THE LAO JOURNAL O~-'AGRICULTURE AND ~-'ORESTRY MSEC special issue No. 17, September 2008 Management of Soil Erosion and Water Resources in the Uplands of Lao P.D.R.
Jo;nt Editors: Dr. O. Ribolzi
Or. A. Pierret
Or. L. Gebbie
Mr. O. Sengtaheuanghoung
Honorarv Editor: Or. M. Chanphengxay
Desiuned and Layout by: Khanhkham Ouneoudom, Information Ccnter, NAFRI
.
eencuu U:;)~ ~'llJii"l l'1£J: 2lJEl., SOJlJSCl.lJ, ~lJ~.lJlJ2"OJ~"lJ, ~fl:n1.J
. ..
Foreword This special issue of the Lao Journal of Agriculture and Forestry aims to highlight some of the recent activities undertaken by the partnership established between the International Water Management Institute (IWMI), an organization funded by the Consultative Group on International Agricultural Research (CGIAR), the Institut de Recherche pour le Developpement (IRD), a French public science and technology research institute under the joint authority of the French ministries in charge of research and overseas development and the Lao National Agriculture and Forestry Research Institute (NAFRI). The partnership was initiated in 2003 when IRD and IWMI took over IBSRAM and the ADB to support the activities of the Management of Soil Erosion Consortium (MSEC) in Lao PDR. In order to alleviate poverty, government agencies in Lao PDR have grouped and resettled upland villagers next to upgraded roads in an effort to provide better access to education, health and markets. To protect biodiversity and preserve some of the last remaining pristine forest environments of Southeast Asia, governmental policy has also reduced the overall area of cultivated land throughout the country. Such land use policies together with the growing market demand have resulted in very rapid land use changes in the uplands of northern Lao PDR. The ensuing need to adequately keep rapid land use changes under control has, in turn, induced a pressing demand for research on the interactions between policy makings, social and biophysical processes and their dynamics in the uplands. In this context, the IWMI-IRD-NAFRI partnership has embarked, through the MSEC, on an ambitious programme of long-term biophysical studies at the catchment scale which resulted in the collection of a unique dataset about soil erosion processes in South-East Asia, which will be publicly accessible at the end of2008. This effort has also been a unique occasion to assess the impact ofland use changes on soil erosion and water quality, and to test alternative farming practices that minimize the negative impacts of these land use changes. Agriculture remains the backbone ofLao PDR's economy: it generates more than 50 per cent of the national GDP and provides jobs to 84 per cent of the population. However, agricultural diversification in Lao PDR remains extremely low to world standards. From a policy making standpoint, it is both tempting and convenient to assume that the adoption of modem farming techniques and knowledge, applicable to a vast array of situations with no or minor alterations, is the way to deal successfully with major environmental challenges posed by rapid land use change. However, such a view does not take into account the diversity of physical and human environments which make up the basement of Lao agriculture.
In a renewed effort to address such burning issues, the lWMl-lRD-NAFRI partnership has recently expanded its activities towards the management of a wider array of environmental services in the context of the uplands ofLao P.D.R. The research reported in this special issue indicates that (i) the combination of biophysical measurements together with socio-economic information represents a promising option to assist policy making processes and (ii) optimizing land use management in the Lao PDR will likely depend as much on recognizing local specificities, limits and potentials as on modernizing farming techniques and practices.
Sincerely,
Dr. Monthathip Chanphengxay General Director, NAFRI September 2008
....
0
tJClUII)
OJ"l
lnOJV1J"l
"""
~~'tl"lUU
tI
fl.t~G~n;~l] Cl.\l ~
G~Ge~]~J, 19n;n;l.ll n;~n;J, 19n;GeGLl~rren;l;llLn]] Ll~~bn;~]] n;l.~~bllm] rr~lgn;llLn]] Ll~Ge~] I
"~Ln]]~~n;~lln;b n;~l.~C]~b ::li~]] lll.I1GryLa~llrCn;cp cl.\lG~n;l.ll~S n;~n;b~n;GeGl1l.'f! n;e~]n;l.llll(]::li~GeGn;Crnll~t
Ilg
l1e~Gl.ny~n;Cn;lg~Lll.rrl.~Ge~::li~
'"Ge~~~llllllllg~b~n;n;l.lll1cy ~n; ~
..' n;~ -
~bl.rrl.tn;~Ge~~~llllll
::lic~n;l.Y n;l.lln;CIl::liGIl>fCr\lG~] b . ll~]Gl.n~ ,
l.~pGe~ G~n;9~Ll~::linrr~ n;~~~b~n;n;r,mGl.Cn;l.ll "Cl.~ n'n'~ GeG eWl]lll.mGl.Lnr n;~l1e~ rrl.C~rr~ ~ll~::lilln;~~n;~LlG]n;b n;~Ll":p ~n;~~~~~ LlQn;:r]] n;~GeGn;e*] ::lil.~ml.ll r3U2(j~l)n
.,:
",
\..JlI
r;~r;BEJ(gLJr;&JJ ()~B~ '§rIn~§ BJE$}B~
'CJJyJrIS;; BJJ{]r;~
n;~()B13n;l.)neL,~ L,n;IT~n;L,U:U:Uln l.)C(;B131 ra
L rON (j(;gJfjJft!J>l(;g
....
ra
0
....0
-c;>
ra
'!'~ c.(l >l(:!JJ c.lf~>lU rrc.(;g>l(:!c.c
A new tension infiltrometer to measure the soil hydrodynamic properties on steep slopes Jean-Pierre VANDERVAERP,2, Olivier RIBOLZII, Christian VALENT/NI, Jean-Marc LAPETlTP,Jean-Marc M/SCIOSCIN and Oloth SENGTAHEUANGHOUNG3
Abstract In Northern Laos, there is increasing concern over soil erosion, an important factor of which is linked with land cultivation on steep slopes. Effective remediation or land use policies require in depth knowledge of the hydro-pedo-biological processes involved in these erosion mechanisms, which can only be achieved through the use of models. However, all models must be parameterized with correct estimates of their driving variables. Among those variables, at least one is always devoted to quantifying the soil hydrodynamic behaviour, generally the soil hydraulic conductivity, which is unfortunately complex and time-consuming to measure. Tension disc infiltrometers are often used to characterise this soil attribute but their use is limited to quasi-horizontal areas. A new approach is presented in this paper which aims to measure the soil hydrodynamic properties on steep slopes. The principle of tension disc infiltrometers measured parameters are related to a known slightly negative pressure head value - was modified in order to combine the advantages of large and small discs. With large discs, a well-conditioned hydraulic conductivity is determined whereas a quasi-homogeneous pressure head condition on a slope is applied with small discs. Thus, the new device was used successfully to estimate hydraulic conductivity in a well-defined slightly unsaturated condition. Fourteen tests were carried out in the Houay Pano catchment in Northern Laos on 35% and 67.5% sloping teak tree stands. Steady state infiltration fluxes, closely related to hydraulic conductivity, ranged between 4 and 22 mm/h indicating a soil a soil predisposed to a high risk of runoff. These preliminary results also indicate that more permeable soils are found on the steeper places in the old teak stands but on the lower slopes in the young teak stands where measurements were made.
Key words:
Disc infiltrometer; Slope gradient; Soil hydrodynamic properties; Teak plantations; Lao PDR
1UR 176 Solutions, IRD-NAFRI-IWMI, Vientiane, Lao PDR 2Laboratoire d'etude des Transferts en Hydrologie et Environnement, UMR 5564 (CNRS, IRD, INPG. U.IF). Grenoble, France ;>NAFRI, Agriculture Land Research Center, Vientiane, Lao PDR nlJFJ~, 2008
73
The Lao Journal ofAgriculture and Forestry, special issue No. 17
is that, by saturating the soil, soil cracks become hydraulically active, increasing Over the last twenty years, tension
hydraulic conductivity (K) by several
disc
become
orders of magnitude (Boolting et al.,
increasingly popular tools for measuring
1991, Schaap and van Genuchten, 2006)
soil hydrodynamic properties close to
compared to soil without cracks, and thus
saturation (Thony et al. 1991; Warrick,
the actual properties df the soil matrix
1992;
are masked during the experiment.
infiltrometers
Hussen
and
have
Warrick,
1993;
Logsdon and Jaynes, 1993 ; Cook and Broeren, 1994). Similar to their wellknown predecessor, the Muntz device (Angulo-Jaramillo et al., 2000), the principle consists in applying a constant pressure of water over a small surface of soil and measuring the amount of infiltrating water, the soil having initially a much lower water pressure head (h) than that applied at the surface. With Muntz devices, the boundary condition
pressure
head
(ha) is positive,
Le. a thin layer of water is maintained above the soil inside a ring which is inserted into the soil to prevent water runoff around the test area. The soil is thus completely saturated. After years of use, soil physicists realised that a major drawback of inserting the ring into the soil is that, in the case of a fragile soil surface, soil crust, roots, stones, etc., measurements will be inaccurate due to damage caused to the soil structure and the creation of artificial macropores. A further negative aspect of the technique 74
Tension disc infiltrometers appeared to solve the two problems cited above. They
impose
a
slightly
negative
pressure head at the soil surface, i.e. the applied water pressure is lower than the surrounding atmospheric pressure. Thus, well-defined soil parameters at the ha pressure head boundary value can be calculated, namely the hydraulic conductivity
K(ha) and the capillary sorptivity S(ha), (Philip, 1957 ; Elrick and Robin, 1981). However, this is valid only when the boundary condition is uniform , which limits the use of disc infiltrometers to approximately horizontal surfaces. When the aim is to characterise the first centimeters under the soil surface, which is frequently the case, itcan notbe dug out to create a flat horizontal area because this removes the soil layer which is of interest. The purpose of this paper is to present a modified infiltrometer design that aims at maintaining a quasi-uniform pressure head boundary condition on
September 2008
sloping soil and measuring the surface
is usually impossible if the soil is not
properties without damaging it at all.
flat and/or covered with vegetation.
Some preliminary results obtained in the
Thus, all aerial vegetation must be cut
Houay Pano catchment (Valentin et al.,
off, leaving roots in place to keep the
2008) in November 2007 are reported.
soil structure intact. To smooth out the irregular soil surface, a layer of fine sand is placed and flattened to receive the disc. The effects of this sand layer were
Tension disc
infiltrometers
(Perroux
extensively discussed by Vandervaere
and White, 1988) are made of a disc
et al. (2000a) who showed that particular
positioned on the soil surface and a
attention must only be paid to it at the
water reservoir closed at the upper end.
beginning of the assessment because
The air is forced into a resistant path,
after
either a Mariotte vase or a hypodermic
infiltration, the sand no longer influences
needle,
the observed flow.
which
maintains
a
slightly
approximately
one
minute
of
negative pressure head at the base of the disc what prevents water from flowing freely out of the device. Consequently, the unsaturated porous media (Le. the soil) on which the disc is placed will pull water out of the reservoir by exerting a capillary force due to the pressure head difference. If the soil is initially at equal or superior pressure head than that applied (which corresponds to a very wet state), no flow will occur.
Because only porous media will make water flow out of the disc, no physical boundary of any kind is required and the disc is simply placed onto the soil. An intimate hydraulic contact is needed between the disc and the soil, which is usually impossible if the soil is not flat and/or covered with vegetation. Thus, all aerial vegetation must be cut off, leaving roots in place to keep the soil structure intact. To smooth out the
Because only porous media will make
irregular soil surface, a layer of fine sand
water flow out of the disc, no physical
is placed and flattened to receive the
boundary of any kind is required and
disc. The effects of this sand layer were
the disc is simply placed onto the soil.
extensively discussed by Vandervaere
An intimate hydraulic contact is needed
et al. (2000a) who showed that particular
between the disc and the soil, which
attention must only be paid to it at the
1711£)'1,
2008
75
The Loo Journal ofAgriculture and Forestry, special issue No. 17
beginning of the assessment because
8 j by soil sampling, determination of K
after
of
requires the estimation of S which can
infiltration, the sand no longer influences
be achieved by analysing the infiltration
the observed flow.
curve at short time intervals (Smettem
approximately
one
minute
et al., 1994; Vandervaere et al., 2000a). At short time intervals, water flows into the soil mainly driven by capillarity due to the difference between the initial soil
Note that, in the case of an infinite disc radius, Eq. (1) would simply reduce to its vertical component:
moisture content (8) and that of the (2)
boundary condition (8 0 ) , This effect is represented by the capillary sorptivity S [LT-1/2] which depends on both
As the sand must not impede
8 i and 8 O' As time advances and water
the flow, a highly conductive type of
progresses into the soil, the capillary
sand is chosen with generally a 100-200
force decreases, as does the observed
IJm homogeneous granulometry. The
flow, until gravity becomes dominant.
hydraulic head H is defined as:
After a theoretically infinite time, all the surrounding soil will reach the 8 0 moisture
H=h+z
(3)
value and the vertical component of the
where z is vertical elevation, is quasi-
flow equals the hydraulic conductivity K
uniform within the whole sand layer
(8 0 ) , However, because of the laterally
since the hydraulic conductivity of the
moving water at the edge of the circular
soil is always much lower than that of the
source (Turner and Parlange, 1974),the
sand. Thus, in the case of a horizontal
total 1low q still depends on S(8 i-
),
soil surface and a thin sand layer, the
Wooding (1968) and White and Sully (1987) showed that the steady value of
pressure head ho is also uniform within the whole sand layer and the calculated
the axisymmetric flow can be expressed
soil parameters will correspond to their
by the so-called Wooding's equation:
well-defined ho values, K(ho) and S(ho)' The applied pressure head ho is freely
00
(
0
(1)
chosen by the experimenter, usually between -200 mm and -10 mm. A value
where r is the disc radius. After
of -10 mm corresponds to a saturated
and estimating 80 and
soil matrix with empty large cracks and
measuring q
76
00
September 2008
macropores (>3 mm). This is the most
1). In such a case, it is very important
frequently used value because cracks
to keep the infiltrometer pressure setting
and large macropores, when present,
always less than -3 cm otherwise the low
do not follow a Darcian behaviour and
points would be at a positive pressure
requires specific treatment within flow
head and water wouId runoff freely at the
models. At a -200 mm pressure head,
soil surface. This illustrates the need for
only pores smaller than 0.15 mm are filled
careful observation of the experimental
with water. Finally, by experimenting with
area before choosing the infiltrometer
several pressure head values on a given
pressure setting, to ensure that the entire
soil, an interesting exploration of the soil
sampled plot remains unsaturated. In the
structure with respect to now properties
example above, a pressure head great
can be achieved.
than -3.5 cm of water should not be set.
However, because of the non-flat nature
The infiltrometer disc size is generally
of the soil surface, in reality the applied
between 5 and 25 cm diameter. Large
pressure head cannot be absolutely
discs are more difficult to maintain
homogeneous over the sampled area.
in good contact with soil surface and
In the case of a microrelief with a 3 cm
require large quantities of water on
difference in elevation between high
permeable soils. Small discs are more
and low points (Figure 1), the pressure
portable but because a smaller area is
head of the soil will vary within a 3 cm
sampled it may not be representative of
range because the hydraulic head H, not
the larger area under study unless many
the pressure head h, is homogeneous
replications
within the contact layer. A 3 cm variation
important limitation of using a small disc
range is the usual pressure
head
is that more water will flow by capillarity
step between measurements made at
at the edge of the disc compared with
different pressure values (Ankeny et
water flowinq vertically and the second
al., 1991; Reynolds and Elrick, 1991)
term at right-hand side of Eq. (1) may
and it is thus considered an acceptable
become dominant over the first one
uncertainty on ho' For example, if hO is set at -5 cm, the applied pressure head
to the point that K cannot be properly
at the soil surface will vary between -8
then becomes a "sorptivitymeter". It is
and -5 cm with a 3 cm microrelief (Figure
generally considered that a 15 to 20 cm
nmJ'J, 2008
are
performed.
Another
estimated. A small disc infiltrometer
77
The Lao Journal ofAgriculture and Forestry, special issue No. 17
diameter is an appropriate compromise
this purpose (Figure 2). Very soft split
(Smetlem and Clothier, 1989; Elrick et
rubber tubing is stuck on the lower side
al., 1990; White et al., 1992).
of the tube guide to improve the contact
On lands with up to 67.5% slope such as those found in the Houay Pano watershed (Valentin et al., 2008), the use of a 15 cm disc infiltrometer is unsuitable because the pressure head difference between the upper end and lower end points of the sampled area would reach 8.5 cm, which exceeds the desired 3 cm range limit corresponding to a reasonable accuracy. Table 1 summarizes
the
characteristics of the different devices
with the irregular soil surface. Each mini parcel has a maximum extension of 45 mm in the direction of the slope to keep the pressure head variations less than 3 cm within the corresponding soil area. In the direction perpendicular to the slope, the mini compartments have the maximum possible extension (Figure 2) so that the area covered by the sum of the 7 mini compartments is very close to a full 15 cm diameter disc.
available for use on a 67.5% sloping soil.
As
The new device proposed here combines
horizontal disc infiltrometers, the contact
the advantages of the classic small and
between the disc and the soil is ensured
large disc infiltrometers.
with fine sand. On sloping land, it is very
described
above,
with
classic
difficult to keep a sand layer parallel to the soil surface without moistening the sand. Moreover, because of the A 15 cm diameter test zone
irregular soil surface, sand would move
is divided into 7 mini compartments
from one mini compartment to another
to each of which a separate mini disc
because the rubber tubing cannot keep
infiltrometer with a hypodermic needle
the compartments totally separated from
air entry is applied. Several needle
each other. This would create hydraulic
diameters are available to choose from
contact above the soil surface between
allowing the boundary conditions to be
the 7 corresponding zones which must
set between -100 and -5 mm. The 7
be avoided (otherwise, the experimental
mini compartments are separated from
conditions would resemble those of the
each other with a tube guide made out
classic 15 cm disc simply posed on a
of PVC which was specially designed for
slope). For this purpose hydrophilic
78
September 2008
cotton is used as the contact material
two manual readings are taken, during
instead of sand (Figure 3). Cotton wool
all experiments, on each of the seven
has enough rigidity to remain within one
reservoirs, to calibrate the seven relations
mini compartment without invading the
between sensors signals (mV) and water
neighbouring one. By gently moistening
levels.
the lower side with a common water spray
every second with a CR1000 Campbell
hose, it is soft enough to mould into the
datalogger. Once the seven water levels
irregular soil surface encountered inside
are calculated, the cumulative infiltration
a mini compartment. The very high
curve for the 15 cm diameter area can
hydraulic conductivity of cotton wool
be establish by adding the seven values,
ensures excellent water transfer from
which can then be analysed using the
the device to the soil (Figure 4). Finally,
same methods developed for classic
note that another important advantage
infiltrometer data. Figures 6 and 7 show
of cotton wool compared with fine sand
the infiltration area before and after
is that it does not fall into soil cracks,
the PVC tube guide was removed and
thus preventing them from becoming
the cotton remaining untouched. It can
artificially
be seen from Figure 7 that the contact
active
under
unsaturated
seven
are
recorded
surface covered with cotton is very
conditions. The
Measurements
individual
infiltrometers
are maintained perpendicular to the surface using a tripod suitable for slopes (Figure 5). The water levels in the seven reservoirs are monitored with pressure sensors installed at their upper ends. The sensors used have a limited range (0.5 psi) so that sufficient accuracy is guaranteed even in the case of a 45° angle between the reservoir and the vertical plane, which would correspond
close to the full 15 cm diameter disc. The area not covered with cotton wool is visually estimated at less than 10% of the total area. Regardless, it is only at the beginning of the test that this has an effect on the infiltration curve. After a few minutes, the seven soil zones become hydraulically connected, each of them, however, remains at its own hydraulic head value because of the resistance to the flow in the soil.
to a 100% slope. To prevent any error
The test duration is not known, a priori,
due to sensor electronic shift with time
since it will depend on the desired
and/or temperature variations, at least
outcome. If the aim is to measure the
nJJEJ'I,
2008
79
The Lao Journal ofAgriculture and Forestry, special issue No. 17
steady rate of infiltration, the test will
are still under analysis. Nevertheless, the
continue until an apparent steady flow
hydraulic conductivity can be evaluated
is reached, which requires a real time
as 60 to 80% of the steady flow values,
survey of the water levels. If only the first
based on field observations of the edge
one or two centimeters of soil is being
front progression at the periphery of
sampled, the test can be shorter than
the test areas. Tests were conducted
20 minutes and it is likely then that only
on two teak stands, old teak (O'T) and
a transient regime of infiltration will be
young teak (YT) and two slopes, 35%
available for analysis (Vandervaere et
and 67.5%. Results presented in Table
al.,2000b).
2 and Figure 9 show markedly different characteristics for each stand. The most permeable area of the four
Cigure 8 gives an example of a water
units was the 35% slope YT soil. In the
level recording. Two flow patterns, a
O'T stand, the less permeable soils were
fast flow at the beginning of the test
found on the 35% slopes whereas in the
(water infiltrating the cotton followed
YT stand these were found on the 67.5%
by capillary-driven flow into the soil)
slopes. The O'T results concur with
followed by a slower gravity-driven near
previous findings by Janeau et al. (2003)
constant flow, can be clearly seen. Short-
in l\Iorthern Thailand regarding runoff
time variations in the signal correspond
decrease with slope. However our finding
to bubbles going through the hypodermic
in the YT plantation do not, suggesting
needle. Each time a bubble is released
a complex relationship between land
in the reservoir, the pressure suddenly
use and soil properties. It is likely that
increases. To facilitate the analysis of
the young teak trees have not yet had a
the transient flow regime, the signal was
marked influence on the soil of this area
smoothed by only keeping the lower
suggesting that the values found in the
points (Fig. 8). This was done manually
YT stand may reflect the previous land
at present but should be achieved with
use, which was as fallow.
a specially designed computing program
In the above analysis, it was assumed that the infiltration bulbs inside the sloping soil are behaving similarly as in horizontal
in the near future. At present, only the steady regime willbe discussed as the sorptivity estimations 80
soil. Indeed, in the sloping geometry
September 2008
the lack of soil material downslope is
state
partially compensated by the upslope
vertical and lateral flow components.
soil surplus which probably makes the
Finally, forthcoming work will aim to
two bulbs quite equivalent, at least in
improve quantification of the effects of
volume. However, this clearly needs
the sloping geometry on the infiltration
further investigating and quantifying in
bulb, to confirm if it can reasonably be
the future through numerical modelling
treated as an axisymmetric situation and
work.
thus be described with the equations
flux
values
by
differentiating
developed in this context.
After important modifications to the classic tension disc infiltrometer design,
The research was conducted within the
a new device was proposed which
framework of the MSEC project, with the
was
financial support of IWMI (International
used
successfully
to
provide
infiltration measurements with unbiased
Water
comparisons between sites with very
IRD (Institut de Recherche pour le
different
infiltration
Developpernent) and EC02CO Cytrix
values ranging from 4 to 22 mm/h were
(project 71, ONDINE). The first author's
obtained on 35% and 67.5% sloping teak
research stay in Lao PDR was granted
stands showing that the relation between
by IRD. Thanks are due to MM. Kee and
slope angle and infiltration capacity is
Kamkhone for help with field work.
slopes.
Steady
Management
Institute),
probably complex and strongly depends on soil use history prior to the tests. These
preliminary
results,
however,
were obtained with a limited number of
Angulo Jaramillo, R., Vandervaere,
observations.
J.-P., Roulier, S., Thony, J.-L., Gaudet, J.-P. and Vauclin, M., 2000.
Although
still
under
processing
Field
at
measurement of soil surface hydraulic
present, the analysis of the first stages
properties by disc and ring infiltrometers:
of infiltration should provide sorptivity
A review and recent developments. Soil
estimations that can be used to calculate
Tillage Res., 55: 1-29.
the hydraulic conductivity from steady
September 2008
81
The Lao Journal ofAgriculture and Forestry, special issue No. J7
Ankeny, M.D., M. Ahmed, T.C. Kaspar, 1991. Simple field
and R. Horton.
Janeau,
J.-L.,
J.-P.
Bricquet,
O.
Planchon and C. Valentin. 2003. Soil
unsaturated
crusting and infiltration on steep slopes
hydraulic conductivity. Soil Sci. Soc. Am.
in northern Thailand. European J. Soil
J.55:467-470.
Sci.,54:543-553.
method for determining
Boolting, D.
H.W.G.,
Gimenez.
J.
and
Logsdon, 5.0., and D.B. Jaynes.1993.
crust
Methodology for determining hydraulic
hydraulic
conductivity with tension infiltrometers.
Bouma
Suction
1991.
infiltrometer for measuring
conductivity of unsaturated soil near saturation. Soil Sci. Soc. Am. J., 55: 566568.
Perroux, K.M., and I. White. 1988.
Design for disc permeameters. Soil Sci.
Cook, F.J., and A. Broeren. 1994.
Six methods for determining sorptivity and hydraulic conductivity with disc permeameters. Soil Sci. 157:2-11. Elrick,
Soil Sci. Soc. Am. J. 57:1426-1431.
D.E.,
Geering,
W.D.
and
Reynolds,
K.-A.
Tan.
Soc. Am. J. 52:1205-1215. Philip,
J.R.
1957.
The
theory
of
infiltration: 4. Sorptivity and algebraic infiltration equations. Soil Sci. 84:257-
H.R. 1990.
Estimating steady infiltration rate times for infiltrometers and permeameters. Water Resour. Res. 26:759-769. Elrick, D.E., and M.J. Robin. 1981.
Estimating the sorptivity of soils. Soil Sci. 132:127-133.
264. Reynolds, W.D., and D.E. Elrick. 1991.
Determination of hydraulic conductivity using a tension infiltrometer. Soil Sci. Soc. Am. J. 55:633-639. Schaap, M.G. and van Genuchten, M.Th. 2006. A Modified Mualem-van
Genuchten Formulation for Improved
Hussen, A.A., and A.W. Warrick. 1993.
Description of the Hydraulic Conductivity
Alternative analyses of hydraulic data
Near Saturation. Vadose Zone Journal
from the disc tension infiltrometers.
5:27-34.
Water Resour. Res. 29:4103-4108.
82
September 2008
Smettem, K.RJ., and B.E. Clothier.
Vandervaere, J.-P., Vauclin, M. and
1989. Measuring unsaturated sorptivity
Elrick,
and hydraulic conductivity using multiple
from tension infiltrometers. 1. The two-
disc permeameters. J. Soil Sci. 40:563-
parameter equation.
568.
J., 64: 1263-1272.
Smettem, K.RJ., J.-Y. Parlange, P.J.
Vandervaere, J.-P., Vauclin, M. and
Ross, and R Haverkamp. 1994. Three-
Elrick, D.E., 2000b. Transient flow from
dimensional analysis of infiltration from
tension infiltrometers. 2. Four methods
the disc infiltror:neter. 1. A capillary-based
to determine sorptivity and conductivity.
theory. Water Resour. Res. 30:2925-
Soil Sci. Soc. Am. J., 64: 1272-1284.
D.E., 2000a. Transient flow Soil Sci. Soc. Am.
2929.
Warrick, A.W. 1992. Models for disc Thony, J.-L., G. Vachaud, B.E. Clothier,
infiltrometers.
and R Angulo-Jaramillo. 1991. Field
28:1319-1327.
Water
Resour.
Res.
Sully.
1987.
measurement of the hydraulic properties
White,
of soil. Soil Tech. 4:111-123.
I., and
M.J.
Macroscopic and microscopic capillary
Turner, N.C., and J.-Y. Parlange. 1974.
length
Lateral movement at the periphery of a
infiltration. Water Resour. Res. 23:1514-
one-dimensional flow of water. Soil Sci.
1522.
118:70-77.
Valentin,
and
time
scales
from
field
White, I., M.J. Sully, and K.M. Perroux. C.,
Phachomphon, Chanhphengxay,
G.,
1992.
Rouw,
A.,
hydraulic properties: disk permeameters,
Pierret,
A.,
tension
Lestrelin, K.,
de
A.,
Measurement
of
infiltrometers
surface-soil and
other
Bourdon, E., Ribolzi, 0., Thiebaux,
techniques. p. 69-103 In G.C. Topp et al.
J.P. 2008. The MSEC project in Lao
(ed.) Advances in measurement of soil
P.D.R. at a glance: biophysical and
physical properties: bringing theory into
socioeconomic backgrounds and the
practice. SSSA Spec. Publ. 30 SSSA,
project's experimental setup. The Lao
Madison, WI.
Journal
of Agriculture
(submitted).
and
Forestry
Wooding, RA. 1968. Steady infiltration from a shallow circular pond. Water Resour. Res. 4:1259-1273.
n1J()'1,
2008
83
The Lao Journal a/Agriculture and Forestry, special issue No. 17
Table 1-
Some characteristics of two pre-existing and the new infiltrometer operating on a 67.5% sloping soil. Classic disc 5 cm
Classic disc 15 cm
New disc 15 cm
diameter
diameter
diameter
no
yes
yes
yes
yes
yes
Pressure head
acceptable
not acceptable
acceptable
homogeneity
(28 mm)
(84 mm)
(28 mm)
Suitable for K estimation Suitable for S estimation
Table 2-
Steady infiltration flow in mm/h on two Houay Pano stands, Old Teak (OT) and Young Teak (YT). Mean (bold), unbiased standard deviation (std dev) and number of measurements (n) in brackets.
35% slope:
67.5% slope:
84
aT
YT
mean:
4.1
14.8
std dev:
1.0
6.2
(n):
(4)
(3)
mean:
12.2
4.4
std dev:
5.3
0.8
(n):
(4)
(3)
September 2008
h =ha
~
~-----------------------------
I SAND I
~
3cm
h = ha + 3 cm Figure 1 -
____________t __
Schematic representation of the soil microrelief and its consequences on the applied pressure head.
September 2008
85
The Lao Journal of Agri culture and Forestry, special issue No. 17
Figure 2-
Tube guide for the 7 infiltrating tubes.
Figure 3-
Cotton wool ensuring contact between the infiltrating tubes and soil
86
n 1HJ6J, 2008
Figure 4-
The new infiltrometer functioning
Figure 5-
The new infiltrometer with tripod on a 67.5% slope ..
September 2008
87
The Lao Journal ofAgriculture and Forestry, special issue No. 17
Figure 6-
Figure 7-
Wet cotton wool after infiltrometer removal.
Wet cotton wool after infiltrometer use and tube guide removal. Note the well separated infiltration zones.
88
n1J()'J. 2008
-28 - , - - - - - - - - - - , - - - - - , - - - - . . , - - - - - - - - - - , - - - - - ,
t
-30
.
-32
! -34
-
s -36
.
Cii
~
.
-38
.
:JJ-ij...J\
-.~~.
.......
.. G-\-~
. ;" L~
End of contact with soil
• -:;~~
:;d..\ V
~ .. :;.~\Y"
r
.:iiiJ.'J.\Y
-40 -42
V
! rough signal i-smoothed signal
Beginning of contact with soil
-44..LC...--------------
o
500
1000
1500
2000
2500
Elapsed time (sec)
Figure 8-
An example of water level recording by pressure sensor, rough signal (points) and smoothed signal (plain line).
fIJ..JE.JGJ, 2008
89
The Lao Journal a/Agriculture and Forestry, special issue No. 17
Mean, Min and Max observed infiltration nux
20
,......~..,----------------
----------------------- ---------
.. VT
~C" 10
+OT ---------------------------------------~
--------------------------f -----------------------1- --------
5
O+----.---~-----,---___,---.,.......--_r_--__r_--__l
o
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
Slope
Figure 9-
Mean, maximum and minimum steady infiltration flux measured within the old teak (red squares) and young teak (blue triangles) stands as a function of the slope.
90
September 2008
·n~(jem~L:l rre~(jC]]G~Il~c~G~!-!2b ~n;Ill.m::in;n;l;l.~Ll~ ::gc:m GnllQeQp Cl.~G~~n;~m] I . . . .
~L;Q£n ~~(j ~ l.§Pucrn . n;~Si b1{n;Lll.t:.~U2e:: ~ rrQP::g~(jtP rn.u : ....
b
...
Ib
Ft5:P
I
I
(jl.8~ rr~lnGl.ln(ji]~e::n;
UIT
l'b
I
G~lf c!~'m (jn;~rrLll.t:.(ji]~e::n;QUl.t:.Llen; . n;~G~(j6m bn;~Ge6n;e~] ~l.8ml.Lln;e!! G~gL:l] LlQ
~e::]] n;~n;eLl~L:l (9 ~ n;l.GG~~ Ge6emm~]G~Lll.t:.li(j~~n;2Lll.t:.Ge~:p ~L:l~n;n;b~n;Q~Ll§ (tr I
I
...
~ ~n;rrmml.L:ll.[p ~e:: rrmr.ll.r.lU2::g6n;l.LlLll.t:.l.rrGe~:p
lb
G~gL:lmQe~]
I'"
(~
£:
•
~ IJlGlm~ rr ~l.n;::ie~6~
~e::]] r.le~m~~n;n;l.rr~~nn;e!! G&gL:l]LlQe~] Ge6n;egmon;l.Ll ~e::]] ~L:l~n;n;b~n;Qn;8]]~Ll§ (G
~~n;rrm] ~b (j~LleL;(j~ml.LlLll.t:.l.rrGe~;p G&rr~rrl.CL;IT ~e::]] gL:l] LlQe~] IT (~ :~G~IT
~n;Ill.m~n;n;l;~Il~~Lln;8n;Q] ~G~
~e::]]
r.l~] ~n;Ge6(j~:p ~Lln;l.Ll
·~n;G~(j~]Ge6r.ler(jo
ru,u ~e::]] l.n;rrm] lnn;mGl.llLll.lln;l.Ll 'l.m l.e:: rrmGL:ln;l.Ll :n;8] n;L;Ge6l.LlL:l(jm ~~]] (jl.8~ rr e U;';> v 0 n ro v 0 (.) (:) ...
b ' "
I
b
1
~lnr.l(j~(jQ~Iln;l.Lln;CIl~6 Gl.~U2~~ n;(jrr~rrl.CL;Lll.t:.l.rrGeQ]n;rTI ~n;Ill.m~n;n;L;Ge6n;cr
nTIf1l.Ll ·~n;r.ll.~Ge61ll.m::g~ ~~TI n;G~n;n;l.rr~::inGe6r.ll.t:.~Lln;l.Ll '~n;l.ll~~Il~n;B nTIn;~ n;r]] ~n;Ill.m~n;n;l; l.~n;\[,pl.~Ll~n;l.Lln;8 ·~n;n;br.le~n;C6]meLl~L:l~~]] gL:l] LlQe~] ~n;Q~ I
Ll§ : n;~] rrl.rt.a::g~Lll.t:.GQ~c~n;rr.?8b~n;r.l(j ~ n;~ rr8 (j6]1l~~~n;b~n;Ill.m~n;n;l;n;IT]~new] n;r]] ~G~l.~Ll~n;l.LlGe6G~~n(j~ .9 ~ n;B(j' 0.025). Contrasting this,
rainfall-runoff
T, EC and pH values were lower during
during and after such events, there are
storm flow while Eh was significantly
often significant increases in turbidity and suspended solid loads, which are frequently interpreted as an indication of bacteriological contamination. Table 2 also shows that CC37 were considerably higher during the storm flow sampling
higher. Dilution
of stream
flow
by
rainwater (lower T, EC and pH; higher Eh) explains the differences observed. Unsurprisingly, SL was much higher during the storm flow. These higher
usually not carried
survey.
events.
These
out
Even
during though,
observations
are
values result from various soil erosion
consistent with those of George et al.
processes
areas
(2004) who reported that fecal coliform
(Chaplot et al., 2007), rills and gullies
bacteria were linked to particles in small
(Chaplot et al., 2005) and the washing-
streams and that the fraction increased
out of free aggregates and part of the
with suspended sediment content.
affecting
September 2008
inter-rill
101
The
observations
presented
above
the question of water quality and the
were conducted along a perennial third
potential associated health risks are an
order stream passing villages close
important area of study.
to the IVlekong corridor but having no direct access to this main river. The objective was to undertake a preliminary assessment of stream water quality at the community-level and based on in situ measurements of several operational parameters (i.e. oxygen content, pH, electrical conductivity, suspended sediment load, total bacteriological flora).
The expansion of Luang Prabang city and its population growth pose a major challenge for city planners. In the near future, it will lead to an increased demand for sanitation infrastructures and freshwater resources, notably for irrigated peri-urban market gardens. The current expansion is characterised by a dynamic from the ancient peninsula city,
This study confirmed that, due to poor
following the stream water paths and
sanitation conditions, a high degree of
spreading progressively up hill slopes.
bacteriological contamination occurs in
The following
stream water passing villages and peri-
suggested in order to reduce or mitigate
urban areas. The most common source
potential negative impacts on water
of pollution was from urban household
quality:
wastewater,
which
may
recommendations
are
potentially
contain pathogens (but also nutrients). Pollution of water sources from industrial effluent was not as common.
• Riparian zones along streams and rivers should be managed in an environmentally friendly and sustainable manner. A strip of natural vegetation on the bank should
The downstream section ofthe catchment mostly includes irrigated pari-urban gardens (via natural streams or small diversion channels) which supply Luang Prabang city with vegetables. These market gardens rely largely on surface waters that could be increasingly polluted by the rapidly growing population. Thus
102
be preserved and protected in order to maintain the. sediment-filtering function of this zone. Encroachment of stream banks by residences, non strengthened fish ponds or informal hydraulic infrastructures should be strictly limited and controlled in orderto prevent material or even human losses generated. by
nVEJ'I,
2008
The Lao Journal ofAgriculture and Forestry, special issue No. 17
landslides or flood hazards.
city and the surrounding villages be
• Environmental discharge thresholds should be estimated and then base on these estimates, water which is extracted from the river system for manufacture activities
and
irrigation
should
be
regulated following a seasonal schedule taking into account rainfall variability and upstream land use. Over-extraction of stream water will place freshwater resources under stress. •
Authorities
should
organise
and
level decentralised sanitation systems. not,
direct
the Payment for Environmenta} Services (PES) concept:
rural dwellers could
loosen the pressure on riparian areas in return for which the urban citizens could finance sanitation infrastructures upstream via, for example, the taxation of profits made on certain tourist activities in Luang Prabang. Finally, our study raises the issue of
encouragethedevelopmentofcommunity If
implemented. This agreement may follow
domestic
wastewater
discharges will rise and stream water passing villages and Luang Prabang will turn into sewage. Local inhabitants and tourists would not only be inconvenienced by bad smells but also health concerns if the polluted water is still used to irrigate food crops.
the spatial
scale
relevance of field
observations regarding the question that needs to be answered, Le. do upland people of northern Lao PDR have access to good quality surface water? Strategies that consist in monitoring large rivers generally provide a smooth integrated fingerprint of entire watersheds (e.g. UN, 1998). This is unquestionably useful for global water resource management atthe regional scale. However this approach may mask system internal variability and
water and
hence part of the local community level
sanitation services to pert-urban areas
reality. Conclusions from such large
and neighbouring villages in order to
scare studies should be considered with
reach the poorest people is of the utmost
the greatest care.
Extending basic
importance to
drinking
prevent
outbreaks of
cholera and otherwater-related diseases. To
support
the
above
mentioned
recommendations, we suggest that an agreement between the Luang Prabang
September 2008
103
Nationale Superieure, Paris, France. by
the
George, I., Anzil, A., Servais, P., 2004 -
ORV (Departernent
des
Quantification offecal and coliform inputs
Ressources Vivantes) and the research
to aquatic systems through soil leaching.
unit SOLUTIONS of IRO (Institut de
Water Research, 38, 611-618.
Recherche pour le Developpement), and
Hua, H.S., 1990 - Accurate methode for
IWMI (International Water Management
calculation of saturation DO. Journal of
Institute).
Environmental Engineering. 116(5), 988-
This
work
department
was
supported
990. Huon, 5.,
Ribolzi,
0., Aubry, E.,
NSC, 2005 - The 2005 population
Bounsamai,
census. Vientiane: National Statistics
Angeli, N., Sengtaheuanghoung, 0.,
Center of the Lao POR). http://www.nsc.
2008 -Iron and manganese concentration
gov.la/PopulationCensus2005. htm.
levels in watercress cultivated within
Chaplot, V., Coadou le Brozec, E., Silvera, N., Valentin, C., 2005 - Spatial and temporal assessment of linear erosion in catchments under sloping
5.,
Longchamp,
M.,
the main stream of the Houay Pano catchment, Northern Lao POR. The Lao Journal of Agriculture and Forestry, This issue.
lands of Northern Laos. Catena, 63
Lamaningao, P. and Sugiura, Y. 2004
:167-184.
- New ways of looking into health and
Chaplot, V.,Khampaseuth, X., Valentin, C., Le Bisonnais, Y., 2007 - Interrill erosion in the sloping lands of northern Laos submitted to shifting cultivation. accepted Earth Surface Processes and Landform. 32, (3) 415-428. Galvez, M., 2007 - Non-conservative processes in a mountainous catchment of Lao P.O.R. - Master thesis, Ecole
104
hygiene promotion activities in Lao POR. In: S. Godfrey (edit.), People-centre approaches to water and environmental sanitation.
30th WEOC International
Conference, Vientiane, Lao POR, 111114. Maniphousay, N. and Souvanthong, B. 2004 -. Application of a household water storage chronination project in the RWSS Programme of Lao POR. In: S. Godfrey
n'1JEJ'I.
2008
The Lao Journal ofAgriculture and Forestry, special issue No. 17
(edit.), People-centre approaches to
commission for Asia and
water and
sanitation.
UN, Economic and Social Commission
30th WEDC International Conference,
for Asia and the pacific. New York, USA.
environmental
the Pacific.
Vientiane, Lao PDR, 656-659. Valentin, C. Lestrelin, G., PhachomRandall, W. G., McCarthy, J., Layton,
phon, K., de Rouw,A., Chanhphengxay,
A., McKay L. D., Williams, D., Koirala,
A., Chaplot, V., Bourdon, E., Briquet,
S. R. and SaylerG. S., 2006 - Escherichia
J.P., Marchand, P., Pierret, A., Ribolzi,
coli Loading at or Near Base Flow in
0., Thlebaux, J.P., 2008 - The MSEC
a Mixed-Use Watershed. Journal of
project in the Lao P.D.R. at a glance:
Environmental Quality, 35:2244-2249.
biophysical
Ribolzi,O.,Silvera,N.,Xayyakummanh, K.,
Latchachak,
K., Tasaketh,
S.,
and
socio-economic
background and project experimental set up. The Lao Journal of Agriculture and Forestry, this issue.
Vanethongkham, K. 2005 - The use of pH to spot groundwater inflows along
Vigiak, 0., Ribolzi, 0., Pierret, A.,
the stream of a cultivated catchment in
Sengtaheuanghoung, 0., Valentin, C.,
the northern Lao PDR. The Lao Journal
2008 - Trapping efficiencies of cultivated
of Agriculture and Forestry. 10: 72-84.
and
natural
riparian
vegetation
of
northern Laos. Journal of Environmental Silvera, N., Ribolzi, 0., Xayyathip,
Quality, 37(3), 739-1316.
K., Sengtahevanghoung, 0., Pierret, A., 2007 - Relative performance of four methods for gauging stream discharge of headwater catchments under low flow conditions. The Lao Journal ofAgriculture and Forestry, 14: 100-112. Stumm, W. and Morgan; JJ., 1981 - Aquatic Chemistry, Wiley, New York, 780pp. United Nations (UN), 1998. Sources and nature of water quality problems in Asia and the Pacific. Economic and social
September 2008
WHO and UNICEF, 2006. Meeting the MDG (Millennium Development Goals) drinking water and sanitation target: The urban and rural challenge of the decade. 1.Water resources development. 2.Potable water - supply and distribution. 3.Sanitation. 4.Water supply. 5.Millennium development goals. I.WHO/UNICEF Joint Monitoring Programme for Water Supply and Sanitation. II.World Health Organization. III.UNICEF. WHO Library Cataloguingin-Publication Data, WA 675. 105
Table 1-
Selected hydro-morphological characteristics of the Houay Xon catchment: surface (A) and planar (Ap) areas; perimeter (P); minimum (Zmin), maximum (Zmax) and mean altitude (Zmean); length of the main stream (L); Gravelus index of compactness (Gc) which is the ratio of the perimeter to the perimeter of a circle which have the same surface area; drainage density (Dc); mean slope gradient (S).
A
Ap
p
Zmin
Zmax
Zmean
L
Gc
Dc
s
592
10605
1.6
1.4
31
m 22.4
106
20.7
26529
280
1336
nm)'), 2008
The Loo Journal ojAgricu/f/ln.:'
Table 2-
LlI7J
Fo/"e,\'/IT. spccia! ISS/le: No. /7
Descriptive statistics of some hydrological characteristics of the Houay Pano stream during baseflow and storm flow periods for 11 selected observation points: median; arithmetic mean; standard deviation (sd); minimum (min) and maximum (max) values of stream discharge; temperature (T); electrical conductivity at 25°C (EC); pH ; redox potential (Eh), dissolved oxygen content transformed to oxygen saturation (DO-sat); suspended load (SL); total colony count at 37°C (CC37).
Flow regime
Discharge
(date) BASE (28 May 2007)
STORM (4 Jun 2007)
T
pH
EC l
Eh
DO-sat
SL
CC37
mV
%
g.l 1
CFU.mr '
lIs
°C
median
0.4
25.6
388
8.3
163
70.6
023
808
mean
0.4
25.7
374
7.8
158
67.1
0.38
1152
sd
0.1
0.8
31
25
21.3
0.50
861
min
0.3
24.4
309
6.9
88
12.0
0.06
186
max
0.6
26.9
417
8.5
181
91.8
1,85
2760
median
5.0
23.9
195
7.6
227
81.2
8,66
19000
mean
4.4
23.9
196
7.6
226
73.1
8.80
42469
sd
1.4
0.2
16
8
23.8
3.98
55984
min
2.4
23.7
170
216
23.9
2.42
3840
93.4
18.72
183000
max
5.3
24.5
J-lS.cm
232
7.5 7.9
238
Differences between base flow and storm flow (threshold of significance: aJ2 = 0.025) are significant for all the parameters (Wilcoxon test, paired samples) except DO-sat. Mean pH is calculated from H+ concentrations. CFU.ml- l , colony-forming units per ml.
September 2008
107
Hou ay Xon catch ment (Northern Laos , Luang Prabang province)
Main stream
Trlbutaries Contour line (m )
•
Old prol l>Ctedfort? st ( 33. 1%)
•
Sec ondor)l IcfG~~UL:m~n;IDfl~m ::lI~rr~r3c!!2r3~~n;c] ~Q~un~ ~n;u~n;b~] m]~uGQ' ::lI~n u@m~lJlGBGrr~::lI~n;~u~~1l~9 ~ Ul)[] °r3Bl)rrmGm b n;*rr~l)~Il~BrTfTIUQ' ::lI~]] rn~~n;r3c!!2~l)~:puQ'G~ ~n;u~ °fl~m::llG&~ fl~G~lJln;B!!2::l1~n;8G~l)~B ~n;u~ :n;*] l)G::lIaG~flu~n;b~uu.a::llLIl~tGBGc8l)!:1~nn;~u f'3Wl)~l)n
..!
6
'{j§J~{g
"
....
~£~B~Jt..(J~{gJJ {j~B~ 'o~u~[f '~~~Be [ft..e[J 6 '" Wi,. «erta '~gucn fIBJfIt..rrBJ '§n~§ BJg§B~ 'fIC£ I
~CJJ~§
(;l,naru(;cw(;cGn ~l,nr3C!!l (;~~ (;l,Srun ~rtU~ ..
I
I
I
..
rtl~rtQ~ ~1rtnl,u(;!! a~n Um1Ul,W(;SGrt~IT~1ITl,Cl3nQa~
.
-
!. rON ()(gJ'!Jnrp'I(g !'~ ~(7 «e» ~if~::lU rr~(g::le~c
Iron and manganese concentration levels in watercress cultivated within the main stream of the Houay Pano catchment, Northern Lao PDR Sylvain HUON', Olivier RIBOLZI2, Emmanuel AUBR'fI, Bounsamai SOULlLEUTH2, Melanie LONGCHAMP', Nicolas ANGELI' and Oloth SENGTAHEUANGHOUNG3
Abstract The bio-availability of metals to currently eaten aquatic herbs such as , watercress is of important interest because of its potential impact on human health. Watercress grows in clear running water and is suited to the moist climates of mountainous tropical regions. In this study, the concentrations of Fe and Mn in watercress, stream bottom sediments and ambient stream water were measured in order to evaluate the impact of local environmental conditions on plant Fe- and Mn- accumulation levels. Two sites in the Houay Pano catchment in northern Laos (Luang Prabang province) where watercress is cultivated directly in the stream waters were selected for their contrasting environmental conditions (oxic vs. dysoxic to suboxic oxygen levels). Total, dissolved, exchangeable and potentially exchangeable Fe- and Mn- concentrations were measured. The results indicate that local environmental factors are determinant factors in the availability of these ionic species. Both the total metal content in the sediments and the metal ion speciation status (availability of dissolved species) determine the level of Fe and Mn uptake by watercress. The concentration levels in watercress increased by a factor of 7.5 for Mn and 2.2 for Fe in the dysoxic to suboxic site (swampy area) compared to the oxic site (upstream running water). The exchangeable and potentially exchangeable Fe- and Mn- concentrations in bottom sedtments' and the dissolved Fe- and Mncontents of ambient water at both sites were consistent with the accumulation levels measured in the watercress.
September 2008
113
The bioaccumulation of these metals should not pose a direct threat in itself, because these elements are not actually very toxic, but rather it is indicative of the local physico-chemical conditions that could be potentially favoring the mobility and accumulation of toxic metals such as the so-called atmophile elements (e.g. Cu, Cd, Zn, As). These elements are the subject of an ongoing study. Because watercress appears to easily accumulate metals it could be used to clean contaminated water by soaking up the toxic elements. Key words:
Aquatic herbs; Metal bioaccumulation; Mountainous environment; Bioremediation, Lao PD. R
'Universite Pierre & Marie Curie (UPMC) - UMR 7618 Bioemco - Case 120.4 place Jussieu
- 75252 Paris cedex 05, France (
[email protected]) 21nstitutde Recherche pour le Oeveloppernent (IRD), International Water Management Institute (IWMI), National Agriculture and Forestry Research Institute (NAFRI) - clo Ambassade de France - BP 06 Vientiane, Lao PDR (ribolzi@irdJr) 3NAFRI, Soil Survey and Land Classification Center (SSLCC)
114
n1JEJ'J,
2008
roots to reduce these metals to their mobile forms (Le. Fe 3 + to Fe2 + and Mn4 + or Potentially toxic metals may accumulate
Mn 3+ to Mn 2 +). Although these processes
in soils and sediments along the banks
are controlled by microbiological activity
of river systems. Metals are either
they may be monitored by environmental
derived directly from man made waste
parameters:
accumulation
(redox
sewage
(Le.
sludge
industrial
potential);
dissolved
pH,
Eh
oxygen
fertilizer
concentration and turbidity, which affect
application on crop fields) or, indirectly
the behavior of metals in natural waters
removed
occurring
(Stumm and Morgan, 1996). From a
metals, located within mineral structures
purely geochemical point of view, the
or sorbed onto mineral surfaces. Their
extent of metal adsorption by plants is
natural concentrations in soils define
controlled by: (1) the total metal content
part of the so-called "soil geochemical
and, (2) the metal ion speciation status
background",
the
in the soil or sediment (Kabata-Pendias,
composition of the rocks of the geological
1993; Alloway, 1995). Both parameters
basement.
contaminate
are influenced by the "free metal ion"
surface and ground waters connected to
concentration in the pore water (Le. the
the river system (Naiman et al., 2005).
Fe 2+ concentration with respect to the
Health authorities recommend very low
total Fe concentration), which is linked
metal concentrations in drinking water,
to the presence of soluble organic and
for example in France these values are 5
inorganic complexes (Benedetti et al.,
~g
1- , 10 ~g 1- , 50 ~g 1- and 200 ~g 1- for
1996). Both adsorption onto Fe-Mn-
Cd, Pb, Mn and Fe, respectively (Decret
oxides and hydroxides surfaces and
2001-1220, December 12, 2001). Metal
complex formation with organic matter
bio-availability generally refers to the
reduce the amount of "free metal ions"
ability of a given element to be transferred
in the water and thus, the extent of
from soil (or sediment) to water. Plants
metal uptake by aquatic plants. Several
take up most of their trace nutrients
studies have shown that the reactivity
from water and can therefore potentially
of trace metals is strongly linked to the
accumulate metals if available. Uptake
presence of oxides and hydroxides in
of Fe and Mn by plants depends mainly
soil and sediments (e.g. Kabata-Pendias
on soluble pools and the ability of plant
and Pendias, 2001; Dumat et al., 2001).
from
naturally
inherited
Metals
1
September 2008
disposal,
waste,
temperature,
1
may
1
from
1
115
The Lao Journal ofAgriculture and Forestry, special Issue No. 17
Antagonist interactions between metals
the total Fe- and Mn- accumulations
also occur, Le. excess
amounts of
in watercress grown in this watershed
Mn in plants, reduces absorption and
and to compare these concentration
translocation of Fe, resulting in a decrease
levels with those determined for bottom
of chlorophyll and photosynthesis.
sediment and ambient water. These
The issue of metal bio-availability to aquatic herbs, in particular watercress (Nasturtium officina le), is of important interest because this cultivated plant is
determinations provided
a first-order
estimate of potential metal accumulations in sediments and plants in relation to the local environmental conditions.
currently eaten raw as salad or cooked as a vegetable. Metal concentrations in commonly consumed watercress are usually very low, ca. 0.009 - 0.3 g kq' and ca. 0.002 - 0.03 g kg-1 for Fe and Mn, respectively (Cumbus et al., 1980; Mohamed et al., 2003; Kawashima and Soares, 2003). However, much higher amounts were reported, up to 7.6 ± 0.7g kq' for Fe and to 08 ± 0.2 g kq' for Mn, for iron-contaminated sites in China (Wong, 1985). Watercress grows in clear running waters and is suited to temperate environments or the relatively cool moist climates of mountainous tropical regions (Herklots, 1972). In the province of Luang Prabang (northern Laos), watercress is usually grown from cuttings and planted in waterlogged areas. On the Houay Pano catchment near Luang Prabang, local cultivation is set up directly in the main stream. The aim of this study was to measure
116
The biophysical and socio-economic characteristics of the study area are described in detail by Valentin et al. (this issue). The geological basement of the catchment is composed of Permian to Upper Carboniferous sedimentary to low grade metamorphic rocks (schists, mudstones and fine-grained sandstones) overlaid by limestone cliffs. Soils are cultivated on steep slopes (average slope: 60 %) by slash-and-burn with no fertilizer input. The soils of the area are dominated by entisols (18.5 %; clay soils with medium fertility, pH = 6.4), ultisols (33.1 %; clay soils with medium fertility, pH
= 5.5),
and alfisols (48.5 %; heavy
clay soils with medium fertility, pH
=
6.2). The stream in which watercress is grown is a third order tributary of the Mekong, located 10 km south of Luang Prabang. The permanent flow is fed by
i11J(J~,
2008
groundwater (i.e. 1.6 - 5.2 I s'; Ribolzi et
suspended solid material from dissolved
ai, 2005) during the dry season. During
(and colloid) material and; (2) in-situ
the rainy season, the water level is
selective metal extractions, based on a
controlled by the amount of rain and the
DGT procedure (Diffusive Gel Transfer,
soil saturation status.
Zhang et al., 1995). The DGT device
Cultivated watercress, bottom sediments and water samples were collected in March 2007 during the dry season from two different sites, located in the upper (site A) and central part (site B) of the catchment (Figure 1). Both sites are included in a monitoring survey of groundwater inflow and outflow in the main stream of the Houay Pano catchment (Ribolzi et al., 2005). The environmental
parameters
measured
the day of sample collection (3/7/2007, 11 am - 2 pm, Table 1) indicate that site A has "oxic conditions" ([02] > 2.9 mg
1- 1 , Tyson and Pearson, 1991). However, site B is located in a swampy zone where "dysoxic" to "suboxic" conditions prevail (2.9 < [02] < 0.0 mg 1- 1 ; Tyson and Pearson, 1991). For site B, it is of note that watercress only grows in areas where oxygen levels are high (samples B1 - 86, Table 2).
for water sampling: (1) a conventional with
water
collection
in
plastic bottles and filtering through a 0.2 IJm acetate filter, in order to separate September 2008
impregnated in a hydrogel to accumulate the metals. Ions must diffuse through the filter and the diffusive layer of hydrogel to reach the Chelex resin. The DGT devices were immersed in the stream at sites A and B, very close to watercress cuttings and were recovered after 3 days. This procedure accumulates "free metal ions" on the resin and, thus, very low concentration levels may be determined in natural waters, in particular when these levels are bellow the detection limits of currently used chemical analyzers, Fe and Mn concentrations were calculated using
the
procedure
described
in
Zhang et al. (1995) with temperature corrections required for the application of diffusion gradients. Total Fe- and Mn- concentrations were determined by Atomic Absorption
Spectroscopy
(Unicam 989 OZ and Unicam AA series, detection limit: 6.5 ppb).
Two different techniques were applied technique
uses a specific layer of Chelex resin,
Watercress
samples
were
dried,
weighed and grounded. Aliquots (200 mg) were dissolved in 2 ml of 0.43M HN03 and 1 ml of 30 % H202 for 24
117
The Lao Journal ofAgriculture and Forestry, special issue No. 17
h at 100°C. After dissolution, solid vegetation residues represented less than 10 % of the initial sample weight.
Fe and Mn concentrations in water-
The solutions were centrifuged at high
cress and bottom sediments
speed to remove any solid material and
IngeneraltotalFe-and Mn-accumulations
filtered through a 0.2 IJm acetate filter.
in plants show remarkable variations
Sediment samples were finely grounded.
depending on the plant species, soil type,
Selective extractions were carried out
stage of growth, as well as ecosystem
on two replicate sample aliquots (2 g)
(Kabata-Pendias and Pendias, 2001).
using procedures described in Dumat
The minimum Mn level for most plants
et al. (2001). Exchangeable (= soluble
is 0.015 - 0.025 g kg,1 but may increase
in water) Fe and Mn were extracted
above 1.0 g kg'1 for some species. A
with 0.01M CaCF during 24 hours at
large range is also observed for Fe in
20°C. Potentially exchangeable metals
plants, ranging from 0.02 g kg,1 up to
(= bound to oxides and/or hydroxides)
3.5 g kq'. Because watercress grows
were
replicate
in waterlogged or aquatic systems, I'v1n
sample aliquots (2 g) with 0.43M HN03
and Fe are more easily available and
and 6M HCI for 2 hours at 20°C. The
plant concentrations should fall in the
solution recovered for both extraction
upper range. It was observed that the
procedures were centrifuged and filtered
total Fe and Mn contents in watercress
as described for the vegetation samples.
were higher for site B (swampy area)
The total organic carbon (TaG) and
than for site A (upstream). Fe and
total nitrogen (TN) concentration
of
Mn concentrations ranged from 14.1
sediments were determined using a
- 46.5 and 21.1 - 26.6 g kq' and 4.4
Carlo-Erba elemental analyzer (Girardin
- 19.6 g.kg'1 and 1.0 - 10.0 g.kg'1, for
and Mariotti, 1991). All measurements
sites B and A, respectively (Table 3). A
(Tables 2 - 4) were carried out in the
simple explanation for this difference
laboratories of UMR Bioemco (Universlte
is the contrasting redox status of each
Pierre & Marie Curie) in Paris.
environment.
extracted
from
two
However,
significant
variations linked to the environmental parameters are also observed locally (Table 2). The dissolved oxygen content
118
n'JJEJ'J,
2008
and the Eh values are higher for site A
10.0 - 24.0 g kg-1 vs. 2.2 - 5.7 g kq',
(56.4 % < [02] saturation < 81.1 %; 166 mV < Eh < 270 mV) than for site B (1.4
respectively. For Mn, the same trend was observed with 4.2 - 20.4 g kg-1 and 0.9 -
% < [02] saturation < 27.3 %; -2 mV
4£££4££££ i&;4f'i.,
Accountability
Villagers are believed to be responsible for water quality degradation. No clear opinion re. changes in water quantity.
Willingness to implement change
I Population is ready to pay for waste collection system but not to invest in upstream land use change although WTP sufficient to abate I soil erosion.
nJJC)'J,
2008
The Lao Journal of Agriculture and Forestry, special issue No. 17
Land use before PES
Modified land use with PES
Beoefll~ 10
land n~'~
Costs to downstream popu tauons
Figure 1 -
Schematic representation of the PES principle (After http://www.itto.or.jp//ive/Live_Server/2869/18_Sander.pdf).
September 2008
143
The Lao Journal ofAgr iculture and Forestry, special issue No. / 7
To Vientiane
Legend -
o
Figure 2-
/ 44
Road
nO'3
Houay Xon 1 (Ill
2
~eters
Tributaries
Map of the watersh ed and location of villages
n 1JtJ'J,
2008
The Lao Journal ofAgric ulture and Forestry, special issue No. 17
Perception of water quantity Very bad
5%
7016
27%
Perception of changes In water quantity 5%
1:1 Bad
lJ negative
Cl A_age
Cl none
Cl Good
•
lJ pos itive
Very good.
35%
43%
• No opinion
Figure 3-
Map of the watershed and location of villages
Perception of changesin water quality
Perception of water quality •
Very t.l
very negative
Cl Slid
Cl negative
lJ A_age
Cl none
Cl Good
44%
Figure 4-
September 2008
• very positive
•
Vety good
Cl posllive 33% • very positive
Map of the watershed and location of villages
145
The Lao Journ al of Agriculture and Forestry, spec ial issue No. / 7
Farmers
+ Legend Major areas for eachuser type
o o
VLllagas
Impact of activities by eachuser type ~
Vlllagas
Faxmers
Fas:mers .
HouayXoo
Fish fanners
Fish f'armers
Tributaries
GardeDers
GardeDers
Figure 5 - Spatial relationship between the different water users along the Houay Xon.
/ 46
ii1JV~, 2008
The Lao Jo urnal ofAgriculture and Forestry, special issue No. 17
Figure 6 - Garden plots in Houay Xon.
Sep tember 2008
147
The Lao Journal ofAgriculture and Forestry, special issue No. 17
DOWNSTREAM
Household wastes
Water & tractio n ......
Figure 7 - Complex water use relationships in the Houay Xon watershed.
148
n1JV6'J,
2008
The Lao Journal ofAgriculture and Forestry, special issue No. 17 I
.,
tuClU t CltJU~t~Bmn6)ua ;)6)£}6)S2a~ClU LUC2Cla6)~tci~ W6)ncuua 2a~ ~.tJ.tJ ;)6);)
n6)utJ~dIU£}6)'c ;)tJ~n6)UU;)JJU~ I
CBJfJ1JebB1J U;Jncb, '\I'
oow U6JG}W, B1J6JG}JJ 6J;J,nB'f7~, 9
U1J~~CWJ ~@dJCJ, 9 9
B%)1J~~m5CJ 91JCW,Ct:?, BG}cdJ, [JCBCS, !Bf:JSCB g~U1J~ 9
U~l c.1C1U ~muu [[W~~~f1U lun')uC1SU ~!::::V1Js~U~lUiLLrlUt:g"lnsu usn'OJ"lniJ LL@;) U~lb1C1U eJ~g;)tH~bllUi lKl,)tJu~Ui;)tJ 'c [lLuuDnnt:E1 'c C)C1!::::[lsbl~l ~blU!::::E~2s~n,)u~uf15') [[JJU LitSUt: LDu5tJLLUU~~@S~ 2 gt:1Jbl LitSfl~U;)Ut\') 'c @tJt:L;)@')n"lUU;)JJ1J~2S~E1u'OJ')n
U~lc.1C1U~U, 1 bl'UtJg~ liln')uuJu [[u~~2s~E1uLuuwn, 5Uin')u~~n');) 'c clWbllJ1t:u')JJ')"V')n B lavet (2000) lU L2bls')@nn')C1"lL;)uEin LL @t: ~~ mmmu;)') 5mn')u ~~n"l;)
lun,)1J1J~l~
tlL2blWblStJ 'U 'U
2S~ lKl.U.U
a-io
LL~ ULmJ')t:EJJ
L~1Jb1T;)nu. 5U1n,)1J~~n');) U!::::nsuc);)tJn')u 0)
(@!::::C1U1J~lc.1C11J, ~~UIJ1"l~cl"l1Jn"ltJ!::::1lJ,)u, ~2S~C11J) ifl! clJJ,)'OJ')nW"ln~t:V1J,)JJ [[@t: 'OJ')n5U 'U L[uu'OJ~@s~. n')umC1@s~ 'c Cl L~bltl'U 2 'OJbl • L~1JrJ"lC1 o iflDfl ;),).l..HLC1n~,)~Ul')~e),)1J 1lJJJ~U'tl,)'L1 LL@!:::: @t:[)UU~lUC11J. 'OJblUiiJ~ [[~UL2m5SJJWC1~ 1J1J~2~ifi~lKlS~2"l~ LU1J 1 151JW ~1J LL@a::; 'OJblmlKls~ LLD u L2 m5SJJ 1lJv Lmu iJ1~~S~2,)~LUU1lJ~U[[uumu1u1J. n,)1J~mJ"lLiJS~ 61J ! cl! ~ [[tJnfl;),)JJ~~iU1J@t:m;],)~~2S~C11J LL@!:::: ~,)~i::; L~tJ2S~:n,)UU;)JJ1J~l1JC1U dtSWbl Ul!::::U"l5U 'U LLUUm~"ltJ bl')tJ ~~@un,)Uf1~1J;)un')ugJJEi;)2S~C1U l1J L2bl1lJblStJ. 'U UJUIJ1JU£J.J1J
'l
III
11
IlIJ
Cl!
Cl!
Ow(
The Lao Journal a/Agriculture and Forestry, special issue No. 17
Semi-quantitative evaluation of waterlogging duration using two models based on soil colour in a representative upland catchment of northern Lao PDR Emmanue/ BOUROON1, Oidier BLAVEP, Oun/am LUANKONGKAM3, Soulileuth BOUNSAMA/l, A/ounsavath CHANHPHENGXAY4, A/ain P/ERRET', Olivier R/BOLZJ1.
Abstract Groundwater is a vital resource for rural populations in tropical areas who depend on seasonal rainfall. Groundwater is often the only source of water feeding streams, allowing them to flow throughout the dry season. The aim of this study was to evaluate two models for estimating the average duration that soil is waterlogged by shallow groundwater table. These models, based on variations in soil colour, were developed by Blavet et al (2000) from observations and measurements made in a semi-arid environment in West Africa. Therefore there is a need to evaluate whether these models are also pertinent in a mountainous context of northern Laos. Our approach consisted in comparing data obtained from field measurements (water table level, morpho-pedological features including soil colour) with predictions made by the models. This study was carried out along two transects with contrasting characteristics in terms of the landscape morphology as well as the soil hydrodynamic: the first was in an open swampy valley with convex hillslopes, the second was in a steep-banked and narrow valley with convexo-concave hillslopes. Preliminary results from our study identified relationships between soil colour and the mean rate of soil waterlogging and are a first step for developing an inexpensive and simple method to predict soil saturation in this environment. Key words:
Groundwater resources; Water/ogging; Soil cotour; Mountainous stream; Lao POR
'lnstitut de Recherche pour le Oeveloppement (IRD), International Water Management Institute (IWMI), NationalAgriculture and Forestry Research Institute (NAFRI) - c/oAmbassade de France - BP 06 Vientiane, Lao PDR 176 Solution IRD IWMI-NAFRI, Vientiane, Laos (
[email protected] ). 2UR 179 SeqBio IRD, Montpellier, France 3National University of Laos - Faculty of Agriculture, Vientiane, Lao PDR 4NAFRI, Agriculture Land Research Center (ALRC), Vientiane, Lao PDR. 150
nJJE)'1,
2008
The Loo Journal ofAgriculture and Forestry, special issue Vu. 17
Introduction
these indicators often requires expensive
Groundwater is a major resource for
electron
people living in tropical zones with
studies have shown that the colour of
seasonal rain. In arid and semi-arid
the soil is strongly controlled by the time
environments, it is the only permanent
course
water
humid
linked to the presence of the water table
it feeds streams
(Schwertmann, 1993; Franzmeier et al.,
resulting in stream flow during the dry
1983; Veneman et al., 1998; Faulkner
season.
and Patrick, 1992). These cycles have
reserve,
and complicated equipment (i.e. scanning
whereas
mountain climates,
in
microscope). A number of
of oxidation-reduction
cycles
an impact on the mobility of elements Our study is focused on the uplands of
such as iron, manganese (Fanning and
northern Laos where groundwater plays
Fanning, 1989) and humic compounds
a key role in the hydrology of the area.
(Duchaufour and Souchier, 1977) which
This water resource is used directly in
are one the principal factor affecting soil
shallow village wells for domestic use
colour in tropical environments (Segalen,
or indirectly from the permanent stream
1969). Several authors also described the
for small scale irrigation or to fill ponds
relationships between oxido-reduction
for fish farming. Within this context and
processes and soil colour using different
with the future perspective of predicted
models (Simonson and Boersma, 1972;
global changes in the Southeast Asian
Megonigal et ai, 1993; Genthener et al.,
region (i.e. global warming, conversion
1998; Blavet et al., 2000).
of annual crops to cash crops requiring more water from the underground water
The overall objective of this study was
table) political decision makers and land
to evaluate soil colour as a simple
managers need to make decisions based
pedological indicator, which is easy to
on sound scientific data.
determine in the field, as a rapid semiquantitative diagnostic of the water
A
number
of
morpho-pedological
table
resources
in the mountainous
characteristics have been linked with
agro-ecosytems in northern Laos. Two
the hydrological functioning of the soil
models developed by Blavet et al. (2000)
such as the redistribution of carbonate
predicting soil waterlogging based on
(Bouzigues et aI, 1997) or the presence
soil colour were applied in the Houay
of oxides (Vizier, 1974). The analysis of
Pano catchment.
September 2008
151
The Lao Journal ofAgriculture and Forestry special issue No. J7
was measured daily during the rainy season and weekly during the dry season. This study was carried out in the Houay
These measurements were made using
Pano
a probe with an acoustic alarm.
catchment,
the
biophysical
characteristics of which are described in detail by Valentin et al (this issue). In order totestthe models incontrasting situations, two
observation
and
measurement
transects were chosen based on of the known hydro-geochemical data of the stream (Ribolzi et al., 2005), morphopedological characteristics of the soil (NAFRI, 1999) and existing piezometers (Ribolzi et al., 2008 this issue). The first transect, RIB 48, is characterised by a convex open valley with a swampy area and a stream channel with gentle slope. The second, RIB 33, is in a deep and narrow convexo-concave valley. Monitoring water table levels The
catchment
was
progressively
equipped with a network of piezometers from 2002 to 2007. The piezometers are made of PVC with a 5.5 cm diameter and are perforated in their base for a distance of 50 cm. The depth at which these were installed depended on the depth of the water table which was determined by auqer surveys. These were installed along transects perpendicular to the axis of the river (including RIB33 and RIB48). The depth of water in the piezometers
Data collected from 2003 to 2006 were used for a descriptive statistics analysis. The minimums, first quartile maximums , medians, 3rd quartiles and averages were calculated to characterise the water levels measured in the piezometers throughout this
period.
Furthermore,
histograms and cumulated distributions were calculated with a 10cm increment. This analysis was used to determine the
time-course
of
soil
saturation.
Correlations were sought between these quantitative data and the observed soil colour. Soil sampling and description A soil survey with hand augers was carried out during the 2006 dry season. Soil samples were collected every 10 cm along vertical profiles in the vicinity
(-1 m) of each piezometer in the two selected transects. For each sample the morphology of the material was described in terms of conditions - wet colour, texture, porosity and stoniness. Morphological
features
associated
with hydromorphy were also described (Le. redoxic coloured spots and lines, pellicular coatings, concretions). These
152
n'1JEJ'l,
2008
The LaD Journal ofAgriculture
LlIIJ
Forestry. special Issue No, 17
pedological characteristics were used
with the hue (e.g. 10 for 10YR ; 5 for
to delimit the volumes of soil along
5YR etc). Table 1 shows the major hue
transects which had been affected by
and their corresponding angular notation
the waterlogging (Figure 1).
Ho measured in the field. HO of each hue, V and C recorded for the soil are then
Soil colour determinations
The soil colour was determined using Munsell colour chart (Munsell, 1976) including
Munsell
Hue
(i.e.
shade,
noted Hue), Munsell Value (Le. colour saturation, noted V) and the Munsell
combined to give the redness rating (RR) proposed by Torrent et al (1983). This ratio describes the forms of iron oxides in the soil and in particular the hematite content (Schwertmann, 1993). It is defined by the following formula:
chroma (Le. brightness level, noted C). The Hue is a mixture of one primary colour, white (saturation) and black (brightness). The letters referring to each Hue indicate the following colours: red (R), yellow (Y), green (G), blue (B), purple (P). The colours correspond to alphanumerical values (e.g. 10R 3/6, 10YR 3/4, 2.5Y 4/4) spread around the circle on the Munsell colour cylinder. The
HR =
(360 - 5 HO) * C \
(2)
18V Table 1 shows the correspondence between HO and RR. Correlations were then sought between these two variables and the cumulated distributions of the water levels measured in the piezometers to establish the prediction models.
shades are converted into an angular coordinate, Ho. By convention the colour
Relationship between soil colour and
Red Purple (1 ORP) is equal to 360 . The
water table levels
formula used for this conversion is the
In experiments carried out in the semi
following:
arid environment of West Africa on
0
HO
= 36 :+:
.' 12 \ \ I 11 + - - ) \. 10 J I
granito-gneissic bedrock, Blavet et al (1)
where /1 is the numerical coordinate of the hue (Le. R
=0; YR=1; Y = 2; GY =
3; G = 4; BG = 5; B = 6; PB = 7; P = 8; RP = 9) and 12 is the number associated
September ]008
(2000) showed that the cloud of points obtained by comparing the variables Ho and RR with the mean annual rate of soil water/ogging (in %) could be described by a sigmoid type curve using the following equation:
/53
The Lao Journal a/Agriculture and Forestry, special issue No. /7
interquartile deviation around the median
100
of 25 cm. The depth of the water table (3)
where x is the value of Ho or RR, a and b are the model constants obtained by adjusting the models using the least squares method (Table 2). Confidence intervals associated to the models follow the normal distribution were estimated as follows:
varied from 25.7 and 138.7 cm, thus with a maximum difference of 100 cm. Next to the stream (piezometer T3A2), the average depth of the water table was 12.7 cm (median = 12.4 cm), the interquartile deviation was only 10 cm and the water table level fluctuated between a depth of 24.4 cm and a height of 11 cm above the soil surface, thus a maximum difference between the extremes of 35 cm. On
J
EViTLG 0 9 x] = 1,96g(x)
(4)
withthe hypothesisofa normal distribution of ; g(x) being the function that allowed to calculate the 95% confidence error margin from all the x values observed for
Ho or RR. Correlation coefficients were obtained by comparing the
% values
from the models with the measured values.
the slope in RIB 33 (Fig.1b, piezometer T1A4), the water table is found at an average depth of 445.4 cm (median
=
436 cm), with an interquartile deviation of 20 cm and the difference between the extremes reached 200 cm. Next to the stream (piezometer T1A 1), the average depth was 41.8 cm (median = 42.4 cm), the interquartile difference was only 5 cm and the difference between the extremes was only 69 cm (the level of the water. table fluctuated between 15 and 54 cm).
Water table level behaviour
Characterisation of the soils near the
Figure 1 illustrates the characteristics
ptezometers
of the water table depending on the
On the slopes (Figure 2a), the soil close
topographic position of the two transects
to the piezometer T3A3 (transect RIB48)
included in the study. On the slope in RIB
was reddish grey (2.5 YR 3/2, 4/2) to
48 (Figure 1a, piezometer T3A3), the
reddish brown (2.5YR 4/4) down to a
water table is found at an average depth
depth of 70cm while, near piezometer
of 118 cm (median
/54
= 110 cm),
with an
T1A4 (transect RIB33), it was reddish
n1Jv'J, 2008
The Lao Journal a/Agriculture and Forestry. special issue No. 17
brown (5YR 3/4 to 5YR 4/4) down to a
the year). The particle size distribution
depth of 200 cm. The soil texture at these
of the material is from a silty-c1ay to a
two piezometers is silty clay. These soils
sandy-clay texture. Below the first 20 cm
are saturated for 0 to 36 days out of 365
the two profiles showed morphological
(0 to 10 % of the year).
characteristics
Deeper into the soil horizon, from 70 cm for piezometer T3A3 and 200 cm for
suggesting
reduction.
The shades are generally greenish grey (5GY 4/1 ) to bluish grey (5G 4/1 ).
the piezometer T1A4, the particle size
Correlations between soil colour and
distribution of the material was more
water table level variations
heterogeneous with a sandy-silt to sandy-
Correlations were calculated using the
clay texture. The processes of oxidation
values obtained from field observations
and reduction are more pronounced and
for Ho, RR and WLG and those predicted
are indicated by shades of dark brown
by the models. A first comparison (Table
(7.5YR 5/6) to greyish brown (10YR
2) using all the values obtained for HO and
5/2), yellowish brown (10YR 5/8) and/or
RR resulted in linear correlations with r2
a5
Y 5/1)
values of 0.82 and 0.87, respectively.
associated with reddish brown colours
This result was obtained using the
(2.5YR 4/6). These soils are saturated for
coefficients a and b from Blavet et al
169 to 202 days out of 365 (46 to 55% of
(2000) and those adjusted using the data
the year). The reduction processes are
from the transects. A second comparison
indicated by shades of greenish grey to
taking into account the topographic
bluish grey (5GY, 5B).
position of the piezometers calculated
yellowish olive brown (2.5Y
linear correlations near the stream, with In the proximity of the stream (Figure 2b) at piezometer T3A2 the oxidation and reduction processes appeared from 10 cm in depth and from 20 cm at piezometer T1A 1. These processes are highlighted by brown (7.5YR 4/2), reddish grey (2.5 YR 4/2) to dark grey shades (1OYR 3/1, 5/1). These soils are saturated for 40 to 193 day in the year (11 to 53 %
September 2008
of~
correlation coefficient r2 values of 0.93 for HO and 0.98 for RR. On the slopes these ccefficients were 0.57 and 0.54, respectively. By adjusting the a and b constants from the two models with the data from the piezometers located on the slopes the linear correlation coefficient, (2,
increased to 0.78 for HO and 0.72 for
RR (Table 2).
155
The Lao Journal ofAgriculture and Forestrv, special issue No. 17
In Figure 3 the HO and RR values which
Correlation between the measured
were in agreement with the prediction
RR and the model predictions
models
a
The piezometric measurements (Figure
function of the topographic location and
1) indicated that in the first case the water
the duration of soil saturation at the
table was at an average depth of 445±20
piezometers.
cm with a RR= 0 between 420 and 440
are
indicated.
This
was
Correlation between the measured Ho and the model predictions Near the
stream
(Figure
cm for a saturation period of 28 to 46 %. In the second case for RR=O at 50 cm the water table was at an average depth
3a),
the
following Ho values were the same as
of 41.8 cm ± 10 cm and a saturation period of 36%.
those proposed by the model: 63° for a saturation period of 11 %; 72° for a
Nearthe stream (Figure 3b), the following
saturation period between 6% and 53%;
values agreed with those proposed by
126°, 162° and 234° for a saturation
the model: RR=O for a saturation period
period between 92 and 100%.
of 36%; RR=1.3 for a saturation period of 11%; RR=-3.75 and -6.3 for a saturation
On the slopes (Figure 3a), the following Ho values were the same as those proposed
by the
model: 90° for a
measured saturation period of 50 to
period from 96 to 100 %. However, the values did not fit closely with those of the model for RR = 0 and a saturation period of less than 6% or equal to 53 %.
60% is close to the model; 72° for a saturation period of 1 to 46%. However,
On the slopes (Figure 3b), the following
an angular value of 45° does not agree
values
with that proposed by the model for a
by the model: RR=O for a saturation
soil saturated period greater than 20 %.
period between 21 and 46 %; RR=5.0
This analysis carried out on the dominant
for a saturation period of 3 to 7%: RR=-
shades of the soil indicated wide ranges
1.0 which corresponded to a saturation
of WLG between the angular value 72°
period of 55%. However, the following
and 45°.
values did not agree with those proposed
agreed
with those
proposed
.
by the model: RR=11.3 for a saturation period of 16 to 39%; RR=Ofor a saturation period of less than 21 %.
156
n'1JEJGJ,
2008
The Lao Journal ofAgriculture and Forestry, special issue No. 17
saturation period ranging from 21 to 46%, the angular value and RR are about 72 On the slopes, our observations of
and 0 respectively. This indicated that for
the dominant colours and piezometric
these two situations, the soil colour had
measurements agree in part with Ho
the same dominant shade (i.e. 10 YR)
predicted by the model adjusted with
but a different hue and chroma: on the
observations from transects. But we
slope the soils were a light yellow brown
could improve the model by taking
colour (i.e. 10 YR 5/8) whereas near the
into account other parameters such as
stream they were greyish brown (i.e. 10
the proportion of the dominant shade
YR 5/1).
compared with and the presence of spots and/or gleyic volumes (Figure 2a). This would complete the correlations between measured values and those predicted by the model.
In this study, we only took into account the annual average duration of soil saturation. In a study carried out on a toposequence in North Carolina, He et al (2003) showed that soils must be
When we recorded the colour in the field
saturated for 21 consecutive days in the
an inaccuracy may have been introduced
year for iron reduction to be seen as
by the fact that all the soil profiles were
lasting morpho-pedological feature.
not examined
in exactly the same
sunlight conditions. This bias could be avoided, at least partly, by using a field spectrophotometer.
Our study showed that the prediction models developed on data obtained in
The model based on the redness rating
a semi-arid African environment can be
can be used to specify the degree of
applied to a mountainous tropical region
colour saturation (Value) and brightness
of northern Laos. Significant correlations
(Chroma) for each soil colour shade, which can have a strict correlation with the period of waterlogging experienced
between soil colour and piezometric measurements have been identified, which allowed us to use existing models
by the soils due to fluctuations in water
of waterlogging prediction based on
table level. At RIB 48 (Figures 2b and 3),
colour indicators. This work represents
on the hillslope and near the stream, for a
September 2008
a first step towards
developing an
157
The Lao Journal ojAgriculfure and Forestry, special issue No. 17
inexpensive, simple and indirect method
Bouzigues, R, Ribolzi, 0., Favrot, J.C.,
for predicting soil waterlogging. Future
Valles, V., 1997. Carbonate redistribution
research should focus on the cycles
and hydrogeochimical processes in two
of reduction or oxidation in the soils
calcareous soils with groundwater in a
throughout the year on the morpho-
Mediterranean environment. European
pedological characteristics in the dry
Journal of Soil Science, 48:201-211.
season.
Duchaufour,
P.,
Souchier
B.,
1977. Pedologie 1 - Pedoqenese et classification. Masson, Paris. This study was carried out within the
Fanning,
framework of the MSEC (Management
1989. Gleization. P. 110-125. In Soil
of Soil Erosion Consortium) project. The authors thank NAFRI (Soil Survey
D.S.,
Fanning
M.C.B.,
morphology, genesis and classification. John Wiley & Sons, New York.
and Land Classification Institute), IWMI Management
Faulkner, S.P., Patrick W.H., 1992.
Institute) and IRD (Institut de Recherche
Redox processes and diaqnostic wetland
pour le Developpernent). We would also
indicators in bottomland harwood forest.
like to thank Keo OUDONE and Seng
Soil Sci. Soc. Am. J., 47, 1196-1202.
(International
Water
KEO for setting up and monitoring the piezometer network.
Franzmeir,
D.P.,
Yahner
J.E.,
Steinhardt G.C., Sinclair H.R, 1983. Colour patterns and water table levels in some Indiana soils. Soil Sci. Soc. Am. J.,
Blavet, 0, Mathe, E., Leprun, J.C.
47, 1196-1202.
2000. Relation between soil colour and
Genthener,
wateriogging duration in a representative
Hodges RL., and Thomas P.J., 1998.
hillside of the WestAfrica granito-gneissic
Redoximorphic features and seasonal
bedrock. Catena, 39, 187-210.
water table relations, Plain,
M.H.,
Virginia.
p.
Daniels
W.L.,
Upper Coastal
43-60.
In
M.C.
Rabenhorst et al. (ed.) Quantifying soil hydromorphology. SSSA Spec. Publ. 54. SSSA, Madison, WI. 158
n1JtJ.."
2008
The Lao Journal ofAgriculture UI/J Forestrs, special issue No. 17
He. X, Veraskas M.J., Lindho D.L., and
Megonigal,J.P, Patrick, W.H., Faulkner,
5kaggs R.W.. 2003. A method to predict
5.P., 1993. Wetland identification in
soil saturation frequency and duration
seasonally flooded forest soils: Soil
from soil color. Soil Sci. Soc. Am. J., 67,
morphology and redox dynamics. Soil
961-969.
Sci. Soc. Am. J., 57, 140-149.
NAFRI, 1999. Biophysical and socio-
Munsell Color, 1976. Munsell Book of
economic
Color. Matte Finish Collection, Baltimore
inventories.
catchment,
KM10
Huay
Village,
Pano Luang
Prabang Province. Soil survey and Land Classification
Center,
Management
of Soil Erosion Consortium (MSEC). Vientiane, Lao PDR: 17p. multi.
K., Latchachak, K., Tasaketh, 5., Vanethongkham, K. 2005. The use of pH to spot groundwater inflows along the stream of a cultivated catchment in the northern Lao PDR. The Lao Journal of Agriculture and Forestry. 10, 72-84. 0.,
Thlebaux,
Torrent,
J.,
5chwertmann,
U.
H., Fechter and
F. Alferez. 1983.
Quantitative relationships between soil color and hematite content. Soil Science,
Ribolzi,0.,5ilvera,N.,Xayyakummanh,
Ribolzi,
MD.
J.P.,
5engtaheuanghoung, 0., Bourdon, BE., Chaplot, V.,de Rouw, A., Huon, 5., Mouche, E., Pierret, A., Briquet, ..I.P.,
136 (6), 354-358. 5egalen, P., 1969. Contribution
a
la
connaissance de la couleur des sols asesquioxyde de la zone intertropicale: les
sols
jaunes
et
rouges.
Cah.
O.R.S.T.O.M. ser, Pedol., vol. VII, no 2. 5chwertmann,
U.,
1993.
Relation
between iron oxides, soil calor, and soil formation. In Soil Color. Soil Sci. Soc. Am. Spec. Pub., 31. Madison, WI.
Marchant, P., Robain, H., Soulileuth,
5imonson, G.H., Boersma L., 1972.
B., Valentin, C., 2008. Effect of fallow
Soil morphology and water table relation:
regrowth on stream water yield in a
11. Correlation between annual water
headwater
table fluctuations and profile features.
catchment
under shifting
cultivation in northern t.ac PDR. The
Soil. ScL Am. J., 36, 649-653.
Lao Journal of Agriculture and Forestry.
Valentin et. all this issue
This issue.
September 2008
159
The Lao Journal ofAgriculture and Forestry, special issue No. J 7
Veneman, Lindbo
and
P.L.M.,
D.L.,
Spokas
1998.
redoximorphic
Soil
L.A.,
moisture
features:
A
historical perspective. p. 1-42. In M.C. Rabenhorst et al.(ed.) Quantifying soil hydromorphology. SSSA Spec. Publ., 54. SSSA, Madison, WI. Vizier, J.F., 1974. Recherche de relations
rnorphoqenetiques
existant
dans
un
type de sequence de sols hydromorphes peu hurniferes du Tchad. Zerne partie: Dynamique de l'eau et du fer dans les sols de la sequence. 3eme partie: Conclusions sur l'evolution actuelle des sols et hypothese sur la formation de la sequence. Cahiers ORSTOM Serie Pedoloqique, 12(3), 211-266.
160
n1JV'l,
2008
The Lao Journal ofAgriculture and Forestry, special issue No. 17
Table 1 -
Correspondences between the alphanumerical values for the soil colour observed in the field (i.e. Hue, value and chroma), with their angular notation (HO) and red ratio (RR).
Hue
Value
Chroma
Ho (degrees)
RR
10 R
3
6
36
20
3
2
45
5
4
2
45
3.75
4
6
45
11.3
4
4
54
5
4
2
63
1.3
5
6
63
3
4
1
72
0
3
1
72
0
5
8
72
0
2.5Y
5
1
81
-3
5Y
5
1
90
-1
5GY
4
1
126
-3.75
58
4
1
234
-11.25
2.5YR
5YR 7.5YR
10YR
September 2008
161
The Lao Journal ofAgriculture and Forestry, special issue No. 17
Table 2 -
Coefficients a and b from the model by Blavet et al (2000) and adjusted
values using data from RIB48 and RIB33. Comparison of the correlation coefficients (r2) for the two transects (s & rb), on the hillslope (s) and near the stream (rb). The input value used in the models was either the angular coordinate Ho or the red ratio (RR). Variables
a
b
r
n
Position
Coefficients a and b from Blavet et al (2000)
Ho
RR
382353
2,07
- 0.16
1
40
0.82
s & rb
29
0.57
s
11
0.93
rb
40
0.87
s & rb
29
0.54
s
11
0.98
rb
Coefficients a and b adjusted
Ho
RR
162
- 0,16
40
0.82
s & rb
- 0,14
29
0.78
s
- 0,16
11
0.93
rb
2.07
1.2
40
0.87
s & rb
3.9
1.4
29
0.72
s
2.07
1
11
0.98
rb
382353
i71J(J~, 2008
The Lao Journal ofAgriculture and Forestry, spec ial issue No. 17
a)
o
-11,6 "10
o 12,7
r===1
-10
•oi 118••
-20
24,4
J;: -120
138,7
-25 cm
- --
o
-160 on
,,
, T3A3
-------_
r3A2
Stream
I
..
i
left bank
Right bank 20m
o
10
20m
10
b)
o
200
'6
277,0
. ·4 ) ~.,! 54
,, ,
-500 on
----
,, ,,
o
10m
0
Zone of redox features
Figure 1 -
473.5
-SOon
-300
10
20 m
I Piezometef
Schematic showing a transversal section of the interaction zone between
the stream and the water table for the traverses RIB48 a) and RIB33 b); Location of the piezometers (T3A2, T3 A4, T1A 1, T1A4) along the transects; Map of the subsurface zones showing redoximorphic features. Whiskers plot (1st quartile, mediane, 3rd quartile, minimum and maximum) of water table levels measured in the two piezometers at each site.
September 2008
163
The Lao Journ al of Agr iculture and Forestry, sp ecia l issu e No. 17
a) T3A3 FrequencyC1l 0
20
0,4
10
0,2
a. (J)
0 (J)
.c Qj
50
f-
0,0
0 1990 Time of work -
Figure 1-
1995
2003
Number of workers - •• Average rice yield
Annual cultivation average work time, number of workers and upland rice yield, per hectare and per year, 1990-2003.
September 2008
201
The Lao Journal ofAgriculture and Forestry, special issue No. 17
10
. _--
9
o Cropping period o Fallow period
-
8,6
8 ,-...
en ....
co
7
C
6
Q)
c 0
~
co ....
5
r--
~
0 ""0 0
4,8
4
co---
3,8
·C
Q)
0...
3
3,2
2 1
0
-
1,8
F 1970
Figure 2-
-
F 1990
F 1995
2003
Average fallow and cropping periods for the fields under shiftin g cultivation,
1970- 2003.
202
nvv'l,2008
The Lao Journal ofAgriculture and Forestry, special issue No. 17
Stage 1
Stage 2
Stage 3
Stage 4
Stage 5
Soil surface (colour) Vegetation cover (density and colour) Sto nes (size) Weeds (species) Weeds (density) Rills (density) Gullies (density) Landslides (risk)
40 years
D Figure 3 -
i
10 to 40 yea rs
• t
Houay Pano
in 2004 Main indicator of the stage
Perceived stages of land degradation in the Houay Pano catchment
(adapted from Pelletreau 2004) . Note: Mi, Mimosa invisa; Co, Chromolaena odorata; le, Imperata cytiruirice; Mc , Microstegium ci/iatum ; Es, Erigeron sumatrensis; Tm, Thysanoleana maxima.
September 2008
203
The Lao Journal ofAgriculture and Forestry, special issue No. /7
Short fallow periods
Long agricultural history
~ Cultivation on steep slopes
and high elevation areas Intense rainfalls
Figure 4-
Land degradation-related causality linkages perceived by the farmers in Ban Lak Sip.
204
n1JE)6),
2008
The Lao Journal of Agriculture and Forestry is a journal of Ministry of Agriculture and Forestry. It aims to disseminate technical information relating to agriculture and forestry within the Lao PDR. The journal is published twice a year. The publication of the journal is being supported by the Upland Research and Capacity Development Programme (URCDP). The journal is distributed among all agencies of the Ministry of Agriculture and Forestry. Subscriptions are available to all other agencies wishing to receive the journal on a regular basis. Abstracts of papers are expected reproduced in various international abstracts. Papers in both Lao and English will be accepted for publication; an abstract in the second language (Lao or English) is also to be published. Only papers with specific relevance to the Lao PDR will be considered for publication.
The following types of papers will be accepted: 1.
Technical papers reporting the results of research in agriculture, forestry, fisheries, environmental sciences, etc;
2.
Reviews of specific subject areas in the above general fields;
3.
Articles of between I to 4 pages on specific technical topics but which may have not been the focus of a research program.
Guidelines for Authors: Technical papers and reviews should generally not exceed 2,500 words (up to 10 double spaced A4 size pages), including tables, figures and references; articles will usually be shorter (1-4 pages). One hard copy, together with a disk copy or CD, should be submitted. Technical papers should follow the usual break up into abstract, introduction, materials and methods, results, discussion, references and acknowledgement.
Title: This should be short and specific.
Author (s) Name: Initials should be used for given names, which precede family name (s). The organization to which the author (s) is attached should be given in a footnote.
Headings in Text: These should be short and accurately reflect the text to follow.
Acronyms: Acronyms should be spell out the first time they appear in the text, followed by the abbreviated form in parentheses, e.g. National Agriculture and Forestry Research Institute (NAFRl). Thereafter, only the acronym should be used in the text.
September 2008
205
Units: The metric system for units of measure should be used. Abstract: Abstractof no more than 200 words in both the English and Lao languages and key words should be provided.
References: The author-date system (Harvard system) should be used. References should be listed alphabetically according to the author's surname. Where an article has more than one author, each name should be listed with the surname first, initials second. References with the same author should be listed. Each reference should be cited in the text by the author's surname and year of publication. Examples of listing of references are as follows:
Anandhana, B. 1989. The amount and distribution of soil organic matter in Thailand. Technical Bulletin No. 168. Soil Survey and Land Classification Division. Land Development Department. Ministry ofAgriculture and Cooperative. Thailand. Anon. 1993. Basic statistics about the socio-economic development in the Lao PDR 1992. State Statistical Center. Brown, J 1997. 'Native plant cultivation', [on-line] Plantnews,
[email protected]. au, 28 November. Brook, R.M. 1992. Early resultsfrom an walley cropping experiment in the humid lowlands ofPapua New Guinea. Nitrogen Fixing Tree Research Reports, 1 0,73-77p. Lao-IRRI Project 1992. National Rice Research Program, 1991 Annual Technical Report, 156p. Schiller, J.M., Phoudalay Lathvilayvong and Dr. Ty Phommasack 1993. Green Manure (INSURF) study in the Lao PDR. Proceeding ofINSURF Planning Meeting, Fuzhou/ Guangzhou, Peoples Republic of China, 14-21 June 1993. Tables: Tables should be numbered consecutively and referred to the text. A brief title for each table should be provided. Each table should be typed on a separate page. Column headings and descriptive matter in tables should be kept brief. Only tables that complement rather than duplicate information in the text will be accepted for publication. Combining tables that present different types of information should be avoided.
Figures: Photographs can be submitted for publication in black and white or color; they should be referred to as 'Figures' in the text. Only well focused photographs with strong contrasts will be published.
206
17m)'J, 2008
Use and designation ofNAFRI Materials
The ideas and opinions expressed here are those of the individual authors and not those necessarily ofNAFRI or the sponsoring agencies. NAFRI encourages use and adaptation of its materials for public use. If materials are reproduced or used, NAFRI requests that the authors and NAFRI are properly cited and informed. Reproduction for commercial use or sale of these materials is strictly prohibited.