• Premise of the study: Roots play an important role in strengthening and stabilizing soils. Existing models predict that tensile strength and root abundance are primary factors that strengthen soil. This study quantified how both factors are affected by root developmental stage. • Methods: Focusing on early development of Avena fatua, a common grassland species with a fibrous root system, we chose three developmental stages associated with major changes in the root system. Seeds were planted in rhizotrons for easy viewing and pots to allow root growth surrounded by soil. Tensile strength was determined by subjecting root segments to a progressively larger pulling force until breaking occurred. Root abundance at two depths was characterized by the cross‐sectional area of the roots divided by the area of the soil core (i.e., root area ratio). Shear strength of 50 mm saturated soil columns was determined with a modified interface direct shear device. • Key results: Tensile strength increased by a factor of ≥15× with distance from the root tip. Thus, soil‐strengthening properties increased with root cell development. Plants grown under dry soil conditions produced roots with higher maximal tensile strength (41.9 MPa vs. approximately 17 MPa), largely explained by 33% thinner diameters. Over 7 weeks of root growth, root abundance increased by a factor of 4.8× while saturated soil shear strength increased by 24% in the upper soil layer. • Conclusions: Root development should be incorporated into models of soil stability to improve understanding of this important environmental property.

International audienceIn-situ x-ray tomography has been used to follow deformation processes in 3D during two triaxial compression tests, one on a specimen of bio-cemented Ottawa 50-70 sand and the other on a specimen of the non-cemented sand. The global stress-strain responses show that the bio-cementation process increases the shear strength (peak deviator stress is approximately doubled), and causes the material to exhibit a linear behaviour up until peak, as well as increasing the dila-tancy angle. The residual strength of the two samples is very close at large strain. Quantitative 3D digital image analysis (porosity, cement-density and strain field measurements), reveals that a dilatant shear band gradually develops pre-peak in the reference material. The cemented sample however undergoes an abrupt change of deformation mechanism at peak stress: from homogeneous deformation to localised dilatant shearing, which is associated with a local loss of cementation

No abstract