Computed tomography-based root structure variables determines root decomposition given a broad range of root and soil variables

Abstract

Previous studies have identified the critical roles of root structure in root decomposition, but the complexity of root structure may not be fully captured by traditional measures, leading to critical uncertainty in quantifying root structure and understanding its effects. To address the knowledge gap, we used new measures based on root-structure scanning with Computed Tomography (CT) and compared the performance of the CT-based root structure variables with the traditional ones in explaining the variation in root decomposition. In addition, we considered the effects of other root and soil factors in the analysis. Given the large number of variables used, we first applied principal component analysis (PCA) to represent the variation in root attributes (structure and connectivity) and soil properties (physical, chemical, and biological ones) with a few PCA axes. Our results showed that: (1) Root variables had greater relative importance than soil variables in root decomposition; (2) The root effects were dominated by root structure variables, and more variance for the root decomposition was explained by adding the CT-based variables which even had slightly greater relative importance than the traditional structure variables; (3) The soil effects on decomposition were dominated by the biological properties where soil catalytic hydrolase was more important than soil phosphatase. Therefore, we validated the capability of CT-based root structure variables to determine root decomposition, given that CT scanning has few destructive effects on the soil-root environment and its variables can capture fine and complicated changes in the root structure. In addition to root decomposition, CT-based root structure variables are potentially applied to understanding other biogeochemical processes.