The dark colored organic soils represented by A horizon of Dark loessial soils, called “Heilutu”, are extensively distributed on the Loess Plateau, China. They show that the Plateau was formerly covered with dense vegetation. On the other hand, in the upper drainage basin of the Qingshui-he River, which flows into the Huang-he at the western part of the central Loess Plateau, Holocene river terraces are widely developed and their deposits consist of gravel bed, peat layers and loessial flood loam bed in ascending order. The peat layers intercalated in the terrace deposits indicate that a peatland was widely expanded along the Qingshui-he during a period of Holocene time. The authors examine the stratigraphical relationship between the peat layers of Holocene terraces and the dark colored organic soils distributed on mountain slopes, and discuss the environmental changes based on the changes in facies of terrace and slope deposits as well as 14C dates of peat and soils. The upper drainage basin of the Qingshui-he, whose altitude ranges from 1, 500 to 2, 900 m above sea level, is situated in a semi-arid region with annual precipitation of 400 to 650 mm, and a part of the basin is occupied by a present peatland with small area. The peat layers of the present peatland are divided into the following four layers; the lower, the middle, the upper and the uppermost, by intervening clay layers. The four layers are different in color, thickness and in content of organic matter and clay. The middle peat layer is the most developed with dark color and rich in organic matter. Three peat layers except the uppermost layer are recognized in the Holocene terrace deposits. The dark colored soils on the mountain slopes and the Pleistocene terraces are stratigraphically correlated to the middle and upper layers of the peat deposits. The 14C dates show 11, 800±200 B. P. (TK-826) for the lower part of the lower peat layer, 7, 980±80 B. P. (TK-827) for the lower part of the middle peat layer and 5, 040±70B. P. (TK-828) for the upper part of the upper peat layer at Kaicheng of 1, 930 m in altitude. The lowest part of dark colored organic horizon of Heilutu distributed on the mountain slope at Mt. Yunwu-shan (2, 148 m a.s.1.) shows 6, 280±70 B. P. (TK-829). The dark colored organic horizon are proved to have been formed mainly during the period from 8, 000 B. P. to 5, 000 B. P., meaning that the Loess Plateau was covered with dense vegetation during the Hypsithermal. The uppermost horizon of loessial flood loam covering over the Holocene terrace consists of thin and light colored humus, indicating that the formation of organic soils with weak activity was continued after 5, 000 B. P. The changes in facies of terrace and mountain slope deposits, and in vegetation density suggest the environmental changes as follows; The gravel bed of the lowest layer of Holocene terrace deposits, forming alluvial fans at mountain foot, indicates that the stream was powerful, suggesting that heavy rainfalls occurred frequently before 12, 000 B. P. During the period from 12, 000 B. P. to 8, 000 B. P., the periglacial debris layers were deposited on the mountain slopes, and the clayey deposits without gravel along the river courses. The former shows the cold and wet condition with sparse vegetation, and the latter the small precipitation. Therefore, the cold and dry climate but the wet soil condition are inferred. During the period from 8, 000 B. P. to 5, 000 B. P., the dark colored organic soils were developed even in the present arid zone with annual precipitation less than 350 mm, indicating that it was more humid than the present. The periglacial action on high mountains ceased under dense vegetation cover, indicating that it was warmer.
It is the purpose of this paper to clarify relationship between forest distribution and its controlling factors such as cumulative air-temperature (degree-day), landform characteristics and snow-cover duration using remote sensing images, DEM (Digital Elevation Model) and climatological data. The outline of procedure of investigation will be described below. The spring, summer and autumn LANDSAT TM/MSS images were used to analyze seasonal change of vegetation and snow-cover. At first all images were converted into images resampled on a 56/58 m grid of DEM by using the nearest neighborhood method. Then the image data were processed in order to eliminate shadows caused only by azimuth and dip of each grid of DEM and direction of sunbeam. This processing is useful to improve the accuracy of landcover classification for a high relief mountainous region like this study area, the northernmost part of the Hida Range, Central Japan. Finally five categories ; dwarf pine, evergreen coniferous forests, deciduous broad-leaved forests, alpine meadow, bare grounds were interpreted, and a digital map of the estimated date of snow disappearance was obtained from the multi-temporal images. Air-temperature for each pixel of the DEM was calculated using daily surface air-temperature at a local AMeDAS station (Automated Meteorological Data Acquisition System) and daily lapse rate in free air at the nearest aeorological observatory. The estimated values were tested comparing the values observed at temporal observatories of mountain huts during summer season. The cumulative air-temperature is defined as an integrated excess of 5°C of daily air temperature, and it is widely admitted that it controls primarily the distribution of forest plants in rainy Japanese Islands. The other controlling factors such as snow-cover duration and slope azimuth and dip were discussed statistically. In conclusion ; (1) Dwarf pine (Pinus pumila) is distributed in gentle and westward slopes of ridge zones with the maximum occupation ratio at ca. 2, 500 m. This is replaced by alpine meadows or bare grounds in and around snow patches durable to late spring or summer. (2) Evergreen coniferous (Abies mariestii, A. veitchii, Tsuga diversifolia and Picea jesoensis var. hondoensis) forests are distributed with the maximum occupation ratio at ca. 2, 000 m. The coniferous forests will prefer a habitat which is West- to North-facing slope with inclination of less than 25. In the higher zone than 2, 000 m, these forests compete with dwarf pine or birch (Betula ermanii) forsts for occupation according to controlling factors. The coniferous plants are seemed to be snow-intolerant and easily damaged by a loss of the cumulative air-temperature due to snow-cover duration, and by mechanical and physiological injury by snow-cover itself. (3) Deciduous broad-leaved (Fagus crenata, Quercus mongolica var. crispula, Betula ermanii etc.) forests are widely distributed in this area. The Occupation ratio curve has bi-modal peaks at 2, 300 m and 1, 500 m. The lower one indicates the normal altitudinal distribution of forest zone according to thermal condition. The higher one is introduced by occupation of Betula ermanii in alpine-subalpine transitional zone. Betula ermanii is snow-tolerant and distributed in mountain slopes of all directions with the maximum frequency in South to East. The east-facing slope is leeward of the winter snow-bearing winds, but receives much snow accumulation by the blowing effect. Betula ermanii will occupay a habitat unfavorable for the coniferous mainly because of much snow accumulation.
The southern part of Kii Peninsula faces to the Nankai Trough where the Philippine Sea Plate is subducted under the Eurasian Plate. In this region, we have experienced a lot of uplifts associated with great earthquakes in the historical period. The authors aim to clarify the mode of tectonic emergence in Kii Peninsula during the Holocene based on geomorphic and biological sea-level indicators. The Nankaido earthquake (M=8.1) of 1946 is the most recent great earthquake along the Nankai Trough. The mode of tectonic movement before and after the Nankaido earthquake was known by the precise re-leveling measurement. Before the earthquake, the northern part of the peninsula had been uplifted and southern part had been subsided slowly. On the contrary, at the time of the earthquake, northern part was subsided, while the southern part was emerged abruptly. After the earthquake, tectonic movement turned into the similar mode as before the earthquake. YONEKURA (1968) estimated the residual uplift during one great earthquake cycle at Kushimoto as approximately 0.2-0.3 m. He also suggested that great earthquakes such as the Nankaido earthquake of 1946 had recurred at an interval about 110 years, and the residual uplifts during earthquakes and intervening period had been accumulated through the late Pleistocene, forming emerged marine terraces. In Kii Peninsula, evidence for former sea levels is recognized as notches, wave cut benches and the calcareous remains of attaching organisms living in the tidal zone. Based on the recognition of highly concentrated zone in the vertical distribution of these former sea level indicators, six former sea levels are distinguished in the southern part of Kii Peninsula : (I) 5.8 m, (II) 4.1 m, (III) 3.3 m, (IV) 2.8 m, (V) 2.0 m and (VI) 0.8 m above the present mean sea level. These former sea levels are dated : (I) 6, 000-5, 500 yr BP, (II) 5, 000-4, 000 yr BP, (III) 3, 800-2, 600 yr BP, (IV) 2, 400-2, 000 yr BP, (V) 1, 800-800 yr BP and (VI) 600-200 yr BP by the radiocarbon method. We may regard the cause of drastic drops of sea levels as coseismic uplifts named event 6 to event 1 in chronological order. Event 1 to event 6 must have occurred at 200-0 yr BP, 800-600 yr BP, 2, 000-1, 800 yr BP, 2, 600-2, 400 yr BP, 4, 000-3, 800 yr BP and 5, 500-5, 000 yr BP respectively. It can be said that these earthquakes have recurred at an irregular interval of more than several hundred years during the late Holocene, and the residual uplift during each earthquake cycle might be significantly larger than the value of 0.2-0.3 m suggested by YONEKURA (1968). Earthquakes such as the Nankaido earthquake of 1946 might be inter-plate type, and the residual uplifts during these earthquake cycles could not be accumulated from the view point of the elastic rebound theory. By the way, the value of 0.2-0.3 m as the residual uplift associated with Nankaido earthquake of 1946 type estimated by YONEKURA (1968) is unreliable because the precise re-leveling measurement was not done throughout a whole interval of earthquakes. The earthquakes reconstructed in this paper, i.e. event 1 to event 6 are not of inter-plate type, but of intra-plate type which have resulted in the cumulative uplifts of Kii Peninsula. The patterns of uplifts caused by the events can be classified into two types : (1) amounts of uplift of event 6, event 4 and event 2 are larger in the southwestern part of the peninsula, whereas (2) those of event 5 and event 3 are larger in the southeastern part. It is difficult to distinguish the amount of uplift associated with event 1 from the present residual uplift of the Nankaido earthquake of 1946.
Lateral change of left-lateral strike-slip displacement along the Kaminirogawa fault was clarified by using a displacement-distance method. The Kaminirogawa fault is one of the major strike-slip faults in the Outer Zone of Southwest Japan, and measures 90 km in length on land-area from west of Tokushima to east of Kochi. The fault cuts and displaces the ENE-trending Sambagawa, Chichibu, Kurosegawa, Sambosan, and Shimanto Terrains. Eight values of left-lateral displacement along the fault were obtained by correlating geological markers on both sides of the fault, such as fold axial surface, faults, and bedding surface. They were plotted in the displacement-distance graph. The distance is measured along the fault from the arbitrally located reference point to the middle point between displaced geological markers on both sides of the fault. The graph shows that the displacement distribution is a triangle-type. The displacement is maximum, 12.9 km, at the center of the fault, and it decreases toward both northeast and southwest at a constant rate. The rate of displacement change is 0.206, which means about 2 km displacement increase along 10 km fault trace. Northeast and southwest fault tips of the fault are inferred to be situated at west of Tokushima, to the south of the MTL, and at 30 km south of Kochi in Tosa Bay respectively. Thus, the fault measures 125.2 km in total length. The Kaminirogawa fault is not a Riedel-shear type fault associated with the MTL. The displacement distribution of the triangle-type is well explained by a dislocation model of fault propagation and displacement increase. The model requires linear positive correlation between length and maximum displacement of strike-slip faults. Formerly clarified their correlation (RANALLI, 1977) is nonlinear positive, and is explained by the changes in the growth rates of displacement and length. Transform, transfer, and trench-linked transcurrent faults should be excluded in considering correlation between length and maximum displacement of strike-slip faults, because they abruptly decrease displacement at their fault “tips”.