechanisms to trigger and intensify the air circulation system. Moreover, this circulation shift and intensification, simultaneous with a shift in Milankovitch periodicity, may have contributed to large global climate changes such as the 15 % increase in global ice volume at ca. 800 ka.

Key words: Rock magnetism, grain size, Asia monsoon, air circulation, Tibetan Plateau Uplift

Biomorphic analysis is the study of the macro and micro remains of biota (biomorphs) in the context of the conditions of their origin. The main perpose of biomorph analysis are the diagnostics of modern and past conditions of pedogenesis, revealing of evolutinary trends of soils and natural and anthropogenic pedosediments, soil-forming rocks, and soil cover. These problems can be soved on the basis of data of biomorph morphology, mainy via the reconstruction of vegetation conditions. Some of the biomorphs are: phytoliths, spores and pollen, diatoms, sponge spicules, cuticle casts, detritus and so on.Paleoenvironmental reconstructoins based on biomorph analysis were done in the humid, semihumid and arid zones of Russian Plain. Early Holocene (7520+ 150 B.P.) darc brown and black soil lenses were studied in the Komi Republic (North-East of Russian Plain, middle taiga). Biomorph analysis had showed woody and herb detritus and bark. Phytoliths are rare, all of them being typical of the taiga flora. Abundant are spicules of fresh water sponges. We suggested that these lenses had resulted from natural catastrophic events which caused strong fires in the territory followed by intense erosion. All soil processes seemed to take place under invariably cold climate conditions yet with a changing wetness. The following chronosequence in the early Holocene was done: cold and wet stage dominated by open spruce-birch moss forest; cold and dry stage with frequent fires; and cold and wet stage with intense erosion which buried the fire traces.

Analysis paleosols of the Late Pleistocene (about 22000 - 24000 B.P.) was done in the Branskaia oblast (broad-leaves dicedeous forests; middle Russian Plain). The results allowed to picked out the next 3 stages of the environment reconstruction: first - cold and wet, dinamic, loesses formation; second (short) - cold and dry, stabilization, soil formation, vegetation similar to forest-tundra; third - again wet, dinamic, water erosian, loesses buried soil.

Analysis of soils buried under barrows of different age (4000 - 2000 B.P.) in the semidesert region of the Russian Plain South (Kalmikia) had showed that after hot and wet period of the bronze age was hot and dry sarmats stage. Then again worm and wet climate was. Modern climate is dry and hot.

Gong Zi-Tong, Liu Liang-Wu

Institute of Soil Science, Academia SinicaP.O.Box821, Nanjng210008, P. R. China

China has vast in territory, complication in natural conditions. The type of paleosols are diversified from the coast of the East Sea to the Tibet Plateau, and from the hills and mountains of South China to the Loess Plateau and the desert of north China. And the paleosols and environment in high-cold region of Tibet Plateau, in arid and semi-arid regions of Northwest China and in east monsoon region have various features, respectively.

High-cold region of Tibet Plateau The uplift of the Tibet Plateau has an effect on the distribution pattern of paleosols in this region and their development. Tibet Plateau has undergone 3-4 glaciations under the influence of global climate change, and the paleospls of mid Pleistocene, late Pleistocene and Holocene are reserved in 4 sets of tills. They are characterized by compact structure, gaily colour, intensive weathering and iron-stained matrix. According to paleosol feature, botanical fossil and pollen analysis, the buried soils in the region can be divided into 3 types: ustic luvisols, ustic cambisols and cryic cambisols. At the same time, the formation of the paleosols in interglacial stage faithfully records the environment change of Quaternary, such as the relation between paleosol and climate, paleosol and neotectonic movement, paleosol and vegetation change, etc.

Arid and semiarid region of Northwest China the paleosols within the sediment of loess and wind drift sand are a specific style in the region. Loess with deep sediment is widespread in China, and there are several layers, even several decades of buried soils in it. The paleosols can differentiate into apparent argillic and carbonate illuvial horizons. Of them a loess-paleosol sequence of Luochuan site is a representive of environment change of Quaternary for the loess region.

The buried soils covered by eolian sands are distributed from Hulunbeier sand land via keerxin sand ground extending to Shanbei and Gonghe basin of Qinghai provience. They have reflected evolution of soil environment for several thousands of years under the action of natural and human factors. Besides, the formation of relict solonchaks and relict bog soils indicates the change of climate, relief and hydrologic conditions here, as well as the influence of human activities.

East monsoon regionthe paleosols on fluvial and lacustrine deposits are dominant in the northern part of this region. The Huang-Huai-Hai Plain has undergone 3 sedimentation-soil formation cycles since late Pleistocene, where formed two great types of paleosols : relict vertisols and buried vertisols. It is a good indicator for ground water with climate fluctuation that several layers of calcareous concretion formed in different geologic periods and their chemical properties change with time. And paleosol horizon with calcareous concretion and dark paleosol horizon are soil-forming product in late Pleistocene and mid Holocene, respectively.

The ancient red earths are found quite extensively in the south part of the region. Using stable isotope and soil magnetic susceptibility, the study reveals paleoclimate change in the region, especially paleotemperature change, showing some peaks and valleys in the curves of¹³C and MS against soil profile. Of course, the peaks formed by paleosol reflect that the weather at that time was warm and humid, plants grew luxuriantly, while the valleys are contrary. Quaternary paleoclimate change was conjectured by properties of buried soils on basalt in Leizhou peninsula. At the beginning of middle Pleistocene, the paleoclimate was colder and mean annual temperature was about 13 C less than present one. In mid part of middle Pleistocene the paleoclimate had become warmer. During late Pleisticene the climate continued becoming warmer and was as same as present one. Above-mentioned fact shows the formation of paleosols here has never interrupted since late Pleistocene.

The loess-palaeosol sequences of the last 1.2 Ma in China have recorded two kinds of climate extremes: the strongly developed S4, S5-1, and S5-3 soils (corresponding to marine stages 11, 13, and 15, respectively) evidence three episodes of great warmth; two coarse-grained loess units (L9 and L15, corresponding to marine stages 22-23-24 and 38, respectively) indicate severest glacial conditions. The climatic and geographical significance of these events are still unclear, and the cause remains a puzzle.

Palaeopedological, geochemical and magnetic susceptibility data from three loess sections (Xifeng, Changwu and Weinan) suggest that the S4, S5-1 and S5-3 soils were formed under sub-tropical semi-humid climates with a tentatively estimated mean annual temperature (MAT) of at least 4-6 C higher and a mean annual precipitation (MAP) of about 200-300 mm higher than for the present-day, indicating a much strengthened summer monsoon. The annual rainfall was particularly accentuated for the southern-most part of the Loess Plateau, suggesting that the monsoon rain belt (the contact of the monsoonal northward warm-humid air mass with the dry-cold southward one) might have stood at the southern part of the Plateau for a relatively long period each year. The loess units L9 and L15 were deposited under semi-desertic environments with tentatively estimated MAP and MAT of only about 1.5-3 C and 150-250 mm, indicating a much strengthened winter monsoon, and that the summer monsoon front could rarely penetrate into the Loess Plateau region.

Correlation with marine carbon isotope records suggests that these climate extremes have large regional, even global significance, rather than being local phenomena in China. They match the periods with greatest/smallest Atlantic-Pacific 13C gradients, respectively, indicating their relationships with the strength of Deep Water (NADW) production in the North Atlantic. These results suggest that the monsoon climate in the Loess Plateau region was significantly linked with the North Atlantic thermohaline circulation on timescales of 10⁴ year.

Two main kinds of models have been proposed to explain the mechanism of the magnetic enhancement of paleosol developed on the Chinese loess, i.e. depositional model and pedogenic model. The dilution of a relatively constant of fine-grained strong magnetic dust from remote sources by weaker magnetic eolian silt of local source during cold episodes was suggested by depositional model. Pedogenic model, however, emphasizes the contribution of in situ production of ultrafine grained strong magnetic particles by chemical, biochemical or biogenic actions during soil forming processes. A technique of grain separation is used to study the grain size distribution of magnetic minerals in loess and paleosol. The representative samples of S1 and L2 taken from Jixian, Xifeng, and Xining along E-W transect of Loess Plateau were subdivided into eight grain size groups by both depositional method and centrifugal method. The products of grain separation were carefully weighed, and then the percentages of each group to the total sample were calculated. Systematic measurements of magnetic parameters were carried out on each group. The results of magnetic measurement show that:

There are several mountain ranges in the Northwest of China, such as the Tianshan mountains, the Qilian mountains, the Arerjinshan mountains,and the Kuen-Lun mountains, all with the elevation of 4000¡“5000 meters. In the alpine zone large quatities of glacier tills extend several kilometers from the end of existing glaciers. After the morains breaking away from the glaciers, the microbe communities appear immdiatelly on the surface of the glacial boulders, which start the soil-forming processes of alpine soils. From the appearance of microbial and algae communities, then lichen and moss grew,finally plant evolved into alpine meadow. with same pace of vegetation changes, the development of alpine soils is original soil firstly, then alpine desert soil, and alpine meadow soil in the end. The alpine soils have negative temperature and ice body, which undergoes a freezing-theawing-freezing process with state changes of ice-water-ice. Frost heaving and stone polygons developed on the soil surface, and soils have weak humification and decomposition of minerals.

This article mainly discuss the climatical and glacier changes as well as the pattern of the alpine soils evolution, and reveal the process of the soil evolution and lts characteristics in phisics and chemistry.

If no substantial erosion or deposition takes place, soil-forming processes modify exposed glacier sediments. These processes involve the vertical movement of carbonate, iron oxides, silicates and Other Chemicals in soil solutions. Compared to weathered tills, fresh tills were deposited in front of the existing glaciers, formed during the last 30 years. These tills extend outward 100-150m from the end part of the existing glaciers and harbour buried stagnat ice. Annual air temperature of the region of the fresh tills is about -7.0 - 7.5 ^oC with >400mm of precipitation. Because the fresh tills are a comparatively recent formations, there is no advanced plant lived on the surface of these tills. Original soil occurs in the above mentioned pedogenic settings.Pedogenic environmant of alpine original soil is characterized by the developmeut of microbial, algae communities and bryophyte and the beginning of weak humification on the surface of the debris. when the melttill is completely exposed from ice, microbial communities oppear immediately on the surface of the debris. In the initial stage of the life cycle of the communities one gram of soil substance contains about 10⁵ microbical bodies, and the amount of the living things increases double and redouble year by year.

The microbial communities involves 95% of no-gemma bacteria, 0.05% of gemma of bacteria, 0.04% of fungus and others.The accumulation of organicmatter created by the communites lead to the flourishing of lichen, such as Protococeus, Gloeocapsa, chrococcus and Stigouema etc., and thus the surface of the fresh debirs shows brown.Because of photosynthesis and holding water of algae,organic matter in the original soils measures by 3-5 g/kg, PH value of the soils at 8.0~8.1. Soil clay minerals mainly are composed of illite element.

The second group of glacier tills is present at the 500-1000m lower from modern glaciers. These are probably Little Ice Age sediments based on ¹⁴C dating(100-400 a B.P.) Compared to fresh tills, little Ice Age morains are Characterized by exist on their surface more abundant living things, including not only microbial and algae communities,but also lichen,bryophyte and padflora. Alpine desert soil, developed in the above mationed settings, has made great advances in their soil forming processes.It is a matter of common observation that there is a steady increase in the accumulation of humus, nitrogen-fixing and the cation exchange capacity of the soil. The amount of Al₂O₃, Fe₂O₃ and P₂O₅ in alpine desert soil is much more than that in original soil.Although soil clay minerals mainly are composed of illite, such as in original soil, but there is a great increase in the component of montmorillonite,wohich suggests that there is a stronger process of soil weathering in alpine desert soil.

The third group of glacier tills is located at 3000-5000 m far from the second (Little Zce Age tills),the radiocarbon age of which 2800-6000 a B.P refers to the climatic Optimum Period. The Neoglaciation tills are distinguished by on which high plant covering level (80-90 %) and powerful soil forming processes. The succession of living things evolved to the climax of alpine meadow flora, correspondingly developed the alpine meadow soil,the profile of which can be divided into three layers: upper is 8-10cm thickness of mattic epipedon (sod), it is formed due to unbalanced accumulationand decomposition of soil organic materials, and also it is suitable to be used as the diagnostic horizon of alpine meadow soil for Chinese Soil Taxonomic Classification. Midlle layear is about 10cm thickness of humus horizon,The amount of soil organic materials is about 120-240 g/kg. Lower is weathering horizon of Neoglaciaton tills or soil parent material. The quantity of CaCO₃ in the soil is about 28-40 g/kg with pH value 6.6-8.3.

The last proup of goacier tills is compared with the deposits of the Last Glaciatis Maximum(LGM), Located at 3000-5000 m far from the Neoglaciatios tills. It is coverd completly by alpine meadow and subalpine meadows flora (covering level >80 %).In the distribution region of LGM tills annul air tempratune is about -2.5-3.0 C, and annul precipitation is about 400 mm. Soil profile can be devided apparently into sod,humus,weak illuvial and weathering parent material layers. The amount of soil organic materials is about 187 g/kg, CaCO₃ 4-5 g/kg, PH value 7.5-7.8.

In the western part of Qilian mountains and Arerjin mountains,because the amount of pricipitation in alpine zone is very small (<300mm), the vegetation in alpine zone is mainly composed of steppe flora. In the above metioned environment, development of alpine steppe soil replaces that of alpine meadow soil. Alpine steppe soil is characterized by dry and cold conditions for soil forming processes,very little accumulation of organic materialsin soil(5-6g/kg) and very much content of CaCO₃ (100-150g/kg).

On the edge of the existing glaciers in alpine zone in high mountains of North western China Four groups of glacier tills have been deposited during the late pleistocene, corresponding the evolution of soil-plant systems can be divided into four stages. In the initial stage, microbial and algae communities appeared firstly and formed original soil.During the second stage,Lichen,moss and pad flora started their life and at same time developed alpine desert soil, and in the final stage, they evolved into alpine meadow or alpine steppe flora /soil.

It has been shown that the progressive uplift of the Qinghai-Xizang plateau and surrounding mountains resulted in the progressive process of dry in the Northwestern China since the late pleistocene.For example, only for the last 400 a, the air tempreature in alpine zone has rised by 1.0-1.3 C,at same time , the precipitation decreased by 50-60 mm, which led to the large quatities of glaciers disappeared and retreated. The evolution of alpine soils is closely related to the climatical and glacier changes.

THE YELLOW CAP ON QUATERNARY RED CLAY OF JIUJIANG, JIANGXI PROVINCE

The Quaternary red clay of Jiujiang, Jiangxi Province, always wears a "yellow cap", which actually is a layer of the Xiashu Loess. Several typical profiles were studied, and it was found that the modern soil developed on the Xiashu Loess was yellow-brown earth, the ratios of SiO₂/Al₂O₃ and SiO₂/(Fe₂O₃+Al₂O₃) of its clay(<0.002mm) were 2.98 and 2.30 respectively, free and active iron oxide degrees were 59.07% and 47.50% respectively, and the dominant clay minerals were hydromica, vermiculite and kaolinite, as well as some of montmorillonite; while the paleosol developed on the buried Quaternary red clay was red earth, the ratios of SiO₂/Al₂O₃ and SiO₂/(Fe₂O₃+Al₂O₃) of its clay(<0.002mm) were 2.40 and 1.85 respectively, free and active iron oxide degrees were 84.3% and 5.3% respectively, and the dominant clay minerals were badly crystallized kaolinite, as well as a little of hydromica and vermiculite. It indicated that the buried Quaternary red clay and its paleosol suffered higher weathering than the Xiashu Loess and its yellow-brown earth, and the climate of the former was warmer and more humid than that of the latter.

According to the principle of soil genesis, the mean annual temperature and annual rainfall suitable for the yellow-brown earth formation are 14.5-16.4^oC and 800-1300mm respectively; while those suitable for the red earth formation are 17.4^oC and 1710mm respectively. Hence, it was considered that the mean annual temperature when the buried red earth was formed were 1-3^oC higher than the modern climate when the yellow-brown earth was formed; and the annual rainfall 400-900mm higher.

Also, according to the chronology study, the Quaternary red clay was formed under several interglacial climates from the late Early Pleistocene to the end of the Middle Pleistocene; while the Xiashu Loess was formed under the last glacial stage of the Late Pleistocene, and it was ealian deposit formed in the dry and cold environment. However, Because the Last Glacial Stage was followed by the best suitable stage with warm and humid climate, the ealian loess suffered strong weathering afterwards, and under present post- glacial climate, it has developed to yellow-brown earth.

Paleosoils are known from several places in the Eastern forelands of the Harz mountains. By correlation with neighbouring areas, they are dated mostly as Eemian. No age determinations were made. By geological mapping, a new locality of paleosoils was found in an opencast sand mine near Hedersleben, about 8 km ENE of Eisleben, district-capital. The opencast mining is situated on the southern slope of the Laweke-valley leading to the Salza- and Saale-meadows. Within the slope, the opencast mining is located in a flat hollow of about 300 m in diameter wich is penetrated by two joining erosion channels. On the surface, calcaric regosols are developed in sand loess. They are associated with chernozems. The climate is dry and warm. Annual temperature amount to 8.5 Celsius degrees in average, annual percipitation is about 450 mm. From top to bottom, the Quaternary sequence consists of sand loess, solifluction and water displaced loess covering Saalian till, solifluction till and glaciofluviatile sand wich are underlayed by presaalian glaciolacustrine sand. Thickness of the singel beds is quite varying. The whole thickness of the Quaternary deposits in the outcrop is unknown. It reaches 22 m in a drilling 350 m to the North. In the mentioned well section the Quaternary sediments overlay sandstones and siltstones of the Upper Buntsandstein deposits wich contain four gypsum beds. The Upper Buntsandstein beds dip gently to the Northeast. They are cutted by the slope surface. Morphological position and increasing thickness of solifluction and water displaced loess as well as of glaciolacustrine sand in direction to the central part of the hollow make subrosion origin probable. Subrosion could be caused by solution of the gypsum beds. The subrosion depression led to the conservation of the paleosoil sequence.

The investigated paleosoil sequence is located on the eastern wall of the opencast sand mine. The wall shows a complete section through the Weichselian deposits of the subrosion depression. The paleosoil sequence is complex by means of erosion, deposition, soil development and solifluction. It is superimposed by calcareous impregnation. A cyclic succession within the loess sequence was observed. A complete cycle contains the following sediments and phenomena from bottom to top:

• coarser grain at the base; enrichment of gravel, pebble or stones of various origin often containing wind-shaped pebbles
• aeolian deposits: loess, sand loess, loess sand, slightly silty sand and sand
• soil development
• erosion, destruction of the soils and redeposition of solum material
• formation of compact solifluction beds (Fließerde), cryoturbate alteration

The observed paleosoil sequence is situated in the late Saalian to Weichselian periglacial region. That's why it is supposed that sediment cycles are caused by climate changes. This assumption corresponds with observations in lacustrine environments. The soil represents the most optimal climate conditions within each cycle. So the soil development reached stages from luvisol and podzoluvisol to chernozem.

A first age determination (14C, made by GEYH, Niederschsisches Landesamt Bodenforschung Hannover) results in 27,400 years b.p. for redeposited humous sediments at the top of a thick luvisol/podzoluvisol sequence.

Loess up to 75 m thick with over 21 interstratified palaeosols spanning the last 1.1 Ma years and perhaps the whole of the Quaternary covers more than 20 000 km2 of the Columbia Plateau, Northwest USA. The loess stratigraphy potentially contains one of the most detailed, high resolution proxy records of climatic change in this region during the Quaternary.

Repeated episodes of Pleistocene cataclysmic outburst floods from ice dammed glacial lake Missoula (Northern Montana) episodically surged across the region during the retreat of the Cordilleran Ice Sheet depositing extensive fine grained slackwater sediments. These were then deflated, transported and deposited as loess by regional westerly winds. The sub horizontal loess-palaeosol sequence of the Palouse being topographically above the Lake Missoula floodwaters did not incur significant erosion and represents the main body of loess on the Columbia Plateau. The sequence progressively thins towards the east the further the distance from the loess source regions along a strong present day precipitation gradient.

Research to date has primarily concentrated on the field identification and regional correlation of loess and palaeosol units of the last glacial cycle (Wisconsin) within the semi arid, high resolution sequences proximal to the source areas.

The main aim of this study is to provide a detailed palaeoenvironmental reconstruction of the evolution of the loess-palaeosol sequence at one site proximal to the sediment source areas, thus establishing a baseline stratigraphy beyond that of the last glacial cycle. Micromorphological observations utilising a pedosedimentary approach form the main basis of the study along with supporting palaeoenvironmental proxies such as particle size, bulk CaCO3, magnetic susceptibility and stable isotope analysis of pedogenic carbonate. Recently obtained thermoluminescence dates and known dated intercalated tephra layers provide a tight chronological control to the sequence.

This paper outlines a detailed study to obtain data on the age and paleoenvironment of Vertisols in order to reconstruct the history of the landscape and soils of the North Caucasus region. Data was obtained from a sequence of three Vertisol pedons - the depression, the microslope, and the mound of a gilgai complex. These soils formed on neogene marine clays in steppe of the North Caucasus, Russia. The gilgai of the area is normal with an amplitude of 0.3 to 0.5 m.

The soil pedons were examined and described in the field together with geobotanical examination of the site. Radiocarbon dating was obtained on carbonate nodules and on selected horizons from the three pedons. Phytolith analysis was performed on bulk samples > 2 mkm from the main genetic horizons across the microcatena. Stable isotopes of carbon and oxygen were used as indicators of carbonates and humus formation.

Contemporary wetness conditions in the microcatena is reflected in the C-14 dates from the deepest horizons (5610�180 in the mound, 3720�180 in the microslope and 2290�430 in the depression). The similar trend was found for carbonates nodules: 1880�120 (mound) and 1300�130 (depression). These Vertisols should have an age of at least 2500 to 3000 years more old if the coefficient of humus rejuvenation proposed by Alexandrovsky and Chichogova (1997) is applied.

The increasing of phytolith in abundance suggest a buried surface at a depth of 36-55 cm in the mound, 70-105 cm in the microslope, and 40-65 cm in the depression. The absence of diatoms in the mound pedon conforms that the site was not waterlogged prior to gilgai formation. Phytoliths of forest cereals were found up to 30 - 80 cm depth. Comarum palustre was found by geobotanical examination. This boreal species normally does not exist in the piedmont of the North Caucasus.

Forms of carbonates were investigated. It is believed that soft spots ( ¹⁸O =+23.2...+26.3 %_o;¹³C =-8.8...-11.5 +%_o) were formed in different environments than nodules (¹⁸O = +31.1...+32.2 %_o; ¹³C =-10.3...-11.8 %_o). Stable carbon and oxygen isotopes indicate the contemporary formation of the soft spots. The nodules are expected to be formed under a more humid and warm climate, probably of mid-Holocene.¹³C of humus is -25.5 %_o in the mound, decreasing to -23.3 %_o in the midslope (at 70-105 cm) and -23.0 %_o in the depression (130-150 cm). The last data from depression are characteristic of arid conditions of pedogenesis (Ryskov, 1997).

Analysis of the data shows a complicated paleoecology of the region during Holocene up to the present time including more dry period indicated by humus, more warm and wet period indicated by carbonates, and more humid and cold period indicated by forest species. The paleoenvironment is recorded along the gilgai microcatena. The most complete information is recorded in the microdepression profile. Nevertheless, it is necessary to study all gilgai complex to get the full and reliable data for reconstruction of environmental changes.

After the deglaciation approximately 12000 years B.P. climate on the territory of Eastern Fennoscandia changed radically several times. Both temperature and moisture content varied greately, thus providing vegetation successions and markable changes in soil formation. The following soil profiles are described as the examples for past Holocene climatic periods effect on soil profiles morphilogy.

• Modern Ferric Podzol formed in the top part of the E horizon of an Albeluvisol on moraine till. The Albeluvisol profile is believed to form in the Atlantic period (8000-500 years B.P.). The profile is described in Southern Karelia. Similiar profile in Northern Karelia is described; Ferric Podzol is formed in the E horizon of a Gleyic Albeluvisol on limnoglacial clay.
• Modern Gleyic Podzol profile formed in the top part of the Eg horizon of a Planosol on limnoglacial clay.
• Modern Ferric Podzol with a fragipan subhorizon. This subhorizon is ascribed to past high ground water level. Such a horizon usually formes in soils near the lakes and bogs; under fragipan soil usually has traces of reductomorphism, such as ortstains or iron accumulations. Similiar fragipan horizons are described in Gleyic Podzols on the border of reductive and oxidative conditions. The phenomena is widely presented all over Eastern Fennoscandia.
• Gleyic Podzol on limnoglacial sand profile under a thick (50 cm) oligotrophic peat layer (Bolshoi Klimetsky island, the Onega lake). The observed palludification is connected with the Onega lake level fluctuations.

The poligenetic profiles described formation is ascribed to climatic changes in Holocene on the territory of Eastern Fennoscandia. In most cases the evolution led to hydromorphism reduction. However, big lakes (such as Onega and Ladoga) level fluctuated significantly, and usually the ground water level in soils of low lake terraces even increased.

Vegetation succession seems to be a significant factor in soil genesis changes. We consider that Albeluvisols and Planosols formation was connected with the Atlantic period luxuriant vegetation.

We propose that in most cases climatic, water level, and vegetation changes were relatively rapid, so that a new profile started its development, while the older one was preserved practically untouched.

The latest research demonstrated that the Qingzang (Tibetan) Plateau began rapid uplift at about 3.6 Ma. Subsequent rapid uplift commenced at about 2.6 Ma and 1.7 Ma. These three phases (A, B and C) of the tectonic uplift, called Qingzang or Tibetan movement, have caused the previous planation surface under lowland environment to change into a high plateau with high mountains and deep basins, and the reorganization of drainage. Large drainage systems such as Huang He (Yellow River) and Chang Jiang (Yangtze River) appeared since 1.7 Ma and started to incise the Plateau. Accordingly, fan boulder conglomerates were formed around the rim of the Plateau and feet of mountains on the Plateau. For example, on the northern slope of the intensely uplifted Himalayas, conglomerate deposits called Gongba conglomerate superimposed unconformably on the late Cenozoic lacustrine stratigraphy containing Hipparion, yielding strong evidence to mountain rising. the previous view of regarding this conglomerate as glacioaqueous deposits seems inappropriate. From 1.2 Ma, the Tibetan Plateau experienced again an intense uplift as indicated by shrinkage of early Pleistocene lakes such as those in Kunlun Shan and Qaidam Basin and appearance of new fan conglomerates such as those at Mengda Shan in the Xunhua Basin with ages of 1.2 - 0.8 Ma. We called this tectonic uplift Kunlun-Huang He (abbreviated Kunhuang) movement. It caused strong left-lateral striking slipping along Kunlun Shan and Qilian Shan and rising of Kunlun Shan to high mountains. The uplift contributed by vertical component of this striking slipping reaches over 1,000 m. A large body of evidence has concluded that only after 0.8 Ma, much of mountains on the Plateau rose over snow line, beginning to form large scale of glaciation. However, subsequent uplift has led the Himalayas as a barrier to hinder the inflowing of moisture of Indian monsoon to the Plateau, causing a drying of the Plateau and rising of snow line, thus the scale of late glaciation decreased, rather increased. Up to now, the maximum scale of the earliest (at or < 0.8 Ma) glaciation on the Plateau was only hundred thousands of square kilometers. Even in the source area of Huang He, there may have developed a small ice sheet of 50,000 km². But the whole Plateau was never covered by a big ice sheet. The Quaternary Plateau big ice sheet as proposed by M. Kuhle, Han Tonglin and others was not the case.

Li Xuyong¹⁾, Libaoguo¹⁾ and Zhong Junping²⁾

2)Department of Agronomy ,Xinjiang Agricultural University,Urumqi,830052,China

A compartment model dynamic equilibrium process of carbonate has been developed in Luochuan loess section of China. There are five major components in the simulation model: soil water balance model(including processes of soil evapotranspiration, runoff, infiltration and percolation), carbonate input of eolian dust, carbonate output of erosion, chemical thermodynamic relationships of CaCO₃ and CaCO₃ flux. The daily weather data is generated by a climate stochastic model(CLIMAK). The time scale is obtained from Kukla magnetic susceptibility age model. In the compartment model, the soil column is defined by a vertical sequence of 1-cm²-area compartments, each with a specified texture, bulk density, property of water-holding, pressure of soil-air CO₂, ionic strength, temperature and parameter of vegetation. We derive a group of equations and use them to calculate net carbonate flux, carbonate flux by rain, eolian dust and erosion in each soil compartment. The influence of upper compartments are also determined.

The carbonate distribution and content that our model predicts closely resembles the measured result. This modeling indicates that the mean carbonate fluxes by rain, eolian dust and erosion are 0.14g/cm² ka,1.45g/cm² ka, and 6.89¡Á10^-5g/cm² ka respectively since the Late Pleistocene in Luochuan section. Evidently, eolian dust is the main source (91.55%) of carbonate. The rate of carbonate influx is much higher than that in soils with “non-loess” parent material. The result indicates that in the parts of S0, L1, and S1, the carbonate influxes are 2.12g/cm² ka, 1.53g/cm² ka, and 2.60g/cm² ka. On the basis of this model, we quantitatively deduce the past climate (precipitation and temperature) over intervals of paleosols The model is also a valuable research tool for evaluating the role of state factors on soil CaCO₃ .

Key words: Compartment model, Carbonate, Loess-paleosol sequence

P.O. Box 821, nanjing210008, p. R. China

LIU Xiu-Ming

The loess and palaeosol sequence in Chinese Loess Plateau provides one of best continental geological record for Quaternary climatic historic study. Rock magnetic has proved to be one of most successful technique for the study, as the magnetic susceptibility is used as an index of pedogenesis in the record and widely accepted as an indicator of summer monsoon intensity for the past decade. It has been found that magnetite and maghaemite are the two main contributor to the magnetic susceptibility. Maghaemite is a common magnetic mineral in surfacial soils which has very similar physical properties to magnetite. For a long time investigators always simplify to treat maghaemite together with magnetite due to difficulty of separation and similarity in their physical properties. Some investigators once believed that chemical CBD technique can selectively remove pedogenic maghaemite and suggested that magnetic susceptibility signal was dominantly contributed by maghaemite rather than magnetite in Chinese loess. This CBD technique with theory of selectively remove pedogenic maghaemite has been proved to be not true by late compared studies. However maghaemite has been reported to be a thermally unstable magnetic mineral, perhaps to be the only exception to be distinguished from magnetite. By applied thermomagnetic technique to Xifeng and Luochuan samples from the top to L2, thermally unstable component can be systematically separated, which is about 10 (in soils) to 35 (loess) percentage of magnetisation, associated with a kink appears in about temperature 350 °C in both SIRM stepwise demagnetization curve and susceptibility-temperature curve. The thermally unstable component in Chinese loess and palaeosols is therefore assumed to be maghaemite in this study. Although maghaemite occurs in greater amounts in palaeosols than in loess units, its relative contribution to magnetisation is always low in soils and high in loess units, suggesting that at least some of the maghaemite is detrital, from the aeolian source area, while some maghaemite could be formed during pedogenesis. The magnetic susceptibility of loess and palaeosol sequence is therefore actually a comprehensive indicator of reflecting both source area and deposition area palaeoenvironments.

Department of Geography, Lanzhou University, 730000

Quaternary red earth covers an extensive area of eastern China. The typical Quaternary red earth, which consists of a sequence of modern soil horizon, homogeneous red earth horizon and net-like horizon, range from several meters thick to more than ten meters. The Quaternary red earth were sampled at 20 cm vertical in terrace of Qiantang River near Hangzhou. Magnetic susceptibility, frequency dependent magnetisation, saturation isothermal remanent magnetisation (SIRM) and coercivity (Bck) of SIRM were measured on doils. A selective dissolution was used to dissolve secondary ferrimagnetic from the samples. The x was measured before and after selective dissolution to determine the role of pedogenesis in the enhancement of ferrimagnetism in the paleosols. The magnetic susceptibility and SIRM was highest in the paleosols, and lowest in the net-like horizon. According to magnetic properties, paleosols can be divided into several layer with different magnetic characterization, indicating pedogenic environment changes. These differences are considered effective indicator for the quantitative study of the paleoclimate. Results also revealed that the content of single domain and superparamagnetic material was higher in the paleosols. This increase can be explained by the possibility of the in situ pedogenic formation of ferrimagnetic materials. This conclusion is based on the followings (1) magnetic susceptibility enhancement with increasing soil development: (2) frequency dependent magnetization; (3) the loss percentage of x by selective dissolution. The low x of net-like horizon can be explained by the reduction of ferrimagnetic minerals. There is no obvious difference in total iron and free iron content among Quaternary red earth profile. Therefore, magnetism studies on Quaternary red earth are undoubtedly useful to Quaternary environmental changes.

Zones of incipient soil development are often observed within relatively thick proximal loess of Late Pleistocene and Holocene age in central North America. Similar soil zones are present in part of the Oligocene Brule Formation in western Nebraska, U.S.A., which is interpreted as volcaniclastic loess. These zones apparently represent centennial- or millennial-scale fluctuation in the balance between loess sedimentation and pedogenic processes; thus, they may provide important evidence of climatic change at these time scales. Incipient soil zones typically display relatively dark colors, slightly higher contents of clay and organic matter, pedogenic granular or fine blocky structure preserved to varying degrees (sometimes evident macroscopically only as a subtle fabric), pore coatings, burrows, and other evidence of faunal activity. One “end-member” interpretation of these zones is that they entirely reflect local increases in temperature, moisture, and/or primary production, leading to enhanced organic matter accumulation, faunal activity, and weathering. At the other extreme, they could entirely be the result of a decrease in the rate of dust flux from a local loess source, possibly but not necessarily driven by climatic change. This would lead to a higher proportion of far-traveled fine silt and clay in the accumulating loess, and more time for organic matter accumulation and faunal activity prior to deep burial through ongoing loess deposition. Higher clay content would also help preserve organic matter and pedogenic structure after burial.

High-resolution particle size analysis of the silt fraction, using electroresistance or laser diffraction techniques can help test these explanations of incipient soil zones. The Jules Soil zone, formed in the Peoria Silt (= Peoria Loess) of southwestern Illinois during Marine Isotope Stage (MIS) 2, is associated with a pronounced shift toward a finer primary grain size mode as determined by high-resolution analysis. A prominent “shoulder” of fine silt suggests greater contributions from distant dust sources during formation of this zone, and exceptionally coarse silt in the loess just above it may represent high dust flux from the local Illinois River valley source, perhaps helping preserve the soil zone through rapid burial. Zones of incipient pedogenesis in the Peoria Loess near Omaha, Nebraska, U.S.A., display a pronounced bimodality. The coarse mode probably reflects dust influx from the nearby Missouri River valley while a fine silt-coarse clay mode may represent farther-traveled dust from the Great Plains. Less pedogenically-altered loess above and below these zones is unimodal or much less bimodal, suggesting that the soil zones formed during periods of reduced local dust flux from the Missouri valley. Loess of the Gilman Canyon Formation (MIS 3 and early MIS 2) in the same area is pedogenically-altered throughout. Some parts of this formation display bimodality but they are not always those with the best preserved pedogenic features. Incipient soil zones in the Loveland Loess (probably MIS 6), also near Omaha, are not consistently associated with changes in loess sedimentology. Ongoing research is applying high-resolution particle size analysis to incipient soils within the Bignell Loess (MIS 1) and the Oligocene Brule Formation.

1. In many discussions of palaeosols no great distinction is drawn between the palaeosol and the regolith, and it is commonly assumed that they are the same thing. In Australia this is clearly not so. Deep weathering profiles are extensively developed, hundreds of metres deep in places, and these are very different from soil profiles.

2. The weathering profiles are ancient, the soil on top usually young or modern. Palaeomagnetic dating of regolith goes back 100 million years, and stratigraphically dated profiles to the Eocene.

3. The weathering profiles are out of equilibrium with the present environment, having formed under a different climate, vegetation and hydrological regime. The soils are approximately in equilibrium with present conditions.

4. The regolith profile is often complicated by an unconformity, with saprolite below and disturbed material above. Even the upper material may be of considerable antiquity.

5. The concept of regolith as parent material is not always simple. In one situation weathering zones that were approximately horizontal crop out on slopes and give rise to a catena that is due to different parent materials rather than modern slope, hydrological or soil processes. In other situations even the old regolith had topographic variability, which is inherited in more complex ways.

6. Where lava flows or duricrusts are present, inversion of relief is extremely common. This produces a characteristic suite of landforms and regolith, reflected in the modern soils. In some places inversion has been repeated several times.

7. Slope mantles affect the distribution of soil and palaeosols. In ancient profiles the disturbed layer is often a footslope deposit, sometimes separated from saprolite by a stoneline at the base. Some Quaternary and modern mantles may be periglacial, though this process is rare in Australia outside Tasmania. Other, resembling periglacial deposits, are probably related to swelling clays.

8. The ancient regolith was so deep that it affected the nature of derived sediments. Early Tertiary streams had a coarse bedload of only quartz because the entire landscape was deeply weathered. Younger Tertiary and modern sediments are polymict, as more and more bedrock was exposed to erosion as the old regolith was stripped.

9. Regolith stratigraphy requires much more complex analysis than the simple law of superposition used in conventional stratigraphy.

10. Despite high latitude Australia had a warm and wet climate in the Mesozoic/ early Tertiary when the deep weathering profiles were formed. In the Miocene increasing climatic variability and aridity, coupled with tectonic uplift, led to widespread erosion of ancient regolith, and the beginnings of younger regolith and modern soils.

Thick loess deposits and alluvial sediments with inclusive paleosols form the most complete terrestrial stratigraphic sequences in the geologic record. Precise interpretation and correlation of loessial or alluvial sequences require careful and complete field observations and descriptions. Soil horizonation, weathering zones, and post burial changes are superimposed on sediment bodies. Detailed descriptions of stacked sequences must independently, and in concert, portray the relationships of soil horizons, fine geologic structure, weathering zones, changes in facies, and major stratigraphic boundaries (both pedostratigraphic and lithostratigraphic) including unconformities. The description recorded at a specific site should also be directly comparable to other sites of the same sequence. Furthermore, scientists of various backgrounds (e.g., geologists, pedologists, and archeologists) and geographic proclivities work cooperatively, but also independently in the study of loessial or alluvial sequences. To enhance both cross discipline and global communication we propose that a common convention and nomenclature be established for field description. Our system combines soil horizon nomenclature, terms for weathering zones, and stratigraphic terminology. Present nomenclature, however, can not denote all the relationships that exist in stacked sequences. We propose adding several terms that denote the occurrence of paleosols, geosols, and both major and minor stratigraphic changes. New soil horizon descriptors are also needed to indicate post burial changes and/or mixing of soil horizons during burial, and as indicators for paleosol classification. We will present examples of both loessial and alluvial sequences in the mid-continent USA to explain our description system.

The global small-scale paleoclimatic reconstruction for the end of Permian and the beginning of Triassic periods show that warm nonglacial type of climatic zonaliy determined on the Earth not in the beginning of Trias, but essentially earlier, namely by the end of Permian. The huge belt of red calcareous beds and related specific gley-glossic paleosols within them covered almost all medium and low latitudes of the land. Outcrops of the late Permian Polder and Salarevo formations (northern part of Russian plain) studied deposits consist of alternation of three kinds of members: gray colored sandstones, non laminated homogeneous red colored mudstones and brightly patchy colored stratified mudstones containing about 5-10 individual paleosols directly superimposing each other. The parent material for each paleosol within every soil profile is rather alike: red, nonlaminated, carbonate-rich mudstones. The superimposing stratification of paleosols allow to assume the pulsatory regime of sedimentation of the deposits enclosing these paleosols. All field-studied paleosols occuring within the multiprofile red colored mudstones have a similar type of profile, i.e. the near set of pedogenic horizons. The most part of these paleosols did not retain their surface organic horizons: they were either completely eroded or mineralized. The core repeating morphotype of studied late Permian paleosols consists of: three main types of pedogenic horizons, superimposing each other from top to bottom and holding the most part of soil record or memory. Eg-horizon, 10 (15) cm thick, the most bleached in profile, white when dry and olive-gray when moist; some pale-red residual spots or even “shadows” of the bleached former pieces of bright-red parent material occur on such background; the lower boundary of horizon is typically glossic with the deep penetration of bleached whitish-olive-gray silty-clay carbonate-rich material along the root channels into the deeper carbonate red beds. This is very unusual type of glossic phenomena: vertical zones around and along the paleo root channels form a very complex and fractal network of bleached tubes narrowing with the depth. EgB -50 (70)cm thick, the most striped and patchy, fractal organized horizon, consisting of predominating bright silty-clay carbonate-rich red beds penetrated by riotous dendroidal network of white or olive-gray tubes (diameter from 3-5cm to 1-3mm) similar by texture and carbonatization with the background. Some of these glossic tubes can pick not only this horizon but the whole 2m.thickness of individual soil profile and enter into the underlaying one; the red background of horizon have complex platy-blocky-prismatic structure and moderately developed illuvial argillans and stress cutans - slikenslides. BC -50(120)cm thick, rather homogeneous in color, red carbonate-rich mudstones with rare thin olive-gray tubes (along the deep roots). Paleobotanists described in these paleosols two new species of root systems (Radicites erraticus and R. sukhonesis Aref. & Naug., 1998) belonging to the grass plants with the deep abundant subvertically dendroidal roots and having some features of the perennials ephemeroids. All studied samples of these soils have very low content of Corg., alkaline to strong alkaline pH (8.5 - 9.6), very variable content of carbonates (8 -50 %) and exchangeable cations. The smectites predominate in the clay fraction which contents also some amounts of chlorites and illites; the kaolinites are completely absent. The content of Fe2O3d. in bleached tube zone substance ( about 0.3 %) are in order less then in red zone substance (3.0-3.6 %). At the same time the suit and composition of the phyllosilicates persist practically the same in these two contrasting zones of paleosols. It means that these glossic bleached root tubes were formed mainly by the “pure” gleyic-stagnic transformations of the free iron oxides without any alkaline aggression against the phyllosilicates. We think that it is the main biota-connected pedogenic process that defines the most bright morphological features in studied paleosols, so these features can be used both in paleopedogenic and stratigraphical purposes. The second bright process is the carbonate impregnation of all paleosol profiles, but up to now it is not clear if it is the real pedogenic process, or lithoinheritanse, or result of some postpedogenic processes.. The other important pedogenic processes in these paleosols are: the formation of complex multiordered pedality in EgB and BC horizons, lessivage and vertic processes. The field study of the multistage paleosoils within the red-colored members of late-Permian Polder and Salarevo formations allow us to frame the working hypothesis of the successive rotation of the pedogenesis and sedimentation in this period of time. The red carbonate sediment accumulation fulfilled discontinuously and the non laminated red beds about 1-2 m thick was deposited in each discrete cycle. After it , during the discontinuity in sedimentation, the subaerial epigenesis of this sediments was beginning, i.e. in situ weathering and pedogenesis. The land surface was covering by grasses with deep root systems that formed step by step the each particular profile with the gley-glossic bleached tubes deep fingering into the red parent material of sediments. The combination of three main features in studied paleosoils and sediments (total carbonatization, bright red color and stagnic-gleyic bleached tubes along the deep root systems) allow us to suppose that these paleopedosediments were developing in very contrast climatic and hydrological environment. The sedimentation properly intervened in hot arid climate and in oxidizing geochemical conditions mainly by sheet wash of red silt and clay particles and its accumulation in shallow lagoons and plain deltas. But after sedimentation the posterior pedogenesis was governed not only by general climate aridity but also by rainy seasons of climate and by periods of surface and subsurface inundations and waterloggings. The excess of water coming from atmosphere during wet seasons in combination with additional water coming from higher continental slopes and/or from the nearest see or lake shores created the seasonal excess of moisture within the soil, particularly around the roots as the source of reducing-agent, i.e. organic matter . Such combination of seasonal climatic and geomorphic hydromorphism in soils of arid regions is rather ordinary for recent pedosphere, but the specific imprinting of the hydromorphism in a form of studied gley-glossic red calcisols is rather rare and, as it seems, has no close analogs among described recent soils. The main problems for further investigations are the following: to determine the age and duration of pedogenesis in every individual paleosol, to understand the fate of iron leaching from gleyc-stagnic bleached tubes, to study the silicates weathering and carbonate generations to discriminate the paleosedimentation processes and paleopedogenesis properly.

The Lanzhou-Linxia area and Hezuo, located respectively in the western Loess Palteau and the NE Tibetan Plateau, have developed very thick loess-paleosol sequences favorable for study of high-resolution climatic change. The precipitation is not so high, but varies very largely both seasonally and trans-annually. Carbonate leaching is very sensitive to this process, thus can be served as a proxy for precipitation change. 21-point running average of densely-sampled carbonate curves from the shajinping and jiuzhoutai sections in lanzhou, beiyuan section in linxia, and hezuo section on the ne tibetan plateau show a very comparable pattern to the stacked marine isotope record during stages 1-6 (martinson et al., 1987), suggests that carbonate as a proxy for climatic change at tens of thousands of year scale can give convincing result.

High-resolution carbonate records from those sections have revealed at least 21 valleys running at millennial scale, most of which seem to be correlated with most warm peaks of the GRIP isotopic record. This suggests that carbonate as a proxy for revealing high-resolution climatic change is also practicable.

Moreover, high values with high frequency and large amplitude of fluctuation of carbonate content are observed for the early last glacial, particularly between isotopic warm peaks 18 and 19 in the GRIP record. This may indicate that moisture kept at a high level and varied very largely in the early last glacial. Thus a cold-humid environment is suggested for that time. Since magnetic susceptibility is very low for this part of the stratigraphy, it seems that carbonate could become a more sensitive and practicable proxy for climatic change in cold-humid environment than susceptibility.

Soils and palaeosols are the reliable climate indicators and records. But the soils developed in different environments possess different characters. In the aspects of palaeoclimatic reconstruction, finding out suitable climate proxy is one of the paramount and most difficult tasks.

The processes of soil development are not only controlled by the temperature, precipitation and materials, but also the intensity of bioactivity and bioproductivity, ground water level and physical property of the materials. Therefore, the soils developed in different environments are various and their is no common suitable climate proxy.