Revealing soil microbial ecophysiological indicators in acidic environments laden with heavy metals via predictive modeling: Understanding the impacts of black diamond excavation

Abstract

Appraising the activity of soil microbial community in relation to soil acidity and heavy metal (HM) content can help evaluate it's quality and health. Coal mining has been reported to mobilize locked HM in soil and induce acid mine drainage. In this study, agricultural soils around coal mining areas were studied and compared to baseline soils in order to comprehend the former's effect in downgrading soil quality. Acidity as well as HM fractions were significantly higher in the two contaminated zones as compared to baseline soils (p < 0.01). Moreover, self-organizing and geostatistical maps show a similar pattern of localization in metal availability and soil acidity thereby indicating a causal relationship. Sobol sensitivity, cluster, and principal component analyses were employed to enunciate the relationship between the various metal and acidity fractions with that of soil microbial properties. The results indicate a significant negative impact of metal bioavailability, and acidity on soil microbial activity. Lastly, Taylor diagrams were employed to predict soil microbial quality and health based on soil physicochemical inputs. The efficiency of several machine learning algorithms was tested to identify Random Forrest as the best model for prediction. Thus, the study imparts knowledge about soil pollution parameters, and acidity status thereby projecting soil quality which can be a pioneer in sustainable agricultural practices.