Evaluation of soil quality in major grain-producing region of the North China Plain: Integrating minimum data set and established critical limits
Introduction
A significant degradation in soil quality (SQ) is occurring worldwide through adverse effects on these soil physical, chemical and biological properties and inorganic and organic contamination caused by human activities (Arshad and Martin, 2002, Sánchez-Navarro et al., 2015, Nehrani et al., 2020). Hence, addressing this challenge requires a comprehensive understanding of SQ status, for the purpose of maintaining both considerable SQ level and crop production in agricultural regions.
Up to the present, many methods of SQ evaluation have been developed (Karlen et al., 1993, Nosrati, 2013, Pulido Moncada et al., 2014, Rahmanipour et al., 2014, Shukla et al., 2006), such as soil quality index (SQI) method (Doran and Parkin, 1994, Doran et al., 1996), dynamic variation of SQ models (Larson and Pierce, 1991), SQ cart design and test kit (Ditzler and Tugel, 2002), geostatistical methods (Sun et al., 2003) and multiple variable indicator kriging methods (Qi et al., 2009), to improve land use management and identify high risk areas for the sustainable use of land resources (Bindraban et al., 2000, Andrews et al., 2004, Li et al., 2019). Among these, the SQI method has become an effective way to evaluate the heterogeneity of SQ (Andrews et al., 2002, Askari and Holden, 2014, Brejda et al., 2000, Li et al., 2019). In the world, SQI has been broadly carried out to evaluate SQ under different agricultural management at different scales and in many regions (Chahal and Van Eerd, 2019, Cheng et al., 2016, Jahany and Rezapour, 2020, Qiu et al., 2019, Viaud et al., 2018, Wang et al., 2019). However, how to objectively integrate quantitative and qualitative SQ indicators (such as soil physical, chemical and biological properties) to calculate comprehensive SQIs is still the major challenge, especially for large-scale SQ assessemnent (Askari and Hloden, 2014; Guo et al., 2017, Obade and Lal, 2016, Yu et al., 2018), which is mainly due to the expensivelabour and capital needed to acquire enough data for accurate assessment in large-scale soil monitoring campaigns. Selecting the most representive indicators is therefore important to reduce costs and data redundancy (Lima et al., 2008, Raiesi, 2017, Bünemann et al., 2018, Wu et al., 2019), while providing enough data for a reliable and accurate assessment. Previous attempts, for example, Larson and Pierce (1991) first suggested a minimum data set (MDS) consisting of a number of variables to describe the quality of a soil. Some authors also suggested using the Delphi data set (DDS), which was selected according to the importance of given variables to overall SQ based on expert opinion, to reduce the number of variables (Herrick et al., 2002). However, Qi et al. (2009) found MDS to be more accurate than DDS due to the subjectivity of expert opinions in the establishment of DDS. Moreover, regardless of the method used, some useful information from the total dataset (TDS) will be lost. Therefore, it is necessary to select the most representative index as a MDS and quantify the degree of information loss when using the MDS.
The SQI method typicallycomprises a four–step procedure: choosing appropriate variables, converting variable values to scores, determining variable weights, and integrating variable scores into an SQI (Rezaei et al., 2006, Yao et al., 2013). During the scoring of indicators, standardized scoring functions (SSFs) are often used to score soil variables and normalize the measured variable values (Diack and Stott, 2001) before they are integrated into an SQI (Glover et al., 2000, Masto et al., 2007, Qi et al., 2009, Askari and Holden, 2014). Linear and non-linear methods are usually applied to transform the key soil properties (Rahmanipour et al., 2014, Yu et al., 2018). For all SSFs, the shape of the SSF curve is determined by function types and critical values. Therefore, the accurate estimation of critical values is crucial for SQ assessments using SSF. Unfortunately, the establishment of critical values of indicators is difficult for a given study area. Critical values are soil property values where the crop yield reaches high productivity when the measured soil property is at an optimal level or has low productivity when the soil property is at an unacceptable level (Glover et al., 2000, Merrill et al., 2013, Biswas et al., 2017). In many cases, critical values based on expert opinion, published values, or measured values observed under near-ideal soil conditions for a specific site and special crop species were directly extended to large areas (Bhaduri and Purakayastha, 2014, Manna et al., 2005, Li et al., 2013). However, previous studies comparing some indices under different standardized methods and found the existing models that performed better to be much site-specific (Rahmanipour et al., 2014, Obade and Lal, 2016, Guo et al., 2017). Given the complexities of crop yield response to critical values, perhaps, the best we can do is to build these values of key soil indicators based on local crop yield.
As one of the most important agricultural regions in China, the North China Plain accounts for approximately 19% of the nation's total agricultural area (Liu et al., 2001). Wheat-Maize double cropping system played a very important role in agriculture production in this area, supplying more than 75% and 32% of the nation's total wheat and maize, respectively (Chang et al., 2020). Crop yield is considered here as a suitable integrator of the SQI because it is a primary concern to farmers in most cases (Merill et al., 2013; Bai et al. 2018). The crop yield in specific condition (ranging from low to high) could be useful in developing a guideline for interpretation of individual key soil indicator (Lopes et al., 2013). Thus, a reliable and accurate envaluation of SQ and an exploration of restricting indicators based on crop yields could provide a theoretical basis for further developing a win–win strategy combining profitable crop production with improvement of SQ. Since SQ evaluation based on the establishment of critical limits of indicators in the farmers' field conditions is limited in North China Plain. Here we developed a framework to evaluate and compare SQ in Yucheng County in the North China Plain using a TDS and some MDSs of physical and chemical soil variables. The specific objectives of this study were: (a) to objectively identify the MDS consisting of key soil variables; (b) to obtain critical values for SSFs based on local crop yields in the real field condition; (c) to quantify SQI according to the MDS and critical values; and (d) to evaluate the restricting indicators that affect crop productivity. We hope this study can improve our understanding of the SQ status in North China Plain and provide an underlying basis to select indicators for future monitoring of SQ and production potential in other agricultural areas. And the finding can enable policy making and policy implementation in relation to SQ in the future.
Section snippets
Study site description and sampling
Yucheng (36° 40′ − 37° 12′ N, 116° 22′ − 116° 45′ E) is a conventional agricultural region with a predominantly wheat (Triticum aestivum L.) and maize (Zea mays L.) rotation cropping system within a single year and is a major grain-producing region on the North China Plain. The total area is 990 km2 with an arable land area of 530 km2. Yucheng has a warm monsoon climate with a mean annual temperature and precipitation of 13.1 °C and 593 mm, respectively (SSOYC, 1985). The topography of Yucheng
Soil properties and SQ distribution
The soil pH ranged from 8.03 to 9.14, with a mean of 8.56, indicating that the soils in this region are alkaline. The surface soils in this area had low clay with an average value of 8.98% and middle SOM contents with a mean of 15.92 g kg−1. The soil EC ranged from 60.0 to 690.0 μS cm−1, with a mean of 120.0 μS cm−1, reflecting the light salinity levels in this area. The soil CEC content was low, with an average value of 120.43 mmol kg−1, ranging from 54.62 to 262.47 mmol kg−1. The soil
MDS variable selection methods and establishment of critical values
Three MDS variable selection methods were evaluated for their accuracy in estimating crop yields response to SQIs. Almost all of the variables chosen for the three MDSs can be found in previous SQI research (Doran and Parkin, 1994, Yao et al., 2013, Yu et al., 2018), especially studies applying MDS variable selection methods (Masto et al., 2007, Qi et al., 2009). However, other variables such as stoniness (Singer and Ewing, 2000); certain biological variables including active carbon, potential
Conclusions
Although MDS methods are effective for the evaluation of agricultural SQ, the process of variable selection can significantly affect its accuracy of SQ assessment. The accuracy of SQI estimations decreases with a decrease in the number of variables when variables are selected within each PC in terms of variable loading. However, when norm values are used to select variables, the SQI can accurately predict SQ and productivity even with the decreased numbers of variables. When published critical
CRediT authorship contribution statement
Kang Tian: Conceptualization, Writing - review & editing. Beier Zhang: Investigation, Software, Writing - original draft. Haidong Zhang: Investigation, Data curation, Visualization. Biao Huang: Supervision, Methodology. Jeremy L. Darilek: Software. Yongcun Zhao: Investigation. Jingsong Yang: Investigation.
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgements
The authors are grateful for funding from the Project funded by China Postdoctoral Science Foundation [2018 M642350], the National Science and Technology Support Program of China [2012BAD05B05-2] and the Knowledge Innovation Program of the Chinese Academy of Sciences [KSCX1-YW-09-02]. Thanks are also extended to Yucheng Integrated Agricultural Experimental Station, Chinese Academy of Sciences, for their help with sample collection.
References (69)
- et al.
A comparison of soil quality indicesmethods for vegetable production system in northern California
Agric. Ecosyst. Environ.
(2002) - et al.
Identifying critical limits for soil quality indicators in agro-ecosystems
Agr. Ecosyst. Environ.
(2002) - et al.
Indices for quantitative evaluation of soil quality under grassland management
Geoderma
(2014) - et al.
Effects of agricultural management practices on soil quality: a review of long-term experiments for Europe and China
Agr. Ecosyst. Environ.
(2018) - et al.
Long-term tillage, water and nutrient management in rice–wheat cropping system: assessment and response of soil quality
Soil Tillage Res.
(2014) - et al.
Land quality indicators for sustainable land management: proposed method for yield gap and soil nutrient balance
Agr. Ecosyst. Environ.
(2000) - et al.
Establishment of critical limits of indicators and indices of soil quality in rice-rice cropping systems under different soil orders
Geoderma
(2017) - et al.
Soil quality – a critical review
Soil Biol. Biochem.
(2018) - et al.
Quantifying soil quality in a horticultural-cover cropping system
Geoderma
(2019) - et al.
Impacts of nitrogen management and organic matter application on nitrous oxide emissions and soil organic carbon from spring maize fields in the North China Plain
Soil Tillage Res.
(2020)