Relationships among soil microbiological attributes and abiotic factors in phytophysiognomies influenced by iron mining

Abstract

The Quadrilátero Ferrífero is an area known worldwide for its iron ore deposits. In this region two large biomes of Brazil meet, Cerrado and Atlantic Forest, in addition to being found several areas modified by mining and areas of ironstone outcrops, known as canga. Thus, in addition to the economic importance, the area is a hot spot of diversity. Soil is a system where there are many ecosystem services that guarantee the maintenance of life on the planet. In this way, the maintenance of soil quality is important for the maintenance of these services. This work aimed to evaluate the influence of abiotic factors on soil microbiological attributes and to spatialize some of these in phytophysiognomies of the QuadriláteroFerrífero. Soil was sampled in phytophysiognomies of Atlantic Forest, Cerrado, Canga, Eucalyptus, and in areas in rehabilitation altered by iron mining. In this work the diversity of arbuscularmycorrhizal fungi (AMF) in trap cultures and the potential of mycorrhizal inoculum were evaluated. Soil microbiological indicators biomass microbial carbon, basal soil respiration, metabolic quotient (qCO2), microbial quotient (qMic), hydrolysis activity of fluorescein diacetate (FDA), urease, acid phosphatase, alkaline phosphatase and β-glycosidase were also predicted and spatialized. In order to predict these last mentioned indicators, soil fertility and texture data, element contents obtained by portable X-ray fluorescence equipment (pXRF) and terrain attribute data were used. The highest potential of the mycorrhizal inoculum was found in canga. The use of trap cultures increases the diversity of AMF species captured. To Consider phytophysiognomy and season in the model improves the prediction of microbiological indicators of soil quality. Soil fertility and texture can predict biomass microbial carbon, basal soil respiration, qCO2 and qMic. Terrain attributes are the best predictors of basal soil respiration. Elements obtained by pXRF, soil physicochemical properties and terrain attributes provide good predictive models for soil enzymes activity. The spatialization of the enzymes activity and of the other microbiological attributes allows a better overview of the variability of these in each phytophysiognomy and season.