Methods to measure the mechanical behaviour of tree roots: A review

The presence of forests on hillslopes significantly reduces the slopes susceptibility to rainfall triggered shallow landslides. This is due largely to the reinforcement of the hillslope soil by tree roots which increase the shear strength of the soil, and in some instances, anchor the soil mantle to the underlying bedrock by deeply penetrating roots. Quantifying the reinforcing effects of tree roots within soils and the evaluation of hillslope stability using geomechanical and numerical models relies on a realistic representation of the characteristics of tree roots distribution within the hillslope and the mechanical strength of those roots. The variety of experimental methods that have been developed since the 1960s and are used to generate these root strength and rooted-soil shear-strength data are reviewed. The majority of these studies have focused on determining the tensile strength of individual roots by loading the root in a pulling device until it breaks and/or determining the shear-strength of rooted soil in comparison to non-rooted soil in a Coulomb-type shear-box test. These studies have also generally either examined mature root systems in the field or relatively young plants grown in special containers specifically designed for tensile tests or laboratory shear-box tests. A particular difficulty that most studies have encountered is fixing or securing the ends of roots in the attachment device of the testing apparatus (so called root-pullers or tensile testing machines) as the various styles of clamping employed can easily damage the root which reduces the measured strength or otherwise results in an unrealistic test result. Laboratory shear-box tests encounter a similar difficulty in that the roots are not generally fixed or constrained at the base of the shear-box; field shear-box tests tend to avoid this problem as the roots are present in their natural anchoring characteristics in the soil and rock substrate. A result universally reported in rooted soil shear-box test studies is that the peak shear-strength of rooted soil significantly exceeds the peak shear-strength of that soil in a non-rooted condition and that the rooted-soil peak strength is typically recorded at a shear-displacement distance several times that of the non-rooted soil. This result fundamentally explains the reduced susceptibility of forested hillslopes to shallow landslides. A variety of solutions developed to deal with the difficulties that root and rooted-soil tests present are outlined. A set of suggested protocols for conducting root tensile tests and field pullout tests are also presented. It is intended that the adoption of these protocols will enable more effective and direct comparisons of test results and more confident interpretation with respect to the similarities and differences between test results generated from different species and field sites.