Reprinted from Agricult:ure, Ecosystems &...
Reprinted from
Agricult:ure, Ecosystems &
Environment
Rehabilitation of a ‘semiarid ecosystem in Senegai.
2. Earm-plot experiments
. .\\
~ --_.
1_1--.-.-”
-
-
-.
-1-m-1.1
-I---H
A@cu.Rure
:
Ecospkms&
Enironrnent
An International Journal for Scientific: Research on the Relationship
of Agricuiture
and Food ProdcC :XI P f.7~ fiiosphere
-
Aims and scope. This journil is concerned with the interaction of methods of agricultural produ
tion. agroecosystems and the
environment. Topics covered
include the comparison of different methods of production (intens ~3. extensive, linear, cyclic) in
terms of their ecology; how,agricultural production methods affect pollution of soil, water alid ; il; the quality of food. and the
use of energy and non-renqwable resources; the effect of industrial pollutants on agriculture:
ind the policy rssues involved
in the change and developmient
of agriculture. The journal aims to serve as a focus for scientls!s
n açrlc::lltrre,
food productior
forestry and the environme& as well as for administrators and policy-makers concerned with t E:S~ IF sues in research estab-
lishments, government, indtstry and international organisations. It publishes original scientlfi
: papers. review ariicles and
occasional comment papers. From 1994 onwards a section of this journal, Appbed
Soi/ fcology i
publ~~+ed se;!.qrately.
Applietl
Soi/ Ecology
addresses the kole of soit organisms and their interactions in relation to: agriclflt:Ji
3 ;:rc-, ,‘i!/,.r
::llrierit cyc11-11.
and other soi1 processes, thie maintenance of soi1 structure and fertility, the impact of hun;:,:i _’
1.
: .’ ‘C, 82 p $, ;,
ecosystems and bio(techno)(logical
contIroI of soil-inhabiting
pests, diseases and weeds
EDtTORS-IN-CHIEF
J.W. Sturrck, 500 Johns Road, Fernside. RDl, Rangiora, North Canterbury, New Zealand
M.R. Carter, Research Branch. Agriculture and Agri-Food Canada, PO. Box 1210. Charlottetown. f 3
C a n a d a
Book Review Editor
B.A. Auld, Agricultural Resebrch Gen!re. Forest Road, Orange, N.S.W. 2800. Australia
EDtTORtAL ADVISORY BOARD
J.i? Aeschlimann, Montpelliei,
France
J. Huber, Darmstadt. Gernlan!
B . A . Auld, Orange, N.S.W., Australia
M . J . J o n e s , A l e p p o , S y r i a
J . N . B . B e l l , Ascot, U K
J.V. Lovett. Queen V i c t o r i a Teri ice. ACT, Austraiia
H . B o u w e r , P h o e n i x , A Z , US A .
G.A. Norton, St. Lucia, Qld , A l:;tralia
K . J . B r e n t . B r i s t o l , U K
D.M. Olszyk, Corvallis. OR, US
G.T. Brooks. Reading, UK
C.K. Ong, Nairobi, Kenya
A . E l T i t i , S t u t t g a r t , G e r m a n y
D . P i m e n t e l , I t h a c a , N Y . U . S . A .
L.O. Fresco. Wageningen, Netherlands
G . H . L . R o t h s c h i l d , M a n i l a . IF%11
M. Giampietro, Rome, Italy
M . S h i y o m i . Mito. Japan
S . R . G l i e s s m a n , S a n t a C r u z , C A . U . S . A .
R.E.H. Sims. Palmerston Nortl
PJ. Gregory, Reading, UK
L.M. Skarby, Goteborg.
Swedf
S . K . H a h n , B r o a d v i e w H t s . , Ci)H, U . S . A .
E . M . A . S m a l i n g . W a g e n i n g e n . detherlands
1. Haque, Lahore. Pakistan
P I . Sunarjo. U j u n g P a n d a n g . I I donesia
J . H e l e n i u s , H e l s i n k i , Finland,
M. Tbth, Budapest, Hungary
M . H o r n u n g , Grange-over-Sends, U K
J . M . W e b s t e r . B u r n a b y , B . C . . C inada
Publication information: Ag[icu/ture,
Ecosystems & Environment (ISSN 0167-8809). For 1998 ~UITWS 67-71 are scheduletl
for publication. Subscribers’to Agriculture. fcosystems & Environment Will have the opportuni I to subscribe to Applied SO.~
Ecology at a speciat reducel subscripticrn rate. Subscription
prices are available upon request In m the PubIIsh~?r. Subscrtption<;
are accepted
on a prepaid ibasis only, and are entered on a calendar year basls. Issues are
ent b!/ surfacr rnarl excepi t:
the following
countries where air delivery by SAL mail is ensured: Argentina. Australia.
Brazil, C:
.lada, Hong Kong, india, lsrae
Japan. Malaysia, Mexico, d e w Zealand.
Pakistan, PR China, Singapore, South Korea, Taiwar
Thailanc. USA For all other
countries airmail rates are: avaiiable upon request Claims for missing issues should be rr lde \\niithin six months ot 1311’
publication (mailingf date.
Orders, ciaims, and produd enquiries: jplease contact the Customer Support Department at thc ?! Regional
Sales Office nearest
you:
N e w Y o r k : Elsevier Science,; f?O. B o x 9 4 5 . N e w Y o r k , N Y 1 0 1 5 9 - 0 9 4 5 , U S A ; T e l . (+l) 2 1 2 - 6 3 3 - 3 7 : C. [Tell free nal‘lber for North
American Customers: l-888/-4ES-INFO
<437-4636)],
Fax. (+ 1) 212-633-3680,
E-mail: usinfo-fui’el
3E:vier icom
Amsterdam: Elsevier Scien
e . P O . B o x 2 1 1 , 1 0 0 0 A E A m s t e r d a m , T h e N e t h e r l a n d s : T e l : ( i - 3 1 ) 2 0 - 4 G375ï. Fax (t-31) 2C -485343;
4
E-mail: nlinfo-f@elsevier.nl
/
Tokyo: Elsevier Science, 9- 5, Higashi-Azab:)
1-chnnie, Minato-Ku.
Tokyo 106-0044.
-lapan Tell,
(+8!: 3-5561-5033.
Fax. i rC:
3-5561-5047, E-mail: info@ Isevier.co.jp
Singapore: E&vier Science,. No. 1 Temasuk
~\\ver;~~r:.
:! !i-01 Millenia ;Ower. Singapore
@3911??.
i(?,y ( .L 55) 433. &;‘y, i:ax. ; ,-(:
337-2230. E-ma!l. arialnfc@’
1 ..lsevier.com.s!:
Agriculture
Ecosystems 4%
Enwonment
Agriculture, Ecosystems and Envimnment 70 (1998) 19-29
Rehabilitation of a semiarid ecosystem in Senegal.
?. Farm-plot experiments
P. Pereza*“‘, J. Albergelc, M. Diattab, M. Grouzkf, M. Seneh
a CIRAD, BP 503.5 34~932,‘Montpellier
Cedex 1. France
’ ISRA, BP 199. Kaolack, Senegal
’ ORSTOIM,
B P 504.5 3 4 0 3 2 , Montpellier Cedex 1, F r a n c e
Accepted 23 October 1997
_.-~-___-----_--_-. -..-..- - _---.-...--. ---..-~-.-
Abstraet
In the Sudano-Sahelian zone of western Africa. characterized by semi-extensive agriculture and subhurnid clim,lte, soi1
degradation and water shortages are widespread features. Soi1 and water conservation practiccs are introduced to the farmers.
but often abandoned thereafter. Some interesting results have been obtained, on a hillside scale. in the southern part of the
cropping basin of Senegal. As a supplement to the experimental design, two small watersheds (2.5 ha) were deliner.ted and
equipped in representative hillside locations. In 1988, bath the watersheds were planned and submitted to an hydnalogical
survey. One of them, located on a colluvial/alluvial terrace. was also submitted to soi1 water storage and grain yield
momtoring. Results highlight a decisive effect of soi1 and relief features on the efficiency of conservation mrasures. F!cle\\ant
results were obtained on the terrace, but upstream areas still generated marked soi1 and warer losses. These phenomena. in
addition to socioeconomic constraints. partly explain farmers’ behaviour noticed on the hillside bcale. i‘ 1998 Published by
Elsevier Science B.V. Al1 rights reserved.
Kqr’rtinrnlr: Soi1 conservation: Runoff; Erosion; Water balance: Watershed management: Millet; Senegal
1, Introduction
be stabilized and runoff reduced on a11 slopes (Perez
and Sene, 1995)
In the cropping basin of Senegal, environmental
In fact, soi1 infiltrabiliity increases from the Upper
degradation is found in a11 landscape units, as shown
part of the hillside to the lowland area (Perez, 1994).
by hiph water erosion, involving sheet erosion in the
Thanks to this natural tmnd, it is possible to control
upper parts of the toposequence.
gully erosion at
overland flow by reducing its velocity and avoiding its
nickpoints and sand deposits in the lowland areas.
concentration. Depending on the local conditions, this
For landscape rehabilitation and as a prelude to any
cari be achieved with a network of filtering ob:rtacles
agricultural intensification, erosion phenomena must
such as stone bunds or live-hedges (La1 and Siewart.
1990; Roose, 1994).
~---
The marked tendency of the local soils to form
*Correaponding author.
surface seals is the result of their weak stability. These
~1167-8X0’~19Xi0;19.0(~ J
1998 Pubiished hy Elsevier Science R.V. All rights reserved.
/‘II SOl(>7-8809(97)OOi57-6
--
-
~_--- ---
ferruginous loamy sands are characterized by yery low
infiltrability shows high spatial variabi1it.y (Perez,
clay and organic matter contents (Charreau aid Nicou.
1094).
1971). Hence, there are very few solutions to increase
In 1988, soi1 and water conscrvatio.n
practiccs wcrc
long-term infiltrability in the soi1 profile. Cropping
implemented in the NI. They included one live-hcdge.
practices creating and temporaril:y maintaining sur-
established in the middle bfthe hasin and douhled with
face roughness permit an increase surface storage and
an upstream graminae lin<: (Ptrrri(~ri/l rrlrrsir!rlr!ll). antl
ponding time (Morin ef al., 1984; Lamachère, 1991).
cight filtering barriers (r/tc>nc pavements and brusb-
Soi1 surface covers of trop canopies or residuc
wood dams) across the \\hiaterM,ay:;. Ihrthcrrnore. XV
mulches cari reduce surface sea1in.g and flow velocit)
eral improved cropping p@tices were introtlucecl into
(Box and Bruce, 1996). Under local conditions. it i\\
the four farm-plots: contour cultivation. dry sca~)n
necessary to promote rapid trop establishment
antl
dccompacting, shallow r d.gin;r and localizcd mantlrc’
adequate canopy growth. This cm bc donc by cou-
application. These
i
techni LWS ;Ire tlc>~crib~tl ‘11 Perc/ ~1
pling water and organic matter management practices
al. (1997).
(Roose et al., 1’992; Sessay and Stocking, 1995).
The 2.5 ha Yaranc Katkrshcd (YA) i4 lo~:~tcti in 111~’
In the light of these observations, rehabilitation
Upper part of the hillslopj on thc cdge of thc: cropping
operations were conducted on the basis of local eco-
area. Because of the ge/n.tle relief, ~:he limits werc
logical features and human uses. The local effect of
dclineated with an cartljen ri(Jgc. Thc: \\aluc of thc
conservation measures and the mechanisms
involved
regular slope is nearly iI<6, xith n o evidcncc ot’ ;~n
were studied in two small watersheds (2.5 ha). corrc-
hydrological network, except i.or a downstream
shal-
sponding to an intermediate scale Inetween thc hillsidc
low wide waterway. Ovekland tlou anci shect crosion
area (1 km’) and the experimental plot ( 100 m’) nncl
characterize t h i s arca. T h c \\vatershc‘d i,, cnrircl)
constituting a relevant sized soi1 I-mit.
cropped and di\\ided intc lier 11irm-plots ~Fi~. 1 th 1;.
The characteristics of both watersheds
ancl thc
The soi1 comprises col uvial dçposits and fine grav-
survey methods used are describcd, followed by thc
cls eroded from the upp :r plateau. Thc tir:<t horizon
results from the hydrological
survey, the water baluncc
(O-20 cm dcpth) is sandy (with 10% clay content) ancl
monitoting and the trop yield study. The discussion
presents 5--30% femc ;cs ravels and the structure is
highlights the consequences
of thc rcsult\\ t’o~
continuous and fragile. Deeprr. the texture rapidly
watershed management on a hillsidc scalc.
bccomes loamy with 50- 10% tcrric gravcls. Belou ;II~
a\\‘erage depth of 50 cm ferric nodules and gra’~.cIs
account for 80% of tlhe oil volume. Under dry con-
2. Material and methods
ditions. this horizon is like a hardpan. b u t wcttcd
material turns crumbly. Gurface sealing is a general
2.1. Watershrd descriptim
feature of the watershed but the strength of the seal
depends on the depth of the hardpan.
The 2.5 ha Ndiba watershed (hl) is located down-
In 1988, soi1 and wattri conservation practiccs wcre
slope on the colluvial/alluvial terrace. A contouring
implcmented in the YA. They are the samc as those
track and natural relief mark the limits of the basin.
established in the NI Ir lsin:
)Y
one live-hedge, three
The upstream slope is nearly 0.5%. wliereas it reaches
filtering obstacles and irn provcd cropping practices.
2.0% in the downstream area where rills are concen-
trated into a widening gully. The watershed is entirely
cropped and divided+o ‘fow-@ym plots (Fig. l(a)).
It has a leached and dis&bed ‘feimginous soil.
In 1985, the NI outle wus equipped with a rain
The first horizon (O-60 c& depth) is sandy and friable
gauge and a water stage :Corder that were set up in a
with a continuous structure. Clay (5-10%) and
concrete-lined ditch. Sec iment loads wcre manually
organic matter (0.5%) contents are very small. Deeper.
collected from 1985 to 1 192. Since 1988, 4 m’- plots
the texture gradually b ‘cornes loamy and ferric spots
were established in diffi rent parts OF the ,watershr:d
:
or gravels appear from .5 m depth. As a consequence
(Fig. la) and runoff vc urnes measured after each
of sheet erosion and dolluvial deposjfs, the topsoil
rainfall event. One yea after installation, filtering
/
-
l? Ptw: et ai. /Agr-iculture,
Ecoqstems
and Envirvnrnent
70 (1998) 19-29
'tj - - field limit
I-1 - - Ii-e hedge
obstacles were eqaipped with 30 marked stakes for
initial stake height (above the soi1 surface) and tht.
measuring upslope sedimentation. During the rainy
actual value.
season, each IO-day period, the cumulative sediment
In 1986, the YA outlet was equipped with a rain
depth was estimated from the difference between the
gauge and a water stage recorder that werl; set up in a
. , .1qm1--..1-7-----<--
-
.-,p-
-
concrete-lined ditch. It iq only since 1988 that the
during 1988 - thc rainiest 1 ar of the decade - and 154
sediment loads were maneally collected. Since 1988,
aftcl planning ( I 0X9- 19’ II. Rain intensities w e r e
3 m”-plots were establishbd in different parts of the
computed t’or a11 t‘\\ ent!, 1 zater then L,=8 mm (15~:
watershed (Fig. l(b)) and runoff volumes were mea-
min depth); in this case.
7, 32 and 93 rainstorms,
sured after each rainfall. A fourth plot was installed in
rzspcctivcly. werc: anal y: d for each period. Fre-
an adjacent brushwood z$ne. As in the Ni basin, 20
quency distribution\\ of w depth (Fig. 2), ma ximuni
marked stakes wcre use8 ‘for measuring sediment
10 and 30 min intcnsitics I 10 ancl 1&, erosivity index
dcpth upstream of the filtkring obstacles.
tKJ. wcrc thc zanic bcforc
md after planning. It was
thus possible to compare tl
hydrological
results from
hoth tinie \\eries.
.~lrll~lti!Jll \\(?Il!:' 111,1X! !‘L ords were lest because 01‘
Seven neutron probe acbess tubes were installed in
Llic Ic~hnica! !)i.~)l-‘!Pn15. ‘1
clevant liydrological data
the NI watershed. includitg the 4 m’-plots, for mon-
4c’t was huilt. Il contain l 40 flood eVents bcforc
itoring the soi1 moisturei The neutron gauge was
planning ( IX-V- 1087). 1 I durin’> 1988 and 36 after
a:
calibrated fc.lr each access’tube
and each specific soi1
planning ( 19X9- I(W). 0 rail, the finit period had a
layer (Perez. 1994). Meayurements
‘were done every
total rain depth of 2057
m and a nmoff depth of
IO-day period during the kainy season.
X7 nim. Thc iinal pcriod
lad a total rain dcpth of
TO strengthen the stud of the soi1 water storage
2310 111111 ancl 2 lull~~ff d
th of 60 mm. The mean
spatial variability, 53 sam4ling spots were located on a
runoff coefficicnr thus SI Eted liom ,4.2% to 2.6%
20x20 m’ grid within the. NI basin (Fig. l(a)). From
(TabIc 1,. In l’ac’i. inter-a nual ,variability was high
1988 IO 1993. samples tiere obtained with a shell
during bath thc ~L’I ic&+. I’U 11 the occurrence of violent
augcr (O- 150 cm tlcpth) atjthe end of the rainy scason.
r;tinhtornis. that \\omctinic
represented upto 60% of
Geostatistic;il concepts webe used for the data analysis
;lllnual IC\\'Cl\\.
(Burgcss and Wchster, 1TXO; Chopart. and Vauclin.
Conccrning rairlstorms. lordy 31% of the events
1990) and the kriging phcedure was used for soi1
initiated runoff during t!xz fi st period and 26% during
water storage.
the final OI~L’. Threshold
célues for rainfall depth
The same studies were dlanned in the YI. Hokvevcr.
( L,I,,,,), maximum 10 min jntensity (ll(,li~~>) and ero-
aerious problems were edcountered
during the data
sivity index (li,,,,,). helow &Ch there [vas no runoff.
analysis, because of the, presence of gravels and
\\\\‘ere computcd. When the ,I+,,,, value remained the
nodules in the soi1 profile. phe results were considered
sxiic during thc tv.0 pci-icxlsi (LPllr,=71 mm). the f ,Olim
unreliable and therefore ace not incl;uded here.
vuluc inzrenscd t‘rorrl :!3 I llrn/h before planning tr)
36 mm/h atier planning. i
In the same way, RI,,,, rose
2.4. Crop yeld rlronitorir~g
from 4.1 to 7.9 (US units) The watershed manage-
ment effect wac relevant
b u1 ,the global rtmoff ,volume
From 1988 to 1992, thei 53 nodes of the 20x20 m’
suvirigs werc 10~.
grid were alro uscd for d&ermining trop yield com-
Table 7 givcs thc ~IMN~ hydrological
results from
ponents and studying their/spatial variability. Ground-
thr square-metcr Plots. Abnual runoff coefficients
nut (Amchis h!pogea) w& harvestecl on 12 m’ arca
ranged frorn 1 O<‘i to ?S%i. During tht: samc y e a r ,
plots and Pearl millet (&rl$isetum tiphoiiies) on 20 ni3
runoff depth somt:time\\,
d ubled between the plots.
area plots. This experimen’t was only conducted in the
Beyond thc trop (over
i
cffe-ts, this spatial variability
NI.
was quite constant: S44 pilot bas the worst infiltr-
abilily. On this \\cale. th Lplim threshold values
3 . Results
rangcd from 6 to 10 m
and zI()lirn
from 18 to
24 mm/h. These results indi ate that the annual outlet
3 . 1 . Hydrologiccll sunvv
flow represented lO-20% :
f the square-meter
runoff
estimations. This means ,that water distribution
Concerning the NI basi , 129 rainfall events were
processes within the
1
wate s.hed
were much higher
recorded before watershecliplanning (198%1987), 50
than the losses to thc outside. Similar results
/
Fil; 3. Comprison of min depth frequency distribution curvë\\. bcIv.wu lu0 pwiod,~. twfwe (lY85--1987) and tifter (1989-1992) plannm;. NI.
I’eri*~tl
N o . o f
No. I>f
.r0t;d
-1 >l.L/
Tut:11
ih?ean ru;l.lfl’
rains
Iloods
rain!al! (mm,
:,ltlGvit! index
nJnoff (mm)
coefficient (ci 1
._--.--...-----
-
-
-
-
-
__-----
----_--
_
__ _~. _._---^-------~~
..-.
.\\ 1 !dS!ll
Hclore (1985---l 987 1
179
10
2057.7
000
x7.1
3 . 2
f%mning ll9R8)
50
18
931.5
45
10.1
4.3
:4111-r (19X9-1991)
154
3 6
‘33’1.7
IllY
6 0 . 2
7 . 6
I;I Iurirf
firl‘ore ( 19X6--19P7)
8 8
3 3
1361.J
5-l:
159.X
11.7
Pl;mning ( I Y8Xj
49
1.3
917.3
131
68.3
7 . 4
,Aftrr (1989-1992)
1 5 4
36
22X2.3
ii)46
22 I .7
4 . 7
--.-
- -
Table 2
k~~nl’all. eroswily irldex and annual runoff balances for 4 ml-pIots located in the NI Iwin
_--I----_--
-
_----_-~__-.
YCil
Total rainfall (mm)
Erosiv. indes
Annual runcltf‘ (mm)
I 9x9
752.1
258
6S.7
60.2
135.2
128.7
IYY(l
&A
237
67.0
5 1 . 9
105.1 i-i.5
1991
505. I
267
5 5 . 7
6X.3
82.0 68.8
IYYZ
5Y4.I
320
x 5 . 0
96.2
159.6
69.2
--_-
-
-
-
i-
--_
~- ~.- --
were obtained in Africa by Thebe (1987): Miller
1987) and 28 for the final one (1989-1992). Although
( 1992).
annuel balances were not feasible, variations in overall
hlany values were missing in the sediment load data
losses between the two periods were rather substan-
set. because of the sampling mistakes. Only 25 rele-
tial: 13955 kg before planning and 2967 kg after
vant recoyds were available for the first period ( i 98%
planning. The same difference was noted between
-_-.
- -
. _ . - _ -_-___l--.----
- .
--.-
2 4
I) Pere: et ai. /Agriculture, Ecosy~mr and Environme~rr
70 ( IYYH) /Y-$
the most erosivc events of each period: 4924 kg (July
response of the YA basin F:;IS net substantially mod-
1986; R=96 US unitsj before planning and 912 kg
ified and thcrc vicrc still ~:rtcl- I,wcs. l’he cumulative
(July 1990; R-98 US units) after planning. Theore-
runoff depth ratio (Lr(Nl/j!‘Lrr YA,) \\Y..IS nearly 41 cil
tical specifïc erosion thus decreased from 1.9 t/ha/year
during the 1986-1987 p ‘riod, and dropped to 27%
to 0.3 t/ha/year, as a result of the watershed manage-
during the final period ( 11139-l 993j.
ment programme.
Table 4 gives the armudl hydroll.>gical results from
Conceming the YA basin, 88 rainfall events were
square-meter plots. Annual runoff‘ coefficients ranged
recorded before planning (1986- 1987), 49 during
from 18% t<:l 30%. A4 in thc NI hasinl spatial varia-
1988 and 154 aticr planning (1989-1992). Rain inten-
bility was qoite constant: thc S55 plot had the worst
sities were computed for all evcnts greater than
infiltrahilit>. On Ilii, \\,:;./I:,. 111: i
/
1 Irl-c~ll~>ld ~Slll~~l
L,=8 mm: in this case., 46, 28 and 93 rainstorms.
respectiveiy.
were analyzed from each period.
Although the freyuency distributions for rain depth
were similar during both periods. maximum 10 min
intensity (Ilo) frequency curves are quite different
before and after planning. Only i!O% of the events
had 110 values greater than 40 mm/h during the first
period, with 55% during the final (one.
The hydrological data set contains 33 flood events
before planning ( 1986-I 987), 13 during 1988 and 36
aftér planning (1989-1997). Overnll,.The first period
had a total rain depth of 1% 1 mm and a runoff depth
of 160 mm. The final period had a total rain depth
‘htal
value of 2285 mm and a runoff depth of 222 mm. The
mean runoff coefficient thus dropped from 1 1.7% to
IYYO
06/lO 11.1 07/l 7
9.7% (Table 3). As in the NI basin, intcrannual varia-
071 i x to 07/i I
bility was high during both the periods from the
08/0 1 t u 0x/1 6
WI 7 tt.1 0x/3 I
occurrence of violent rainstorms.
Threshold values for rainfall depth (I-plirn), max-
'li~tdl
imum 10 min intensity (Il~lli,,,) and erosivity index
(Rli,), below which there is no runoff, were computed.
The values were nearly steady during the two periods:
Lplim remained the same (13 mm), IlOlim increased
from 19 to 24 mm/h and Rlint from 2.3 to 2.7 (US
units). In comparison with the NI basin, the watershed
management effcct was less relevant, even the global
runoff volume savings were similar. The hydrological
Table 3
Rainfall, erwivity index and annuul runofl‘ balances liw 4 m’-plots Iocat?d in thc YA hasin
Total rainfall (mm )
Erosiv. index
Annual runoff (I~I~I)
Sji
SS4
s55
SS7
-
-
-
--. - - - -
-----.~---__~
1 9 8 9
740.2
254
171.7
139.7
1 9 9 0
433.8
247
9X.6
86.1
1991
505.6
268
128.4
102.5
1 9 9 2
603.0
I
278
108.8
1 2 0 . 4
--~~-
.-~
- - . -~_
F! Perez et al./Agriculture, Ecosystems and Environment 70 (199X) 19-29
2s
ranged from 6 to 7 mm and IlOlim from 11 to 24 m.m/h.
Normal distribution :f’unctions titted a11 the soi1 water
The Eatter value corresponds to the fourth plot, which
storage (O-150 cm dcpth) data sets. cxcept for the yt:ar
was in the brushwood zone. These results indicated
1988 which was delctcd from the :subsequent geosta-
that the annual outlet flow represented 40-50% of the
tistical analysis. A spherical mode! was uscd to com-
square-meter
runoff estimations. These proportions
pute parameters of the normalized semi-variogram
were higher than those issued from the colluviall
fimctions (Burgess and Webster, 1980). !Soi1 watel
alluvial terrace plot<. This is mainly because of the
storage presents an i,Mropic spatial structure with a
gcntle and uniform slope, and also to the low mean soi1
steady 50 m range valuc.
infilrrability within the YA basin, according to B#ader
The kriged contour maps highlightzd thc same
( 1994): Terri ( 1996,~.
characteristic arcas. c\\‘;>n 11 ben thc‘ actual watel
Although ther’: was no scdiment load data before
storage values diffcr.*. ‘X ~!own\\tr~;lI:l c.onfiucncc
planning. it \\V;~S possible to compare the global ,sedi-
z o n e was ,hown to c’.x~.~v(I thc inliltr;~~ion values.
ment losses in the Y.4 basin after planning, estimated
while the central axis zouc’ exhibited a chronic
to 1 1096 kg (35 flood events). with the NI basin losses
deficit (Fig. 3). Obviously. u’atcr accumulation \\~US
during the same period. The global sediment load ratio
t h e result o f t h c rclicf and cnhanced by filtering
(I,s( N1,1/ts( YA)) reaches nearly to 27%. Watershed
barriers. Further topsoii texture and soit surface
management obviously had limited effects on soi1
feature studies, accordin:: to thc mcthod of Caserlave
slabi!ization of thc flelcls located in the upstream part
and Valentin ( 1989). cor tirrncd thut thti central axis
of thc hillsidc.
zone was characterized hy higher silt and very fine
sand contents and unstable superficiai structure
(Perez, 199-I). Hencc. \\vatcl-hcd management did
not bave a marked ef‘fect (111 the lvater storage spatial
A di;tributicm titting proccdure vras established for
variabilitj \\vhich. dcpc.ndcJ up~m topogrqhy and soit
each 20x20 ni’ grid data set from the NI basin.
characteristics.
r7-c \\\\
/-
_ w i
i c’y
4
n
40
0
26
40
6 0
80
100
1 2 0
7 4 0
160
100
2 0 0
Fig. 3. Kriged contollr map of the soi1 water stkge (O-150 c:m depth). NI hasin, sûmplin g gid 20 x 21) m’. snmpline Lis.te Novemlxr l 1. 190 1
- . .
. -
- _ I ” - - - - _ -
- - . - - -
- - -
26
However, some field observations partly invalidat~,l
\\V~S determined from 1Jcriods without runofl‘ y*nd
these results, obtained at; the cnd of the rainy SC;I\\OZ:
comparcd to that o f ti&: crops dcfinecl aho\\~.. t\\
from a 20x30 m3 samjpling grid. In particular. .I
ratio of II .3 was thus jobtained in I’avour of thc
ponding area regularly appeared abovc thc livc-hed~e.
hedpe: it was appli~zd dhereafter fol all situations.
but its spatial extension was limited.
In 1991 and 1!)92, auger samples were thus col-
At the beginning of the tainy jeason. thç infiltration
lected along three tradsects perpendicular to thc
gain above the hedgc \\~LIS $rountl 1 IX mm in 190 I and
hedge. The sampling sites were syrnmetrical
and
S-l mm in 1992. In t 99 1, blobal watcr stora-c varia--
located at 5 m, 2 m, 1 ti and 0.5 m from the central
tioni. rneasured at the cnd /of the rainy season. wcre in
point (Fig. 4). Sampling was carried out on a t Na!
tine with the lirst infiltratidn gains. Howcver. for t 992.
‘oasis. Mean moisture levels at a given depth \\vLxrc
a simple study of global +ater storage variations did
analyzed, with transects considered as replicatcs. ‘1’1~~‘
not highlight the fifterinb rote of thc hcdge. Thc
water status was determiried by dividing the spacc into
infiltration gain mainly
ct the needs of the shrubs
two units.1 .
and Graminae species at t t: end of the season. Morc-
over,
1
marked stake rnonit ring cnabled assessment of
1. the first unit refers to the cropping area on caitl
annual sedimentation
ups ‘rcam from the ‘hcdge. This
side of the hedge, characterized by thc watc‘r
sedimentation
was found t be about I .8 crn/year after
suppiy measured at th( -5 m and +S m abscissac:
installation of the hedgc. 1 and levelled o f f at about
the mean runoff deptgs from S41 and S43 square-
05 cm/year thereafter.
’
meter plots were Subtracted from thc dail!
Concerning the filteri g barriers, from 1989 to
precipitation;
1992, a comparison
was
ade between water storage
2. the second unit refer,’ to the hedge, characterized
4
values from the S47 neutro probe access tube. located
hy the water supply In easured at the A1-0.5 m and
in the main gully and thc S4l and S42 watcr storage
ir 1 m abscissae; the live-hedge evapotranspiration
values (Fig. I(a)). Soi1 water storage could be corn-
P: Pere; et ul. /Agricuiture. Eco.~y~tms ami Environmcr~t
70 (1998) 19-29
“_ 7
Table 5
hedgé: upstream (13 spots) and downstrcam
( 1 i spot:,)
Grain yields for a Pearl millet (var. SOUNA III) croppsd in thc NI
part, belonging to the same farmer.
basin in 1988, 1990 and 1992 (millet/groundnut roration)
~----
-
Table 5 gives the results ofthe grain yicld J :u.iatiorrs
Location
Grain yield (k#ha)
recorded in 1988, 1990 and 1992 with a Pearl millet
-
(var. Souna III) trop. The production level XX high in
Mean
S D
Coefficient of
variation (!FI
comparison
to nearby fields (Perez ct ai., iY97). Thi$
- - - -
- -
--~-
was partly because of the soi1 chara’cteristic:,
but also
19,??8
Upstream
1143
3 9 7
34.7
to the current adoption of improved tcchniqucs
h;; thc
Downstream
891
.3l;*
1J.3
farmer. The downstream
area rcached the same puten-
i9w
Upstream
tial as the upper part. Even tho@I annuai ~:ij:n;,l~c
Downstrcam
variations interfered with evaiuation of thc agric[~~-
tural results, local farmer5 strr5scil 111~ !;IL~ iii:li ! .
IYY2
Upstrcam
816
52
30. I
surface savings and field homugeneir>, Y\\ t’r~ IK (1 ~CIL
Downstream
1 1 7 7
4 19
35.6
- -
-
vant benefits.
Variations due to the location, relative to the Ii\\~-hcdgc position.
Harvest spots: 20 m2.
4. Discussion
puted on a per-year basis until the last measurement
level(25O cm depth) was reached by the wetting frrlnr.
A c c o r d i n g t o Amir ( 1996~. soi1 rchabiiitaticrn
In fact. this was net a major constraint as most of thc
attempts are dependent on the existing climntic CO~I-
surface runoff was trapped at the beginning of !h<
ditions, cropping systems ancl thc socic~tlconomi:
rainy season, when, trop caver was sparse and violent
environment. In the case of’ western Africa. wit?
rainstorms occurrcd. During the monitoring period.
semi-extensive agricul[urts and subhumid ciimatt .
infiltration gains bithin the gully ranged from 101 tc)
few technical references
are available, aven thou@l
127 mm (Table 5). Thc mean sedimentation above rhe
many extension programs havc developed !~C\\C soii
lilterinp barriers was found to be about 16 mm/year
and water conservation pracrices
Serpentic anil
after installation. and levelled off at 13 mm/year there-
Lamachere (1990) improved wzter infiltr,ltion b:t
after. These confirm former results obtained by Ruellc
combining soit ploughing and stone hunds. in
et al. Cl 990).
1000 rn’ plots located in northern Burkina Faso. in
the same country, Van Duijn et al. ( 1994) c’:)nfirmed
3.3. Crop yield mcmitoring
the advantage of stone bunds for water management iu
the local food trop system. Most authors ackr.ou’lcdg~
The effect of the improved cropping practices on
that the trop response is often moderdte because 01
soi1 and water management and then on the yicld
subsequent leaching processes or unbalanced watc‘l
components were studied separately (Perez et al..
and minera1 supply (Reyniers and Forest. 1990~
1996). The ficld survey within the NI basin high-
The NI, located on the colluviaYailu\\;ial ttrrrace. i.
lighted the spatial variability in the trop r’esponse to
characterized by a good soi1 infitrability and a doti n
the soit and water conservation measures. Bcfore
stream gully system. Before planning. thts annu::l
planning, the downstream
widening guily was’under-
runoff coefficient was low (4.2%) and corr~spondcd
going erosion and topsoil crusting. Some 2500 m’
to a marked water deficit betwecn the square-mele!
were progressively abandoned by the farmer, but ia
runoff potential and the outlet fow. Overall. despk:
1988 the entire arca was cropped, becausc of thc
the low absolute values. watershed managrmccl
sediment deposits above the filtering barriers and
allowed a reduction of 40% in water losses, and
dry sea.son soit decompacting.
sediment loads were six times lower. Within thc
As two plots were under an alternate trop rotation
watershed, water distribution was not greatly !modifieci
(544 and S46 locations), rather than carrying out a
according to the water storage spatial variability.
geostatistical analysis the remaining area was divided
However, iimited areas Bocated above the fïlterin?
into two blocks relative to the location of the live-
barriers, concentrated water infiltration and trapped
-
-“....
* *v\\
._--
-
-----
-____
sediments. Consequentli,
the topography and soi1
rangzlands uzhile neglect-
surface features of the downstream zone were con-
:tices. except for the most
siderably modified. Thede improvements, associated
with the new tillage tee d niques and manure applica-
tershed management in the
tion, favoured sustainablk cropping of the entire area.
pping basin o f
Senegal:
The YA. located in thel upper part of the hillside, is
l;ltiorl processes along the
characterized by a low
nfiltrability and a uniform
rcclaimation, social con-
relief. Before planning
annual runoff coefficient
p organization.
was nearly 12% and re
sented 40% of the square-
meter runoff potential.
r planning, water and soi1
losses remained high. ObGiously, the filtering effect of
the conservation measur s was not efticient enough.
b
The absence of a well ide, tified drainage net led to the
creation of large fluctuati$g waterways. Surface runoff
thus bypassed the fdte$ing barriers and sediment
dcposits wcre small (live-/hedge: 0.5 cm/year; filtering
iquc dc ruissellement ;L stockage
barriers: 1 .O cm/year).
’
+ ct extrapolation
sl:r vcrwnt
Moreovcr. hecause of tbe soi1 constraints, improved
39-:V~2.
cropping practiccs were ess efficient than applied on
‘Ix efrect of surface covêr on
1
the colluvial/ alluvial terr ce. For exemple, dry season
: Agassi, M . < E d . ) , Soi1 Erosion,
<I~I. Mürecl Dckker,
New ~0rL.
dccornpactmz
created II 1
cm deep subsoiling in thc
downalopc si:ndy soils i,ut
5 only 7 cm deep in thc
NI. CIptimal interpolation
and
upslope gra\\‘elly soils. S$l surface fcntures also chan-
[‘-rtic.. 1: thc scm-1 arigram and
ged more rapidly under rzzindrop impact (Perez, 1994).
1533 1.
Ix\\ Ctats de surface de la %nnc
rnfïltrxtion.
Coll. Didactiques.
.-
i*,. Fra xc. 226 pp.
5. Conclusion
:smél oratior, I&I profil culturel
t+argilcux
de la zone tropicale
On a farm-plot scale, lhe two experimental water-
lice\\ :tgronomiqucs
Apron. Trop.
sheds were representativk
of the bral environmental
\\‘<atcr hnlancc c<timarion mndel:
constraints and the land dse features. The poor quality
\\Xi\\ So~l Sci. Soc. Am. J. 51.
and the crusting tenden y of upslope soils were net
favourable for the
1
estab,ishment
of crops or Young
rforc\\trv
for snil management in
shrubs. Greater effort wa also requi red from the oxen
HA. (13s. ). Soi1 degradation. A
t
for soi1 tillage. Often far f/rom the Vi:llage and rcnted to
York, USA. pp XIII-XVII.
nr ruiswllcmeni
et a l’iniiltration
outsiders, the fields locbted on th.ese soils are not
rrcla~c I n : S,ivakumar.
hl.V.li..
priorities fQr farmers. 111 contrast. downslope soils.
:IL). Soi1 Water Balance in thc
deep and easy to till,! allow rapid development.
.x. IASH workshop,
Fcbr-uar!
because of surface savinbs and the high yield poten-
0, 199. pp. IOY-120.
.
tial.
.ire4ows fircml farmeri’
fïelds in
-.
\\lanagcmcnt
Rescarch
Program.
In the light of these pbenomena,
reinforced by the
t 992 SADCC puhl.. Gaborone.
diffcrent tcchnical result% described in this paper. the
natural trend will probabiy lead to the developed bclts
xl. W.B.. Bcnyamini.
\\i. 1983.
located along the low and axes and topped by
water cnnservation
in the semi-
degraded hillsides.
1
Thisi tendency could explain the
ling ai a tool for conservation
s. 1, 215-221..
behaviour of farmers de+ribed in F#erez et al. (1997).
illement sur lcï -01s cultives du
Although developing l$llsides -requircs collective
diqx~htic
à I”aména~em~nt
de
work, the samc pcople, /as individual farmers, were
SA Montpellier France, 250 pp.
?? Perez et al./Agriculture, Ecosystums and Envinmnwnt 70 (1998) 19-20
29
Prrez. l?, Serre, M., 1995. Evolution des structures agraires el
Serpentie, G.. Lamachere. J.M.. 1990. Valorisation a,@oIc des
érosion dans Be sud Saloum (Senegal). Reseau Erosion, No. 15,
eaux de ruissellement et lutte contre l’érosion sur champ\\
ORSTGM
Montpellier France, pp. 59-68.
cultivés en mil en Ione \\ouclano-suh~licilne. In: K$ergreis. A..
Perez, P., Boscher, C., Sene, M., 1996. L’amélioration des
Claude, J. (Eds.), Utilisation rationnelle de I”eau des petits
techniques culrurales pour une meilleure gestion de l’eau
bassins versants en zone aride. Ed. AUPELF UREE John
pluviale (sud Saloum, Senegal). Agriculture and Développe-
Libbey Eurotext, Montmuge, France, pp. 275-286.
ment, :No. 9, pp. 20-30.
Sessay. M.F.. Stocking. MA.. lY95. Soi1 productivity ;.nd fcttility
Pcrcz. P., Albergel, J., Diatta. M., Grouzis, M., Sene, M.,
maintenance of a dcgradcd Clxisol in Sierra Leone. In: Ganr>.
1997. Rehabilitation of a semiarid ecosystem in Sencgal.
F.. Campbell, B. (Eds. I. Sustainable land management tn
1. Experiments at the hillside s’cale Agri. Ecot. Env.,
African semi arid and subhumid rcgions. Proc. of thc SCGPE
in mess.
workshop. November 1093. Dakar. CIRAD-CA, Montpellier.
Ils;$ mers, EN.. Forest, F, 1990. La pluie n’est pas le seul remède à
pp. 1?9- i N.
la s&here<se en Afrique. Les flux hydtiques dans le système
Tbcbc. B.. 1987. Hydrodynamique dc quelques sol,; du Nord
WI -I:ulturc-atmospiltre en zone intertropicale. Sichercsse. I
C~mcrcwm. Ba\\ins verhants dc Mouda. Contribution à l’étutlc
( 1). 36-W.
de\\ II anstcrts d’échelle\\. Doctoral Thesis. USTL Montpellier-.
Ko w. E., 1995. Introduction à la gestion conservatoire de l‘eau, de
France. il)6 pp.
la b:omasac et de la fertilité des sols (GCES). Bull. Pedol. de la
Terri. D.. 1996. Slr~pe. ,t\\ptct and surface storage. In: Apassi. ht.
FAC). No. 70. Roma, Italia, 420 pp.
(Ed.), Soi1 Et-o\\ion. C«nwr\\-drion and Rehabilitati:)n.
Ma:-c,,l
RWSZ. E., Duguc, P., Rodriguez, L.. 1992. La GCES, une nouvelle
Dekkcr, NCA. York. USA. pp 777107.
stwegie de lutte ;tnti-érosive appliquée à f’aménagement de
Van DuiJn. H.J.H’.. Van Dricl. V:,F., Kaborc. 0.. 1993. Influencé
terroir en zone soudano-sahélienne du Burkina Faso. Bois and
dei c~wlotts pierreux sur la relation entre le bilan hydrique et le
Fin&\\ des Tropiques, No. 233. pp. 49-63.
~rcndsment dans le bassin versant de Gualaga à Namsigua
Ku:llc. I?, Senr, .M., Juncker, E.. Diatta, MM., Perez, P. (Eds.), 1990.
l Burkina Faso). In: Re:;nier~,.
1-N.. Nctoyo. L. (Eds.). Btlan
I)ci’cn!.e et restauration des sols. Coll. Fiches Techniques. vol.
111 driquc :~grizolc et &herew en ,Al’rique tropicakr. EJ. J~IIw
1. No. 1, ISRAIUNIVAL. Dakar Senepül. 63 pp.
I.ibbcy Imrotnt. Paris. Fran~c. pp. 275-288.
-
E~nvibnment
Notes to contrfbutors
Submission of manuscripts. Manuscripts should be submitted in triplicate to the Editorial Office of Agricuhe. tcosysfems &
Environment,
PO. Box 181,100O AD Amsterdam, The Netherlands.
Electronic mnnuscripts. Electronic manuscripts have the advantage that there is no need for the rekeying of ta?xt. thereby
avoiding t+te possibility of introducing errors and resulting in reliable and fast delivery of proofs.
For the initial submission of manuscripts for consideration,
hardcopies are sufficient. For the processing of accepted papers.
electronic versions are preferred. After final acceptance, your disk plus two final and exactly matching priited versions should
be submitted together. Double density (DD) or high density (HD) diskettes (3.5 or 5.25 inch) are acceptable. It is important that
the file saved is in the native format of the wordprocessor program used. Label the disk with the name of the computer and
wordprocessing package used, your name, amd ihe name of the file on the disk. Further information may be obtained from
the Publi&er.
Authors in Japan please note: Upon request, Elsevier Science Japan Will provide authors with a list of people who cari check
and improve ‘the English of their paper (befcre submission).
Please contact our Tokyo office: Elsevier Science Japan, l-9-15
Higashi-Azabu, Minato-ku, Tokyo 106-0044;
Tel. (03)-5561-5032;
Fax (03)-5561-5045
Enquiries c:oncerning mantiscripts and proofs: guestions arising after acceptance of the manuscript. especially
WISB relating
to proofs, should be directed
to: Elsevier Science Ireland Ltd., Bay 15K, Shannon Industrial Estate, Shannon, CO. Clan?, lrelanc.
T e l . (+353) 61 471944, Fax (+353) 6 1 4 7 2 1 4 4 .
Advertising information. Advertising orders and enquiries may be sent to: International: Elsevier Science, Advertising
Department, The Boulevard, L.angford Lane, Kidlington. Oxford 0X5 lGB, UK, Tel. (+44)(O) 1865 843565; Fax 1+44)(O) 1865
843976. USA and Canada: Elsevier Science Inc., Mr. Tino DeCarlo.
655 Avenue of the Americas, New York, hY 10010-5107.
USA,Tel. (+l) (212) 633 3815; Fax (+l) (212) 6 3 3 3 8 2 0 . J a p a n : Elsevier Science Japan, Marketing Service, 1-9-15 Higashl-Azabu
Minato-ku, ‘Tokyo 106-0044, Japan, Tel. (+81)-3-5561-5033;
Fax (+81)-3-5561-5047.
US mailing notice, Agriculture, Ecosystems & Environment
(0167-8809) is published monthly by Elsevier Science E%V
(Molenwerf 1, Postbus 211, 1000 AE, Amsterdam). Annual subscription price in the USA US$ 1327.00 (valid in No:th. Centrai
a n d S o u t k America o n l y ) , i n c l u d i n g a i r s p e e d d e l i v e r y . S e c o n d class p o s t a g e p a i d a t J a m a i c a , N Y 1 1 4 3 1 .
USA POSTMASTERS:
Send address changes 10, Agriculture, Ecosysfems 8 Environment
Publications IExpediting,
Inc., 200
M e a c h a m A v e n u e , E l m o n t . N Y 1 1 0 0 3 . AIRFREIGHT AND MAILING i n t h e U S A b y P u b l i c a t i o n E x p e d i t i n g .
Agriculture, Ecusystems, & Environment has no page charges
For a full and complete Guide for Authors, please refer to
Agriculture; Ecosysstems, & Environment, Vol. 67, No. 1, pp. 91-94.
The guide car? also be found on the World Wide Web:
access under http://www.elsevier.nl or http://www.elsevier.com
Copyright II 1998, Elsevier Science BM AII rights reserved
0169-8809/98/$19.00
@ The paper used in this publication meets the requirements of ANSI/NISO
239.46-1992
(Perman#ence of Paper).
P r i n t e d i n ‘rhe Netherlands.
_ ___._
__-.-~--,---
__-
-..
.~
P_I
__, __ _-.
.-
NE$V FROM ELSEVIER SCIENCE !
I
APPLIED SOIL ECOLOG$’I
A Section !of Agriculture, Ecosystems & Enviranmen(
1
Editors-in-Chief:
C.A. Edwards, The Ohio State Univers@, Dept. of Entomoiogy,
’
Suiiding, 1735 Neiidvenue, Columbus. OU 4321 ii. f22U, US& i
Agdcu/kral Universify, Dept. of Plant Ecokq & Soi/ B~/ogy,
Bkrnsesfeeg 69, 6708 PD Wageningen, The Netheriands
AIMS AND SCOPE
Applied Soi! Ecology a d d r e ’ ses
the role of soil organisms aR d
their interactions in relationito:
a g r i c u l t u r a l productivity,
nutrient cyding and other sbil
processes, fhe maintenant of
soi1 structure and fertility,
t tl,e
i m p a c t of human activities çnd
xenobiotics on soi1 ecosystbms
and bio(techno)logical
contIol
of soil-inhabiiing
pests,
diseases and weeds. Such
issues are the basis of
sustainable agricultural and
fores@ systems and the
long-term conservation of +its
i n b o t h t h e t e m p e r a t e a n d
tropical regions.
The disciplines covered inc(ude
Editorial Advisory Board:
t h e f o l l o w i n g , a n d preferenh
T.H. Andereon, InstiM fi Bodenbiokgie.
Will be given to articles whioh
Braunschweig, Germany, Il. Alkin!ton,
are interdisciplinary and
Sdrod of Agi-Mure, Aberdeen, UK,
integrate two or more of thdse
M.H. Beare. New Zealand Inst. ~CY Crop and
disciplines.
FOC~ Research, chrislchurch, New Zealand
- soi1 microbiology and
C. Chenu, INRA, Vmai//es, France.
m i c r o b i a l e c o l o g y
E.T. Chriiensen, Research Cenfre
- soil invertebrate zoology $d
Foulm. qe/e, Denmark, D.C. Coleman,
ecology
Unwersity of Gw@a, Atiens, GA, USA.
- root and rhizosphere ecolagy
KM. DOU~E, CSIRO, Gien Osmond, SA,
-@l science
Aus&a/ia D.W. Freckman. Cdoradu Stafe
,+oil biotechnology
Univers@, F M cdlins, CO, USA,
“- ecotoxicology
S. Hansen, 7he Royal Veferinaryand
- nematology
Agn’uMa! Unywsily, Fredetiksbsrg,
” entomology
Denmark, H.kJ. Hoilink, Chi0 SraC
- plant pathology
- agronomy and sustainable
agriculture
- nutrient cycling
- ecosystem modeillng
ana
f o o d w e b s
The journal Will pubkh origihal
papers, review articles sfx$
communicatxxs,
v;evwpoint$,
edito~ials, book revinw and:
announcemw-its
Agricult:ure, Ecosystems &
Environment
Rehabilitation of a ‘semiarid ecosystem in Senegai.
2. Earm-plot experiments
. .\\
~ --_.
1_1--.-.-”
-
-
-.
-1-m-1.1
-I---H
A@cu.Rure
:
Ecospkms&
Enironrnent
An International Journal for Scientific: Research on the Relationship
of Agricuiture
and Food ProdcC :XI P f.7~ fiiosphere
-
Aims and scope. This journil is concerned with the interaction of methods of agricultural produ
tion. agroecosystems and the
environment. Topics covered
include the comparison of different methods of production (intens ~3. extensive, linear, cyclic) in
terms of their ecology; how,agricultural production methods affect pollution of soil, water alid ; il; the quality of food. and the
use of energy and non-renqwable resources; the effect of industrial pollutants on agriculture:
ind the policy rssues involved
in the change and developmient
of agriculture. The journal aims to serve as a focus for scientls!s
n açrlc::lltrre,
food productior
forestry and the environme& as well as for administrators and policy-makers concerned with t E:S~ IF sues in research estab-
lishments, government, indtstry and international organisations. It publishes original scientlfi
: papers. review ariicles and
occasional comment papers. From 1994 onwards a section of this journal, Appbed
Soi/ fcology i
publ~~+ed se;!.qrately.
Applietl
Soi/ Ecology
addresses the kole of soit organisms and their interactions in relation to: agriclflt:Ji
3 ;:rc-, ,‘i!/,.r
::llrierit cyc11-11.
and other soi1 processes, thie maintenance of soi1 structure and fertility, the impact of hun;:,:i _’
1.
: .’ ‘C, 82 p $, ;,
ecosystems and bio(techno)(logical
contIroI of soil-inhabiting
pests, diseases and weeds
EDtTORS-IN-CHIEF
J.W. Sturrck, 500 Johns Road, Fernside. RDl, Rangiora, North Canterbury, New Zealand
M.R. Carter, Research Branch. Agriculture and Agri-Food Canada, PO. Box 1210. Charlottetown. f 3
C a n a d a
Book Review Editor
B.A. Auld, Agricultural Resebrch Gen!re. Forest Road, Orange, N.S.W. 2800. Australia
EDtTORtAL ADVISORY BOARD
J.i? Aeschlimann, Montpelliei,
France
J. Huber, Darmstadt. Gernlan!
B . A . Auld, Orange, N.S.W., Australia
M . J . J o n e s , A l e p p o , S y r i a
J . N . B . B e l l , Ascot, U K
J.V. Lovett. Queen V i c t o r i a Teri ice. ACT, Austraiia
H . B o u w e r , P h o e n i x , A Z , US A .
G.A. Norton, St. Lucia, Qld , A l:;tralia
K . J . B r e n t . B r i s t o l , U K
D.M. Olszyk, Corvallis. OR, US
G.T. Brooks. Reading, UK
C.K. Ong, Nairobi, Kenya
A . E l T i t i , S t u t t g a r t , G e r m a n y
D . P i m e n t e l , I t h a c a , N Y . U . S . A .
L.O. Fresco. Wageningen, Netherlands
G . H . L . R o t h s c h i l d , M a n i l a . IF%11
M. Giampietro, Rome, Italy
M . S h i y o m i . Mito. Japan
S . R . G l i e s s m a n , S a n t a C r u z , C A . U . S . A .
R.E.H. Sims. Palmerston Nortl
PJ. Gregory, Reading, UK
L.M. Skarby, Goteborg.
Swedf
S . K . H a h n , B r o a d v i e w H t s . , Ci)H, U . S . A .
E . M . A . S m a l i n g . W a g e n i n g e n . detherlands
1. Haque, Lahore. Pakistan
P I . Sunarjo. U j u n g P a n d a n g . I I donesia
J . H e l e n i u s , H e l s i n k i , Finland,
M. Tbth, Budapest, Hungary
M . H o r n u n g , Grange-over-Sends, U K
J . M . W e b s t e r . B u r n a b y , B . C . . C inada
Publication information: Ag[icu/ture,
Ecosystems & Environment (ISSN 0167-8809). For 1998 ~UITWS 67-71 are scheduletl
for publication. Subscribers’to Agriculture. fcosystems & Environment Will have the opportuni I to subscribe to Applied SO.~
Ecology at a speciat reducel subscripticrn rate. Subscription
prices are available upon request In m the PubIIsh~?r. Subscrtption<;
are accepted
on a prepaid ibasis only, and are entered on a calendar year basls. Issues are
ent b!/ surfacr rnarl excepi t:
the following
countries where air delivery by SAL mail is ensured: Argentina. Australia.
Brazil, C:
.lada, Hong Kong, india, lsrae
Japan. Malaysia, Mexico, d e w Zealand.
Pakistan, PR China, Singapore, South Korea, Taiwar
Thailanc. USA For all other
countries airmail rates are: avaiiable upon request Claims for missing issues should be rr lde \\niithin six months ot 1311’
publication (mailingf date.
Orders, ciaims, and produd enquiries: jplease contact the Customer Support Department at thc ?! Regional
Sales Office nearest
you:
N e w Y o r k : Elsevier Science,; f?O. B o x 9 4 5 . N e w Y o r k , N Y 1 0 1 5 9 - 0 9 4 5 , U S A ; T e l . (+l) 2 1 2 - 6 3 3 - 3 7 : C. [Tell free nal‘lber for North
American Customers: l-888/-4ES-INFO
<437-4636)],
Fax. (+ 1) 212-633-3680,
E-mail: usinfo-fui’el
3E:vier icom
Amsterdam: Elsevier Scien
e . P O . B o x 2 1 1 , 1 0 0 0 A E A m s t e r d a m , T h e N e t h e r l a n d s : T e l : ( i - 3 1 ) 2 0 - 4 G375ï. Fax (t-31) 2C -485343;
4
E-mail: nlinfo-f@elsevier.nl
/
Tokyo: Elsevier Science, 9- 5, Higashi-Azab:)
1-chnnie, Minato-Ku.
Tokyo 106-0044.
-lapan Tell,
(+8!: 3-5561-5033.
Fax. i rC:
3-5561-5047, E-mail: info@ Isevier.co.jp
Singapore: E&vier Science,. No. 1 Temasuk
~\\ver;~~r:.
:! !i-01 Millenia ;Ower. Singapore
@3911??.
i(?,y ( .L 55) 433. &;‘y, i:ax. ; ,-(:
337-2230. E-ma!l. arialnfc@’
1 ..lsevier.com.s!:
Agriculture
Ecosystems 4%
Enwonment
Agriculture, Ecosystems and Envimnment 70 (1998) 19-29
Rehabilitation of a semiarid ecosystem in Senegal.
?. Farm-plot experiments
P. Pereza*“‘, J. Albergelc, M. Diattab, M. Grouzkf, M. Seneh
a CIRAD, BP 503.5 34~932,‘Montpellier
Cedex 1. France
’ ISRA, BP 199. Kaolack, Senegal
’ ORSTOIM,
B P 504.5 3 4 0 3 2 , Montpellier Cedex 1, F r a n c e
Accepted 23 October 1997
_.-~-___-----_--_-. -..-..- - _---.-...--. ---..-~-.-
Abstraet
In the Sudano-Sahelian zone of western Africa. characterized by semi-extensive agriculture and subhurnid clim,lte, soi1
degradation and water shortages are widespread features. Soi1 and water conservation practiccs are introduced to the farmers.
but often abandoned thereafter. Some interesting results have been obtained, on a hillside scale. in the southern part of the
cropping basin of Senegal. As a supplement to the experimental design, two small watersheds (2.5 ha) were deliner.ted and
equipped in representative hillside locations. In 1988, bath the watersheds were planned and submitted to an hydnalogical
survey. One of them, located on a colluvial/alluvial terrace. was also submitted to soi1 water storage and grain yield
momtoring. Results highlight a decisive effect of soi1 and relief features on the efficiency of conservation mrasures. F!cle\\ant
results were obtained on the terrace, but upstream areas still generated marked soi1 and warer losses. These phenomena. in
addition to socioeconomic constraints. partly explain farmers’ behaviour noticed on the hillside bcale. i‘ 1998 Published by
Elsevier Science B.V. Al1 rights reserved.
Kqr’rtinrnlr: Soi1 conservation: Runoff; Erosion; Water balance: Watershed management: Millet; Senegal
1, Introduction
be stabilized and runoff reduced on a11 slopes (Perez
and Sene, 1995)
In the cropping basin of Senegal, environmental
In fact, soi1 infiltrabiliity increases from the Upper
degradation is found in a11 landscape units, as shown
part of the hillside to the lowland area (Perez, 1994).
by hiph water erosion, involving sheet erosion in the
Thanks to this natural tmnd, it is possible to control
upper parts of the toposequence.
gully erosion at
overland flow by reducing its velocity and avoiding its
nickpoints and sand deposits in the lowland areas.
concentration. Depending on the local conditions, this
For landscape rehabilitation and as a prelude to any
cari be achieved with a network of filtering ob:rtacles
agricultural intensification, erosion phenomena must
such as stone bunds or live-hedges (La1 and Siewart.
1990; Roose, 1994).
~---
The marked tendency of the local soils to form
*Correaponding author.
surface seals is the result of their weak stability. These
~1167-8X0’~19Xi0;19.0(~ J
1998 Pubiished hy Elsevier Science R.V. All rights reserved.
/‘II SOl(>7-8809(97)OOi57-6
--
-
~_--- ---
ferruginous loamy sands are characterized by yery low
infiltrability shows high spatial variabi1it.y (Perez,
clay and organic matter contents (Charreau aid Nicou.
1094).
1971). Hence, there are very few solutions to increase
In 1988, soi1 and water conscrvatio.n
practiccs wcrc
long-term infiltrability in the soi1 profile. Cropping
implemented in the NI. They included one live-hcdge.
practices creating and temporaril:y maintaining sur-
established in the middle bfthe hasin and douhled with
face roughness permit an increase surface storage and
an upstream graminae lin<: (Ptrrri(~ri/l rrlrrsir!rlr!ll). antl
ponding time (Morin ef al., 1984; Lamachère, 1991).
cight filtering barriers (r/tc>nc pavements and brusb-
Soi1 surface covers of trop canopies or residuc
wood dams) across the \\hiaterM,ay:;. Ihrthcrrnore. XV
mulches cari reduce surface sea1in.g and flow velocit)
eral improved cropping p@tices were introtlucecl into
(Box and Bruce, 1996). Under local conditions. it i\\
the four farm-plots: contour cultivation. dry sca~)n
necessary to promote rapid trop establishment
antl
dccompacting, shallow r d.gin;r and localizcd mantlrc’
adequate canopy growth. This cm bc donc by cou-
application. These
i
techni LWS ;Ire tlc>~crib~tl ‘11 Perc/ ~1
pling water and organic matter management practices
al. (1997).
(Roose et al., 1’992; Sessay and Stocking, 1995).
The 2.5 ha Yaranc Katkrshcd (YA) i4 lo~:~tcti in 111~’
In the light of these observations, rehabilitation
Upper part of the hillslopj on thc cdge of thc: cropping
operations were conducted on the basis of local eco-
area. Because of the ge/n.tle relief, ~:he limits werc
logical features and human uses. The local effect of
dclineated with an cartljen ri(Jgc. Thc: \\aluc of thc
conservation measures and the mechanisms
involved
regular slope is nearly iI<6, xith n o evidcncc ot’ ;~n
were studied in two small watersheds (2.5 ha). corrc-
hydrological network, except i.or a downstream
shal-
sponding to an intermediate scale Inetween thc hillsidc
low wide waterway. Ovekland tlou anci shect crosion
area (1 km’) and the experimental plot ( 100 m’) nncl
characterize t h i s arca. T h c \\vatershc‘d i,, cnrircl)
constituting a relevant sized soi1 I-mit.
cropped and di\\ided intc lier 11irm-plots ~Fi~. 1 th 1;.
The characteristics of both watersheds
ancl thc
The soi1 comprises col uvial dçposits and fine grav-
survey methods used are describcd, followed by thc
cls eroded from the upp :r plateau. Thc tir:<t horizon
results from the hydrological
survey, the water baluncc
(O-20 cm dcpth) is sandy (with 10% clay content) ancl
monitoting and the trop yield study. The discussion
presents 5--30% femc ;cs ravels and the structure is
highlights the consequences
of thc rcsult\\ t’o~
continuous and fragile. Deeprr. the texture rapidly
watershed management on a hillsidc scalc.
bccomes loamy with 50- 10% tcrric gravcls. Belou ;II~
a\\‘erage depth of 50 cm ferric nodules and gra’~.cIs
account for 80% of tlhe oil volume. Under dry con-
2. Material and methods
ditions. this horizon is like a hardpan. b u t wcttcd
material turns crumbly. Gurface sealing is a general
2.1. Watershrd descriptim
feature of the watershed but the strength of the seal
depends on the depth of the hardpan.
The 2.5 ha Ndiba watershed (hl) is located down-
In 1988, soi1 and wattri conservation practiccs wcre
slope on the colluvial/alluvial terrace. A contouring
implcmented in the YA. They are the samc as those
track and natural relief mark the limits of the basin.
established in the NI Ir lsin:
)Y
one live-hedge, three
The upstream slope is nearly 0.5%. wliereas it reaches
filtering obstacles and irn provcd cropping practices.
2.0% in the downstream area where rills are concen-
trated into a widening gully. The watershed is entirely
cropped and divided+o ‘fow-@ym plots (Fig. l(a)).
It has a leached and dis&bed ‘feimginous soil.
In 1985, the NI outle wus equipped with a rain
The first horizon (O-60 c& depth) is sandy and friable
gauge and a water stage :Corder that were set up in a
with a continuous structure. Clay (5-10%) and
concrete-lined ditch. Sec iment loads wcre manually
organic matter (0.5%) contents are very small. Deeper.
collected from 1985 to 1 192. Since 1988, 4 m’- plots
the texture gradually b ‘cornes loamy and ferric spots
were established in diffi rent parts OF the ,watershr:d
:
or gravels appear from .5 m depth. As a consequence
(Fig. la) and runoff vc urnes measured after each
of sheet erosion and dolluvial deposjfs, the topsoil
rainfall event. One yea after installation, filtering
/
-
l? Ptw: et ai. /Agr-iculture,
Ecoqstems
and Envirvnrnent
70 (1998) 19-29
'tj - - field limit
I-1 - - Ii-e hedge
obstacles were eqaipped with 30 marked stakes for
initial stake height (above the soi1 surface) and tht.
measuring upslope sedimentation. During the rainy
actual value.
season, each IO-day period, the cumulative sediment
In 1986, the YA outlet was equipped with a rain
depth was estimated from the difference between the
gauge and a water stage recorder that werl; set up in a
. , .1qm1--..1-7-----<--
-
.-,p-
-
concrete-lined ditch. It iq only since 1988 that the
during 1988 - thc rainiest 1 ar of the decade - and 154
sediment loads were maneally collected. Since 1988,
aftcl planning ( I 0X9- 19’ II. Rain intensities w e r e
3 m”-plots were establishbd in different parts of the
computed t’or a11 t‘\\ ent!, 1 zater then L,=8 mm (15~:
watershed (Fig. l(b)) and runoff volumes were mea-
min depth); in this case.
7, 32 and 93 rainstorms,
sured after each rainfall. A fourth plot was installed in
rzspcctivcly. werc: anal y: d for each period. Fre-
an adjacent brushwood z$ne. As in the Ni basin, 20
quency distribution\\ of w depth (Fig. 2), ma ximuni
marked stakes wcre use8 ‘for measuring sediment
10 and 30 min intcnsitics I 10 ancl 1&, erosivity index
dcpth upstream of the filtkring obstacles.
tKJ. wcrc thc zanic bcforc
md after planning. It was
thus possible to compare tl
hydrological
results from
hoth tinie \\eries.
.~lrll~lti!Jll \\(?Il!:' 111,1X! !‘L ords were lest because 01‘
Seven neutron probe acbess tubes were installed in
Llic Ic~hnica! !)i.~)l-‘!Pn15. ‘1
clevant liydrological data
the NI watershed. includitg the 4 m’-plots, for mon-
4c’t was huilt. Il contain l 40 flood eVents bcforc
itoring the soi1 moisturei The neutron gauge was
planning ( IX-V- 1087). 1 I durin’> 1988 and 36 after
a:
calibrated fc.lr each access’tube
and each specific soi1
planning ( 19X9- I(W). 0 rail, the finit period had a
layer (Perez. 1994). Meayurements
‘were done every
total rain depth of 2057
m and a nmoff depth of
IO-day period during the kainy season.
X7 nim. Thc iinal pcriod
lad a total rain dcpth of
TO strengthen the stud of the soi1 water storage
2310 111111 ancl 2 lull~~ff d
th of 60 mm. The mean
spatial variability, 53 sam4ling spots were located on a
runoff coefficicnr thus SI Eted liom ,4.2% to 2.6%
20x20 m’ grid within the. NI basin (Fig. l(a)). From
(TabIc 1,. In l’ac’i. inter-a nual ,variability was high
1988 IO 1993. samples tiere obtained with a shell
during bath thc ~L’I ic&+. I’U 11 the occurrence of violent
augcr (O- 150 cm tlcpth) atjthe end of the rainy scason.
r;tinhtornis. that \\omctinic
represented upto 60% of
Geostatistic;il concepts webe used for the data analysis
;lllnual IC\\'Cl\\.
(Burgcss and Wchster, 1TXO; Chopart. and Vauclin.
Conccrning rairlstorms. lordy 31% of the events
1990) and the kriging phcedure was used for soi1
initiated runoff during t!xz fi st period and 26% during
water storage.
the final OI~L’. Threshold
célues for rainfall depth
The same studies were dlanned in the YI. Hokvevcr.
( L,I,,,,), maximum 10 min jntensity (ll(,li~~>) and ero-
aerious problems were edcountered
during the data
sivity index (li,,,,,). helow &Ch there [vas no runoff.
analysis, because of the, presence of gravels and
\\\\‘ere computcd. When the ,I+,,,, value remained the
nodules in the soi1 profile. phe results were considered
sxiic during thc tv.0 pci-icxlsi (LPllr,=71 mm). the f ,Olim
unreliable and therefore ace not incl;uded here.
vuluc inzrenscd t‘rorrl :!3 I llrn/h before planning tr)
36 mm/h atier planning. i
In the same way, RI,,,, rose
2.4. Crop yeld rlronitorir~g
from 4.1 to 7.9 (US units) The watershed manage-
ment effect wac relevant
b u1 ,the global rtmoff ,volume
From 1988 to 1992, thei 53 nodes of the 20x20 m’
suvirigs werc 10~.
grid were alro uscd for d&ermining trop yield com-
Table 7 givcs thc ~IMN~ hydrological
results from
ponents and studying their/spatial variability. Ground-
thr square-metcr Plots. Abnual runoff coefficients
nut (Amchis h!pogea) w& harvestecl on 12 m’ arca
ranged frorn 1 O<‘i to ?S%i. During tht: samc y e a r ,
plots and Pearl millet (&rl$isetum tiphoiiies) on 20 ni3
runoff depth somt:time\\,
d ubled between the plots.
area plots. This experimen’t was only conducted in the
Beyond thc trop (over
i
cffe-ts, this spatial variability
NI.
was quite constant: S44 pilot bas the worst infiltr-
abilily. On this \\cale. th Lplim threshold values
3 . Results
rangcd from 6 to 10 m
and zI()lirn
from 18 to
24 mm/h. These results indi ate that the annual outlet
3 . 1 . Hydrologiccll sunvv
flow represented lO-20% :
f the square-meter
runoff
estimations. This means ,that water distribution
Concerning the NI basi , 129 rainfall events were
processes within the
1
wate s.hed
were much higher
recorded before watershecliplanning (198%1987), 50
than the losses to thc outside. Similar results
/
Fil; 3. Comprison of min depth frequency distribution curvë\\. bcIv.wu lu0 pwiod,~. twfwe (lY85--1987) and tifter (1989-1992) plannm;. NI.
I’eri*~tl
N o . o f
No. I>f
.r0t;d
-1 >l.L/
Tut:11
ih?ean ru;l.lfl’
rains
Iloods
rain!al! (mm,
:,ltlGvit! index
nJnoff (mm)
coefficient (ci 1
._--.--...-----
-
-
-
-
-
__-----
----_--
_
__ _~. _._---^-------~~
..-.
.\\ 1 !dS!ll
Hclore (1985---l 987 1
179
10
2057.7
000
x7.1
3 . 2
f%mning ll9R8)
50
18
931.5
45
10.1
4.3
:4111-r (19X9-1991)
154
3 6
‘33’1.7
IllY
6 0 . 2
7 . 6
I;I Iurirf
firl‘ore ( 19X6--19P7)
8 8
3 3
1361.J
5-l:
159.X
11.7
Pl;mning ( I Y8Xj
49
1.3
917.3
131
68.3
7 . 4
,Aftrr (1989-1992)
1 5 4
36
22X2.3
ii)46
22 I .7
4 . 7
--.-
- -
Table 2
k~~nl’all. eroswily irldex and annual runoff balances for 4 ml-pIots located in the NI Iwin
_--I----_--
-
_----_-~__-.
YCil
Total rainfall (mm)
Erosiv. indes
Annual runcltf‘ (mm)
I 9x9
752.1
258
6S.7
60.2
135.2
128.7
IYY(l
&A
237
67.0
5 1 . 9
105.1 i-i.5
1991
505. I
267
5 5 . 7
6X.3
82.0 68.8
IYYZ
5Y4.I
320
x 5 . 0
96.2
159.6
69.2
--_-
-
-
-
i-
--_
~- ~.- --
were obtained in Africa by Thebe (1987): Miller
1987) and 28 for the final one (1989-1992). Although
( 1992).
annuel balances were not feasible, variations in overall
hlany values were missing in the sediment load data
losses between the two periods were rather substan-
set. because of the sampling mistakes. Only 25 rele-
tial: 13955 kg before planning and 2967 kg after
vant recoyds were available for the first period ( i 98%
planning. The same difference was noted between
-_-.
- -
. _ . - _ -_-___l--.----
- .
--.-
2 4
I) Pere: et ai. /Agriculture, Ecosy~mr and Environme~rr
70 ( IYYH) /Y-$
the most erosivc events of each period: 4924 kg (July
response of the YA basin F:;IS net substantially mod-
1986; R=96 US unitsj before planning and 912 kg
ified and thcrc vicrc still ~:rtcl- I,wcs. l’he cumulative
(July 1990; R-98 US units) after planning. Theore-
runoff depth ratio (Lr(Nl/j!‘Lrr YA,) \\Y..IS nearly 41 cil
tical specifïc erosion thus decreased from 1.9 t/ha/year
during the 1986-1987 p ‘riod, and dropped to 27%
to 0.3 t/ha/year, as a result of the watershed manage-
during the final period ( 11139-l 993j.
ment programme.
Table 4 gives the armudl hydroll.>gical results from
Conceming the YA basin, 88 rainfall events were
square-meter plots. Annual runoff‘ coefficients ranged
recorded before planning (1986- 1987), 49 during
from 18% t<:l 30%. A4 in thc NI hasinl spatial varia-
1988 and 154 aticr planning (1989-1992). Rain inten-
bility was qoite constant: thc S55 plot had the worst
sities were computed for all evcnts greater than
infiltrahilit>. On Ilii, \\,:;./I:,. 111: i
/
1 Irl-c~ll~>ld ~Slll~~l
L,=8 mm: in this case., 46, 28 and 93 rainstorms.
respectiveiy.
were analyzed from each period.
Although the freyuency distributions for rain depth
were similar during both periods. maximum 10 min
intensity (Ilo) frequency curves are quite different
before and after planning. Only i!O% of the events
had 110 values greater than 40 mm/h during the first
period, with 55% during the final (one.
The hydrological data set contains 33 flood events
before planning ( 1986-I 987), 13 during 1988 and 36
aftér planning (1989-1997). Overnll,.The first period
had a total rain depth of 1% 1 mm and a runoff depth
of 160 mm. The final period had a total rain depth
‘htal
value of 2285 mm and a runoff depth of 222 mm. The
mean runoff coefficient thus dropped from 1 1.7% to
IYYO
06/lO 11.1 07/l 7
9.7% (Table 3). As in the NI basin, intcrannual varia-
071 i x to 07/i I
bility was high during both the periods from the
08/0 1 t u 0x/1 6
WI 7 tt.1 0x/3 I
occurrence of violent rainstorms.
Threshold values for rainfall depth (I-plirn), max-
'li~tdl
imum 10 min intensity (Il~lli,,,) and erosivity index
(Rli,), below which there is no runoff, were computed.
The values were nearly steady during the two periods:
Lplim remained the same (13 mm), IlOlim increased
from 19 to 24 mm/h and Rlint from 2.3 to 2.7 (US
units). In comparison with the NI basin, the watershed
management effcct was less relevant, even the global
runoff volume savings were similar. The hydrological
Table 3
Rainfall, erwivity index and annuul runofl‘ balances liw 4 m’-plots Iocat?d in thc YA hasin
Total rainfall (mm )
Erosiv. index
Annual runoff (I~I~I)
Sji
SS4
s55
SS7
-
-
-
--. - - - -
-----.~---__~
1 9 8 9
740.2
254
171.7
139.7
1 9 9 0
433.8
247
9X.6
86.1
1991
505.6
268
128.4
102.5
1 9 9 2
603.0
I
278
108.8
1 2 0 . 4
--~~-
.-~
- - . -~_
F! Perez et al./Agriculture, Ecosystems and Environment 70 (199X) 19-29
2s
ranged from 6 to 7 mm and IlOlim from 11 to 24 m.m/h.
Normal distribution :f’unctions titted a11 the soi1 water
The Eatter value corresponds to the fourth plot, which
storage (O-150 cm dcpth) data sets. cxcept for the yt:ar
was in the brushwood zone. These results indicated
1988 which was delctcd from the :subsequent geosta-
that the annual outlet flow represented 40-50% of the
tistical analysis. A spherical mode! was uscd to com-
square-meter
runoff estimations. These proportions
pute parameters of the normalized semi-variogram
were higher than those issued from the colluviall
fimctions (Burgess and Webster, 1980). !Soi1 watel
alluvial terrace plot<. This is mainly because of the
storage presents an i,Mropic spatial structure with a
gcntle and uniform slope, and also to the low mean soi1
steady 50 m range valuc.
infilrrability within the YA basin, according to B#ader
The kriged contour maps highlightzd thc same
( 1994): Terri ( 1996,~.
characteristic arcas. c\\‘;>n 11 ben thc‘ actual watel
Although ther’: was no scdiment load data before
storage values diffcr.*. ‘X ~!own\\tr~;lI:l c.onfiucncc
planning. it \\V;~S possible to compare the global ,sedi-
z o n e was ,hown to c’.x~.~v(I thc inliltr;~~ion values.
ment losses in the Y.4 basin after planning, estimated
while the central axis zouc’ exhibited a chronic
to 1 1096 kg (35 flood events). with the NI basin losses
deficit (Fig. 3). Obviously. u’atcr accumulation \\~US
during the same period. The global sediment load ratio
t h e result o f t h c rclicf and cnhanced by filtering
(I,s( N1,1/ts( YA)) reaches nearly to 27%. Watershed
barriers. Further topsoii texture and soit surface
management obviously had limited effects on soi1
feature studies, accordin:: to thc mcthod of Caserlave
slabi!ization of thc flelcls located in the upstream part
and Valentin ( 1989). cor tirrncd thut thti central axis
of thc hillsidc.
zone was characterized hy higher silt and very fine
sand contents and unstable superficiai structure
(Perez, 199-I). Hencc. \\vatcl-hcd management did
not bave a marked ef‘fect (111 the lvater storage spatial
A di;tributicm titting proccdure vras established for
variabilitj \\vhich. dcpc.ndcJ up~m topogrqhy and soit
each 20x20 ni’ grid data set from the NI basin.
characteristics.
r7-c \\\\
/-
_ w i
i c’y
4
n
40
0
26
40
6 0
80
100
1 2 0
7 4 0
160
100
2 0 0
Fig. 3. Kriged contollr map of the soi1 water stkge (O-150 c:m depth). NI hasin, sûmplin g gid 20 x 21) m’. snmpline Lis.te Novemlxr l 1. 190 1
- . .
. -
- _ I ” - - - - _ -
- - . - - -
- - -
26
However, some field observations partly invalidat~,l
\\V~S determined from 1Jcriods without runofl‘ y*nd
these results, obtained at; the cnd of the rainy SC;I\\OZ:
comparcd to that o f ti&: crops dcfinecl aho\\~.. t\\
from a 20x30 m3 samjpling grid. In particular. .I
ratio of II .3 was thus jobtained in I’avour of thc
ponding area regularly appeared abovc thc livc-hed~e.
hedpe: it was appli~zd dhereafter fol all situations.
but its spatial extension was limited.
In 1991 and 1!)92, auger samples were thus col-
At the beginning of the tainy jeason. thç infiltration
lected along three tradsects perpendicular to thc
gain above the hedgc \\~LIS $rountl 1 IX mm in 190 I and
hedge. The sampling sites were syrnmetrical
and
S-l mm in 1992. In t 99 1, blobal watcr stora-c varia--
located at 5 m, 2 m, 1 ti and 0.5 m from the central
tioni. rneasured at the cnd /of the rainy season. wcre in
point (Fig. 4). Sampling was carried out on a t Na!
tine with the lirst infiltratidn gains. Howcver. for t 992.
‘oasis. Mean moisture levels at a given depth \\vLxrc
a simple study of global +ater storage variations did
analyzed, with transects considered as replicatcs. ‘1’1~~‘
not highlight the fifterinb rote of thc hcdge. Thc
water status was determiried by dividing the spacc into
infiltration gain mainly
ct the needs of the shrubs
two units.1 .
and Graminae species at t t: end of the season. Morc-
over,
1
marked stake rnonit ring cnabled assessment of
1. the first unit refers to the cropping area on caitl
annual sedimentation
ups ‘rcam from the ‘hcdge. This
side of the hedge, characterized by thc watc‘r
sedimentation
was found t be about I .8 crn/year after
suppiy measured at th( -5 m and +S m abscissac:
installation of the hedgc. 1 and levelled o f f at about
the mean runoff deptgs from S41 and S43 square-
05 cm/year thereafter.
’
meter plots were Subtracted from thc dail!
Concerning the filteri g barriers, from 1989 to
precipitation;
1992, a comparison
was
ade between water storage
2. the second unit refer,’ to the hedge, characterized
4
values from the S47 neutro probe access tube. located
hy the water supply In easured at the A1-0.5 m and
in the main gully and thc S4l and S42 watcr storage
ir 1 m abscissae; the live-hedge evapotranspiration
values (Fig. I(a)). Soi1 water storage could be corn-
P: Pere; et ul. /Agricuiture. Eco.~y~tms ami Environmcr~t
70 (1998) 19-29
“_ 7
Table 5
hedgé: upstream (13 spots) and downstrcam
( 1 i spot:,)
Grain yields for a Pearl millet (var. SOUNA III) croppsd in thc NI
part, belonging to the same farmer.
basin in 1988, 1990 and 1992 (millet/groundnut roration)
~----
-
Table 5 gives the results ofthe grain yicld J :u.iatiorrs
Location
Grain yield (k#ha)
recorded in 1988, 1990 and 1992 with a Pearl millet
-
(var. Souna III) trop. The production level XX high in
Mean
S D
Coefficient of
variation (!FI
comparison
to nearby fields (Perez ct ai., iY97). Thi$
- - - -
- -
--~-
was partly because of the soi1 chara’cteristic:,
but also
19,??8
Upstream
1143
3 9 7
34.7
to the current adoption of improved tcchniqucs
h;; thc
Downstream
891
.3l;*
1J.3
farmer. The downstream
area rcached the same puten-
i9w
Upstream
tial as the upper part. Even tho@I annuai ~:ij:n;,l~c
Downstrcam
variations interfered with evaiuation of thc agric[~~-
tural results, local farmer5 strr5scil 111~ !;IL~ iii:li ! .
IYY2
Upstrcam
816
52
30. I
surface savings and field homugeneir>, Y\\ t’r~ IK (1 ~CIL
Downstream
1 1 7 7
4 19
35.6
- -
-
vant benefits.
Variations due to the location, relative to the Ii\\~-hcdgc position.
Harvest spots: 20 m2.
4. Discussion
puted on a per-year basis until the last measurement
level(25O cm depth) was reached by the wetting frrlnr.
A c c o r d i n g t o Amir ( 1996~. soi1 rchabiiitaticrn
In fact. this was net a major constraint as most of thc
attempts are dependent on the existing climntic CO~I-
surface runoff was trapped at the beginning of !h<
ditions, cropping systems ancl thc socic~tlconomi:
rainy season, when, trop caver was sparse and violent
environment. In the case of’ western Africa. wit?
rainstorms occurrcd. During the monitoring period.
semi-extensive agricul[urts and subhumid ciimatt .
infiltration gains bithin the gully ranged from 101 tc)
few technical references
are available, aven thou@l
127 mm (Table 5). Thc mean sedimentation above rhe
many extension programs havc developed !~C\\C soii
lilterinp barriers was found to be about 16 mm/year
and water conservation pracrices
Serpentic anil
after installation. and levelled off at 13 mm/year there-
Lamachere (1990) improved wzter infiltr,ltion b:t
after. These confirm former results obtained by Ruellc
combining soit ploughing and stone hunds. in
et al. Cl 990).
1000 rn’ plots located in northern Burkina Faso. in
the same country, Van Duijn et al. ( 1994) c’:)nfirmed
3.3. Crop yield mcmitoring
the advantage of stone bunds for water management iu
the local food trop system. Most authors ackr.ou’lcdg~
The effect of the improved cropping practices on
that the trop response is often moderdte because 01
soi1 and water management and then on the yicld
subsequent leaching processes or unbalanced watc‘l
components were studied separately (Perez et al..
and minera1 supply (Reyniers and Forest. 1990~
1996). The ficld survey within the NI basin high-
The NI, located on the colluviaYailu\\;ial ttrrrace. i.
lighted the spatial variability in the trop r’esponse to
characterized by a good soi1 infitrability and a doti n
the soit and water conservation measures. Bcfore
stream gully system. Before planning. thts annu::l
planning, the downstream
widening guily was’under-
runoff coefficient was low (4.2%) and corr~spondcd
going erosion and topsoil crusting. Some 2500 m’
to a marked water deficit betwecn the square-mele!
were progressively abandoned by the farmer, but ia
runoff potential and the outlet fow. Overall. despk:
1988 the entire arca was cropped, becausc of thc
the low absolute values. watershed managrmccl
sediment deposits above the filtering barriers and
allowed a reduction of 40% in water losses, and
dry sea.son soit decompacting.
sediment loads were six times lower. Within thc
As two plots were under an alternate trop rotation
watershed, water distribution was not greatly !modifieci
(544 and S46 locations), rather than carrying out a
according to the water storage spatial variability.
geostatistical analysis the remaining area was divided
However, iimited areas Bocated above the fïlterin?
into two blocks relative to the location of the live-
barriers, concentrated water infiltration and trapped
-
-“....
* *v\\
._--
-
-----
-____
sediments. Consequentli,
the topography and soi1
rangzlands uzhile neglect-
surface features of the downstream zone were con-
:tices. except for the most
siderably modified. Thede improvements, associated
with the new tillage tee d niques and manure applica-
tershed management in the
tion, favoured sustainablk cropping of the entire area.
pping basin o f
Senegal:
The YA. located in thel upper part of the hillside, is
l;ltiorl processes along the
characterized by a low
nfiltrability and a uniform
rcclaimation, social con-
relief. Before planning
annual runoff coefficient
p organization.
was nearly 12% and re
sented 40% of the square-
meter runoff potential.
r planning, water and soi1
losses remained high. ObGiously, the filtering effect of
the conservation measur s was not efticient enough.
b
The absence of a well ide, tified drainage net led to the
creation of large fluctuati$g waterways. Surface runoff
thus bypassed the fdte$ing barriers and sediment
dcposits wcre small (live-/hedge: 0.5 cm/year; filtering
iquc dc ruissellement ;L stockage
barriers: 1 .O cm/year).
’
+ ct extrapolation
sl:r vcrwnt
Moreovcr. hecause of tbe soi1 constraints, improved
39-:V~2.
cropping practiccs were ess efficient than applied on
‘Ix efrect of surface covêr on
1
the colluvial/ alluvial terr ce. For exemple, dry season
: Agassi, M . < E d . ) , Soi1 Erosion,
<I~I. Mürecl Dckker,
New ~0rL.
dccornpactmz
created II 1
cm deep subsoiling in thc
downalopc si:ndy soils i,ut
5 only 7 cm deep in thc
NI. CIptimal interpolation
and
upslope gra\\‘elly soils. S$l surface fcntures also chan-
[‘-rtic.. 1: thc scm-1 arigram and
ged more rapidly under rzzindrop impact (Perez, 1994).
1533 1.
Ix\\ Ctats de surface de la %nnc
rnfïltrxtion.
Coll. Didactiques.
.-
i*,. Fra xc. 226 pp.
5. Conclusion
:smél oratior, I&I profil culturel
t+argilcux
de la zone tropicale
On a farm-plot scale, lhe two experimental water-
lice\\ :tgronomiqucs
Apron. Trop.
sheds were representativk
of the bral environmental
\\‘<atcr hnlancc c<timarion mndel:
constraints and the land dse features. The poor quality
\\Xi\\ So~l Sci. Soc. Am. J. 51.
and the crusting tenden y of upslope soils were net
favourable for the
1
estab,ishment
of crops or Young
rforc\\trv
for snil management in
shrubs. Greater effort wa also requi red from the oxen
HA. (13s. ). Soi1 degradation. A
t
for soi1 tillage. Often far f/rom the Vi:llage and rcnted to
York, USA. pp XIII-XVII.
nr ruiswllcmeni
et a l’iniiltration
outsiders, the fields locbted on th.ese soils are not
rrcla~c I n : S,ivakumar.
hl.V.li..
priorities fQr farmers. 111 contrast. downslope soils.
:IL). Soi1 Water Balance in thc
deep and easy to till,! allow rapid development.
.x. IASH workshop,
Fcbr-uar!
because of surface savinbs and the high yield poten-
0, 199. pp. IOY-120.
.
tial.
.ire4ows fircml farmeri’
fïelds in
-.
\\lanagcmcnt
Rescarch
Program.
In the light of these pbenomena,
reinforced by the
t 992 SADCC puhl.. Gaborone.
diffcrent tcchnical result% described in this paper. the
natural trend will probabiy lead to the developed bclts
xl. W.B.. Bcnyamini.
\\i. 1983.
located along the low and axes and topped by
water cnnservation
in the semi-
degraded hillsides.
1
Thisi tendency could explain the
ling ai a tool for conservation
s. 1, 215-221..
behaviour of farmers de+ribed in F#erez et al. (1997).
illement sur lcï -01s cultives du
Although developing l$llsides -requircs collective
diqx~htic
à I”aména~em~nt
de
work, the samc pcople, /as individual farmers, were
SA Montpellier France, 250 pp.
?? Perez et al./Agriculture, Ecosystums and Envinmnwnt 70 (1998) 19-20
29
Prrez. l?, Serre, M., 1995. Evolution des structures agraires el
Serpentie, G.. Lamachere. J.M.. 1990. Valorisation a,@oIc des
érosion dans Be sud Saloum (Senegal). Reseau Erosion, No. 15,
eaux de ruissellement et lutte contre l’érosion sur champ\\
ORSTGM
Montpellier France, pp. 59-68.
cultivés en mil en Ione \\ouclano-suh~licilne. In: K$ergreis. A..
Perez, P., Boscher, C., Sene, M., 1996. L’amélioration des
Claude, J. (Eds.), Utilisation rationnelle de I”eau des petits
techniques culrurales pour une meilleure gestion de l’eau
bassins versants en zone aride. Ed. AUPELF UREE John
pluviale (sud Saloum, Senegal). Agriculture and Développe-
Libbey Eurotext, Montmuge, France, pp. 275-286.
ment, :No. 9, pp. 20-30.
Sessay. M.F.. Stocking. MA.. lY95. Soi1 productivity ;.nd fcttility
Pcrcz. P., Albergel, J., Diatta. M., Grouzis, M., Sene, M.,
maintenance of a dcgradcd Clxisol in Sierra Leone. In: Ganr>.
1997. Rehabilitation of a semiarid ecosystem in Sencgal.
F.. Campbell, B. (Eds. I. Sustainable land management tn
1. Experiments at the hillside s’cale Agri. Ecot. Env.,
African semi arid and subhumid rcgions. Proc. of thc SCGPE
in mess.
workshop. November 1093. Dakar. CIRAD-CA, Montpellier.
Ils;$ mers, EN.. Forest, F, 1990. La pluie n’est pas le seul remède à
pp. 1?9- i N.
la s&here<se en Afrique. Les flux hydtiques dans le système
Tbcbc. B.. 1987. Hydrodynamique dc quelques sol,; du Nord
WI -I:ulturc-atmospiltre en zone intertropicale. Sichercsse. I
C~mcrcwm. Ba\\ins verhants dc Mouda. Contribution à l’étutlc
( 1). 36-W.
de\\ II anstcrts d’échelle\\. Doctoral Thesis. USTL Montpellier-.
Ko w. E., 1995. Introduction à la gestion conservatoire de l‘eau, de
France. il)6 pp.
la b:omasac et de la fertilité des sols (GCES). Bull. Pedol. de la
Terri. D.. 1996. Slr~pe. ,t\\ptct and surface storage. In: Apassi. ht.
FAC). No. 70. Roma, Italia, 420 pp.
(Ed.), Soi1 Et-o\\ion. C«nwr\\-drion and Rehabilitati:)n.
Ma:-c,,l
RWSZ. E., Duguc, P., Rodriguez, L.. 1992. La GCES, une nouvelle
Dekkcr, NCA. York. USA. pp 777107.
stwegie de lutte ;tnti-érosive appliquée à f’aménagement de
Van DuiJn. H.J.H’.. Van Dricl. V:,F., Kaborc. 0.. 1993. Influencé
terroir en zone soudano-sahélienne du Burkina Faso. Bois and
dei c~wlotts pierreux sur la relation entre le bilan hydrique et le
Fin&\\ des Tropiques, No. 233. pp. 49-63.
~rcndsment dans le bassin versant de Gualaga à Namsigua
Ku:llc. I?, Senr, .M., Juncker, E.. Diatta, MM., Perez, P. (Eds.), 1990.
l Burkina Faso). In: Re:;nier~,.
1-N.. Nctoyo. L. (Eds.). Btlan
I)ci’cn!.e et restauration des sols. Coll. Fiches Techniques. vol.
111 driquc :~grizolc et &herew en ,Al’rique tropicakr. EJ. J~IIw
1. No. 1, ISRAIUNIVAL. Dakar Senepül. 63 pp.
I.ibbcy Imrotnt. Paris. Fran~c. pp. 275-288.
-
E~nvibnment
Notes to contrfbutors
Submission of manuscripts. Manuscripts should be submitted in triplicate to the Editorial Office of Agricuhe. tcosysfems &
Environment,
PO. Box 181,100O AD Amsterdam, The Netherlands.
Electronic mnnuscripts. Electronic manuscripts have the advantage that there is no need for the rekeying of ta?xt. thereby
avoiding t+te possibility of introducing errors and resulting in reliable and fast delivery of proofs.
For the initial submission of manuscripts for consideration,
hardcopies are sufficient. For the processing of accepted papers.
electronic versions are preferred. After final acceptance, your disk plus two final and exactly matching priited versions should
be submitted together. Double density (DD) or high density (HD) diskettes (3.5 or 5.25 inch) are acceptable. It is important that
the file saved is in the native format of the wordprocessor program used. Label the disk with the name of the computer and
wordprocessing package used, your name, amd ihe name of the file on the disk. Further information may be obtained from
the Publi&er.
Authors in Japan please note: Upon request, Elsevier Science Japan Will provide authors with a list of people who cari check
and improve ‘the English of their paper (befcre submission).
Please contact our Tokyo office: Elsevier Science Japan, l-9-15
Higashi-Azabu, Minato-ku, Tokyo 106-0044;
Tel. (03)-5561-5032;
Fax (03)-5561-5045
Enquiries c:oncerning mantiscripts and proofs: guestions arising after acceptance of the manuscript. especially
WISB relating
to proofs, should be directed
to: Elsevier Science Ireland Ltd., Bay 15K, Shannon Industrial Estate, Shannon, CO. Clan?, lrelanc.
T e l . (+353) 61 471944, Fax (+353) 6 1 4 7 2 1 4 4 .
Advertising information. Advertising orders and enquiries may be sent to: International: Elsevier Science, Advertising
Department, The Boulevard, L.angford Lane, Kidlington. Oxford 0X5 lGB, UK, Tel. (+44)(O) 1865 843565; Fax 1+44)(O) 1865
843976. USA and Canada: Elsevier Science Inc., Mr. Tino DeCarlo.
655 Avenue of the Americas, New York, hY 10010-5107.
USA,Tel. (+l) (212) 633 3815; Fax (+l) (212) 6 3 3 3 8 2 0 . J a p a n : Elsevier Science Japan, Marketing Service, 1-9-15 Higashl-Azabu
Minato-ku, ‘Tokyo 106-0044, Japan, Tel. (+81)-3-5561-5033;
Fax (+81)-3-5561-5047.
US mailing notice, Agriculture, Ecosystems & Environment
(0167-8809) is published monthly by Elsevier Science E%V
(Molenwerf 1, Postbus 211, 1000 AE, Amsterdam). Annual subscription price in the USA US$ 1327.00 (valid in No:th. Centrai
a n d S o u t k America o n l y ) , i n c l u d i n g a i r s p e e d d e l i v e r y . S e c o n d class p o s t a g e p a i d a t J a m a i c a , N Y 1 1 4 3 1 .
USA POSTMASTERS:
Send address changes 10, Agriculture, Ecosysfems 8 Environment
Publications IExpediting,
Inc., 200
M e a c h a m A v e n u e , E l m o n t . N Y 1 1 0 0 3 . AIRFREIGHT AND MAILING i n t h e U S A b y P u b l i c a t i o n E x p e d i t i n g .
Agriculture, Ecusystems, & Environment has no page charges
For a full and complete Guide for Authors, please refer to
Agriculture; Ecosysstems, & Environment, Vol. 67, No. 1, pp. 91-94.
The guide car? also be found on the World Wide Web:
access under http://www.elsevier.nl or http://www.elsevier.com
Copyright II 1998, Elsevier Science BM AII rights reserved
0169-8809/98/$19.00
@ The paper used in this publication meets the requirements of ANSI/NISO
239.46-1992
(Perman#ence of Paper).
P r i n t e d i n ‘rhe Netherlands.
_ ___._
__-.-~--,---
__-
-..
.~
P_I
__, __ _-.
.-
NE$V FROM ELSEVIER SCIENCE !
I
APPLIED SOIL ECOLOG$’I
A Section !of Agriculture, Ecosystems & Enviranmen(
1
Editors-in-Chief:
C.A. Edwards, The Ohio State Univers@, Dept. of Entomoiogy,
’
Suiiding, 1735 Neiidvenue, Columbus. OU 4321 ii. f22U, US& i
Agdcu/kral Universify, Dept. of Plant Ecokq & Soi/ B~/ogy,
Bkrnsesfeeg 69, 6708 PD Wageningen, The Netheriands
AIMS AND SCOPE
Applied Soi! Ecology a d d r e ’ ses
the role of soil organisms aR d
their interactions in relationito:
a g r i c u l t u r a l productivity,
nutrient cyding and other sbil
processes, fhe maintenant of
soi1 structure and fertility,
t tl,e
i m p a c t of human activities çnd
xenobiotics on soi1 ecosystbms
and bio(techno)logical
contIol
of soil-inhabiiing
pests,
diseases and weeds. Such
issues are the basis of
sustainable agricultural and
fores@ systems and the
long-term conservation of +its
i n b o t h t h e t e m p e r a t e a n d
tropical regions.
The disciplines covered inc(ude
Editorial Advisory Board:
t h e f o l l o w i n g , a n d preferenh
T.H. Andereon, InstiM fi Bodenbiokgie.
Will be given to articles whioh
Braunschweig, Germany, Il. Alkin!ton,
are interdisciplinary and
Sdrod of Agi-Mure, Aberdeen, UK,
integrate two or more of thdse
M.H. Beare. New Zealand Inst. ~CY Crop and
disciplines.
FOC~ Research, chrislchurch, New Zealand
- soi1 microbiology and
C. Chenu, INRA, Vmai//es, France.
m i c r o b i a l e c o l o g y
E.T. Chriiensen, Research Cenfre
- soil invertebrate zoology $d
Foulm. qe/e, Denmark, D.C. Coleman,
ecology
Unwersity of Gw@a, Atiens, GA, USA.
- root and rhizosphere ecolagy
KM. DOU~E, CSIRO, Gien Osmond, SA,
-@l science
Aus&a/ia D.W. Freckman. Cdoradu Stafe
,+oil biotechnology
Univers@, F M cdlins, CO, USA,
“- ecotoxicology
S. Hansen, 7he Royal Veferinaryand
- nematology
Agn’uMa! Unywsily, Fredetiksbsrg,
” entomology
Denmark, H.kJ. Hoilink, Chi0 SraC
- plant pathology
- agronomy and sustainable
agriculture
- nutrient cycling
- ecosystem modeillng
ana
f o o d w e b s
The journal Will pubkh origihal
papers, review articles sfx$
communicatxxs,
v;evwpoint$,
edito~ials, book revinw and:
announcemw-its