Abstract
Soil physical properties (SPP) are considered to be important indices that reflect soil structure, hydrological conditions and soil quality. It is of substantial interest to study the spatial distribution of SPP owing to the high spatial variability caused by land consolidation under various land restoration modes in excavated farmland in the loess hilly area of China. In our study, three land restoration modes were selected including natural restoration land (NR), alfalfa land (AL) and maize land (ML). Soil texture composition, including the contents of clay, silt and sand, field capacity (FC), saturated conductivity (Ks) and bulk density (BD) were determined using a multifractal analysis. SPP were found to possess variable characteristics, although land consolidation destroyed the soil structure and decreased the spatial autocorrelation. Furthermore, SPP varied with land restoration and could be illustrated by the multifractal parameters of D1, △D, △α and △f in different modes of land restoration. Owing to multiple compaction from large machinery in the surface soil, soil particles were fine-grained and increased the spatial variability in soil texture composition under all the land restoration modes. Plough numbers and vegetative root characteristics had the most significant impacts on the improvement in SPP, which resulted in the best spatial distribution characteristics of SPP found in ML compared with those in AL and NR. In addition, compared with ML, △α values of NR and AL were 4.9- and 3.0-fold that of FC, respectively, and △α values of NR and AL were 2.3- and 1.5-fold higher than those of Ks, respectively. These results indicate that SPP can be rapidly improved by increasing plough numbers and planting vegetation types after land consolidation. Thus, we conclude that ML is an optimal land restoration mode that results in favorable conditions to rapidly improve SPP.
基金项目
National Key Research and Development Program of China(2017YFD0800502)
National Natural Science Foundation of China(41671510)
国家科技期刊平台
NSTL
维普
万方数据