Developing a soil quality index to compare soil fitness for agricultural use under different managements in the Mediterranean environment

•We developed a SQI to evaluate the interaction between soil quality parameters.•The fertiliser level appeared to drive SQI results.•No effects due to cropping sequence and stubble management were found.•The water extractable N was particularly sensitive to soil management.•The SQI was not significantly correlated with annual grain yield. Due to increasing land-use pressures, soil-quality assessment is in growing demand, thus a standard set of procedures to assign a soil quality index (SQI) would be beneficial. In this study, the effectiveness of different managements in upholding soil quality (for crop productivity) was compared using soil quality indicators incorporated systematically to determine a SQI. The managements included three cropping sequences (wheat based three-year crop rotation, wheat+dry bean double-cropping and continuous wheat), two stubble managements (burning vs. incorporation), and three NPK-rates (nil, intermediate and optimal). Soil physical and chemical parameters were measured, screened through principal component analysis (PCA), normalised, and then integrated into a weighted-additive SQI. Crop sequence significantly affected soil pH and had a measurable effect on plant available P, with lower pH and higher P availability for the legume-based annual rotation. Stubble incorporation enhanced the labile N pool in spring, while ashes generated from burning of residues increased the level of exchangeable cations. Changes in total N were not detectable for any management. The SQI indicated that soil quality was most affected by NPK rate. The correlation between SQI and yield was not statistically significant, suggesting that other soil quality indicators, not measured in this study, were more influential upon yield at the experimental site in 2006. The SQI obtained using the method described herein was able to synthesise the complex information contained in large multivariate data-sets, and therefore would be useful for application at regional and national scales.