Spatial Variability Of Soil Fertility Evaluation In The Natural Forests Of Northeast China

Forty-one plots of natural spruce-fir mixed forests were selected for spatial variability and soil fertility assessment using geo-statistics and principal component cluster analysis method in Jinggouling forest farm in the region of Changbai Mountains. We studied the change law and spatial distribution of soil properties at depths of 0-20cm,20-40cm,40-60cm and 0-60cm, and evaluated soil fertility quality. The overall soil fertility characteristics will provide a theoretical basis for rational management, improving soil fertility status and protecting soil ecosystems of the natural forests in Changbai Mountains. The main results were as follows:(1) With the increase of soil depth, the content of soil moisture, organic matter, total nitrogen, total phosphorus, total potassium, readily available potassium and cation exchange capacity (CEC) decreased, but soil bulk density and pH increased. There existed little change of the variation coefficient of each indicator at varied depths. The coefficients of variation were 0.09-0.81 for all indicators, with the highest of available phosphorus, and the lowest soil pH.(2) The nugget effects of 61% of soil fertility indices ranged 25%-75% at varied soil depths, indicating that the spatial variability of soil fertility was jointly affected by natural factors and random factors. The high uniformity in all positions was illustrated by the fractal dimension close to 1.90 in all directions. Preferential pattern appeared in different position as for each indicator judged by the coefficient of determination.(3) Different interpolation methods were fitted for different fertility indices. In most cases, Co-Kriging had more advantages than Ordinary Kriging and Inverse Distance Weighted method. The variation tendency was the same in the situation of soil fertility spatial distribution. With the increase of soil depth, the color of plaque had the same variation tendency as for each index.(4) Forty-one plots were classified into 4 classes using principal component cluster analysis: the score for the first class was 0.58-0.81 (excellent), including 6 plots; the score for the second class was 0.06-0.58 (good), including 9 plots; the score for the third class was -0.23-0.06 (fair), including 13 plots; and the score for the fourth class was -0.59--0.25 (poor), including 12 plots. Then ordinary Kriging interpolation was carried out using ArcGIS to generate the spatial distribution map of soil fertility quality in the region. Soil fertility quality increased from the north to the south. The Root-Mean-Square Standardized was 0.9544 (in the vicinity of 1), satisfying the interpolation accuracy. The soil fertility quality of the study area was in good condition with 86.29% of the plots falling into excellent, good and fair classes.