High-resolution soil sampling reveals the pattern of biological weathering and soil formation under trees

Trees contribute to bedrock weathering in a variety of ways. However, evaluating their full impact is complicated by a lack of direct observation of unexposed root systems of individual trees, especially when the scale of the analysis goes down to the level of microbiomes. In the present study, we investigated the contribution of tree root systems to bioweathering and soil production at the macro- and microscale. Soil profiles developed under trees on granite bedrock were investigated in two parts of the Sudety Mountains, SW Poland: the Rudawy Janowickie Mountains, and the Stołowe Mountains. Soil profiles were gradually excavated and soil samples collected from pre-defined positions of the root zone: 1) bulk soil, 2) rhizosphere, 3) cracks, 4) topsoil, and 5) control positions. In total, we analyzed 103 samples for soil chemistry and microbiological activity. In addition, we analyzed 19 samples using XRF (X-ray Fluorescence). Four parent rock samples, in the form of thin-sections, were the subject of mineralogical evaluation. Soil analyses included: total organic carbon (C) and nitrogen (N) content, soil pH_H2O, soluble iron (Fe_d), and aluminum (Al_d), non-crystalline (amorphous) iron (Fe_ox), and aluminum (Al_ox). For microbiological analyses, we used a Biolog (EcoPlate) system to determine the functional diversity of soil microorganisms. We evaluated the results on soil chemistry and microbiological activity statistically by principal component analysis (PCA) and redundancy analysis (RDA). Differences between soil sampling positions were assessed using a non-parametric Kruskal-Wallis (K-W) rank sum test and a post-hoc pairwise Dunn test. Trees developed different root architectures, likely shaped by the depth to bedrock and its pre-existing net of fractures and fissures. Tree roots were able to enter bedrock cracks at one study site (at Pstrążna, Stołowe Mountains). The soil profile was too deep for root system penetration at the second study site (Mt Jańska, Rudawy Janowickie Mountains, RJM). The rhizospheric soil along the roots had significantly different chemical properties compared to non-rhizospheric soil types. At Mt. Jańska, soil differed from the crack soil in terms of Al_ox (p_Holm-adj. < 0.0006) and Fe_ox (p_Holm-adj. < 0.004), and from the bulk soil (p_Holm-adj. < 0.02) and topsoil (p_Holm-adj. < 0.007). In addition, at Pstrążna, the soil differed from the control soil in terms of C (p_Holm-adj. < 0.009) and soil pH_H2O (p_Holm-adj. < 0.0008) and from the topsoil in terms of soil pH_H2O. The highest metabolic activity was in cracks at Mt. Jańska and in control samples from Pstrążna. In general, the spatial distribution of soil microbial activity, and the weathering that results from that portion of the soil biome, is spatially heterogeneous and appears to be partially determined by the interaction of root growth and bedrock fracture patterns.