Evaluation of soil quality using multiple lineal regression based on physical, chemical and biochemical properties.

The aim of this work is to obtain an expression using multiple lineal regressions (MLR) to evaluate environmental soil quality. We used four forest soils from Alicante province (SE Spain), comprising three Mollisols and one Entisol, developed under natural vegetation with minimum human disturbance, considered as reference soils of high quality. We carried out MLR integrating different soil physical, chemical and biochemical properties, and we searched those regressions with Kjeldahl nitrogen (N(k)), soil organic carbon (SOC) or microbial biomass carbon (MBC) as predicted parameter. We observed that Mollisols and Entisols presented different relationships among their properties. Thus, we searched different equations for both groups of soils. The selected equation for Mollisols was N=0.448 (P) + 0.017 (water holding capacity) + 0.410(phosphatase) - 0.567 (urease) + 0.001 (MBC) + 0.410 (beta - glucosidase) - 0.980, and for the Entisol SOC = 4.247 (P) + 8.183 (beta-glucosidase) -7.949 (urease) + 17.333. Equations were applied to samples from two forest soils in advanced degree of degradation, one for Mollisols and the other one for the Entisol. We observed a clear deviation in the predicted parameters values related to the real properties. The obtained results show that MLR is a good tool for soil quality evaluation, because it seems to be capable of reflecting the balance among its properties, as well as deviations from it.