Estimation of fine root biomass using a minirhizotron technique among three vegetation types in a cool-temperate brackish marsh

摘要

The minirhizotron technique is a non-destructive method to evaluate fine roots, which converts two-dimensional image data to three-dimensional root biomass data. Recently, conversion factors in soils at 10-cm depth intervals successfully estimated fine root biomass using image data from the minirhizotron method. However, this technique was conducted only at one forest site and did not consider different vegetation types. Therefore, the objective of this study was to verify a method for calibration of minirhizotron data with the core sampling values obtained by direct measurement of root biomass in wetland ecosystems among three vegetation types. Evaluations by minirhizotron technique and soil-core sampling were made at 30-cm soil depth in a cool-temperate brackish marsh in northern Japan. Linear regression was examined between root volume and weight of fine roots in soil core samples, and the fine root biomass on minirhizotron tubes was calculated from their length and diameter. The technique was well adapted for vegetation types dominated by Phragmites australis, Juncus yokoscensis, and Miscanthus sinensis and Cirsium inundatum. Compared with the fine root biomass estimated by the core sampling method, fine root biomass estimated by the minirhizotron method was overestimated in the 0-10-cm layer. Further, we determined conversion factors based on the ratio of the fine root biomass by the core sampling method to that by the minirhizotron tubes. Estimation of the fine root biomass using the conversion factors for each 10-cm soil depth was well adapted in P. australis vegetation and J. yokoscensis vegetation types as a forest ecosystem; meanwhile, M. sinensis and C. inundatum vegetation types were not well adapted. This study suggests that the minirhizotron technique is available to estimate fine root biomass of single-species dominated vegetation in the brackish marsh using conversion factors for each 10-cm depth.