Table 1.
Desirable traits to compare species efficiency for stabilizing slopes and abbreviations used in the text.
Figure 1.
Mean temperature and precipitations in Liuku.
Diagram from Liuku meteorological station situated 30 km south from the field site (source: Meteorological Bureau of Yunnan Province). Data from January 2005 to May 2010. Arrows show the months when roots were collected.
Figure 2.
Soil profiles at the study site.
Soil horizons at a) a reference site with no previous evidence of landslides or erosion; b) the stable site where a shallow landslide had occurred eight years previously and c) the hotspot. Colours were identified using a Munsell colour chart (Munsell 1947). OL: fresh litter, OF: fermenting litter, OH: litter with humic substances and well-transformed organic matter, A: organico-mineral layer, AB: mixture between A and B, B: layer of bedrock alteration, pieces of bedrock are visible, C: bedrock, mineral layer (Legros 2000; Baize and Girard 1995).
Table 2.
Soil textural, chemical and physical characteristics in horizons A and B at the hotspot and stable site.
Figure 3.
Description of the root systems of the nine studied species.
Root systems of the nine species and their rooting depth (m). A. americana's root system was composed of an underground stem from which emerged thin roots; A. hispidus's root system comprised only a few thin roots emerging from the plant collar; A. codonocephala possessed a root system with long lateral roots turning downwards over time; B. championii had long and deep roots, which deviated on encountering an obstacle and which were densely branched; C. anomala possessed a tufted and shallow root system; vertical roots of F. tikoua emerged from creeping stems; J. curcas and P. stricta both possessed taproot systems, but roots of P. stricta were deeper and more densely branched; R. chinensis had a sprouting root system comprising long, deep and scarcely branched roots. Note that the scale for soil depth (y axis) differs between species for easier viewing.
Table 3.
Shoot characteristics and time of harvest for individuals studied (means ± se).
Table 4.
Different statistical tests were used depending on the variables tested.
Figure 4.
Number of stems per square meter of each species present on the unstable hotspot and the stable site in 2009 and 2010. “Others” were composed of the following species: Bidens pilosa L., Celosia argentea L., Elsholtzia winitiana Craib, Indigofera sp., Malvastrum coromandelianum (L.) Garcke, Convolvulus arvensis L., Solanum verbascifolium L.
Figure 5.
Mean individual soil volume standardized by collar diameter (ISV/Dc).
Negative values of ISV/Dc represent downslope orientation.
Figure 6.
Root density of all species and proportion of coarse and fine roots.
(a) Root area ratio (RAR) of all species. Negative values of RAR represent downslope orientation; (b) Coarse and fine roots area ratio (RAR).
Figure 7.
Tensile stress at failure (Tmax) for root diameters from 0 to 2 mm. Logarithmic scales. Fitting curves: Tmax = α*D−β, equations are presented on the graphs with the determination coefficient (R2), parameters in brackets are not significant.
Figure 8.
Ultimate strain at failure (εult) for each species. Vertical bars denote 0.95 confidence intervals, letters indicate significant differences between species (P<0.05).
Figure 9.
Bending rigidity (EI) as a function of root mean diameter for each species where data were sufficient for statistical analyses. Mean root diameter is the mean of depth and width diameters for each root. Note that scales differ between graphs for easier viewing.
Figure 10.
Nitrogen and cellulose contents.
a) Nitrogen and b) cellulose concentrations in coarse and fine roots for each species. Letters indicate significant differences between species (P<0.05) when all (coarse and fine) roots are taken into account. Asterisks denote significant differences between coarse and fine roots within a species. Vertical bars denote 0.95 confidence interval.
Figure 11.
3D diagram of species' scores.
Summary of species' performance according to root abundance in soil, root mechanical resistance and root physiological properties.
Table 5.
Multi-criteria table summarizing performances of each species: traits and related functions.
Figure 12.
Best species depending on the location on the hotspot of instability.
On degraded hotspots, the diversity of root systems will play an important role for slope stability. Roots growing upslope from the stem will have more chance to cross the potential shear plane if the plant grows at the top of the slope. Thus, root systems with desirable traits upslope of the stem will act more efficiently if they are located at the top of the slope, whereas the inverse is applicable for downslope roots. Root systems with desirable traits deeper in the soil will act more efficiently in the middle of the hotspot.