open access

https://hdl.handle.net/2115/32914

Spatial analysis of soil properties in the field located on complex hills

日本語

2008-02-29

北海道大学大学院農学研究院

Memories of the Research Faculty of Agriculture, Hokkaido University

29(2)

215-308

Soil scientists and agronomists have recognized variation of soil properties in time and in space. Especially at the upland field located on complex hills, the problem to enhance soil variability becomes apparent as same as soil degradation by erosion. The same soil management and cultivations can’t resolve the heterogeneity in the field. In some cases the practice rather accelerates the variations and cause environmental pollution by human-introduced materials like pesticides or nutrients. Therefore soil and environmental management systems correspond to spatial variations is searched, the demands for more accurate information on the spatial distributions of soils will increase. Climate, parent material, topography, biotic factors and human activities influence soil formation, but climate often exerts control at coarser scales than small watershed or than one field of interest here. I selected some fields with relatively uniform parent material and cultivation practices, so that spatial soil variation can be attributed to changes of topography. Soil variation has been formed by the movement of water and soil itself through and over the landscape, so soil properties, soil erosion class and productivity are related to landscape position. But the methodology hasn’t established or there has been no application for individual soil environmental conditions or agriculture systems in Japan. This paper describes the spatial variability of soil at typical land use on complex hills in Japan. Furthermore, I estimate the validation of topographical analysis and geostatistics which identify soil properties distributions and reduce soil variability. Soil survey was performed at a 2.8 ha upland field in Kamifurano of central Hokkaido which had been cultivated for about 80 years. Topography of the investigated field was undulant hills, average slope is 0.24 mm-1 and 80% of the field was steep slope over 0.20 m m-1. The maximum annual soil erosion rate was 11.5% of topsoil (29 mm y-1), which was estimated by analysis the inventories of 137Cs in soil profiles. Soil profiles could be identified by the positions on slope: the horizon sequence was observed in which Ap / C - type occupied on the upper and convex areas, contrarily on the lower and concave areas, Ap / A / B - type. At the eroded area of Ap /C -type the thickness of Ap horizon had been kept constant by being mixed with subsoil by tillage. As a result, soil properties on this field suggested relatively high spatial variability compared with those was reported. Average carbon content of Ap horizon was 2.50% and the coefficient variance was 39.0%. Richer zone (over3.5%) accounted for 14% and poor zone (below 1.5%), 12% of entire field. Coefficient variance of pH was relatively low (5.9%)but there were acid soils below pH 5.75 and non acid soils pH over 6.50 which don’t need to be limed for cropping. In this case, the technique to spread some soil improvement materials and fertilizer according to soil condition on sites, such as the precision farming, should be applied. When it is applied, it is essential to grasp the spatial variability of soil properties at the field. Topographical analysis was effective to identify in 2～4 zones according to each soil properties to prepare soil map. The existence of the drift and trend about soil spatial variations disturbed direct application Geostatistics, which is powerful tool to estimate soil distributions, on the upland soil over a complex landscape. Soil survey and sampling were conducted on the hilly pasture in the Hokkaido university Shizunai livestock farm located in Hidaka sub-prefecture. The pasture was reclaimed by horse stocking without earth work. Land surface relief was more complex than the previous upland field. The difference of 137Cs inventory between maximum and minimum slope site was negligible. This result suggested that soil erosion and deposition haven’t happened owing to the grass which has covered soil surface. The variability of most soil chemical and physical properties including thickness of A horizon didn’t exceed the results of upland field or the references. The significant relation between topographical features and some soil properties could be recognized, especially the properties related to soil water movement. The topographical characterization model could explain the distributions of soil water contents, saturated hydraulic conductivity and soil pH. But topographical attributes generally couldn’t agree with the results of soil investigation on the hilly field used as pasture. In this paper, I discussed the realization that topographical analysis contributing to soil management by identifying soil distributions. The grassland for hay selected for soil investigation was located in Nakafurano near the foot of Tokati-dake Mountains. The maximum slope was 0.20 mm-1, which didn’t reach the cases of the upland field or the pasture,and the wave of land surface was moderate. The coefficient variance about soil pH and carbon content of Ap horizon were lower than that recorded at the upland field, but it had the same level with that of the pasture. Regular grass seeding, simultaneous soil tillage and improvements operate to increase the soil variability. Average soil pH was 5.64 and most data didn’t reach 6.50, which is criterion of liming. 2,572 kg ha-1 CaCO3 was estimated for liming according to the result of libratory testing pH and the soil response to some CaCO3. But 60% of the entire field was expected to be over 500 kg ha-1 which is excess or insufficient. It was not sufficiently possible to improve soil acidity with uniform quantity of lime even if a lot of soil samples prepared. The measure to spread adequate mass of lime for each site, like precision farming, have been necessary in this case. The pH was significantly correlated with slope curvature, which is one of topographic attributes, and low pH zones were located on the convexity slope positions. But the topographical characterization model was not enough to enhance accuracy of liming. Instead of topographical analysis, geostatisitics was practical for identification of the spatial distribution of CaCO3. Inadequate area could be decreased up to 40% with detail pH map estimated by kriging method. The soil variation in one field at hilly area, which has been cultivated by the same ways, was recognized to prevent uniformity of crops growths. Soil variation development was highly related to the land use and agricultural practices. In upland field, annual tillage operations encourage soil erosion and deposition occurs in many landscapes in response to the way water or soil materials moves through and over the landscape. Therefore, soil distribution for using soil management could be identified sufficiently by topographical attributes and the topographical characterization model. Grassland soil also showed some variability that was weak relationship to the current topographical attributes about some soil properties. In some cases, geostatistics is superior to estimate soil distribution and inform about soil management to reduce soil spatial variations. In order to acquire conveniently the valid information about soil spatial distributions in arable land, it is important to examine the adaptation of methodology according to the field conditions.

紀要論文

Version of Record

1881-8064