04/10/2026
New publication: Minimum dataset with integrated scoring and indexing methods for soil quality assessment. Authors: Khandakar (Rafiq) Islam ,Arifur Rahman, Warren Dick, Vinayak Shedekar, Javier Gonzalez, Dexter Watts, Norman Fausey, Marvin Batte,T ara VanToai, Randall Reeder, Dennis Flanagan.
Islam K, Rahman A, Dick W, Shedekar V, Gonzalez J, Watts D, et al. (2026) Minimum dataset with integrated scoring and indexing methods for soil quality assessment. PLoS One 21(4): e0346136. https://doi.org/10.1371/journal.pone.0346136
Abstract
Soil quality (SQ) is a key determinant of agricultural productivity and environmental sustainability, yet its assessment is challenged by the diverse functions of soil and the absence of universally accepted indicators. This study aimed to develop a crop yield-correlated minimum dataset (MDSCorr) for SQ assessment and evaluate its performance across multiple U.S. regions.
Over a five-year period, data (n = 576) from geo-referenced composite soils at 0–30 cm depth were collected from gypsum amended cover crop integrated corn-soybean rotation experimental sites at Shorter (Alabama), Farmland (Indiana), and Hoytville and Piketon (Ohio).
Using the available soil and crop yield data, six scoring functions (four linear and two nonlinear) and three indexing approaches (additive, weighted additive, and Nemoro) were evaluated to calculate the SQ index (SQI). The MDSCorr identified a reduced set of key soil properties most strongly associated with corn productivity, including total organic carbon, microbial biomass carbon, active carbon, total nitrogen, and aggregate-related physical indicators explaining SQ. Using different scoring and indexing approaches, the calculated SQI values at the Indiana site, used as a reference ranged from 0.31 to 0.6.
Among the approaches, linear scoring with threshold limits and additive indexing produced the most consistent SQI values, reducing variability to within ±1% compared to the total dataset (TDS). The MDSCorr-based SQI showed strong positive correlations with the TDS-derived SQI (R² = 0.53 to 0.93) and outperformed the principal component analysis-based MDS (MDSPCA) in terms of reliability and consistency.
Based on MDSCorr-derived SQI values, the relative SQ rankings for the four study sites were: Hoytville > Indiana > Alabama > Piketon. While calibration and validation are recommended across geographic regions and cropping systems, the MDSCorr approach, when combined with linear scoring and additive indexing, has the potential to provide a simplified and transferable framework for SQ assessment.
Conclusions
Soil quality indices are sensitive to indicator selection, scoring functions, and indexing methods, and no single approach is universally applicable across sites. Among the evaluated methods, the MDSCorr combined with LSM1 and SQIa consistently produced the most reliable, sensitive, and transferable SQI values across diverse soils and regions.
The TDS approach exhibited high variability (~37%) in SQI values, due to inconsistencies in scoring and indexing methods. To enhance accuracy and consistency, scoring and indexing techniques were optimized using sensitivity analysis, the Nash efficiency coefficient (Ef), and the relative deviation coefficient (ER).
We developed a MDSCorr comprising nine universally applicable soil indicators strongly associated with five years of corn productivity across the sites. SQI values derived from the MDSCorr showed better alignment with TDS when linear scoring was combined with additive or weighted additive indexing methods, compared to PCA-based indexing. Validation across all sites revealed that SQI differences between MDSCorr and TDS were within ±1%, supporting the use of MDSCorr as an effective and simplified alternate to TDS. We developed a MDSCorr comprising eight universally applicable soil indicators significantly associated with five years of corn productivity across the sites. SQI values derived from the MDSCorr showed better alignment with TDS when linear scoring was combined with additive or weighted additive indexing methods, compared to PCA-based indexing.
Validation across all sites revealed that SQI differences between MDSCorr and TDS were within ±1%, supporting the use of MDSCorr as an effective and simplified alternate to TDS. Although the development of CMDS was explored, its higher variability in SQI distribution across sites limited its reliability relative to MDSCorr. Based on these findings, we recommend using MDSCorr in conjunction with linear scoring and additive indexing for robust SQI calculation.
The final SQI rankings across the sites were: Hoytville > Indiana > Alabama > Piketon. Future research should focus on validating the MDSCorr framework across a wider range of cropping systems and soil types, incorporating key biological indicators to further improve sensitivity to management-induced changes, and refining region-specific threshold values using long-term datasets. Integration of MDSCorr with digital soil mapping and decision-support tools would further enhance its applicability for site-specific soil management and monitoring.
Soil quality (SQ) is a key determinant of agricultural productivity and environmental sustainability, yet its assessment is challenged by the diverse functions of soil and the absence of universally accepted indicators. This study aimed to develop a crop yield-correlated minimum dataset (MDSCorr) fo...
