A simple model for partitioning forest soil respiration based on root allometry

Abstract

Partitioning soil respiration (Rs) into heterotrophic (Rh) and autotrophic (Ra) compartments is an important first step in understanding the response of soil carbon flux to changes in climate. This is because Rh and Ra respond differently to changing environmental conditions. However, the methods currently used to partition Rs have high costs and large uncertainties. In this study, we developed a simple model to partition forest Rs based on root biomass, Rs=Rh+∑i=1n[ci×DBHidi×f(Di)], which the latter is modelled as the sum of root biomass density (MT) derived from an existing root allometric equation (MT=c×DBHd, where DBH is tree diameter at breast height) multiplied by a horizonal distance function (f(D)) based on the relationship between in situ Rs and distance (D) from the tree base. We applied this model to a complex tropical natural forest in Southeast Asia by inputting a site-specific root allometric equation and in situ Rs–D function to obtain: Rs=Rh+c′∑i=1n(DBHi2.59×Di−0.452). Compared to studies that used the classical root biomass regression, our results show a stronger linear correlation between Rs and simulated root biomass density (r2 = 0.797). By using the y-intercept (i.e. no root biomass at all), Rh was estimated to contribute 66.9 ± 5.7% to total Rs. We validated the root-allometry-based model by using the trenching method. In our study area, it took about 1 week (6.5 days) for tree roots to die after trenching, and Rh/Rs ratios estimated from our model were about 2.7% higher than those estimated with the trenching method. With the increasing number of allometry equations for global woody plants, the model has the potential of being used to partition Rs in various forests at regional or global scales.