A simple method for estimating the coarse lateral root biomass of shrubs using ground-penetrating radar: Validation by Caragana microphylla Lam. in Inner Mongolia

The potential increases in carbon stocks in arid regions due to recent shrub encroachment have attracted extensive interest among both ecologists and carbon policy analysts. Quantifying the shrub root biomass amount in these ecosystems is essential to understanding the ecological changes occurring. In this paper, we proposed a simple nondestructive method for estimating the coarse lateral root biomass of shrubs based on the root counts obtained from ground-penetrating radar (GPR) radargrams. Root data were gathered via field experiments using GPR with antenna center frequencies of 900 MHz and 400 MHz. Five Caragana microphylla Lam. shrubs of different sizes were selected for measuring objects, and a total of 40 GPR survey lines were established for GPR data acquisition. The soil profile wall excavation method was used to obtain the total root biomass from each radargram. A model for estimating the root biomass was built by establishing the relationship between the root biomass in each profile and the root counts interpreted from the radargrams. According to the mathematical relationship between the root diameter and root biomass, the proxy root radius was derived, which could explain the rationality of the proposed model from the biological mechanism. The established model provided high confidence in estimating the root dry biomass using the GPR data obtained at the two antenna frequencies (R2= 0.73 for 900 MHz and R2= 0.71 for 400 MHz). The leave-one-out cross-validation results showed that the model exhibits satisfactory performance. This study expands the application of geophysical methods in root research and offers a new simplified method for estimating the root biomass from GPR data under field conditions. [Display omitted] •Root counts can efficiently estimate the coarse lateral root biomass of shrubs.•GPR is an efficient and labor-saving tool to obtain root counts.•A proposed index Ra can summarize the characteristics of all detectable roots.•The proposed method has the potential to be used in large regions.