Network analysis was effective in establishing the soil quality index and differentiated among changes in land-use type

• Published in: Soil & tillage research Vol. 246; p. 106352
• Main Authors: Gao, Ming, Hu, Wei, Li, Meng, Wang, Shuli, Chu, Lin
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 01.02.2025
• Subjects: agricultural management; China; cropland; data collection; forests; grasslands; Land use; Network analysis; Principal component analysis; soil; soil quality; Soil quality index; tillage; wind erosion; China; Soil quality index; Network analysis; Land use; Principal component analysis
• ISSN: 0167-1987
• DOI: 10.1016/j.still.2024.106352

Understanding the implications of land-use type on soil quality and function is critical to the adoption of suitable agricultural management practices in a specific region. Principal component analysis (PCA) is a widespread technique for calculating soil quality index (SQI), but it cannot correctly evaluate soil quality in some cases. Network analysis (NA) is a novel and effective technique for calculating SQI for determining susceptibility in different land uses but it is still limited. Moreover, few studies have compared NA and PCA to quantify soil quality. This study aimed to develop valid and accurate SQIs through NA and PCA to estimate the impacts of land-use types (cropland, forest, and grassland) on SQIs in Tongliao and Qiqihar, which are the two regions subject to frequent wind erosion in northeast China. A total of 27 soil physical, chemical, and biological properties were measured for the selection of the minimum data set (MDS), and eight SQI values were determined for each study site using additive or weighted methods and linear or nonlinear scoring functions. Results indicated that most soil attributes and SQIs varied markedly among three land-use types and were greater in grasslands or forests than in croplands. The amount of MDS generated using NA was considerably low, but soil physicochemical and biological properties were comprehensively covered. SOC in Tongliao and SHC in Qiqihar were selected by both methods and were considered the most sensitive soil quality indicators for detecting the effects of land-use types. The soil sensitivity index of the SQI calculated by NA (1.34–2.02) was higher than that of the SQI calculated by PCA (1.30–1.80). Thus, NA was more effective than PCA in computing the SQI and differentiated among changes in land use better as a simple and stable tool. The SQI developed through NA using the weighted method and nonlinear scoring function is a suitable and practical quantitative tool for SQI assessment, which is proposed to be used for soil quality assessment for various land-use types in northeast China. •NA was more sensitive and useful than PCA in the SQI assessment.•SQINA-wnl exhibited the best performance in distinguishing between the SQIs.•Conversion of forests or grasslands into croplands decreased soil quality.