Soil Sustainability in the Anthropocene

A new geological epoch—the Anthropocene—was voted by the 34-member Anthropocene Working Group (AWG) to mark the profound ways in which humans have altered our planet. In the past 200 years of the Anthropocene, human activities have become an important driving force behind the important changes to the Earth’s environment. The pedosphere, as the foundation and central junction of the Earth’s Critical Zone, dominates biogeochemical and hydro-pedological coupling processes and provides the necessary ecological functions that sustain terrestrial life. However, unreasonable anthropogenic activities, such as those associated with intensive agricultural management and rapid urbanization, have caused a series of problems to soils, including soil acidification, salinization, pollution, and erosion. To help address these challenging issues, new technologies have been developed and applied in soil science, such as digital soil mapping, soil remote sensing inversion, proximal soil sensing, geostatistics, spatial analysis, and machine learning.

Therefore, this Special Issue presents new developments and methodologies, best practices, and applications in soil science and includes novel empirical research, reviews, and opinion pieces covering all related topics, including the following:

(1)

Data processing, machine learning, and geostatistical and spatial analysis in soil science;

(2)

Spatial and temporal changes in soil organic carbon, nitrogen, phosphorus, heavy metals, salinity, and others in representative areas;

(3)

The global cycle of soil carbon, nitrogen, and water;

(4)

Digital soil mapping;

(5)

The relationships between soil properties and human activities;

(6)

Inversion of soil properties from single and/or multisource sensor-based data (e.g., multispectral, hyperspectral, thermal, LiDAR, SAR, gas, and radioactivity sensors);

(7)

Climate modeling of soil systems;

(8)

Soils for sustainable agriculture;

(9)

Emerging approaches for characterizing soil carbon and greenhouse gas emissions;

(10)

Soil biodiversity.

In the study of Yang et al. [1] conducted in southeast China, the researchers aimed to investigate secondary anthropogenic contributions and accumulation rates of heavy metals in soils along highways. This study revealed that soils within 150 m of a highway edge were contaminated with heavy metals, which accumulated on the surface and showed a tendency to migrate downwards, with a transfer depth of 10–30 cm. The contribution of heavy metals was highest for Cd, followed by Cu, Zn, and Pb, and ranged from 1.0 to 30.5% at a depth of 30 cm.

The research article of Xu et al. [2], investigated the impact of land-use types on soil organic matter mineralization. The authors selected typical land-use types and systematically examined the role of organic carbon composition and microbial community composition in the mineralization process of organic matter in Ultisol. This study demonstrated that the mineralization process of organic matter followed the order of Bam < Cam < Tea, with Cam and Tea showing a higher carbon content in soluble pools than Bam. Dissolved organic matter (DOM) exhibited a sequence of Bam < Cam < Tea, while the complexity of the chemical structure in DOM followed the opposite trend.

The study by Guo et al. [3] aimed to predict the spatial distribution of soil properties using machine learning. This study evaluated the performance of a Support Vector Machine–Recursive Feature Elimination (SVM-RFE) feature selection method with four common machine learning techniques in predicting and mapping the spatial distribution of soil pH in an urban area of Fuzhou, China. The results of the study identified five environmental variables critical to the prediction of soil pH: mean annual temperature, slope, topographic wetness index, modified soil-adjusted vegetation index, and Band 5. This study provides new insights into the prediction of soil pH at small and medium scales.

The study of Yang et al. [4] used 331 profile images of major soil types from Hubei and Jiangxi provinces, China, to develop a soil diagnostic horizon recognition model. Through the pretreatment and enhancement of image information using the UNet++ architecture, the model provides an efficient, cost-effective, and accurate method for determining soil horizon information. The results of this study can serve as a basis for further research and practical applications in the field of soil surveys.

In the study of Ye et al. [5], soil samples were collected from two layers at 0–10 cm and 10–20 cm. The authors investigated the effects of soil properties, environmental factors, and land use on the stability and erodibility of soil aggregates using the partial least squares method and path analysis. The results showed that slope, soil organic carbon (SOC), elevation, percentage of farmland and grassland landscape, and topographic wetness index (TWI) had significant effects on soil aggregate stability.

Exploring the effects of straw return and environmental factors on the spatiotemporal variation of soil organic matter (SOM) in black soil regions is essential for soil carbon sequestration research. In the study of Yan et al. [6], the authors predicted the spatial distribution of SOM in a black soil area with long-term straw return in the Northern Hemisphere. The conclusion is that straw return can promote SOM accumulation, providing a reference for SOM accumulation measures in black soil areas.

In the study of Zeng et al. [7], vis-NIR spectra and machine learning were used to predict soil fertility classes indirectly and directly. The author found that vis-NIR spectroscopy was fairly successful as a method for soil fertility class allocation for most of the soil properties, using either direct or indirect models.

In the study of Horta et al [8], the aim was to evaluate the accuracy of the DRASTIC-LU index and the Revised Universal Soil Loss Equation (RUSLE). It was concluded that these two models showed accuracy and could be used in early assessment when choosing the best location and improving management practices for OPP systems.

In the study of Sun et al. [9], a loess–paleosol profile derived from thick loess in Chaoyang city of Liaoning province, Northeast China, was investigated as a well-preserved LBP that included various paleosol types. According to this study, the Chaoyang profile can be deemed as a benchmark loess–paleosol profile for the recognition and classification of paleosols and paleoclimate reconstruction in Northeast China. However, the differences in morphological and physical properties between paleosols and loess suggest different soil fertility and agronomic properties and need further studies to assess their functionality with climate fluctuation.

The study of Zou et al. [10] took Conghua District of Guangzhou City, China, as the study area to investigate the effects of topography, soil, land use, and other factors on soil trace elements. The spatial variability of boron (B), manganese (Mn), molybdenum (Mo), copper (Cu), and zinc (Zn) in the farmland soil of a typical red soil area was plotted using the method of geographically weighted regression (GWR). After a cross validation, the GWR Kriging (GWRK) model results of each element were relatively stable, and the order of the fitting coefficient (R2) was Cu > Zn > B> Mn > Mo. This study clarifies the spatial distribution and influencing factors of soil microelements in the studied region.

The research study of Qu et al. [11] used robust geographically weighted regression (RGWR) to explore the spatially non-stationary relationships between wheat yield and its main influencing factors. The authors obtained the best modeling effect (RI = 52.31%). This integrated method based on yield effectively improved soil nutrient management in the region.

The article of Qu et al. [12] proposed robust residual cokriging (RRCoK) with LUTs for the spatial prediction of soil TN. Categorical and point auxiliary data were included in the prediction model. The author compared its spatial prediction accuracy with the ordinary Kriging method (OK), the robust co-Kriging method, and the robust residual Kriging method. The results showed that the RRCoK method obtained the highest spatial prediction accuracy (RI = 43.14%) and could effectively incorporate categorical and point auxiliary data together for a high-accuracy spatial prediction of soil properties.

The article of Kang et al. [13] proposed that the Land Balance Policy has led to the “better land occupied and worse land supplemented” program. To improve cultivated land quality, the authors investigated different treatment methods and found that the treatment of “woody peat + rotten crop straw + biostimulation regulator II + conventional fertilization” had the best effect on the improvement of organic matter, soil aggregates, and soil microbial activity and was the best treatment method.

The study of Han et al. [14] investigated changes in soil organic carbon after wildfires in a boreal forest. The authors found that the ability of soil to sequester carbon decreased after wildfire disturbances. This study provides a quantitative assessment of SOC changes after wildfires, which are useful for forest management and modeling forecasts of SOC stocks, especially in boreal forests.

In all the abovementioned articles, both advanced scientific progresses and recently explored developments are presented. Therefore, this Special Issue can serve as a highly useful reference material for soil researchers.