Soil microbial communities affected by vegetation, topography and soil properties in a forest ecosystem

Samaneh Tajik,Shamsollah Ayoubi,Nicola Lorenz

DOI: https://doi.org/10.1016/j.apsoil.2020.103514

IF: 4.8

2020-05-01

Applied Soil Ecology

Abstract:<p>Microbial communities play a seminal role in biogeochemical cycles and can be influenced by both biotic and abiotic factors. A few studies highlighting the importance of topographic factors influencing soil microbial community. This study was conducted to investigate linear and nonlinear relationships between soil properties, topographic attributes and vegetation indexes with soil microbial communities in Shast Kalate (Bahram Nia) forest at the north of Iran. To analyze relationships between soil microbial community and selected variables, correlation analysis, multiple linear regressions (MLR) and machine learning approaches including artificial neural network (ANN) and random forest (RF) were applied. To validate modeling of soil microorganisms using environmental variables, two input datasets were used including soil properties, topographic and vegetation indexes (STV dataset) and topographic and vegetation indexes (TV dataset). Phospholipid fatty acids (PLFA) were analyzed to measure the structure of soil microbial communities. Linear and nonlinear analyses indicated that the results of the models developed by STV and TV datasets perfectly accorded (r = 0.72; p &lt; 0.01). Furthermore, the ANN (0.68–0.83 minimum and maximum accuracy respectively) and RF (0.60–0.79 minimum and maximum accuracy, respectively) models were more powerful than MLR (0.39–0.69 minimum and maximum accuracy respectively) model to predict soil microorganism communities. Results of sensitivity analysis showed that topographic attributes and vegetation indexes were to be influential parameters in modeling and would be an indication of soil microorganisms for soil quality assessment and management practices in forest soils. Protozoa were closely interrelated with topographic and vegetation indexes and behaved differently rather than other communities such as arbuscular mycorrhiza fungi (AMF), actinobacteria, fungi, general bacteria, Gram⁻ and Gram⁺ bacteria.</p>

soil science