Вы находитесь на странице: 1из 4

Second International Symposium on Plant Growth Modeling, Simulation, Visualization and Applications

Analysis of 3D structural root architecture data of trees grown on slopes

Frdric Danjon1, David H. Barker2, Michael Drexhage3, Alexia Stokes4


1
INRA, BIOGECO, 33612 Gazinet Cedex, France. fred@pierroton.inra.fr
2
Prima Subur Sdn Bhd, Kuala Lumpur, Malaysia
3
INRA, LERFOB, 54280 Champenoux, France
4
INRA, AMAP, 34398 Montpellier Cedex 5, France

Abstract crossing plane, (4) mean root length to the previous


The root architecture of two Quercus alba growing branch and (5) total root length after the intersection
on sloping ground was measured using 3D digitizing in point (Wu, 1995). However, existing slope stability
situ. These detailed data were then incorporated into models generally use only the mean root area ratio
models of slope stability, showing how such data can (RAR) and tensile strength (see [1] for a complete
be used to determine root reinforcement potential. review).
This study aims at providing a chain of tools useful
Keywords: 3D digitizing, coarse root architecture, root
for determining the contribution of tree roots to slope
area ratio, slope reinforcement, landslide
reinforcement.

1. Introduction 2. Materials and methods


One dominant (tree D) and one suppressed (tree S)
The fixation of soil by plant roots is of primary
Quercus alba trees were sampled in two nearby similar
importance for reducing landslide occurrence in natural
mixed stands, 5 km from Athens, Georgia, USA. Soils
or man-made slopes. The reinforcement of a potential
were well drained nutrient depleted regolithic
clays derived from granite and mica gneiss
bedrock. Slope inclination was 12 for tree D
and 18 for tree S. Measurements and analysis
were made according to [2] with the following
additions:
1) Excavation: The two root systems were
progressively excavated using high velocity
air-lances to loosen and remove soil.

Figure 1: Vertical distributions of parameters of


roots intersecting planes parallel to the slope at
10 cm intervals. Upslope and downslope profile
for each tree. CSA : total cross sectional area, f
: mean cumulative root length after the impact.

Soil was driven downslope and either directly


lifted out using hand-tools or blown out of the
resulting excavation through a trench
progressively dug out of its lower side.
To maintain the root systems exactly in
landslide plane by a woody root is mainly a function of their original position they were held by suspending
(1) its tensile strength and of the following major roots from wooden planks spanning from the
characteristics at the point of intersection: (2) root stump to the edges of the excavation.
cross sectional area (CSA), (3) root angle towards the 2) Measurement of root architecture: The root

978-0-7695-2851-9/07
0-7695-2851-1/07 $20.00
$20.00
2007
2007
IEEEIEEE 74
DOI 10.1109/PMA.2006.54
systems were measured in situ, i.e. in the soil in their large taproot which forked at 67 cm and possessed
original positions. The topology and XYZ coordinates about half the root biomass and length of tree D (Table
and diameters of all root origins and of all root 1). Tree D had a fairly homogeneous distribution of
segment ends were measured with a 3D digitizer. The root length i.e. nearly 25% root length in both up- and
positive X-axis was oriented horizontally upslope. For downslope quarters and nearly 50% root length in the
tree D, all roots of size above an initial diameter of 10 sectors perpendicular to the slope. However, there was
mm were measured, this threshold was set to 5 mm for more upslope reinforcement with regard to root dry
tree S. The length of digitised segments often reached weight, but at the expense of the downslope quarter.
40 cm. Each root system required one week for Therefore, Tree Ds roots were thicker upslope than
exhumation and measurement. downslope. Both trees had a small inter-lateral length
3) Determination of root specific gravity: The mean downslope (Table 1). Order 2 axes encompassed more
specific gravities of roots in five diameter classes (2, 5, than half of the root volume, being 60 and 84% of root
10 and 20 mm diameter limits) were determined from a dry weight in tree D and S, respectively.
sample of roots by water displacement. Their values A more detailed assessment of root spatial distribution
were 0.48, 0.48, 0.54, 0.59 and 0.62, respectively. The is provided by the 1D distribution displayed in figure
same method was used to compute root dry weight 1. Tree D possessed more root length between -40 and
from the root volume obtained by digitizing. -110 cm distance beneath the soil surface, both up- and
4) Spatial distribution: To compute spatial downslope. Conversely, Tree S had a discontinuity in
distribution of root dry weight, all the measured root root length and volume distribution at around 60 cm
segments were divided into 1 cm long virtual sub- distance beneath the soil surface, which corresponded
segments. to the limit between shallow horizontal roots and the
5) Axis rotation: To derive root characteristics as a downslope root proliferation below the taproot fork.
function of distance to the soil surface, and to compute
characteristics of root crossing planes parallel to the Potential contribution to soil reinforcement
slope, the root system data coordinates were rotated Three types of figures can be used to visualize the
around the Y axis. spatial distribution of parameters of roots intersecting
6) Assessment of soil reinforcement: The potential planes parallel to the slope:
soil reinforcement by the roots was assessed by 1) The vertical distribution of both upslope and
computing several characteristics of roots crossing downslope means of parameters determining the
planes parallel to the slope surface at any depths i.e. reinforcement potential are displayed in figure 1.
potential sliding planes (fig. 1). The results were Tree S had a very low root CSA upslope whereas
displayed as multiple 2D graph on 4 planes parallel to for tree D, root CSA constantly decreased at 40 cm
the slope at 30 cm depth intervals (fig. 2). Additional depth and was two times larger upslope (600 cm2) than
soil cohesion due to the roots was computed using the downslope in shallow horizons. Mean angle toward the
angle of roots crossing the horizontal planes ([3]): planes parallel to the slope at impact point were around
S = RAR * Tr * K -30 above 40 cm depth and around -45 below 40 cm
where: depth in both trees. These depths corresponded
K = sin + cos * tan respectively to the horizontal shallow roots and to the
RAR is root area ratio, is angle of shear distortion intermediate and deep oblique and vertical roots
of roots crossing the potential slip plane, is soil impacts. The mean length to the previous branch was
internal friction angle. Tr is root tensile strength approximately constant (25 cm) between 30 and
computed from a generic equation ([4]): 110 cm depth whereas a linear decrease in the total
Tr = 28.97x-0.52 root length after impacts occurred between 40 and
where x is root diameter. 120 cm depth in tree D.
K is usually set to 1.2 regardless of the value of
[1].
7) Calculation of Factor of Safety (FOS): The FOS of
the slope can be computed using S as input in an
existing slope stability model, Slip4Ex [5].

3. Results
The two root systems had very different rooting
strategies. Tree D had no taproot and Tree S had a

75
2) Multiple 2D impact maps on planes parallel to
the slope (fig. 2) provide an overview of the spatial
structure of root reinforcement. In tree D, the 30 cm
deep plane showed intersections of large roots within a
radial distance of 50 cm and few impacts of small roots
up to 180 cm radial distance.
3) For tree D, at 30 cm depth, RAR was around
10% within radial distances of 50 cm but was null or
very low in sectors situated at a radial distance between
100 and 200 cm (see [6]). Conversely, the 90 cm deep
plane was reinforced by a large amount of finer roots
evenly distributed within a radial distance of 1 m, RAR
ranging from 0.14% to 3.3% in this specific area.
S within a radial distance of 50 cm of the stem
decreased from about 30 KPa at 30 cm depth to about
3 KPa at 120 cm depth in the planes parallel to the soil
surface.
When = 40, K was approximately 1.26 in both
Figure 2: Multiple 2D distribution on 4 planes parallel to trees and at all depths used in figure 5. However K
the slope at 30 cm intervals, tree S. The (0,0) was close to 1.0 when = 20 and in tree D was 0.75
coordinates correspond to the line passing through the when = 0.
centre of the stump and perpendicular to the soil Depending on the position of the potential slip
surface. a) root impacts represented by a circle
proportional to their diameter. The scale is the same for
surface, the FOS for unrooted soil varied between 2.8
coordinates and for diameters. b) root area ratio (RAR) 3.7 and 1.8 2.0 for slopes with trees D and S,
given in forty surfaces of unequal size determined from respectively (Table 2). When the mean value for S
dividing each plane in eight 45 degrees radial sectors in of roots was included, the FOS increased, depending
each of the five 50 cm radial distance rings spanning on the quantity of roots present (Table 2). Where S
the 0-250 cm radial distance interval. Small dots are the was high and the potential slip surface deepest (1.2
limits between surfaces for which the RAR was m), FOS increased the most (Table 2).
computed. c) Soil cohesion

76
Depth of Factor of Safety function of age and forest type, and could provide as
potential Slope Slope + Slope Slope + output all quantitative characteristics needed for the
slip Tree D Tree D Tree S Tree S evaluation of slope reinforcement by vegetation.
surface
(m) Acknowledgments
0.3 2.85 11.36 1.87 8.20
0.6 2.95 7.04 1.76 3.73 We thank L. Burnel (INRA), D. Porterfield, R. Hendrick and
0.9 3.26 4.81 1.91 2.38 L. Ogden (Warnell School of Forest Resources) for technical
1.2 3.66 3.96 2.03 2.13 support and J. Greenwood (Nottingham Trent University) for
assistance with Slip4Ex. Funding was provided by The Royal
Table 2: Factor of Safety (FOS) for unrooted and rooted soil Academy of Engineers, the Warnell School of Forest
on slopes with trees D and S. The FOS of unrooted soil Resources, UGA, CEG, Pittsburgh and the EU project Eco-
increased when the potential slip surface was deeper, but was Slopes QLK5-2001-00289.
reduced as root additional cohesion decreased with depth in
rooted soil. References

Conclusions [1] Reubens B, Poesen J, Danjon F, Geudens G, Muys B


The chain of tools developed will enable a more 2007 The role of fine and coarse roots in shallow slope
stability and soil erosion control with a focus on root system
accurate description and use of root architectural
architecture: a review. Trees 21:385-402
parameters in standard slope stability analyses. Root
architecture for the same species in a given stand can [2] Danjon F, Fourcaud T, Bert D 2005 Root architecture and
be highly variable [2], as for the two trees in our study. wind-firmness of mature Pinus pinaster. New. Phytol.
Therefore it is difficult to predict how a tree can 168:387-400
reinforce soil in a forest, depending on species only.
Tree D, having many oblique roots, but without any [3] Wu TH 1995 Slope stabilization. In: R.P.C. Morgan and
central taproots, would be more useful for fixing soil at R.J. Rickson (Eds), Slope stabilization and erosion control, a
the top or toe of a slope, where roots crossing the more bioengineering approach, Spon, London pp 221-264
vertical slip surface would reinforce soil better. Tree S
[4] Genet M, Stokes A, Salin F, Mickovski SB, Fourcaud T,
however, with its long, central taproots would fix soil Dumail J, van Beek LPH 2005 The influence of cellulose
better at the centre of the slope, where lateral roots are content on tensile strength in tree roots. Plant Soil 278:1-9
less likely to traverse the slip surface. The functions
developed would be useful for comparing root systems [5] Greenwood J 2006 SLIP4EX - A program for routine
uprooted in a landslide, with intact root systems in an slope stability analysis to include the effects of vegetation,
undisturbed area. This kind of information would be reinforcement and hydrological changes. Geotech. Geol. Eng.
useful in determining how e.g. age, species and 24: 449-46
planting history affect shallow slope stability.
[6] Danjon F, Barker D, Drexhage M, Stokes A 2007 Using
Combined with spatial/temporal models of root
3D root architecture models of shallow slope stability. Ann.
growth, such tools could also be used to forecast the Bot. In press.
development of root architecture on a slope as a

Table 1. Root size and branching characteristics. Entire root systems and % or value in each slope oriented sector: upslope /
perpendicular to slope / downslope (respectively "up", "pp" and "do"), stump removed.
Variable unit Tree D Tree S
Sector total up pp do Total up pp do
Length m 432 24% 56% 21% 172 7.6% 45% 48%
External surface m2 0.17 30% 54% 17% 0.059 6.9% 41% 53%
Volume m3 0.13 39% 48% 13% 0.053 4.8% 28% 67%
Dry weight kg 79.6 39% 48% 13% 32.8 4.8% 28% 67%
Root number n 706 26% 58% 17% 347 6.3% 42% 52%
Mean inter-lateral length cm 10.2 11.1 10.8 6.84 10.7 13.2 11.8 9.6
Mean branching angle degree 46 46.1 46.3 44.9 44.8 51.2 39.9 47.9

77

Вам также может понравиться