STUDY OF CROP ROTATION IN SEN&AL f ...
STUDY OF CROP ROTATION
IN SEN&AL f
RESULTS AND METHODS
by
R,, NICOU
Translated from French t Wtude de Successions Culturales
Résultat* et Méthodesa
L*Agroncmie Tropicale, 1978 No 1
Janvier - Hars, pp.' 51-62
ICT 80-3
J3y R6 P?Icou*
sui .:mPi : In cwder ~CG study the place of new crcps (caize, uplahd rice,
cottonr in crap rotation, experiments were done in Senegal Setween 1965
and 1972 in eight geographical locations in the 800-1300 mm zone. In the
design used, ~11 of the crops could follow one another, and after four years
many combinations could be obtainede
Anafysis of the results led to two conclusions w!:ich are ir,portant
beyond L:;e geograDhica1 fraF?ework of Senegal" hy drawing up a table for trop
compatibility, hinonîal cratps cari beedefineds Sy pairing theso binosials ohor
sen according to socio-economic conditions, a lot of agronorically viable
combinations cari Se obtained. Thus for Senegal, a certain number of rotation
types were defined for the ecological areas ccncerned.
hs for nethodology, t%e design cari be sitnplified by t:eaxs of se-
rialization, thus giving faster and more precise results irrespective of the
yearo
These experiments called "preceding crops" Will not achieve mea-
ningful results if they are carried out in traditional cultivating condi-
tions, In fac%, tke differences are too minimal to develop a cotxpatibility
tabler %erefore, there iq good reason for using recommended verieties, well-
known cropt?ing tec'iniques and for obtaining an adequate level of intensifi-
cation.
IT3Y ‘P233S
: cmj? rotation, experizeRta1 design, ecology, preceding crops)
cropping techr,iques, intensification, Senegal.
In 1965, the need to diversify Senegalese agriculture and in par-
ticular, ixroase cereal production, led I?&T-Senegal to consider research
on crcq rotation, other than that already recoaKended for conmon use, and
which las largely based on the alternation of groundnuts, millet and/or sor-
ghuzr
.
At thc beginning, the new crops to be introduced were cotton and
maizer i. bit later caoe upland rice.
It was decided to do an experiment with the sims :
- of determining in what way and where the new crops could be
introduced in the rotation.
- of detarmining what are the Sest poesible successions aRong t?e
different crops eligible for rotation, with the successive cul-
tivntion of cereal upon cereal being exarzined in particulero
* NICOU (R) : S:UT Resear&h Agronomist, National Agronomie Bescarch Centre
in Sanbey (ISRA - Senegal).
Fr'esently at 1211.1, BP ho 5035, 34032 Montpellier Cedex, France.
Taking into account the demands of the new crops, and setting out
is not possible
from tbe principle $hat diversification with these .species
conducted in
unless one bas a>?propriate techniques, tlese experinents were
intensified agrieultural conditions :
lrdeep" (15 - 20 cm) soi1 tillage
turning under harvest residues whenever Possible
he,avy z:ineral fertilization with initial addition of rock ghos-
'. .
phatos to rcduce thezphosghorus dèficiency, atid *tien addition of.;
comple-entary fertilizers specific to each trop, allowing COE-=
pensation for minera1 exports ;
the use of recoimended varieties when available
t%e best-adasted cropping techniques.
Thc design used was to globally reveal trop compatibility, in or-
der to define useable binomials for trop rotation. Also, by the length of
the experiments, one hoped to show the beginning of evolving phenomena for
continous cropping of the same species.
Research lasted for eight years, from 1965-1972, during which some
experiments were done in several locations, Some had to be abandoned in ?zid-
stream.
A camplete and dctailed plan was drawn up for reference*. From tlne
only results %ich provided a minimum guarantee for trop development.
The current docu:?ezt aims to be more condecsed, by exceeding the
simple results useabla in Senegal, it tries to define a methodological
approach for the study of trop rotation.
The design was adapted from a plan used at the time by fZ.kT in
Dahomey (N. %RTS). The experiment includes eight replications, The number and
nature of crops csn vary clepending on the zone, but the principle rezains t:ie
same.
Year 1. For each replication, n strips 6 metres wide, and 2 metres apart are
cultivated : n represents the number of crops, The length of the strips de-
pends on the nueber of plants cultivated in the second yeara Strips are rahdo-
mized.
Year 2. For each re^lication , p strips 6 metres wide and 5, netres apart are
planted in perpecdicular fasbion t p represents the nuEîber of crops x?ich
cari be different Pro:. ~3. Xt is harvested in plots of 6m x 6m :yhere strips of
the first and second year intersect. T~US, there are n x p successions. The
main information concern the global effects of n preceding crops for each
second year trop-
* Report on eight years of resesrch on preceding crops in Senegal, by RI NICCU.
Himeographed copies available froc ISBA - CNRA in Bambey, Senegal.
-3 -
Year 3. Four replicntions are planted entirely with cereal and four with
cotton or groundnuts. Planting is done in the direction of the first year
strips, respecting the 2 :?etre spaces between rows* The harvest is done in
6m x 6m plots where the strips intersect.
Year 4, The eight re-lications are uniformly cultivated in the second Year
strip direction, and always harvested on the same 6m x 6m plots.
Thus, one obtains 2 x n x p rotations teated with the saxe plant,
which may vary from one geographical location to another. Thus it is possi-
ble to have up tc four years of cereals or cotton, for example.
The 6m x 6m = 36 ma net plot dimension is considered to be the nini-
mum sirie. It is difficult to znake it larger, but that obviously depends on
the nunber of crops* If n = 6 and
= 5 (which was fie-uently the case) each
block rr:easures 38 x 46 m = 1,748 ms , which represents 1.4 ha for eight repli-
cations. Therefore, with interblocks, it is necessary to find a 2 ha homoge-
nous plot of land.
The direction fur statistical analysis are :
- in year %, analysis is possible on the eight replications of
each trop- It is done using the following scheize I
degrees of freedon (d.Zq)
BlOClU
7
Prcceding Crops
n - l
Error
7 (n - 1)
In yecrs 3 and 4, there are two analyses pcr site, with four
replicatiois for each (as a function of the third year trop).
d.f.
Blocks
3
Column Sffects
n - l
Error (a)
3 #n.- 1)
Line Effects
P-1
Error (b)
3 hJ - 1)
Successions
(n - 1) (p - 1)
Error (c)
3 h -11 (p - 1)
Lastly, it should be pointed out that for t"ris first approach
to the problem of trop successions, only yield components 7rere taken into
account. Few observations were done on t%e soi1 or the planés.
GENERAL CCMDITIONS FCR 11~:PLEkENTATION
Since new species ceuld only be used where rainfall is sufficient,
the tria1 locations were spread out in tke South-Central part of the country
(Sine-Saloum), the Bast (enstern Senegal) and the South (Casarnance) (fig. l),
The pertinent data is ahown in Table 1
TABLE 1
:
:
1
:
:
:
:
:
:
RAINFALL
:
:
:
:
REGION
i TRIAL LOCATIOMS ; Annual average i
CROPS
:
t
for years of
:
SOIL
;
:
:
:experiments (mm) :
:
:
:
t
t
f
:
t
:
t
:
:
:
:
D . . -
. .
I Nioro-du-Rip
:
365
:
t
:
:SINE-SALOUK
: Keur-Samba
:
855
:Ploughed-in grass
:Leached
:
:
t
:fallow
:
:
t
:
:
+roundnuts
:Ferruginous:
:
I
:
:
t
:
:
: Maka
:
811t
rl$illet
:
?
: EASTERN
: Sinthion-Mal&ne t
957
tSorghum
:Tropical :
:SENEGAL
: Cotiary
:
761
sMaize
:soil
:
, :
: Missira
t
635
:Cotton
:
:
:
:
:
:
:
:
i
t
s
$
:
;
. :
t
:
tploughed-in grass :
x
.:
:
:
:fallow
:
2
XUPPER
D
t
:Ciroundnuts
:Leached
:
t
: Velingara
a
768*
#illet
:Ferruginous:
rCASAKANCE
:
t
:Maise
;Tropical :
:
2
t
:Mpland Rice
:with stainst
:
I
x
:Cotton
.
.
:
:
:
:
s
:and contre-:
:
:
I
IPloughed-in grass 8
:
:
:
tfallow
':
-*
ttions
:
:CENTRAL
:
:
:Groundnuts
r(beige
i
:
: Séfa
:
1,170
:Millet
: soil)
:
:CPS.&ANCE
:
:
:Corn
:
:
:
:
:
:Upland Rice
I
:
*: The amount of rainfall is available for only 2 years out of four.
The experimental locations are in the 800-1300 mm zone (average over
many years) with the length of the rainy season varying from north to south
from 105 to 125 days. Bowever, one mwt remenber that the climate is especial-
ly cbaracterized by its extreme irregularity, and conditions are frequently
a long way from the averages- At times, drought spells during the trop season
have seriously disturbed certain crops.
Soils are basically characterized by their sandy to coarse loamy
texture in the Upper horizons. Thc clay content generally increases towards
the south which gives t%e:.: r>ore potential fertifity.
GEOGRAPHICAL LOCATION OF THE VARIOUS TRIALS
i
@LOUGA
A Tilmakha
A Forbotte
BAMBEY C N R A
A Tip
A Boule1
LCIJ I LI\\V?
\\ SINE SALOUM ., r-i.L-
v
b
\\1
I
IL
--
i
0 research station
dh experimental fields
Scale : 1/3.000.000
In 1965, ?Jhen this experimentation started, recommended varicties
were only availahle for groundnuts and cotton ; there was also a local millet
composite for the Casamancee
Local varieties were therefore used in the beginning for millet,,
sorghum and maize, but they were replaced fairly quickly :
- the local uorghum : by 51-69
- the local naize : by a hybrid : BD.5
- the local upland rice ; by 1'1 Kong Pao'(IKP)
R,ES,UI.J!S
YEAR 1 (Table )
Groundnut yields were very good (generally an average of 2,391
kg/ha 1 I those of cotton were fairly good for the Allen 333 variety (on the
average 1,617 kg/ha) .>rith the exception of those obtained in Kissira (poor
phytosanitary control). For sorghum, with the exception of Keur-Samba (Striga
attack) the perfornabces were appropriate, especially from the t&me 51-69 was
used (Nioro, 1966).
The yields of Taise were modest in the beginning and improved with
the use of 3X (Séfa no 2 and Vélingara).
For millet the results are very mediocra except in the Casanance
(Sanio Sefa compositejr
In s-ite of certain accic?ents which Will be taken into account
later on (with millet in particular) on the whole, one cari
consider the re-
sults to be representetive.
YJGR 2 (Table III and fig. 2)
The influence of different preceding crops cari be compared by trop
in each location. For detailed results one should refer to the document al-
ready mentioned above,
Here we have again taken the summary table in which the regions were
separated into two groups : the Sine-Saloum plus Eastern Senegal, and the
Casanance. For cach group, the average yield was determined, and the average
percentage obtained for each preceding trop was calculated as a function of
the result of t:le trop succeeding itself. For example, for grouncfnuts, the
average yield obtained with groundnuts/groundnuts ccnstitutes the base 100.
Figure 2 shows these percentages in a more visual form.
The level of yields is :
- Good for groundnuts in a11 zones, as well as for millet, zaize
ancl upland rice in the Casamance.
- Satisfsctory everywhere for cotton and sorghum.
- ?<ediocre for maize in the Sine-Saloum and Eastern Senegal.
- Poor for ::illet in the South-Central and Eastern zones*
TABLE II
YEhR 1 : hVERkGE YI%LDS OBTAIN5D FOR DIFFEX%D!T
.-
CROPS AT EACH ';XPI?,R,Ii%P?TAL LOCBTIOW
-.
$:
1
:
:
:
t
t
:
:
t Groundnutst Millet: Sorghum: Maize : Paddy :
Cotton t
:
: with patds r grain :
grain : grain : Rice
:Cotton grain:
a
: (kg/ha) :(kg/ha)i kg/ha) :(kg/ha):(kg/ha): (kg/ha) :
:
:
:
?
t
,t
1
t
:
:
:
:
t
:
t
:
:Keur-Samba 1965,...0... :
2 , 4 2 9
: 376" : 482* : 1,470 : - a
1$43 :
!'
:
:
:
:
t
:
t,
:
a
t
:
,f
t
t
:Finthiou-XalGme 1965.e. :
2;oqo
:1,271 :1,5.56 :
833 t - :
a,o17 :
:,
t
:
:
:
t
t
f
t
t
*r
z
t
fpissira 1965....~...... t
2,555
: 933* fl,290 t
939 : - :
700* t
t
t
t
:
:
t
:
:
:
t
:
:
:
:
Ira
t
:Séfa no 1 1965~~.....~~ :
2,970
~2,447 : -
: 2,038 : 1,109 :
t
:
t
t
t
t
t
:
t
t
t
1
:
t
:
t
fyioro-du-Rip 1966.,.... :
2 , 1 4 3
: 846
:2,952
:1,442:
-
a
1,470 t
:
:
t
t
:
t
:
t
:'
:
:
2'
:
C
t
.
:Maka 1966 0*.o...r..*s.. 1
2,371
$1,205
t2,287
:1,661:
-
a
1,876 t
a
:
t
t
t
:
:
1
:
t
:
:
t
:
t
' ;Cotiary 1967...0*0ea..m :
2,2Sr
:1,318 :1,639
$1,346:
-
:
1,969 t
t
t
t
t
t
:
:
t
:
:
1
t
f
t
t
t
tSéfa no 2 1969e*m.,.see t
1 , 7 1 4
$2,352 :
-
: 2,486 : 2+794 i
-
:
t
t
t
t
:
:
2
:c
:
t
t
t
t
1
*:t
t
:Vélingara 1970...C...08 t
2,964
:2,542 : -
t 1,976 t
5 9 4 :
2,053 t
:
t
1 f
t
:
:
:
:
*'Yields considered insufficiento
Qhe coeflici-1-0
-*.ts of variation are relatively high, especially for
millet and mai ze. In spite of this, many effects are significant (5C % of the
rcsults). 1-t depends particularly on the trop (30-33 % for groundnuts, millet
and cotton, 60 $4 for sorghum , 75 % for maiae and 100 % for upland rice), with
certain plants showing thenselves to be more sensitive thon others to the
global impact of t%e plant that precedes it.
If one compares the classification of preceding crops in the two
cases (a11 results, and significant results) one sees that tbere is little
variation ; only th-? differences are more or less prcnounced.
Thus discussion cari be based on the percentages which appear in
the summary table.
TABLE III
RECAFITULkTICN OF RESULTS OBTAINE9 IH YZAR 2
:
:
Year lt
:Floughed:Ground- :
t
:
:
:
f
:Number
:Meaningr
Crop
.
.
:
:in grass:
nuts
: MillM:Sorghum : Maize:Cotton : Rice:Average: of
: *fui :
Yesr 2
:t
: Fallow :
:
:
:
i
:
:
:Results:Results:
fSine-Salqum *
:Kg/ha: 2,050
' 2,123 5 6 i i
:
: 1,978
; 2,195 5 2,220
;a,142 : 2,149 * -
:
:
:7
:
'Eastern Senegal..* : % '
104
:
:
:
:
100 ;
111 ;
112 :
108 ;
109 f
:
:
:
:
:
: aa
GTlOUXbNtfr\\S
;
e-3. )C i.
:Kg/fia i 2 $395
i 2,390
i 2,588 5 -
f2,382
:
-
f2,464:
2,444
'
3
i
2
~CasaoanceZZZZZG~:G
t
:
f7
:
:
:
% i
100 i
100 ;
108 ; -
108
;
-
f
103;
i"o
,z
:
:
:
:
:Sine-Saloum +
iKg/ha':
962
i 1,023
i
881 i
953
$005 i 1,014 ;
-
;
973;
6
'
:
2
:
IfZastern Senegal...
i $4 .i
109
i
116
i
100 i
108
f 114 :
115:: "
i
: :
: :
118 :
MILL&i!
.g
-
-
:Casamance.........
;Kg/ha; 2,511
i 2,544
i 2,286 :
:
i2,607 i
f2%529; 2,495 1 ;
3
:
i
1
f : . 9 . .
:
f %
;
110
,;
111
i
100 i
-
114 .j
-
.:
i 111;
:
:
;to
&F&# Lm
:Sine-Saloum
+
:Kg/ha: 2,120
: 2,325
.i 2,143 .i 1,320
il,974 ; 2,307 ;
-
i 2,034 i
6
;
4
j14
:Eastern Senegal...
i 9/5 i
161
;
176
:
162 i
100
i 150 t
175 f
-
;
t
:
:
t
:
t
:tCS
:Sine-Saloum +
.:
'Kg/haf 1,683
: 1,660
f 1,635 :
969
il,248 : 2,046 ;
-
;1,540;
5
:
4
:
:19
fEastern Senegal*.-
i 96 i
135
;
133
f
131 i
78
i 100 ;
164 f
-
' : : :
:
:
ito
M A I Z E
:
.;Kg/h+ 3,431 ; 4,230
;3,377;
-
;3,538 ' -
jll
~Cas~mance......... :
:
;3,399; 3,595 i 3 : 2
:
:
$2
:
:
:
:
:
:
97 i
120 i
95;
-
f
100
i
-
f
96:
:
:
P*
.
.
.
.
.
P c1 CR
-..
ts \\o
.
.
I
*.
r Lu I . .
l
I
”
çr G . .
l
I.
çr
0 *.
r.
.
.
Groundnuts
The variations are relatively small t on the average, less than
12 %*
- Groundnuts prefer to be planted after cereals or cotton
- Floughed-in grass fallow is not always a good precedent
- It is not advisable to cultivate one trop of groundnuts after
another.
In Eastern Senegal there are marked differences, whereas the dif-
ferences are minitial in the Cssamance*
Hillet
In spite of certain marked differences, and taking into account
what was said previously about the value of thc millet results, there is
reason to be Prudent# It cari be observed, however, tkat the results obtained
with the Séfa Sanio in the Casamance, wheo yields were very satisfactory,
are comparable to those obtained further north in more questionable condi-
tions.
By avoiding the cultivation of millet following millet, one keeps
clear of possible problemse A11 other combinations seem to be viable with a
certain preference for groundnuts as the preceding trop*
Sorghum
The differences are ver-y notioeable. Sorghum is a very poor previous
trop, for itself. In certain locations (Nioro-du-Bip) thore was practically
no yield at ail. It prospers very well after groundnuts and cotton and to a
lesser extent, after ploughed-in grass fallow, millet and maibea
It grows well after groundnuts, millet and ploughed-in grass fallow
(at least in the Sine-Saloum and Eastern Senegal).
Sorghum as a preceding cro.', bas a distinct depressive effect on it.
The reoetition of two corn crops is not recommendea, except in Cen-
tral Casaaance where the soi1 and climatic conditions are appropriate for
eaiae on 2aize succesions.
Differonces are not very pronounced but tendancies are repeated.
Floughed-in grass fallow, groundnuts and corn are good preceding
crogs for cotton.
Sorghum as a preceding trop has a depreasive effoct on cotton, even
if, compared to cotton preceding itself.
c, 0 s z C 4 1
8
0
-Y---
.
?J
5
;
llm
OVER
OVER
RICE
RICE
GRASS FALLOW TURNED
GROUNDNUTS
MILLET
SORGHUM
MIAZE
COTTON
IJPLANO
GRASS FALLOW
TURNDED
GROUNDNUTS
MILLET
SORGHUM
MAIZE
COTTON
UPLAND
_*
“.
e
.
;
.
li
Upland Rice
The differences are very pronounced. It should be planted after
groundnuts, and, if really necessary, after millet. It is better to avoid
ploughed-in grass fallaw or naize and not to cultivate strictly rainfed rice
on the same land for two successive years.
Generally, two spectacular effects cari be seen :
- the general depressive impact of sorghum on a11 crops except for
groundnuts (at least in the rcgions examined, for t':is is not the
case furtber north).
- the favourable impact of groundnuts and cotton on a11 cereals,
but especially on upland rice.
It was hoped that years 3 and 4 would provide important additional
information, concerning in particular, the interactions between crops and their
inflriences on yields obtained in the fourth year.
The detailed stady which was done, did not provide this information.
Nost of the time the findings confirmed the effects obtained in
year 2.
- depressive effects of sorghum which cari have repercussicns beyond
the second year. Dut in this residual effect, it is hard to dis-
tinguish :
(3 Chat cornes from the direct impact of the crops : harvest residue.z,
influence on soi1 (rooting, rhizosphere), parasitism etce..,
. What is a consequence of a depressive effect in year 2 : poor
production with little output does not weaken the environment
and cari thus favour the third year trop.
- interest in groundnuts as a preceding trop for maire and upland
rice.
- lack of interest in having crops succeed themselves.
On the other hand, results obtained elsewhere are confirmed : plou-
ghed-in grass fallow has fairly significant residual effects, and its benefi-
cial impact will be even more important in thc third year than in the second.
The so-called succession effects, hard to define precisely, and which
in a wûy, measure tho conseguences of the interaction of two or three years of
cultivation are rarely significant. Design becomes too comglicated and the
coefficients of variation too high to reveal significant effects. It becomes
impossible te compare a11 the successions since there is not the saze number
of geogra-hical replications for each binomial year 3 - year 4,
From this, however, one ascertains an important peint for the defi-
nition of rotations in each ecological zone : too frequent regetition of the
Saxe crops is undesirable , particularly with sorghum. This then, is not favou-
rable to 2 year rotations.
- 13 ”
Here we have an example :
TABLE IV
CGiiP&RISêJJ CF YIELDS GBTAINED IM YEAX 4 3ITH
A 4,YW LND k B YEI;R RCTkTICN
:
tPloughed-in;
t t
:Ploughed-in:
a
tlirst Year Crop
:
grass
t Cotton: :First Year Crop
:
grass
:Sorghumr
t :
:
fallow :
: I
!
fallow :
:
t
t
:
t 2
:
:
:
;Second Year Crop
:
Maize
tsorghum: :Second Year Crop
t
Xaize
; Cotton:
:
:
t
: .i
t
:
:
t
:
:
: :
:
:
:
:Third Year Crop
:
Cotton
: Cotton: :Third Year Crop
t
Sorgbum
:Sorghumt
f
t
:
: :
t
:
:
:
t
:
1 ?
:
:
:
:Foarth Year Sorghum t
2,832
i 2,109 : tFourth Year Cotton t
3,308
: 2,178 :
:Yield (Kg/ha)
t
t
: rYield (Kg/ha)
:
:
:
A two year sorghum-cotton rotation is
The fourth year yields are systematically
compared to a four year rotation of four
higher in the four-yesr than in the two
different crcpso
year trop rotation.
Two things cari be learned from this experiment. Cne concerns the
useable trop rotations in tle sandy and sandy-clay soils of Sencgal, and by
extension in ecclogical conditions similar to the %:cst-hfrican sahelo-sudanian
region. The otO=er, wbich takes a lesson from the experimental design used,
tries to define a reszarch methodology for trop rotation.
Xe wilL examine ther: in turn.
CCNCLUSIGNS GN CRCF X?TkTION
First of a11 , one must remember the tendancies that appear when
crops succeed tkei;selves.
'fJith the exception of the southern part of Senegal where tbe soi1
texture is izore favcurable,, there is rarely any interest (froc t?ne agronomie
point of view) in repeating the same trop.
Thc techniques and varieties used up until 1972 do not particularly.
allow for continuous cereal crops in tbe south-central and eastern parts of
the country. In Central Casamance, one cari consider the succession of Gaize
after maixe, but strict rainfed rice (different from rice that draws water
from t5r wcter table) does not withstand either maize or upland rice itself
as previous crops0
The two following points result from the preceding te;?dancy.
- Grcundnuts and cotton are excellent precedents for a11 cereals-
upland rice, in particular.
- In a- genernl way, and npnrt from tbo exceptionzl case of sorghum,
excxzined elsewhere, cereals are good precedents for grcundnuts and
- 14 ”
The most reEarkable phenomenon is still the depressive effect that
a sorghum ci-02 has on sandy scil. This impact shows up in the sorghum itself,
then with decreasing intensity in the maize, cotton and millet. In thc south
central and eastern parts of the country, where the problem was studied, sor-
ghum shcws itself tc be a favourable precedent for groundnutsp It seems tbat
this is net the case in the north-central and northzrn parts of Senegal where
the soi.1 is sandier. Depressive effects on the groundnut yield vere obsnrved
in Bambey.
Studies undertaken since, on this subject havc allowed us to propose
and verify certain explanations. The decomposition of sorghum harvest residues
(rcots in particular) releases toxic substances which decompose too slowly in
sandy soils with little biological activity. 'Empirically, this phenomenon was
controlled by sy-rreading czanure, which activates the microbes and contains
certain micro-organisms necessary for the rapid elimination of substancesr
Systematic scientific exploration, thanks to the CM3S team in Nancy,
directed by Y. DOI~I':Bl3JES, has led to another solution, aimed at inoculating
a micro-organisn (Trichoderma) which attends to this rapid elimination. This
solution is being tested in Bambey at the present timer
From these geceral rules, and the few observed effects cf.crop
-
successions, two principles mn be devised forcrop rotations designs t
- Consecutive replanting of the same trop must be avoided.
- One must orient oneself towards the use of compatible trop bino-
mials which Will be paired as a function of sacio-economic ne-
cessities and objectives. These rules are even more strict as the
production levels go UP*
From the preceding data, it was possible tc draw up a compatibilit:.
table for crops, a table which is viable in the ecological conditions defined
above, and w?en a certain number of intensive techniques are used te obtain
a given level cf productivity. If one wanted tc apply this table ta tradi-
tional crops with low level yields, there is a good chance that the diffe-
ronces wculd net be apparent, and it is even probable t"at one wculd only
observe a few differences between the various preceding crops.
- 15 -
Five grades were used :
++
very good
good
Pocr
xxxxxxxxxx
Very poor
CRC17 COPPkTIE3ILIT~,, IJ'J CRCF RCYATION
:Plcuçhed-in grass:
:
, Groundnutsi
fallow .
EilletfSorghum
:
:
:
2
:
:
:
I
!
:Groundnutsao.. i /////M/////// i xxxxxxxxx
xxxxxxxxx
f + i + i + : +
:
:
:
:
:
f
t
:
I
z
:
:
t
f
1
:
:
%Millet
xxxxxxx :
.* . . . . . . :
?
+
:
:
+;+: 4
:
:
t
t
: xxxxxxx ;
:
2
2
:
:
:
t
:
:
:Sorghum....... :
xxxxxxx t
II
:
:
:
++
r
3
:
:
:
xxxxxxx :
:
+4 :
Zero :
:
:
:
8
8
:
:
:
*+?4Q5~Q.rrbaaor~ :
:
:
. ,
:
:
:
1
:Cotton
:
+
:
+
XXXxXXX~XXXXXXX*
t
..* . . . . .
I
* r xxxxxx~xxxxxxxx~
:
:
t
t xxxxxxx~xxxxxxx~
t XXXXXx;XXXxxxXX~
:
tUpland Rice*.. : /////////////// f ++ i + i zero i////i *+ :~:~~~~t
:
:
z
:
D
t
3
:
:
(1) except in Casamsrnce
(2) except in Central Casamnnce.
RBS~ARCB DFSIGPT
Inconveniences
They should be reviewed here briefly.
- Too large an area. It depends to a great extent on the number of
different crops, for it is hard to have plots smaller than 36 m2, and spaces
between the strigs less than 2 metres wideo Therefore it is necessary to use
an adequate research structure and mechaniae larming operations as nauch as
possible,
m burat ion*
The findings from yearb 3 an& 4 are minimal comparik to those ob-
taSned in yaar 2. The only interest iS to be able to confirm a limited number
of effects (limited by the number of 4rops tested in years 3 and 4) through
replications in time.
? ? ?
- Very high coefficients ok variation in years 3 and 6,. In year 2,
it CB~ go below 20 % and eometimes even close to 10 7; for a11 arops.
During this seoond year, the coefficient of variation is, of course,
dependent on aleatory factors such as climate or heterogeneity of the land.
It depends es-ecially on the use of recommended varieties and good farming
techniques.
The best trustworthy and most often significant results ?I@re obtei-
ned when advenced technology wcrs appkied throughaut, and yields ?rere hiçh.
This confiras :whaC wcl3 3:rid onrlier t one C=~I? only hopo for siguifiasat effects
with preceding crops, when advanced technology is used.
:Je do not think that thé problem is interesting tc study for tre-
ditional or slightly iapraved cultivatkon.
Findings and &ro:sositions
In S?ite of their imperfections, these exper5ments still provi6ed a
basis for the T:ain trop rotations tested on a large scale in the south-cental,
eastern and sçuthern parts of Senegal. There, they Canfirmed their :rorth.
Fron the findings provided Sy this design, w0 Uan enumerate certain
pr&lFI;iplea which we think should guide research on crlop rotation t
.
rr the existing design should be ieduced $0 %wo yeat*&$ uith the u3e
of aeries s-La gered in time, &n d
or er to bave a yearip &ÊeCt, This is possi.
ble in tww waye t
?? deries exatily the sa--e I
staggered b a year which give complete
results in three ge8Fe a but neeessitafe kwo siFpez?i*asats on the
jand in yea.r 2,
??Staggered 2 year aeu"iea~ whkb cari be different ta the ertent
that series X. SS talcen 5M.o acéount before eplbarking on Saries
EX.
Qbviously it is preferable to hare eight replicatione each tine
which re?resent a good safe numbera Fos+ whaterer reasond one shuuld not have
less than 3ix.
- Varieties and farming techniques should be thoroughly adapted to
each trop, yuaranteeing (except for catastrophic climatic conditions) the
success of the trop at tke technological level in question.
- In Senegai, ocly the yield characteristics -,+Jere retained. Cer-
tainly, resesrch must continue, and a more conplete follow-up than the one
done would be desirable.
Particularly in the zones with irregular rainfall, water problems
are important and it :rould be useful to examine the water resx-ves after each
trop. Tha nreccding trop factor Flays an important role in tha case of drought
spells during the croi season, eapeciallp if the crops differ in length of
growing cycles.
” 17 -
The problem of soi1 nitrogen seeU very important. It is probably
the source of many phenomena and particularly of interest in the alternation
of cereals and legumes, but one must not lose sight of the importance of micro-
biological phenomena : the case of sorghum is the most remarkable.
It is not our purpase to prooide a restrictive list, for numerous
aspects cari be considered. This follow-up cari permit a more rapid grasp of the
global phenomena thzt were registered in the preceding experiment. Koreover,
one cari establish it in Series II as a function of the hypotheses put forth
in Series 1.
As shown, research on preceding crops led
to the establishment of
a trop compatibility table for each ecologîcal zone* Binomials are defined
from these tables. They Will constitute the baaie for possible trop rotations,
but also for studies on possible farming systemse If two-year rotations are
not advisable (at least in the frarrevfork of the previous studies), trop bino-
mia&s provide more flexibility to study integrated systems, including a11
eonstraints, particularly those of the trop ëalendar.
GEPJElUL cOIdcLusIc?IW
From the analysis of results of an experiment on preeeding crops,
carried out for eight years in Senegal, two conclusions can be Urawn.
- Cne concerns Senegal directly. It led ta the main trop rotations
recommended in the south-central,
eastern and soukhorn PEUT~A of Senegal. Let
us briefly review them here*
Sine-Saloum snd Bastern Se,negal :
Year 1 : Ploughed-in grass fallow or maiee or early millet
Year 2 t Cotton or groundnuts
year 3 : Sorghum
Year 4 : Groundnuts
Upper Casamance :
Year 1 : Ploughed-in grass fallow or maize
Year 2 z Cotton or groundnuta
Year 3 : Uplend rice or "Sanie" (long maturity cycle, photosensitive)
millet
Year 4 : Groundnuts
One cari also have Sanie millet in Year 5.
Central-Casamance :
tio possibles cases t
. For farmers who grow millct t
- maire or upland rice
- upland rice or maize
- graundnuts
- millet
- 18 -
E
For farmers who do not grow millet, 3 solutions z
- maire - groundnuts - upland rice - groundnuts
- maize - groundnuts - upland rice
- grcundnuts ; X/2; upland riee - l/zi maire ; 1/2 maire - 1/2 upland
rice.
Thoroughly tested in the studies called %tructure d'exploitation11 (1)
these rotations proved to be valuable.
- The other led ta the definition of a methodology for the establi-
shment of trop rotations adapted to each ecologieal zone*
The purpose in each case is to define a certain number of compatible
trop binomials. Besides these binomials, one cari make the most intcresting
combinations as a function of different socio-eeonomic imperatives, on the
condition that the principle of compatibility is always respectede
These binomials are drawn from a çompatibility table established
after a simplified experiment on preceding crops. Hpwever, one must not lose
sight of tke fact that a technological level must be set at the beginning,
and that the compatibility table is only viable up to this level. Sxperience
seems to show that the influence of one trop on another is particularly sensi-
tive to intermediary levels, when one has succeeded in increasing production
without controllPng a11 the agronomie factors (of the environment).
By setting the basis of the study right at the beginning, the com-
patibility table cari be abtainod after three years, on the condition that an
adequate experimental network has bcen usede Fer each zone, it is preferable to
have a thorough experimentdtion with enough replications in one location, rather
than to increase the tria1 locations to the detriment of accuracy.
fi) transiator's note : Nodel farms operated by hired 'laborers under direct
supervision of agronomists, where t-ime required for eompleting each cul-
turai. operation and labors peak are thoroughly measured on an actual scaleo
- 19 ”
BIBLIOGRAFHY
CHARRZAU (C), NICOU (R) - L'amélioration du profil cultural dans les sols
sableux et sabla-argileux de la zone tropicale sèche Ouest-
Africaine et ses consequences agronomiques. (Improvetnent of
the trop profile in the sandy and sandy-clay soi18 of the dry
West-African tropical zone and its agronomie consequences)r
kgron. Trop., 19711 no 2, 5, 9, Il.
CHOPART (JrL.), NICOU (R) - Effet dépressif de cultures &Pétées de sorgho
dans les sols sableux du Sénégal, Premiers essais d'explica-
tien. (The depressive effect of repeated sorghum crops on the
sandy soils of Senegal. First explanatory experiments)
kfrican Soila Vole XVII, 1/1973.
NICOU (R) - Bilan de huit années d'étude des précédents culttiraux au Sénégal.
(Report of eight years of research on preceding crops in Senegal).
Xeroxed Iloc. IRAT/ISRA - CNRA, Bambey (Senegal).
*
IN SEN&AL f
RESULTS AND METHODS
by
R,, NICOU
Translated from French t Wtude de Successions Culturales
Résultat* et Méthodesa
L*Agroncmie Tropicale, 1978 No 1
Janvier - Hars, pp.' 51-62
ICT 80-3
J3y R6 P?Icou*
sui .:mPi : In cwder ~CG study the place of new crcps (caize, uplahd rice,
cottonr in crap rotation, experiments were done in Senegal Setween 1965
and 1972 in eight geographical locations in the 800-1300 mm zone. In the
design used, ~11 of the crops could follow one another, and after four years
many combinations could be obtainede
Anafysis of the results led to two conclusions w!:ich are ir,portant
beyond L:;e geograDhica1 fraF?ework of Senegal" hy drawing up a table for trop
compatibility, hinonîal cratps cari beedefineds Sy pairing theso binosials ohor
sen according to socio-economic conditions, a lot of agronorically viable
combinations cari Se obtained. Thus for Senegal, a certain number of rotation
types were defined for the ecological areas ccncerned.
hs for nethodology, t%e design cari be sitnplified by t:eaxs of se-
rialization, thus giving faster and more precise results irrespective of the
yearo
These experiments called "preceding crops" Will not achieve mea-
ningful results if they are carried out in traditional cultivating condi-
tions, In fac%, tke differences are too minimal to develop a cotxpatibility
tabler %erefore, there iq good reason for using recommended verieties, well-
known cropt?ing tec'iniques and for obtaining an adequate level of intensifi-
cation.
IT3Y ‘P233S
: cmj? rotation, experizeRta1 design, ecology, preceding crops)
cropping techr,iques, intensification, Senegal.
In 1965, the need to diversify Senegalese agriculture and in par-
ticular, ixroase cereal production, led I?&T-Senegal to consider research
on crcq rotation, other than that already recoaKended for conmon use, and
which las largely based on the alternation of groundnuts, millet and/or sor-
ghuzr
.
At thc beginning, the new crops to be introduced were cotton and
maizer i. bit later caoe upland rice.
It was decided to do an experiment with the sims :
- of determining in what way and where the new crops could be
introduced in the rotation.
- of detarmining what are the Sest poesible successions aRong t?e
different crops eligible for rotation, with the successive cul-
tivntion of cereal upon cereal being exarzined in particulero
* NICOU (R) : S:UT Resear&h Agronomist, National Agronomie Bescarch Centre
in Sanbey (ISRA - Senegal).
Fr'esently at 1211.1, BP ho 5035, 34032 Montpellier Cedex, France.
Taking into account the demands of the new crops, and setting out
is not possible
from tbe principle $hat diversification with these .species
conducted in
unless one bas a>?propriate techniques, tlese experinents were
intensified agrieultural conditions :
lrdeep" (15 - 20 cm) soi1 tillage
turning under harvest residues whenever Possible
he,avy z:ineral fertilization with initial addition of rock ghos-
'. .
phatos to rcduce thezphosghorus dèficiency, atid *tien addition of.;
comple-entary fertilizers specific to each trop, allowing COE-=
pensation for minera1 exports ;
the use of recoimended varieties when available
t%e best-adasted cropping techniques.
Thc design used was to globally reveal trop compatibility, in or-
der to define useable binomials for trop rotation. Also, by the length of
the experiments, one hoped to show the beginning of evolving phenomena for
continous cropping of the same species.
Research lasted for eight years, from 1965-1972, during which some
experiments were done in several locations, Some had to be abandoned in ?zid-
stream.
A camplete and dctailed plan was drawn up for reference*. From tlne
only results %ich provided a minimum guarantee for trop development.
The current docu:?ezt aims to be more condecsed, by exceeding the
simple results useabla in Senegal, it tries to define a methodological
approach for the study of trop rotation.
The design was adapted from a plan used at the time by fZ.kT in
Dahomey (N. %RTS). The experiment includes eight replications, The number and
nature of crops csn vary clepending on the zone, but the principle rezains t:ie
same.
Year 1. For each replication, n strips 6 metres wide, and 2 metres apart are
cultivated : n represents the number of crops, The length of the strips de-
pends on the nueber of plants cultivated in the second yeara Strips are rahdo-
mized.
Year 2. For each re^lication , p strips 6 metres wide and 5, netres apart are
planted in perpecdicular fasbion t p represents the nuEîber of crops x?ich
cari be different Pro:. ~3. Xt is harvested in plots of 6m x 6m :yhere strips of
the first and second year intersect. T~US, there are n x p successions. The
main information concern the global effects of n preceding crops for each
second year trop-
* Report on eight years of resesrch on preceding crops in Senegal, by RI NICCU.
Himeographed copies available froc ISBA - CNRA in Bambey, Senegal.
-3 -
Year 3. Four replicntions are planted entirely with cereal and four with
cotton or groundnuts. Planting is done in the direction of the first year
strips, respecting the 2 :?etre spaces between rows* The harvest is done in
6m x 6m plots where the strips intersect.
Year 4, The eight re-lications are uniformly cultivated in the second Year
strip direction, and always harvested on the same 6m x 6m plots.
Thus, one obtains 2 x n x p rotations teated with the saxe plant,
which may vary from one geographical location to another. Thus it is possi-
ble to have up tc four years of cereals or cotton, for example.
The 6m x 6m = 36 ma net plot dimension is considered to be the nini-
mum sirie. It is difficult to znake it larger, but that obviously depends on
the nunber of crops* If n = 6 and
= 5 (which was fie-uently the case) each
block rr:easures 38 x 46 m = 1,748 ms , which represents 1.4 ha for eight repli-
cations. Therefore, with interblocks, it is necessary to find a 2 ha homoge-
nous plot of land.
The direction fur statistical analysis are :
- in year %, analysis is possible on the eight replications of
each trop- It is done using the following scheize I
degrees of freedon (d.Zq)
BlOClU
7
Prcceding Crops
n - l
Error
7 (n - 1)
In yecrs 3 and 4, there are two analyses pcr site, with four
replicatiois for each (as a function of the third year trop).
d.f.
Blocks
3
Column Sffects
n - l
Error (a)
3 #n.- 1)
Line Effects
P-1
Error (b)
3 hJ - 1)
Successions
(n - 1) (p - 1)
Error (c)
3 h -11 (p - 1)
Lastly, it should be pointed out that for t"ris first approach
to the problem of trop successions, only yield components 7rere taken into
account. Few observations were done on t%e soi1 or the planés.
GENERAL CCMDITIONS FCR 11~:PLEkENTATION
Since new species ceuld only be used where rainfall is sufficient,
the tria1 locations were spread out in tke South-Central part of the country
(Sine-Saloum), the Bast (enstern Senegal) and the South (Casarnance) (fig. l),
The pertinent data is ahown in Table 1
TABLE 1
:
:
1
:
:
:
:
:
:
RAINFALL
:
:
:
:
REGION
i TRIAL LOCATIOMS ; Annual average i
CROPS
:
t
for years of
:
SOIL
;
:
:
:experiments (mm) :
:
:
:
t
t
f
:
t
:
t
:
:
:
:
D . . -
. .
I Nioro-du-Rip
:
365
:
t
:
:SINE-SALOUK
: Keur-Samba
:
855
:Ploughed-in grass
:Leached
:
:
t
:fallow
:
:
t
:
:
+roundnuts
:Ferruginous:
:
I
:
:
t
:
:
: Maka
:
811t
rl$illet
:
?
: EASTERN
: Sinthion-Mal&ne t
957
tSorghum
:Tropical :
:SENEGAL
: Cotiary
:
761
sMaize
:soil
:
, :
: Missira
t
635
:Cotton
:
:
:
:
:
:
:
:
i
t
s
$
:
;
. :
t
:
tploughed-in grass :
x
.:
:
:
:fallow
:
2
XUPPER
D
t
:Ciroundnuts
:Leached
:
t
: Velingara
a
768*
#illet
:Ferruginous:
rCASAKANCE
:
t
:Maise
;Tropical :
:
2
t
:Mpland Rice
:with stainst
:
I
x
:Cotton
.
.
:
:
:
:
s
:and contre-:
:
:
I
IPloughed-in grass 8
:
:
:
tfallow
':
-*
ttions
:
:CENTRAL
:
:
:Groundnuts
r(beige
i
:
: Séfa
:
1,170
:Millet
: soil)
:
:CPS.&ANCE
:
:
:Corn
:
:
:
:
:
:Upland Rice
I
:
*: The amount of rainfall is available for only 2 years out of four.
The experimental locations are in the 800-1300 mm zone (average over
many years) with the length of the rainy season varying from north to south
from 105 to 125 days. Bowever, one mwt remenber that the climate is especial-
ly cbaracterized by its extreme irregularity, and conditions are frequently
a long way from the averages- At times, drought spells during the trop season
have seriously disturbed certain crops.
Soils are basically characterized by their sandy to coarse loamy
texture in the Upper horizons. Thc clay content generally increases towards
the south which gives t%e:.: r>ore potential fertifity.
GEOGRAPHICAL LOCATION OF THE VARIOUS TRIALS
i
@LOUGA
A Tilmakha
A Forbotte
BAMBEY C N R A
A Tip
A Boule1
LCIJ I LI\\V?
\\ SINE SALOUM ., r-i.L-
v
b
\\1
I
IL
--
i
0 research station
dh experimental fields
Scale : 1/3.000.000
In 1965, ?Jhen this experimentation started, recommended varicties
were only availahle for groundnuts and cotton ; there was also a local millet
composite for the Casamancee
Local varieties were therefore used in the beginning for millet,,
sorghum and maize, but they were replaced fairly quickly :
- the local uorghum : by 51-69
- the local naize : by a hybrid : BD.5
- the local upland rice ; by 1'1 Kong Pao'(IKP)
R,ES,UI.J!S
YEAR 1 (Table )
Groundnut yields were very good (generally an average of 2,391
kg/ha 1 I those of cotton were fairly good for the Allen 333 variety (on the
average 1,617 kg/ha) .>rith the exception of those obtained in Kissira (poor
phytosanitary control). For sorghum, with the exception of Keur-Samba (Striga
attack) the perfornabces were appropriate, especially from the t&me 51-69 was
used (Nioro, 1966).
The yields of Taise were modest in the beginning and improved with
the use of 3X (Séfa no 2 and Vélingara).
For millet the results are very mediocra except in the Casanance
(Sanio Sefa compositejr
In s-ite of certain accic?ents which Will be taken into account
later on (with millet in particular) on the whole, one cari
consider the re-
sults to be representetive.
YJGR 2 (Table III and fig. 2)
The influence of different preceding crops cari be compared by trop
in each location. For detailed results one should refer to the document al-
ready mentioned above,
Here we have again taken the summary table in which the regions were
separated into two groups : the Sine-Saloum plus Eastern Senegal, and the
Casanance. For cach group, the average yield was determined, and the average
percentage obtained for each preceding trop was calculated as a function of
the result of t:le trop succeeding itself. For example, for grouncfnuts, the
average yield obtained with groundnuts/groundnuts ccnstitutes the base 100.
Figure 2 shows these percentages in a more visual form.
The level of yields is :
- Good for groundnuts in a11 zones, as well as for millet, zaize
ancl upland rice in the Casamance.
- Satisfsctory everywhere for cotton and sorghum.
- ?<ediocre for maize in the Sine-Saloum and Eastern Senegal.
- Poor for ::illet in the South-Central and Eastern zones*
TABLE II
YEhR 1 : hVERkGE YI%LDS OBTAIN5D FOR DIFFEX%D!T
.-
CROPS AT EACH ';XPI?,R,Ii%P?TAL LOCBTIOW
-.
$:
1
:
:
:
t
t
:
:
t Groundnutst Millet: Sorghum: Maize : Paddy :
Cotton t
:
: with patds r grain :
grain : grain : Rice
:Cotton grain:
a
: (kg/ha) :(kg/ha)i kg/ha) :(kg/ha):(kg/ha): (kg/ha) :
:
:
:
?
t
,t
1
t
:
:
:
:
t
:
t
:
:Keur-Samba 1965,...0... :
2 , 4 2 9
: 376" : 482* : 1,470 : - a
1$43 :
!'
:
:
:
:
t
:
t,
:
a
t
:
,f
t
t
:Finthiou-XalGme 1965.e. :
2;oqo
:1,271 :1,5.56 :
833 t - :
a,o17 :
:,
t
:
:
:
t
t
f
t
t
*r
z
t
fpissira 1965....~...... t
2,555
: 933* fl,290 t
939 : - :
700* t
t
t
t
:
:
t
:
:
:
t
:
:
:
:
Ira
t
:Séfa no 1 1965~~.....~~ :
2,970
~2,447 : -
: 2,038 : 1,109 :
t
:
t
t
t
t
t
:
t
t
t
1
:
t
:
t
fyioro-du-Rip 1966.,.... :
2 , 1 4 3
: 846
:2,952
:1,442:
-
a
1,470 t
:
:
t
t
:
t
:
t
:'
:
:
2'
:
C
t
.
:Maka 1966 0*.o...r..*s.. 1
2,371
$1,205
t2,287
:1,661:
-
a
1,876 t
a
:
t
t
t
:
:
1
:
t
:
:
t
:
t
' ;Cotiary 1967...0*0ea..m :
2,2Sr
:1,318 :1,639
$1,346:
-
:
1,969 t
t
t
t
t
t
:
:
t
:
:
1
t
f
t
t
t
tSéfa no 2 1969e*m.,.see t
1 , 7 1 4
$2,352 :
-
: 2,486 : 2+794 i
-
:
t
t
t
t
:
:
2
:c
:
t
t
t
t
1
*:t
t
:Vélingara 1970...C...08 t
2,964
:2,542 : -
t 1,976 t
5 9 4 :
2,053 t
:
t
1 f
t
:
:
:
:
*'Yields considered insufficiento
Qhe coeflici-1-0
-*.ts of variation are relatively high, especially for
millet and mai ze. In spite of this, many effects are significant (5C % of the
rcsults). 1-t depends particularly on the trop (30-33 % for groundnuts, millet
and cotton, 60 $4 for sorghum , 75 % for maiae and 100 % for upland rice), with
certain plants showing thenselves to be more sensitive thon others to the
global impact of t%e plant that precedes it.
If one compares the classification of preceding crops in the two
cases (a11 results, and significant results) one sees that tbere is little
variation ; only th-? differences are more or less prcnounced.
Thus discussion cari be based on the percentages which appear in
the summary table.
TABLE III
RECAFITULkTICN OF RESULTS OBTAINE9 IH YZAR 2
:
:
Year lt
:Floughed:Ground- :
t
:
:
:
f
:Number
:Meaningr
Crop
.
.
:
:in grass:
nuts
: MillM:Sorghum : Maize:Cotton : Rice:Average: of
: *fui :
Yesr 2
:t
: Fallow :
:
:
:
i
:
:
:Results:Results:
fSine-Salqum *
:Kg/ha: 2,050
' 2,123 5 6 i i
:
: 1,978
; 2,195 5 2,220
;a,142 : 2,149 * -
:
:
:7
:
'Eastern Senegal..* : % '
104
:
:
:
:
100 ;
111 ;
112 :
108 ;
109 f
:
:
:
:
:
: aa
GTlOUXbNtfr\\S
;
e-3. )C i.
:Kg/fia i 2 $395
i 2,390
i 2,588 5 -
f2,382
:
-
f2,464:
2,444
'
3
i
2
~CasaoanceZZZZZG~:G
t
:
f7
:
:
:
% i
100 i
100 ;
108 ; -
108
;
-
f
103;
i"o
,z
:
:
:
:
:Sine-Saloum +
iKg/ha':
962
i 1,023
i
881 i
953
$005 i 1,014 ;
-
;
973;
6
'
:
2
:
IfZastern Senegal...
i $4 .i
109
i
116
i
100 i
108
f 114 :
115:: "
i
: :
: :
118 :
MILL&i!
.g
-
-
:Casamance.........
;Kg/ha; 2,511
i 2,544
i 2,286 :
:
i2,607 i
f2%529; 2,495 1 ;
3
:
i
1
f : . 9 . .
:
f %
;
110
,;
111
i
100 i
-
114 .j
-
.:
i 111;
:
:
;to
&F&# Lm
:Sine-Saloum
+
:Kg/ha: 2,120
: 2,325
.i 2,143 .i 1,320
il,974 ; 2,307 ;
-
i 2,034 i
6
;
4
j14
:Eastern Senegal...
i 9/5 i
161
;
176
:
162 i
100
i 150 t
175 f
-
;
t
:
:
t
:
t
:tCS
:Sine-Saloum +
.:
'Kg/haf 1,683
: 1,660
f 1,635 :
969
il,248 : 2,046 ;
-
;1,540;
5
:
4
:
:19
fEastern Senegal*.-
i 96 i
135
;
133
f
131 i
78
i 100 ;
164 f
-
' : : :
:
:
ito
M A I Z E
:
.;Kg/h+ 3,431 ; 4,230
;3,377;
-
;3,538 ' -
jll
~Cas~mance......... :
:
;3,399; 3,595 i 3 : 2
:
:
$2
:
:
:
:
:
:
97 i
120 i
95;
-
f
100
i
-
f
96:
:
:
P*
.
.
.
.
.
P c1 CR
-..
ts \\o
.
.
I
*.
r Lu I . .
l
I
”
çr G . .
l
I.
çr
0 *.
r.
.
.
Groundnuts
The variations are relatively small t on the average, less than
12 %*
- Groundnuts prefer to be planted after cereals or cotton
- Floughed-in grass fallow is not always a good precedent
- It is not advisable to cultivate one trop of groundnuts after
another.
In Eastern Senegal there are marked differences, whereas the dif-
ferences are minitial in the Cssamance*
Hillet
In spite of certain marked differences, and taking into account
what was said previously about the value of thc millet results, there is
reason to be Prudent# It cari be observed, however, tkat the results obtained
with the Séfa Sanio in the Casamance, wheo yields were very satisfactory,
are comparable to those obtained further north in more questionable condi-
tions.
By avoiding the cultivation of millet following millet, one keeps
clear of possible problemse A11 other combinations seem to be viable with a
certain preference for groundnuts as the preceding trop*
Sorghum
The differences are ver-y notioeable. Sorghum is a very poor previous
trop, for itself. In certain locations (Nioro-du-Bip) thore was practically
no yield at ail. It prospers very well after groundnuts and cotton and to a
lesser extent, after ploughed-in grass fallow, millet and maibea
It grows well after groundnuts, millet and ploughed-in grass fallow
(at least in the Sine-Saloum and Eastern Senegal).
Sorghum as a preceding cro.', bas a distinct depressive effect on it.
The reoetition of two corn crops is not recommendea, except in Cen-
tral Casaaance where the soi1 and climatic conditions are appropriate for
eaiae on 2aize succesions.
Differonces are not very pronounced but tendancies are repeated.
Floughed-in grass fallow, groundnuts and corn are good preceding
crogs for cotton.
Sorghum as a preceding trop has a depreasive effoct on cotton, even
if, compared to cotton preceding itself.
c, 0 s z C 4 1
8
0
-Y---
.
?J
5
;
llm
OVER
OVER
RICE
RICE
GRASS FALLOW TURNED
GROUNDNUTS
MILLET
SORGHUM
MIAZE
COTTON
IJPLANO
GRASS FALLOW
TURNDED
GROUNDNUTS
MILLET
SORGHUM
MAIZE
COTTON
UPLAND
_*
“.
e
.
;
.
li
Upland Rice
The differences are very pronounced. It should be planted after
groundnuts, and, if really necessary, after millet. It is better to avoid
ploughed-in grass fallaw or naize and not to cultivate strictly rainfed rice
on the same land for two successive years.
Generally, two spectacular effects cari be seen :
- the general depressive impact of sorghum on a11 crops except for
groundnuts (at least in the rcgions examined, for t':is is not the
case furtber north).
- the favourable impact of groundnuts and cotton on a11 cereals,
but especially on upland rice.
It was hoped that years 3 and 4 would provide important additional
information, concerning in particular, the interactions between crops and their
inflriences on yields obtained in the fourth year.
The detailed stady which was done, did not provide this information.
Nost of the time the findings confirmed the effects obtained in
year 2.
- depressive effects of sorghum which cari have repercussicns beyond
the second year. Dut in this residual effect, it is hard to dis-
tinguish :
(3 Chat cornes from the direct impact of the crops : harvest residue.z,
influence on soi1 (rooting, rhizosphere), parasitism etce..,
. What is a consequence of a depressive effect in year 2 : poor
production with little output does not weaken the environment
and cari thus favour the third year trop.
- interest in groundnuts as a preceding trop for maire and upland
rice.
- lack of interest in having crops succeed themselves.
On the other hand, results obtained elsewhere are confirmed : plou-
ghed-in grass fallow has fairly significant residual effects, and its benefi-
cial impact will be even more important in thc third year than in the second.
The so-called succession effects, hard to define precisely, and which
in a wûy, measure tho conseguences of the interaction of two or three years of
cultivation are rarely significant. Design becomes too comglicated and the
coefficients of variation too high to reveal significant effects. It becomes
impossible te compare a11 the successions since there is not the saze number
of geogra-hical replications for each binomial year 3 - year 4,
From this, however, one ascertains an important peint for the defi-
nition of rotations in each ecological zone : too frequent regetition of the
Saxe crops is undesirable , particularly with sorghum. This then, is not favou-
rable to 2 year rotations.
- 13 ”
Here we have an example :
TABLE IV
CGiiP&RISêJJ CF YIELDS GBTAINED IM YEAX 4 3ITH
A 4,YW LND k B YEI;R RCTkTICN
:
tPloughed-in;
t t
:Ploughed-in:
a
tlirst Year Crop
:
grass
t Cotton: :First Year Crop
:
grass
:Sorghumr
t :
:
fallow :
: I
!
fallow :
:
t
t
:
t 2
:
:
:
;Second Year Crop
:
Maize
tsorghum: :Second Year Crop
t
Xaize
; Cotton:
:
:
t
: .i
t
:
:
t
:
:
: :
:
:
:
:Third Year Crop
:
Cotton
: Cotton: :Third Year Crop
t
Sorgbum
:Sorghumt
f
t
:
: :
t
:
:
:
t
:
1 ?
:
:
:
:Foarth Year Sorghum t
2,832
i 2,109 : tFourth Year Cotton t
3,308
: 2,178 :
:Yield (Kg/ha)
t
t
: rYield (Kg/ha)
:
:
:
A two year sorghum-cotton rotation is
The fourth year yields are systematically
compared to a four year rotation of four
higher in the four-yesr than in the two
different crcpso
year trop rotation.
Two things cari be learned from this experiment. Cne concerns the
useable trop rotations in tle sandy and sandy-clay soils of Sencgal, and by
extension in ecclogical conditions similar to the %:cst-hfrican sahelo-sudanian
region. The otO=er, wbich takes a lesson from the experimental design used,
tries to define a reszarch methodology for trop rotation.
Xe wilL examine ther: in turn.
CCNCLUSIGNS GN CRCF X?TkTION
First of a11 , one must remember the tendancies that appear when
crops succeed tkei;selves.
'fJith the exception of the southern part of Senegal where tbe soi1
texture is izore favcurable,, there is rarely any interest (froc t?ne agronomie
point of view) in repeating the same trop.
Thc techniques and varieties used up until 1972 do not particularly.
allow for continuous cereal crops in tbe south-central and eastern parts of
the country. In Central Casamance, one cari consider the succession of Gaize
after maixe, but strict rainfed rice (different from rice that draws water
from t5r wcter table) does not withstand either maize or upland rice itself
as previous crops0
The two following points result from the preceding te;?dancy.
- Grcundnuts and cotton are excellent precedents for a11 cereals-
upland rice, in particular.
- In a- genernl way, and npnrt from tbo exceptionzl case of sorghum,
excxzined elsewhere, cereals are good precedents for grcundnuts and
- 14 ”
The most reEarkable phenomenon is still the depressive effect that
a sorghum ci-02 has on sandy scil. This impact shows up in the sorghum itself,
then with decreasing intensity in the maize, cotton and millet. In thc south
central and eastern parts of the country, where the problem was studied, sor-
ghum shcws itself tc be a favourable precedent for groundnutsp It seems tbat
this is net the case in the north-central and northzrn parts of Senegal where
the soi.1 is sandier. Depressive effects on the groundnut yield vere obsnrved
in Bambey.
Studies undertaken since, on this subject havc allowed us to propose
and verify certain explanations. The decomposition of sorghum harvest residues
(rcots in particular) releases toxic substances which decompose too slowly in
sandy soils with little biological activity. 'Empirically, this phenomenon was
controlled by sy-rreading czanure, which activates the microbes and contains
certain micro-organisms necessary for the rapid elimination of substancesr
Systematic scientific exploration, thanks to the CM3S team in Nancy,
directed by Y. DOI~I':Bl3JES, has led to another solution, aimed at inoculating
a micro-organisn (Trichoderma) which attends to this rapid elimination. This
solution is being tested in Bambey at the present timer
From these geceral rules, and the few observed effects cf.crop
-
successions, two principles mn be devised forcrop rotations designs t
- Consecutive replanting of the same trop must be avoided.
- One must orient oneself towards the use of compatible trop bino-
mials which Will be paired as a function of sacio-economic ne-
cessities and objectives. These rules are even more strict as the
production levels go UP*
From the preceding data, it was possible tc draw up a compatibilit:.
table for crops, a table which is viable in the ecological conditions defined
above, and w?en a certain number of intensive techniques are used te obtain
a given level cf productivity. If one wanted tc apply this table ta tradi-
tional crops with low level yields, there is a good chance that the diffe-
ronces wculd net be apparent, and it is even probable t"at one wculd only
observe a few differences between the various preceding crops.
- 15 -
Five grades were used :
++
very good
good
Pocr
xxxxxxxxxx
Very poor
CRC17 COPPkTIE3ILIT~,, IJ'J CRCF RCYATION
:Plcuçhed-in grass:
:
, Groundnutsi
fallow .
EilletfSorghum
:
:
:
2
:
:
:
I
!
:Groundnutsao.. i /////M/////// i xxxxxxxxx
xxxxxxxxx
f + i + i + : +
:
:
:
:
:
f
t
:
I
z
:
:
t
f
1
:
:
%Millet
xxxxxxx :
.* . . . . . . :
?
+
:
:
+;+: 4
:
:
t
t
: xxxxxxx ;
:
2
2
:
:
:
t
:
:
:Sorghum....... :
xxxxxxx t
II
:
:
:
++
r
3
:
:
:
xxxxxxx :
:
+4 :
Zero :
:
:
:
8
8
:
:
:
*+?4Q5~Q.rrbaaor~ :
:
:
. ,
:
:
:
1
:Cotton
:
+
:
+
XXXxXXX~XXXXXXX*
t
..* . . . . .
I
* r xxxxxx~xxxxxxxx~
:
:
t
t xxxxxxx~xxxxxxx~
t XXXXXx;XXXxxxXX~
:
tUpland Rice*.. : /////////////// f ++ i + i zero i////i *+ :~:~~~~t
:
:
z
:
D
t
3
:
:
(1) except in Casamsrnce
(2) except in Central Casamnnce.
RBS~ARCB DFSIGPT
Inconveniences
They should be reviewed here briefly.
- Too large an area. It depends to a great extent on the number of
different crops, for it is hard to have plots smaller than 36 m2, and spaces
between the strigs less than 2 metres wideo Therefore it is necessary to use
an adequate research structure and mechaniae larming operations as nauch as
possible,
m burat ion*
The findings from yearb 3 an& 4 are minimal comparik to those ob-
taSned in yaar 2. The only interest iS to be able to confirm a limited number
of effects (limited by the number of 4rops tested in years 3 and 4) through
replications in time.
? ? ?
- Very high coefficients ok variation in years 3 and 6,. In year 2,
it CB~ go below 20 % and eometimes even close to 10 7; for a11 arops.
During this seoond year, the coefficient of variation is, of course,
dependent on aleatory factors such as climate or heterogeneity of the land.
It depends es-ecially on the use of recommended varieties and good farming
techniques.
The best trustworthy and most often significant results ?I@re obtei-
ned when advenced technology wcrs appkied throughaut, and yields ?rere hiçh.
This confiras :whaC wcl3 3:rid onrlier t one C=~I? only hopo for siguifiasat effects
with preceding crops, when advanced technology is used.
:Je do not think that thé problem is interesting tc study for tre-
ditional or slightly iapraved cultivatkon.
Findings and &ro:sositions
In S?ite of their imperfections, these exper5ments still provi6ed a
basis for the T:ain trop rotations tested on a large scale in the south-cental,
eastern and sçuthern parts of Senegal. There, they Canfirmed their :rorth.
Fron the findings provided Sy this design, w0 Uan enumerate certain
pr&lFI;iplea which we think should guide research on crlop rotation t
.
rr the existing design should be ieduced $0 %wo yeat*&$ uith the u3e
of aeries s-La gered in time, &n d
or er to bave a yearip &ÊeCt, This is possi.
ble in tww waye t
?? deries exatily the sa--e I
staggered b a year which give complete
results in three ge8Fe a but neeessitafe kwo siFpez?i*asats on the
jand in yea.r 2,
??Staggered 2 year aeu"iea~ whkb cari be different ta the ertent
that series X. SS talcen 5M.o acéount before eplbarking on Saries
EX.
Qbviously it is preferable to hare eight replicatione each tine
which re?resent a good safe numbera Fos+ whaterer reasond one shuuld not have
less than 3ix.
- Varieties and farming techniques should be thoroughly adapted to
each trop, yuaranteeing (except for catastrophic climatic conditions) the
success of the trop at tke technological level in question.
- In Senegai, ocly the yield characteristics -,+Jere retained. Cer-
tainly, resesrch must continue, and a more conplete follow-up than the one
done would be desirable.
Particularly in the zones with irregular rainfall, water problems
are important and it :rould be useful to examine the water resx-ves after each
trop. Tha nreccding trop factor Flays an important role in tha case of drought
spells during the croi season, eapeciallp if the crops differ in length of
growing cycles.
” 17 -
The problem of soi1 nitrogen seeU very important. It is probably
the source of many phenomena and particularly of interest in the alternation
of cereals and legumes, but one must not lose sight of the importance of micro-
biological phenomena : the case of sorghum is the most remarkable.
It is not our purpase to prooide a restrictive list, for numerous
aspects cari be considered. This follow-up cari permit a more rapid grasp of the
global phenomena thzt were registered in the preceding experiment. Koreover,
one cari establish it in Series II as a function of the hypotheses put forth
in Series 1.
As shown, research on preceding crops led
to the establishment of
a trop compatibility table for each ecologîcal zone* Binomials are defined
from these tables. They Will constitute the baaie for possible trop rotations,
but also for studies on possible farming systemse If two-year rotations are
not advisable (at least in the frarrevfork of the previous studies), trop bino-
mia&s provide more flexibility to study integrated systems, including a11
eonstraints, particularly those of the trop ëalendar.
GEPJElUL cOIdcLusIc?IW
From the analysis of results of an experiment on preeeding crops,
carried out for eight years in Senegal, two conclusions can be Urawn.
- Cne concerns Senegal directly. It led ta the main trop rotations
recommended in the south-central,
eastern and soukhorn PEUT~A of Senegal. Let
us briefly review them here*
Sine-Saloum snd Bastern Se,negal :
Year 1 : Ploughed-in grass fallow or maiee or early millet
Year 2 t Cotton or groundnuts
year 3 : Sorghum
Year 4 : Groundnuts
Upper Casamance :
Year 1 : Ploughed-in grass fallow or maize
Year 2 z Cotton or groundnuta
Year 3 : Uplend rice or "Sanie" (long maturity cycle, photosensitive)
millet
Year 4 : Groundnuts
One cari also have Sanie millet in Year 5.
Central-Casamance :
tio possibles cases t
. For farmers who grow millct t
- maire or upland rice
- upland rice or maize
- graundnuts
- millet
- 18 -
E
For farmers who do not grow millet, 3 solutions z
- maire - groundnuts - upland rice - groundnuts
- maize - groundnuts - upland rice
- grcundnuts ; X/2; upland riee - l/zi maire ; 1/2 maire - 1/2 upland
rice.
Thoroughly tested in the studies called %tructure d'exploitation11 (1)
these rotations proved to be valuable.
- The other led ta the definition of a methodology for the establi-
shment of trop rotations adapted to each ecologieal zone*
The purpose in each case is to define a certain number of compatible
trop binomials. Besides these binomials, one cari make the most intcresting
combinations as a function of different socio-eeonomic imperatives, on the
condition that the principle of compatibility is always respectede
These binomials are drawn from a çompatibility table established
after a simplified experiment on preceding crops. Hpwever, one must not lose
sight of tke fact that a technological level must be set at the beginning,
and that the compatibility table is only viable up to this level. Sxperience
seems to show that the influence of one trop on another is particularly sensi-
tive to intermediary levels, when one has succeeded in increasing production
without controllPng a11 the agronomie factors (of the environment).
By setting the basis of the study right at the beginning, the com-
patibility table cari be abtainod after three years, on the condition that an
adequate experimental network has bcen usede Fer each zone, it is preferable to
have a thorough experimentdtion with enough replications in one location, rather
than to increase the tria1 locations to the detriment of accuracy.
fi) transiator's note : Nodel farms operated by hired 'laborers under direct
supervision of agronomists, where t-ime required for eompleting each cul-
turai. operation and labors peak are thoroughly measured on an actual scaleo
- 19 ”
BIBLIOGRAFHY
CHARRZAU (C), NICOU (R) - L'amélioration du profil cultural dans les sols
sableux et sabla-argileux de la zone tropicale sèche Ouest-
Africaine et ses consequences agronomiques. (Improvetnent of
the trop profile in the sandy and sandy-clay soi18 of the dry
West-African tropical zone and its agronomie consequences)r
kgron. Trop., 19711 no 2, 5, 9, Il.
CHOPART (JrL.), NICOU (R) - Effet dépressif de cultures &Pétées de sorgho
dans les sols sableux du Sénégal, Premiers essais d'explica-
tien. (The depressive effect of repeated sorghum crops on the
sandy soils of Senegal. First explanatory experiments)
kfrican Soila Vole XVII, 1/1973.
NICOU (R) - Bilan de huit années d'étude des précédents culttiraux au Sénégal.
(Report of eight years of research on preceding crops in Senegal).
Xeroxed Iloc. IRAT/ISRA - CNRA, Bambey (Senegal).
*