h~tih~t Senegdais de Recherches Agricdes ‘,
University of Hawaii
Theme : Ikproving and Smtaining Food and Raw Materai production in West Afrka :
Reveming Soi1 Acidification, Loss of Otgunàc matter, and Erosive Runoff in Food Pkmktbn
System
Proceeding of the Wokshop of the West Group : Cape Verde, Gambia, Mali and Senegal
Jamwy 11 - 14 th 1999, Kaohck Senegal
By Adnata Niane Badiane
With the collaboration of Mateugue Diack and Mamadou Khouma
January 2001
--

---

Lmproving and sustaining Food and Raw Material production in West
Africa : Reversing Soil Acidification, Loss of Orgauic Matter, aud
Erosive Runoff in Food production Systems
West group:
Cape verde, Gambia
Mali, and Seuegal
Proceedings of the West Group Workshop
januav Il-14 ” 1999, Kaolack Senegal (w. Africa)

---

2
CONTENTS
Foreword ..........................................................................................
.
Opening Ceremony (Speech of the Director General of ISRA). ....................... .
1. Restoring Natural Resources for Food Security and Income . . . . . . . . . . . . . . . . . . . . .7
Russell Yost, .dminata N. Badiane, Aboru Berthe, Mohamed Kebbeh, Isawinda Baptista,
Modou Sene, Ricahrd tiblan
2, Improving. Food Production in Rural Areas of the Sahel : A Review of
Potential Technologies.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .._..................................17
Richard A. Kablan, Russell S. Yost, Abou Berthe, Manmdou Dauwbia, Aminata N. Badiane,
Isaztri?zda Baptista, Kevinn Braman and Isabel Anahog
THEME 1. FARM PROGRAMMING AND COUNTRY POLICY
3. Farmer to Farmer visit... _.. _._ ,._ ,.. .._ .._ ___ __ _._ ,. ., __ ___ _. _. .__ __. ._. . . . 2 7
Isaurinda Baptista, Isabel Anahory
4. Adoption and Farm Ievel Impact of Lmproved Fertility management
Technologies in the Soudano Sahelian Zone of the Gambia.. . . . . . . . . . . . . . . . . . . .3O
Mohmed K e b b e h
5. Lmproving and Sustaining Food and Raw Material Production in West
Africa : A participative Rapid Rural Appraisal in Fansirakoro.. . . . . . . . . . . . . . . .40
Abou &erthe, CI. Doumbirr, MI. ,!$&a~ M.K. Rjitteye, H.A., Mâiga, ,R F. Traoré. Richard A.
Kablan. Russell S. Yost
THEME 2. SOIL AND WATER CONSERVATION
6. Improving Food Crop production in Subsistence Farming in Fansirakoro ami
N’tétoukoro.. . . . . . . . ,-, . ., ,. , .., ,., ,,. . , , . . . ,. . . . . . . . . . . . . . . . . . . . . . _. . . . . . . . . . . . . . . . . . .5-I
Abou ti. Berthe, S.F. Traoré, B. Traoré, C.D. Kanrissoko, B. Guinda, Richard A. Kablan armd
Russell. Yost
7. Effect of Manure and P-Source Fertiliser on the Optimization of Soi1 Water
and Nutrient use for the Main Cropping system in Senegal Peanut Basin . . . .6O
Modo~ Sene, hfateugue Diack, Aminata N. Badiane
THEME 3. NUTRIENT MANAGEMENT
8. Effesit of Qrganic and Inorgarr*P ~~~+;risers on. the Nvtrilpnt Wta~s and Yielrt of
Drylands crops. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-l
Isaurinda, Baptista, Isabel Anahoy
-------
---
- 1 - .

---

3
9. A5 Lntegrated Fertihzation Study of the Groundnut - Millet Rotation
Cropping system.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . _. _ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
Babotc Jobe
10. Piliostigma reticulatum used for Soil Qrganic Matter Build up : Effects on the
Soif Quality and Crop Yield in the Peanut Basin of Senegal.. . . . . . . . . . . . . . . . . . . . . . .96
Mateugue Diack, Modou Sene, Aminata N. Badiane
11. Phosphogypsum Efficiency to correct Soi1 P Deficiency and/or Soi1
Acidity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
.._.................
100
Ma&que Diack, Modou Sene, Aminata N. Badiane
12. Phosphorus Buffer Coefficients of Seiected soifs of West Africa . . . . . . . . . . . . . . 105
Aminata Sidibé, M Keita, O.B. Coumqr4, Mamadou Dozlmbia, André Bationn, Richard A.
Kablan, RS.
13. Nutrient Balances under Contrastïng Millet cropping system.. _ . _ . . . . . . . . . .116
Mamadou Doumbia, Aminata Sidibé, M. Koné, B.S. Couliba&, N. Couliba&, A, Couliba&,
RA. Kablan and Iùcssel S. Yost
THEME 4. ENVIRONMENTAL IjUPACT
14. Drinking Water Quality as influenced by the cropping System.. . . . . . . . . . . . . ,128
Mamadou Doumbia, Aminata Sidibé, M Koné, A. COulibaly, Richard.A. Kablan arul Russel
s. Yost
ANNEX
Report of the field trip (Koutango and Nioro sites). ................................. .137
Report on sessions.. ........................................................................
.139
Minutes of the Pienary session of InterCRSP......................................... ,148
Country workplan ..........................................................................
.150
ListofParticipants.. ................................ ....................................... .153
Program of the InterCrsp West Group workshop.................................... .156

---

4
Poreword
‘Ibis workshop is a realization of a long term effort supported by the USAID AFican
Bureau, The many Collaborative Research Support Progt-ams (CRSPs), The Institute du
Sahel (INSAH), and more importantly the individual research organizations of each of
the member countries in the InterCRSP project. As such each of these organizations and
their directors and scientists should be proud of bringing together and sharing of their
best thoughts and suggestions to address the severe food security, low income, and
sustainabihty challenges fac4 by sub-Saharan agriculture today.
The InterCRSP efforts represent an attempt to take the best technologies of the
CRSPs and, in fact, from anywhere in the world to address these challenges.
As will be apparent in the papers, there is some cause for optimism, although fat-mers
are u.sing extremely low levels of nutrient fertilizers, they seem well aware of the value of
fertiliser and the need for it. They just seem unable to purchase it at least for the
subsistence crops of sorghum and millet. Fertilizer use on irrigated rice, cotton, and
maize is profitable and use on these crops is growing. Experience in The Gambia shows
that even the subsistence fat-mers respond quickly to changes in fertilizer prices,
unfortunately, dramatically higher costs due to recent structural adjustments have
resulted in drastic reductions in fertihzers use. While fertihzer use is extremely low in
Mali, it is apparent that IER scientists are aware of the issues and are pursuing a number
of initiatives to improve the policy environment. While there are some technical
adjustments to increase yield and efficiency of fertihzer use under study by the
InterCRSP project, it is apparent that much of the deficit in fertilizer use relates to policy
and economic issues.
One of the interesting initiatives underway in Senegal is the
government subsidized application of rock phosphate and probably more importantly of
phosphogypsum, a heretofore unwanted waste product that has great value in restoring
nutrients and reversing a portion of the widely recognized nutrient mining taking place in
sub-Saharan Afiica. Other countries in the tropics have recently noted the value of this
resource and it has become a standard commercial practice in the acid, low fertility soils
of Brazil.
As will become apparent in the scope of presented papers, studies on nutrient
management predominate at this stage of the project. Tbis represents a strong underlying
concern throughout the project for the severe nutrient mining that is gradually reducing
both trop yields but also production potential. We expect tbis theme to be joined with
new studies on water and soil conservation and more economic evaluation and analyses
as their role becomes more apparent and the proper studies and personnel are identitied.
We note that among the papers presented some of the first studies of’ impact of
nutrient management on village water supplies have been conducted. Based on the
initial results, it seems that fertilizer application rates are SO low that there was Little
increase in nitrate concentrations of village drinkng water Wells. We caution that the
resuhs are preliminary, however, and the situation may be completely different in tu-ban
margins where prices of vegetables are SO higb that exorbitant amounts of physical
labor and fertihzer are applied to meet the demand. With the extremely sandy soils,
highly intensive agriculture, and shallow water tables, nitrate contamination of
ground~~di~~ I~,i.,Gns a concern and hu& fither study.
Lastly, we must thank the Institut Senegalais de Recherches Agricoles, and the
personnel of the Kaolack Experiment Station for both organizing and hosting this first
worksbop, snd also for assembling thesc proceedings into such a useful document.

---

5
This bas, indeed, been an exemplary workshop and we would like to thank the
Coordinator Dr. Aminata Badiane and her staff and colleagues for providing such a
bigh standard for subsequent workshops.
Russell Yost
Feb. 28,1999

---

6
Openîng Speech by the Director General of ISR4
Dear Representative of INSAH,
Dear Director General ofNAR1,
Dear Coordinator of the West Inte$RSP Group,
Bear Participants,
It is a pieasure for me to be bere with you for the fh-st workshop of the Western
group of InterCRSP which inciude Cape Verde, the Gambia, Mali and Senegai.
I need to welcome you flrst in Senegai, particularly in Kaoiack, the capital City of
the Peanut Basin.
1 would take this opportun& to extend to you my greetings for the New Year.
The collaborative research project, whose data will be reviewed during this workshop, is
very important to us. In fact, restoring and increasing the productive capacity of soils in
Sub-Saharan Afi-ica, especially in West Afiica, are key components to promoting a
sustainabie agriculture in this region.
The objectives sought by tbis coliaborative research project, çombating soi1
acidiflcatioa, ioss of organic matter, iosses du-e to soi1 erosion, and their impact on
production systems, should bring a slgnificant contribution to increasing food production
in West Africa.
It is very comforting to notice that these objectives are major components clearly
deflned in .the strategic worlcplan of’ ISRA The implementation of this workplan starts
this year.
The strategic workpian of ISRA (1999-2003)
is the first phase of the
impiementation ofthe enterprise project of TSRA. This project det%nes the major strategic
orientations and options of ISRA for year 2015.
Dear participants,
ISBA, which is proud to host this workshop, is going through deep changes. in
fa&, the institute is now, by the vote of the Pariiament: registered by the law on
SciLqtific and Technoiogic Public Sewices, ( ‘EPST-f’).
In addition, the presidentiai açt approving the organization and fimctioning of
ISRA has been signed recentiy. Another presidential act on the estabiishing ruies of
ISRA wiil .be signed very soon. In 1.998, several procedures were adopted to ailow an
optimai fhnctioning of the institute. There were procedures for scientific management
managing research stations, accounting systems, etc. These changes should ailow better
management of new or ongoing projects such as this invoiving Cape Verde, the Gambia.
Mali. Senegal, INSAH, Virginia Te&. and the Umiversity of Hawaii.
Dear participants,
Buring four days, you Will review and screen activities implemented through the
coliaborative research program on soil, water and nutrient management. Based on your
qualifications, 1 have no doubt that the recommendations fi-om this workshop will
improve the petiormance of this project.
With this hope, 1 declare open the fh-st workshop of the West InterCRSP group

---

7
Restorhg Natural Resources for Food Security and licorne
Russell Yost’, Aminata Badiane2, Abou Berthe, Mohamed Kebbeh3,-Isaurinda Baptista4,
Modou S&I~~, and Richard Kablan’
1 University of Hawaii, 19 10 East W&t Road, Honolulu, HJ 96822
2 Isra-DG, Route des Hydrocarbures, Bei Air, BP 3 120, Dakar, Senegai
3 National Agricultural Research Institute, PMEI 526, Serrek.unda, the Gambia
4 INIDA, CP 84, Praia, CAPE VERC(E
Abstract
One ofthe most serious constraints to food security in West Af%a has been the steady
deciine in agricuiturai productivity per unit area. The decreasing productivity of the food
production system in West Afnca must be reversed in order to achieve food security and
sustainable productitity. The InterCRSP/West project is dedicated to reversing this trend
through improved management of the regions natural resources. One of the most
important keys to the improved management is another resource - the human resources.
that is the talent, experience, and dedication of key scientists of the region. The many
shnilarities in problems, experiences with the problems, and the solutions to the problems
form a knowledgebase of major value for the region. With such a knowledgebase, that
which has been iearned in some couutries need not be reiearned in others. We suggest
that the exchange of this knowledge and .the raishlg of conscientiousness of the scientists
with these skills is one of the most promising and exciting ways to reverse the
productivity decline. As a result of the first two years of research we idente the
following activities as integral to our strategy to reverse the decline in productivity.
1 .‘%ross vi&,” the coordinated visits of a key, experienced scientists fiom one
country to another and vice versa, are suggested as one of several methods to improve
the use of knowledge and experience of scientists in ail countries in the InterCRSP/West
group. This may be a method to provide temporary expertise to initiate programs or to
fil1 short-term or initial, exploratory research needs. A procedure is being developed to
facilitate and improve this sharing of knowledge and skill.
2. Conservation of soil and water is needed more than ever. Major losses of
water, soii, and nutrients routineiy occur at present. Stone lines and various steps in their
enhancement with multiple-use vegetation is being suggested and tirther tested as
effective strategies to better utilize water before it is lost to runoff.
2. Nutrient management effIcienc.y centers on animal manure and to meet this
* challenge a ‘knanure extender” hypbthesis is heing tested. The technique seeks to
enhance the efGects of manure tith inorganic fertilizers. particularly in sandy. poorly
buffered soils. The technique is based on a current practice by farmers in the Dougouba
village of Mali.
4. On-farm tests. based on a thorough. initial ex7loratol-y partiçipatory appraisai.
acquaints bath the extension personnel, researchers. farmers, and the community to
constraints and opportunities for improvement.
5. Current fertilizer and food pricing policy is not conducive to providing
incentives and raw materials to produçers to increase productivity. Research in Gambia
shows that fertilizer use was reduced dramatically with price increases. On-i-ànn studies
indicate that the constraint is not the farmer’s awarencss of the need for nutrients but
rather the policy goveming its pricing and supply. While some technical improvements in

---

fertjlizer efficiency appear possible with improved management of manure and fertilizers
as suggested above, it is clear that ruttrient use efficiencies are not as limiting as
marketing and policy constraints.
1. Cross visits
Surely one of the most important of the resources of the West region of the
InterCRSP effort and, indeed, all of the countries participating in the InterCRSP project,
is the human resources represented therein. It is, therefore, one of the goals of this
project to acquaint ,s+ntists of each institute with these potential resource persons and
to develop a minimal cost protocol for financing the travel of these persons to assist
other countries in their speciality. An ex:ample of a recent “cross visit” was the August
visit of Dr. Mohamed Kebbeh, NAR& The Gambia to accompany a visit to the
InterCRSP/Mali and the SM-CRSP/Mali experimental sites and subsequently to visit the
InterCRSP/Senegal site. The purpose of this vi&, Fom InterCRSP objectives, was to
take advantage of the short ter-m researcb that Dr. Kebbeh undertook wbile studying with
John Sanders in Purdue University. Dr. Kebbeh provided a senrinar while at IER/Mali
and also gave a summary of his work while in Senegal. While the main purpose was to
provide the results of the work in Pur-due with the mode1 developed by Economist John
Sanders, Dr. Kebbeh also had the opportunity to share some of his experiences with
fertilizer availability, use, and cost policy in Tbe Gambia. For example, Dr. Kebbeh
pointed out that prior to structural adjustment the rate of application of chemical
fertihzer to crops in The Gambia was one of the highest in Afiica. After structural
adjustment of the Gambian currency, fertilizer prices became quite high and did not corne
down with time. Fertihzer use has sbarply dropped off reflecting the adverse “ price of
product” / ‘price of input” ratio. Ifthe farmers are similarly responsive to fertilizer cost
in other countries as they appear to be in The Gambia then it suggests a strong possiiility
that reducing fertilizer price may help in increasing fertihzer use, which seems imperative
to reverse the nutrient mining that cantinues to erode the food production capacity and
endanger food security.
Another example of cross visits is the expansion of soil and water conservation
research in The Gambia. Under consideration is the possible cross visit of a Senegalese
scient& with extensive experience in soil and water conservation in conditions of soil
and climate very similar to those in some areas of The Gambia. Such sharing of expertise
and building on past experience in quite similar conditions cari accelerate progress and
avoiding many years of costly experimentation.
Other examples of local expertise are apparent, but in many cases the presence
of this expertise is not known and it cari be difficuh for home institutions to justitqr
visits when they already have persons with job descriptions similar to those of the
experts. Perhaps one of the difficulties is in recognizing the unique talents in various
InterCRSP institutions and rewarding and stimulating that talent.
During the August 1998 visit to Kaolack, Senegal, this issue was discussed and a
recommendation to the InterCRSP coordinators was formulated. We suggest that
persons with the spectic talent be identified and the minimal costs of travel (air ticket
and per diem) be divided between the receiving institution and the InterCRSPWest
proje.ct. 711~ LrEti~il visit by Dr. KebLh vras completely finan& bj the Project but ti,L,
cannot be continued.
Several nronosed “Cross visits” include:

---

9
1. Modou Sène visit to The Gambia to assist in establishing a soil and water
conservation program in that country.
2. Modou Sène visit to Mali to assist in the improved utihzation of stone lines,
their vegetation, stabilization and preservation. Also to estimate costs and advisability of
locating water retention structures of the type developed by the French NGO near
Fansirakouro.
3. Further visits by M. Kebbeh to Mali to update and complete a survey of
fàrmer information that suggests subsistence fàrmers in the Cinzana region are,
contrary to expectations, applying chemical fertilizer to their subsistence crops and
how they are achieving this.
4. A visit by a soil and water conservationist to Cape Verde is under discussion.
5. A sharing of expertise on phosphogypsum (Badiane?) and rock phosphate
(Doumbia?) is needed between Mali and Senegal.
“Cross visits” cari be use&.l from the perspective of local scientists because it
illustrates the broad redevance of their programs in others countries and the interest of
other courttries in those issues and approaches. Such visits cari generate the necessary
enthusiasm to stimulate I%esh approaches and ideas to solving the age-old problems.
2. SoiI and Water Chservation
One of the more striking observations of the agriculture in the OHVN (Upper
Niger River Watershed) is the extreme drought that seems prevalent, yet as much as 800
mm of rain is received annually. Also noticeable are the steep slopes and fiequent
exposure of impermeable rock and soil -- the result of extensive surface soil erosion and
exposure of hardened underlying materials. The result appears to be a major loss of
rainfall, especially when large, intense storms occur. As a result, it is clear that major
efforts in recovering and improved utihzation of the rain are needed.
During the initial Msit to the Kati region near Bamako, stone barriers were seen in
Tourodo and Fansirakouro. These structures had been extremely useful in slowing water
overflow and accumulating rich, but highly erosive soit. Others have noted that stone
barriers are beneficial, but oflen their beneficial effects are lest when the stones are
removed and reused for other purposes. Also stone lines are only effective in trapping
larger soil particles while dissolved nutrients and organic mater& continue to be lest.
Consequently, alternatives are being developed including testing species of grasses or
trees that secure the highly erosive soil ;accumulated behind the stone barriers, lending
some permanency to the practice and also provide for forage, timber, and firewood. One
species has been suggested for this purpose, Guiera senegalensis, but others have
advantages depending on the location and household needs. Investigations of alternative
species include testing species designed :for animal forage selected by the International
Livestock Center in Africa (ILCA) being tested at Bamako, Mali. One noteworthy
species is ZiTpht1.s mauritiaca used both (as a forage and for fruits that children relish. In
addition, cropping options that capture w,ater before it leaves the field are needed. Some
,photos of these technologies are available on the InterCRSP website:
(http:/fagrss.sherman.hawaii.edulyostlintercrsplercr~2~eb-Technologies.htm).
In somc villL,c., laig~ trees have .bt~ r,~..6~~~tirtJy choy+J &,y.,, LU produce
timber necessary for fencing livestock. An alternative, living faces, are being tested at
several villages. Svecies under test are Jatrwha SPP.. and Zi~vhzrs rmcromta (which

---

10
Another technology that substantially improves food security are small dams of
very small size that retain water throughout much of the year. Such a water control
structure was developed by the Associat:ion Volontaire du Progrès (a French NGO) in
the vicinity of Fansirakouro, Mali. ‘Ibis structure maintains water throughout the dry
season and has measurably increased ,the water table, both above and below the
structure, according to the delighted villagers. The Fansirakouro fàrmers showed the
structure to InterCRSP scientists requesting assistance in developing additional
structures of a similar des@. Such structures could be of use in many countries of the
Sahel. These small structures, also iIlustmted at the above website, are currently under
study for adaptability to wider implementation and use. The small structure does not
appear to me extensive maintenance and thus should not risk falling into major
disrepair. A characterization of these types of technologies will occur afier the
workshop with surveys by lER/Mali and Virginia Tech soil and wtaer conservation
engineer. Experience gained in Senegal is also thought to be of major use in improving
SO~I and water conservation practice in the InterCRSP/West network (Perez et al.,
1997).
0ne of the major steps in preparing for improved soil sud water conservation in
Mali has been a brief survey of the soils of the eqerimental sites of Fansirakouro and
N’tetekouro carried out by LaboSEP (Laboratoire de Sol, Eau, et Plante) of the
Institut d’Economie Rurale (IER), M+l.i (Table 1).
Table 1. Survey of soils represented in on-farm experiments in N’tentoukouro and
Fansirakouro, Mali, LaboSEP, IER, Mali.
(Survey conducted by 0mar Doumbis, IER)
‘A plinthite layer is especially critical as if exposed to oxidizing conditions and allowed
to dry because it cari harden into stone forming laterite. This may have occurred in
some soils where extreme erosion has exposed the pliuthite present in the lower
horizons.
Soil and water conservation efforts have been underway in other InterCRSP/W
countries, most notably Senegal and Cabo Verde. As we’ll see during the workshop
vi& near Kaolack, Senegal, there are several examples of excellent water conservation
efforts
in several watersheds of southem Senegal. Some of these results are published in Perez
21 al. (1997) and ciLSLiL Llir: followiug ittLyL&s.
1. Contour cultivation
2. Dry season de-compacting. This technique corresponds to a 10 cm deep
subsoiling done by ~&til traction with a singlz excavating pi& Bo~;auje there is no
--

---

11
available time during the beginning of the rainy season (seeding operations), it is the
only solution to help infiltration of water fiom the first, and often violent rainstorms.
3. Shallow ridging before emergence, for a groundnut trop, sometimes ridging
sd over a line of manure (Sène, 1995).
4. Shallow earthing up, for the millet trop. This might include covering manure
with the ridged soil.
5. Lastly, brushwood checkdams and stone pavements were used to stabilize
the main guIly and transform it into a permanent waterway.
The extensive, steep slopes of Cape Verde present a severe challenge to all soil
and water conservation efforts. However, it is clear that much effort has been invested
in soil and water conservation. Some techniques implemented in&& ,ontour-hedge
rows, planting on the contour and vegetating with drought Te@stant species. Extensive
work has been invested in stabWng stream channels. It is clear that improved
conservation of soil and water remain at the tore of improved natural resource
management in the InterCSRP/W project. New, more effective conservation measures
are needed while existing methods should be more widely disseminated.
3. Nutrient management efficienq
As numerous papers have imiicated, nutrient mining and nutrient depletion are
widespread in the region and probably are the single most important reason for the
consistent decline in trop productivity of the region. The InterCRSTYW project is
working on this problem Fom two perspectives and disciplines: that of the policy level
and farm level economic progamming and secondly by encouraging more estimates of
the amount of nutrient mining by encouraging routine assessments of the nutrient
balance status of all proposed or improved cropping systems.
Fertilizer pricing policy. The analysis at the policy level revolves around the
combined work of Mohammed Kebbeh (The Gambia) and John Sanders (Purdue
Elniversity) (Kebbeh, 1998). A farm-level economic programming mode1 developed by
Ousmane Coulibaly (1995) was adapted to predict conditions in The Gambia (Kebbeh,
1998). Not surprisiugly a large number of studies identifl the extensive nutrient mining
as a result of ins&icient fertilizer use, in some cases caused by unfàvorably high prices
of fertilizer. Van der Pol (1993), for example, estimates that 40% of the fàrm income is
derived fi-om mined nutrients. The factors directly leading to the nutrient mining and
negative nutrient balances are thus proba.bly not as much agronomie as socioeconomic
and include pricing policies and the interaction of pricing policies with socioeconomic
conditions which have reduced the use of fertilization. Before structural adjustment-
fertilizer use in The Gambia was among the highest in Sub-Saharan Atica. After the
adjustment and sharp price increases, ftiilizer
use has fallen to among the lowest in
the region. The Coulibaly model, developed as part of a Malian student’s thesis at
Purdue University, replicates this observation, when adapted with coefficients
representative of conditions in The Cambia. For example, Kebbeh (1998) iltustrates
that a 25% increase in fertilizer pric.e is predicted to result in a nearly 4-fold reduction
in fertilizer use. Further studies of the fertïlizer price increase indicate that as much as
30% of the fertilizer price in The Gambia is simply due to taxes on imported fertrlizer
*~.*~+n,~h?s. It seems unWe~: +h+ those in chzrgc ~C+e;;“3 this p&y 3re CO,- :-.2~!: of
the extreme sensitivity of fertilizer use to price fluctuation and, in turn, probably are
not aware of the extremely negative effects on sustainability and long term productivity
of Gambian agriculture of thiç tax on imported fkrtili7cr. Further analysis with current

---

12
data is needed to prepare a case for policy change to stimulate fertilizer use as a hasic
component of food security and sustainability.
The Coulïïaly mode1 appears likely to be usefùl in examining policy
implications of pricing of fertilizers and food products in Mali as welL Scientists at the
Institut D’Economie Rurale, Mali observe that the current ratio of price of grain
product to price of fertilizer is so low that fàrmers cannot and should not be applying
chemical fertilizers. For example, the ratio of the price of a kilogram of millet to the
cost of a kilogram of urea N is about 0.2:, which is well below the recommended value
of 2 ,to 4 to expect adoption @‘DC, 1988). A participatory rural appraisal of fàrmers in
the Cinzana region, however, revealed that 22% of the f&mers interviewed said that
they were, in fact, purchasing and applying L.Gzers to millet (SM-CRSP, 1998).
Two hypotheses may explain this unexpected purchase and application of inorganic
fertilizer in face of such an unfavorable benefit/cost ratio:
- Millet is the staple trop in the Cinzana region and is an important source of
family pride and food security. Being able to produce the staple food to nourish the
family and provide for the household is obviously highly regarded. Thus the value of
provîding for one’s own fias@ is not reflected in market prices partially because this is
a subsistence item and the availability of millet ou the market is sporadic, usually only
available when trop yields are unusually great and excesses occur (Kebbeh, persona1
communication, 1998). According to this hypothesis, the ratio does not reflect the true
value of millet for the food security of the Cinzana household.
- The second factor might be technological, that is the iârmers may be
practicing a particularly efficient method of fertilizer application such that the trop
response per unit input might be much larger than that assumed in calculating the IFDC
ratio of 2 to 4. This seems to be the case at least with manure applications in the village
of Dougouba, near C&ana, Mali. A common practice of fàrmers in this region is to
localize the application of manure. As will be discussed later regarding the “manure
extender,” there is some reason to expect unusually efficient nutrient utilization when
inorganic fertilizers are physically mixed and placed with manure.
Furth.er studies of the 22% of the farmers in the Cinzana region are needed to
confirm this surprisingly fkequent use of inorganic fertilizer in spite of low market
prices for millet. Further studies are needed of the C&ana (Sahelian climate) and the
Bamako (Sudano-Sahelian climate) production systems using the Coulibaly mode1 with
updated prices and more accurate fertilizer response fùnctions. With further analysis in
hand, discussions and seminars with poliqy-makers is suggested as a next step to bring
the results to the attention of the Miiistry of Agriculture and other appropriate
authorities in Mali. The Gambian example of sharply reduced fertiliser use in response
to high prices tends to validate the dynamics and predictions of the Coulibaly model. It
remains to be seen if reversing fertilizer prices Will also reverse the decreases in
fertilizer application following structural adjustment and restore the food production
potential of the country.
The exercise of analyzing fertilizers use by subsistence farmers when price/c.ost
ratios are less than 0.5 has illustrated the need for holistic assessments of the problems
of food security and productivity decline. It has al.so illustrated the growing talent and
experience of scientists in the region and their potential to effect change if efforts are
convdinated and concll’rterl GI an interdisciplinary, collahorative fzrhion ~n1ong the
countries. This exercise has been a powerful example of the value and enthusiasm that
çan be generated with carefùlly selected and conducted cross-visits of experienced

---

1 3
scientists f?om one country to another. We have leamed a great deal Fom the cross-
visit and will be seekiug to plan and organize a structure for such visits in the future.
Technology to improve the efficiency of fertilizer use.
Mi-mure extender
The “Manure extender” concept is an approach proposed by the
InterCRSP/West project to realize the beneficial effects of inorganic fertilizer while
building on an existing fàrmer practice of using animal manure. The objective of the
approach is to extend the beneficial effect of manure by physically mixing manure and
N, P, K inorganic fertilizer and incorporating it in a localized fàshion in the trop row
such that nlants such as sorghum, millet, and peanut are placed at either end of the
manure -’ IA, 9, K mixture placement. Some farmers in the Dougouba village, Cinzma,
Mali., have been applying compost in such a loc,alized fashion for several years so it
may be relatively easy for them to adapt their current methods by adding mixed
inorganic ferti&er to the manure.
This proposed method of application is thought to offer some potential benefits
over the usual manure versus inorganic compatisons in which the manure and the
inorganic fertilizer are physically separated:
1) The manure, when localized, should provide a higher CEC environment than
the surrounding sandy soil (usually less than 5% clay), which might reduce the
excessive amounts of nutrient in soil solution in such poorly bufhied soiL and reduce
damaging effects of excessive concentrations of nutrients, i.e. salt damage, and thus
possïbly reduce loss through excessive leaching downward of the nutrient cations.
2) The mamue should provide: a bufher against radical temperature and
moisture change, because of the high moisture holding capacity of organic material and
the high water content when at the same soil moisture tension as the surrounding soil.
This should improve the growing environment of plant roots, increase the number of
microorganisms to mineralize the organic material, and improve likelïhood of
inoculation and intensity of mycorrhizal colonization of plant roots.
3) The manure should provide a slow release of nutrients while the N, P, K
provides soluble, immediately available nutrients.
4) The mamue should provide the small amounts of micronutrients necessary to
meet trop requirements. Although macronutrients such as N, P, K, Ca, Mg, and S are
usually deficient, it is also quite likely that micronutrients are deficient in the extremely
sandy soils.
Unfortunately, reliable tests of whether or not the “manure extender” idea
really works have not yet been completed. Consequently, it remains an idea that may or
may not work but seems to have some interesting advantages to manure or inorganic
fertilizer alon e.
Phosphogypwn
The use of phosphogypsum, heretofore, an unused waste product of the soluble
phosphate fertilizer industry, is a direct method of adding calcium and sulfate, both of
which are usually present in inadequate quantities. The importance of the calcium and
sulfate in phosphogypsum becomes apparent when the high costs and low availability
of alternative sources of calcium is known. Lime and rock phosphate are probably the
only other major sources of calcium for the region. The practice of applying gypsutn to
acid scils ic relatively a ne:>, practice, probably only rrall,. becotig a uidc;;.:.zU
commercial practice in Brazil in the 80’s and 90’s (IBRAFOS, 1992). During 1998
year the Senegalese govemment has begun a program of subsidizing the use of
phosphogypsum and rock phosphate by encouraging ‘/2 rock phosphate and X

---

14
phosphogypsum application. It is unclear whether the excellent yields of peanut and
millet in the peanut basin in 1998 are at a11 related to this new practice.
While some farmers seem to worry about the acidifying effects of
phosphogypsum, it is usehd to consider the extensive and widespread adoption of
phosphogypsum together with limestone as a highly successful means to quickly
remediate the surface and subsoil of acid SO& of Brazil (IBRAFOS, 1992).
Rock phosphate continues to spark controversy as the residual benefits seem
difficult to quanti@ in a mariner that cari be satisfactorily analyzed economically
(Jomini et al., 1991), who found that for Niger the price of rock phosphate needs to be
less than ?4 that of soluble phosphate in order for it to be a better choice. One
unresolved issue appears to be that the residual effect of rock phosphate has not been
well-quantified and whether the acidity neutralization and additions of calcium are
significant. Considering the extremely high price of limestone, perhaps one of the few
ways to stabilize acidity might be with the use of rock phosphate. On-going studies in
Mali are eagerly awaited in which the effects of composting with rock phosphate seem
to be a beneficial pre-processing of the rock phosphate. It is unclear at the present time
how much ofthe beneficial effect might be due to the higher CEC of compost and thus
maintaining low soi1 solution calcium driving the dissolution process. or whether the
organic materials might be a source of acidity to accelerate the dissolution of the rock
phosphate or whether the higher water capacity and higher water content of the
organic media permits dissolution reactions to continue longer than in a sandy soi1 at
the same soil moisture tension.
4.On-farm experîments
On-farm experiments are essential to both developing the appropriate
technology as well as in testing and ensuring the adoption of the technology. The on-
farm experiments conducted in Mali, for example, have been highly useful in better
understanding the constraints to adoption of technologies that Will mitigate nutrient
mining, acidification, and in the conservation of both soi1 and water
The on-farm experiments in the villages of Fansirakouro and N’tetekouro
revealed that the farmers are well-aware that fertilizers benefit crops. Fat-mers and
extension agents were, not surprisingly, unfamiliar with experimental protocol and
applied fertilizer treatments of manure and mixed fertilizers in a manner that precluded
statistical comparison of treatments. One farmer, for example, applied the fertilizer
only on the poor section of his field because he knew that it would likely improve trop
growth and yield. This practice, of course.. makes it impossible to compare treatments
to estimate the effect of fertilizer. One of the main results. then of the on-farm
experiments was not the initial, intended purpose, rather it was an indication that the
techniques and methods of efficient fertihzer and manure application are yet to be
leamed. Some fertilizers. for example, nitrogen is mostly applied alter the initial
planting and cari be sut-face-applied, while: manures? phosphates, and liming materials
including rock phosphates should almost always be applied and incorporated into root
zone for efficient utilization. Another important observation was that most farmers had
great difficulty obtaining the 2 tons / ha of manure required for even the small
~+i~h~htal area. This illustiutLS 11~ scarcity of tL l*iiblidlL ùuuxe, thc: iliipus.AluiGi~
of supplying a11 nutrients needs by organic mean alone and emphasizes the need for
inorganic fertilizers to supply the nutrient needs arrest nutrient mining.

---

1.5
Numerous problems were also ençountered in non-nutrient factors that became
confounded with treatment effects making it impossible to test the initial hypotheses of
‘treatment effect. Examples of such confounding factors were heavy shading by trees
and major variations in amount of grave1 in the SO~LS on which treatments were applied.
Dates of pla,nting for all treatments were not the same and treatments were applied
differently. The treatments included four ‘treatments: 1) A controb 2) 2 tons of animal
manure, 3) 2 tons of animal manure + N, P, K, fertihzer, and 4) the N, P, K, fertilizer
alone. The latter N, P, K treatment is necessary in order to test for the presence of a
synergistic effect among the organic and inorganic fertilizers as hypothesized.
The results illustrate that farmers appear to be well aware what fertilizers are
and that they cari be beneficial. Improved efficiency of fertilizer use seems possible
throwgh improved placement, timing, and improved coordination of applications with
plant growth and nutrient requirements. Xt seems prudent to fiuther explore some of
the combined effects of manure and mixed N, P, and K fertilizers as suggested above.
The magnitude of change needed to reverse nutrient mining, however, Will clearly
require more than a simple agronotmic: improvement in fertilizer use efficiency.
Economie constraints appear to be much more ‘limiting and fertilizer pricing policy
seems to be one of the key limiting factors to reversing nutrient mining and restoring
productivity of the land.
5. Current food and fertilizer pricing policy
As pointed out in both a recent book and later publications, Sanders indicates
that it is finally accepted that both orgamc and inorganic sources of nutrients Will be
required to reverse the nutrient mining affecting Sub-Saharan Africa (Sanders and
Ahmed, 1998). Tbis is also the conclusion of a publication focused on nutrient
management in Af?ica (Buresh et al., 1997).
It seems agreed that fertilizer pricing is one of the focal point of limited
application of nutrients and price subsidies are advocated by Sanders and Ahmed
(1998). They argue that progress towards reducing malnutrition, achievement of
greater food se&su&iency, and reduction in environmental degradation cari ail result
fiom increased and improved fertilizer management.
These authors point out that there are synergistic effects between organic and
inorganic nuttient sources and on the poorly buffered soils the combined application
makes sense. Some of the reasons given above for the rationale for the “manure
extender” idea are in agreement with Sanders’ suggestions. The authors suggest that
sub-saharan countries have been rationing foreign exchange and putting a low priority
on fertilizer imports. Low input strategies invobing mixed trop rotations and
livestock, while helpful, are no longer sufficient to ensure positive nutrient-balance
cropping systems, i.e. those systems that no longer use more nutrients than they
contribute to the soi]. These methods are complements to inorganic fertilizer ramer
than substitutes.
Sanders and Ahmed also point out that a national fertilizer strategy needs to
include the revision of economic policy to emphasize the profitability of food
production rather than focusing on maintaining low urban food prices. In the long run
with impro\\ J Jmulogies, opyor NUI&,& and incentives fol iA[ IIALI s, fuod 1~1 ices \\t il1
likely lower. Lowering fertilizer prices seems, therefore: a way of permitting fat-mers to
feed themselves and begins to take up their critical role in providing improved food and
raw material not only for themsclves but also for the entire socicty.

---

16
Li&ature &ed
Buresh, R P.A. Sanchez, and F. CaIhoun. 1997. Replenishing Soil Fertility in Afrka.
Soil Science Society of America Special Publication No. 5 1. Madison,
Wisconsin, USA.
Coulibaly, 0. 1995. Devaluation, New Technologies, and Agricultural Policies in the
Sudanian and Sudano-Guinean zones of Mali. Doctoral Thesis-Purdue
University, West Lafàyette, In-a.
IBRAFOS, 1992. II Seminario sobre o Uso do Gesso na Agricukura. 413 pp. (Second
Seminar on the Use of Gypsum in Agriculture), Uberaba, Minas Gerais, BraziL
Jomini, P.A., Robert R Deuson, J. Lowenberg-DeBoer and Andre Bationo. 1991.
Modelling stochastic trop response to fertilization when carry-over matters.
Agricuhural Economies 6:97- 113.
Kebbeh, M. 1998. Adoption and farm 1eveI impact of improved fertility management
technologies in the Sudano-Sahelian zone of The Gambia, National Agricultural
Research Institute, The Gambia.
Perez, J.A., M. Diatta, M. Grouzis, and M. Sène. 1997. Rehabilitation of a semiarid
ecosystem in Senegal. 1. Experiments at the hikide scale. Agriculture,
Ecosystems & Environment. 65: 95-106.
Sanders, J., B. 1. Shapiro, and Sunder Ramaswamy. 11996. T&e Economies of
Agricultural Technology in Semiarid Sub-Saharan
Apica. John Hopkins
University Press.
Sanders, J. and M. Ahmed. 1998. Developing a fertiliir strategy for Sub-Saharan
A&-ica. Manuscript prepared for presentation at the American Society of
Agronomy, Oct. 17-24, 1998. Baltimore, MD.
Sène, M. 1995. Influence de l’état hydrique et due comportement mécanique du sol sur
I?mplantation et la fructification de l’arachide. Doctoral Thesis, Montpellier,
France.
van der Pol, Floris, and B.Traore. 1993. Soi1 nutrient depletion by agricultural
production in Southem Mali. Fertiliser Research. 36: 79-90.
--..-.-
-
--.-m-m-
. . ..---
-
1--

---

17
Improving fond production in rural amas of the Sahel: a
review of potential technobgies
Rk Kablan, RS. Yost, A. Berthe, M. boumbia, A Badiane, 1. Baptista, K. Brannan and
1. Anahory
Ab&ltXt
Due to increasing number of factors affecting its food production systems, the
capacity to produce food in Afiica hab ,sn declining over the past decade. TO &.rther
identify and understand the most limiting factors to declining trop performance, the
InterCRSP/West team conducted a survey in October of 1997 across four countries of
the Sahel (Cape Verde, Gambia, Mali and Senegal). These countries are involved in a
collaborative research program for improving and sustaining food and raw material
production in Africa. The survey revealed that effective biomass management, water
conservation strategies and prices of fàrm products limit the food producing
capabilities of these countries. Cut-rently, the dominant agricultural practice is mixed
fàrming involving trop and livestock. Millet, sorghum and peanut are the main staple
food grown by farmers. Labor availability is a major concern because of migration of
youth to cities. In most areas tree density is very low due to human usage and high
mortality during drought. Grass caver is also domiuated by ammal species. InsufIicient
land is available for fallow. Mamue production is low because of lack of forages,
which they attribute to low fertility of the soils, decline in rasa11 and lack of fertilizers.
Wind erosion and run-off have also reduced the production capacity of fragile soils
1. Induction
A stable food supply is essential ta the welfàre of a society. In most of West-
Afiica food production per capita has been declining over the last decades. This has
been mainly attriiuted to the rapid population growth and soil fertility depletion.
In the Sahel, increased soil acidity and erosion along with decreased soil
organic matter, seem to be the major constraints to fond production. Because farmers
in the four countries visited are a11 resource poor and in view of the urgent nature of
the agricultural problems, the greater impact on sustaining food production would be
to determine technologies that cari positively affect farming activities. This article will
b.rieIly review technologies that may signifïcantly improve small farmer’s production
systen
2. Ewaluating current techn~s and their relevame ta production
systems in the Sahel
Crop Residue management
Crop residue management strategies deserve special attention in all four
countries. At the end of the survey it became apparent despite their efforts, current
farming practices are neither feeding the peopie, nor helping them emerge fi-om
poverty, while the degradation of the natural resource base continues. In the region, a
common practice is to remove trop residue fiom the farm to be sold. This practice
takes away nutrients that would have been available to plants for the next growing

---

18
season. Of@ biomass capable of restoring soil organiç matter leaves the fie!d as
fùelwood (Fig 1.) or as trop residues (fig.2).
Fig. 1. Biomass leaving the fàrm to the City as fùelwood
(Fansikoro, Mali)
Fig.2. Biomass leaving the field as
trop residues (Kaolack, Senegal)
Because field oRen remain bare between cropping seasons, soil nutrient losses may be
accelerated by runoff and wind storms. In addition, the removal of trees and other
types of vegetation on hilltops often fiuther increases soil nutrient losses during
rainstorm events (Fig.3 and Fig. 4).
__.-.-__
--
.---

---

19
Fig.3. Hilltop vegetation clearing a potential source of erosion (Fansirakoro, Mali)
Fig.4. Typical aspect of a I?~DI aRer trop harvest (Fansirakoro Mali)
Studies conducted elsewhere have often shown that substantial accumulation of soi1
organic matter cari be achieved ifcrop residues are incorporated into the soil aRer trop
harvest or even 1eA on the soi1 surface as mulch. Furthermore, mulch caver is know to
enhance soil biological activity, mod@ soil temperature, improve water conservation,
reduce soil sealing and crushing, and reduce soil erosion (Lai, 1974; Padwick, 1983;
Sivakumar et aL, 1992). Note that Pieri (1989) shows cases where the opposite occurs.
In fact, in the presence of manure (trop residues), it appears that in addition of other
factors not yet known, fertilizer management practices and soiI texture cari result in
either gain or loss of SO~I organic matter over time.
?Vhy biomass is not lefi on the fàrrn is not well understood but the lack of water to
_--
~- -.---,

---

z
o

---

---

---

23
Fig.9. Underground composting (C&ana, Mali)
0ther researchers bave reported that the lack of widespread practice of
composting was due to lack of water and limited availability of plant and trop residues
(Poulain, 1980). Ways to minimize constraints to manuring and composting must be
fùrther researched including how to improve its quahty (nutrient content etc.) in order
to be able to meet plant nutrient requirements. Managing manu-t-e for its P content cari
be also important for the region.
3. IMerCRSP/W& strategies to impmve food pimhction
Manzrre extenakr
Experience gained with the exploratory stnvey has stimulated InterCRSP/West
collaborators to propose new ideas on how to address some the problems mentioned
above. For example, experiments design to look at the synergistic effects of using a
mixture of chemical f%rti&r with manure are being conducted at Fansirakoro and
N’Tetoukoro, Mali. This approach is important for several reasons. 1) We know
fàrmers prefèr chemical fertilise over manure but they always have little to use.
Therefore, an idea to mix the two in small quantities ‘Ymanure extender” may be an
attractive alternative, 2) Chances are that the inorganic fertifizer will be much more
effective with the manure than separate (better water hokiing conditions in tbe manure
etc., 3) ll%ufked changes ~II temperatures should improve root growth, 4) Improved
environment for bioîogical activity in tbe manure-inorganic fert%zer combination, 5)
The manure may be providing other nutients such as micronutrients that are not
present in the inorganic fertilizer, but are needed in small amour@. For example the
requirement of zinc would be very httle and there may be enough in manure to provide
for some excellent growth as long as the N, P, and K are supplied by the inorganic

---

24
fertilizer. In fact, recent findings support the synergistic effect of the mixture of manure
and common fertihzers. Most importantly, this approach does not pit the manure
against the inorganic fertilizer as competitors. It likely should stimulate the improved
management of manure for use as an amendment for better nutrient management.
Finally, it is a low input operation, the only requirement is access to a little inorganic
fertilizer and the added benefit of saving labor fiom heavy mamuing (R Yost, persona1
communication).
Determining the value ofmanure @futrient balances)
CurrentIy, there are no guidelines for det ermining ifmanure applied is adequate
for trop requirements. However, the’aurount of manure to be applied on a given field
is dependent on the nutrient status of the field, the nutrient needs of the trop to be
grown, the nutrient content of the manure. Therefore, the need for simple spread sheet
software to calculate nutrient balances on yearly basis is imperative. This approach Will
help achieve improved nutrient recommendations to supply trop nutrients at a rate that
satisfies their needs and also maintain a balance in the soil. Definitely, without such an
approach the process of nutrient depletion will continue and even become more severe
every year.
Incentives for increasing production (Farm equipment, hflurkets, Prices
of farms products and inflastruc turcs,)
The unavailability of fàrm equipment appeared to be a serious concern for the
farmers surveyed. Farm work is physically demanding and because most farmers are
old, working manually long hours during the cropping season without any equipment
cari reduce overall labor productivity. SO,, we believe access to a minimum level of
modem equipment (cart to carry manure etc.) may boost productivity.
Analyzing experience in northeast Uganda in terms of its implications for future
development, both in Uganda and elsewhere in Af?ica, Carr (1992) pointed out two
striking facts. The first was the rapid positive impact on the productivity of small
farmers of a stable political situation combined with good infrastructure which tends to
encourage efficient and competitive pricing. The second is the underestimation by
govemment over many years of the farmers quick response to real price incentives. As
a result of its apparent lack of understanding of the underlying forces which motivate
farmers, successive govermrrents have been providing initiatives that reduced both the
quality of marketing and the level of economic incentives to tirmers. This author went
to add that this remains a common feature of governments in Sub-Sahara Afi-ica today.
Political exhortation or extension pressure is applied to farmers to produce one trop
while pricing signais encourage them to produce another (Cocoa and coffee in Cote
D’ivoire, Cotton in Mali, Peanut in Senegal etc.). Alternatively, çampaigns are
mounted to encourage the use of purchased inputs while pricing policies are reducing
the benefit to cost ratio of the proposed innovation.
We deeply concurre with the author that this may be the most critical issue for
the improvement of Afi-ican farming systems, but yet it seems to receive remarkably
little attention f?om policy makers and others. It is therefore apparent that research and
policy action should be directed to both at mot only raising the level of food supply but
air.. to increasc: the a. xagc standard of’living ffan;; households.

---

25
4. Conclusion
The problem of nutrient mining appears to be largely a resuh of both inappropriate soil
nutrient management and fertilizer cost policies. Evaluating current technologies used
in the region and adapted to existing farming practice appears to be the best option
there is to increase food output. This will require both fùndamental research and a
greatly increased level of field experiments if the problems of increasing production,
particularly in the drier areas, are to be solved. Any measurable result invariably will
depend on the level of understanding of the real problems of the small farmer face and
,the apphcabihty of these solution to these problems.
Iiterature Cited
Carr. 1982. Farming systems in the African Savana: Chapter Il. p 72-74.
.Lal, R 1974. Role of mulching techniques in tropical soils and water management.
Tech. Bull. 1. Jbadan, Nigeria: Internat. Inst. Of Tropical Agric.
‘Pie+ 1992. Fertility of soils: A &ure ,for fàrming in the West tican Savannah.
Springer Verktg, Berlin. 348~~.
:Poulain, J.F. 1980. Crop residues and traditional cropping systems of West Africa:
Effects on the minera1 balance and level of organic matter in soils; Propo&s for their
better management. P. 38-71. Organic cycling in Afiica. FAO Soils Bulletin 43. Food
and Agriculture Organization of the United Nations (FAO), Rome, Italy.
Sanchez, PA., C.A. P@ L.T. Szott, E. Cuevas, and R Lal. 1989. Organic input
management in tropical agrosystems, p. 125 1976. Properties and management of
SO&. III D.C. Coleman et al. (Ed,) Dynamics of soil organic matter in tropical
ecosystems. NifIa Project, Univ. of Hawaii, Honolulu.
Scai6e, M.k 1971. The long te.rm effects of fertihze farm yard manure and leys at
Mwanhala, western Tanzania. East Ahican Agric. Forestry J. 37:8-14.
Sivakumar, M. V. K., A. Manu, S. M. Virmani, and E. T. Kanemasu. 1992. Relation
between Chmate and soi1 productivity in the tropics. P. 91- 119. In: R. La1 and P.A.
Sanchez. Myths and science of soils tropics. SSSA Spec. Publ.No.29, SS!%, ASA,
Madison, WI.

---

26
THEME 1: FARM PROGRAMMIN GANDCOUNTRY
POLICY

---

27
Farmer to farmer visit
b@sta, 1. and L Anahory, hrida, Praia, Cape Verde
Several activities, which deal with erosion, sahnity and loss of soil organic matter, have
been implemented in Cape Verde; however, there has been a lack of communication
among farmers concerning the benefits or constraints of such activities. In order to
promote communi cation and transfer of experiences between &mers, LNLDA’s
researchers organized a farmer-to-farmer vi& in the Riieira Seca watershed. TO
conduct this vi& a group çomposed of twenty fàrmers, two researchers and a
technician spent a day in difberent areas of the watershed where &-mers explained their
a&vities related to soil salinity, Loss of soil organic matter andlor soi1 erosion.
Although many of these activities were performed in the past projects there was still
much learn fiom those present. The visit was considered a success and fàrmers
suggested that this type of activity is worth repeating. The impact of this visit Will be
monitored as part of the LnterCRSP activities for 1999.
1. rntrodnction
In Cape Verde, soil characteristics, topographie conditions and climate
constraints have produced a unique agricuhural. system The two major systems of
production are rainfèd agri&ure practiced mostly on the hillsides of the valleys
(ribeiras), and irrigated agriculture hr the valley bottoms and near the occasional water
source. The prevailing rainfèd agriculture activity on scarce arable land and population
pressure have led farmers to cultivate on steep slopes. The annual maize/beans
intercropping in these areas has accelerated erosion and soil degradation (i.e. loss of
organic matter). Lrrigated trop production is practiced whenever and wherever water is
available. Water supply is the major constraint to production and there has been an
increase in trop loss due to gradua1 decrease in water quality. S&e Cape Verde has a
short rainy season to replenish the aquifexs fàrmers are forced to overpump their Wells.
The successive water withdrawal causes Salt water intrusion consequently increasing
soil salirùty over time.
Several activities which deal with erosion, salinity and loss of soil organic
matter have been implemented in Cape Verde; however, there has been a lack of
communication among fàrmers concerning the benefits or constraints of such activities.
With this farmer-to-fàrmer visit, INIDA’ s researchers intended to promote transfer of
information and technology between fat-mers related to soil salinity, loss of soil organic
matter and soil erosion.
2. Methodology
Visit prenaration
In order to conduct the fit-mer-to-farmer visit we began by contacting several
farmers of Ribeira Seca watershed with experience in one or ail the areas related to our
objectives. We celected about twenty fqrmers, \\xlith snmething +n tea@ nr le~rn, and
explained to them the objectives of the visit. Knowing the number of people that would
be involved in this activity we rented two vans to accommodate the visitors and
ordered lunch as we intended to stay all day in the field.

---

28
-ne visit
The visit was done on July 27&. We lefi INIDA, S. Jorge at 8 AM with the
farmers, drivers and INIDA’s group composed of two researchers and one technician.
The visit was done in difkrent areas of Ribeira Seca watershed such as Lém Jorge,
Maeati, S. CristovZo, Caiumbra, Mato Ahonso, Mendes Faleiro Cabrai, Godim and S.
Jorge, where both rainfèd and irrigated agriculture are practiced.
At the end of the v-kit and ptir to returning home, we had hmch at a fkmer’s
field while the visitors gave their impressions about this experience.
3. Discussion
a, soil salinity
1. Farmers activities
Salinity was seen in most fields we visited, especially in those where flood
irrigation is common. In order to overcome this constraint several farmers suggested
that this irrigation method be replaced by drip irrigation. Others have been planting
tomatoes closer to the Wells fohowed by bananas and sugar cane since the former
requires fkequent irrigation and the Gloser to the well the less pumping is needed.
2. Researchers’ nersnective
Sahnity of both soil and water cannot be solved ovemight. Although fàrmers
acttiies are satisfactory, there are other actions that need to be considered. For
example:
a. Crop rotation systems using salt resistant crops.
b. Avoid removing sand tiom the coastal areas since it is causing sait water
intrusion and decreasing water quality in the aquifkrs.
c. Use of organic materials to improve soil structure and consequently improve
infiltration and drainage during the rainy season.
b. loss of soif organic matter
1. Farmers activities
Loss of soil organic matter is a constraint throughout the Ribeira Seca watershed.
This problem has been difkult to surpass since, as exphined by farmers, at the time of
harvest they have to leave the soil bare in order to have some feed for the animais.
Some of them, however have been using, when possible, animal manure and/or banana
leaves in the soil to replenish its organic matter content.
2. Researchers’ uersnective
Generally the soils of Santiago island are low in organic matter (4 to 2%) and the
quantities of manure and/or organic materials necessary for major change in soi1
organic matter status are large. However, we would suggest, in addition to farmers
activities, the use of compost made of household trash and other organic material to
improve soil organic mater content.

---

29
c. Soil erosion
1. Farmers activities
In Riieira Seca watershed corn/beans intercropping is the prevailing rainfed a&vity,
but when practiced on steep slopes it results in aggravated soil loss and erosion. Given
this problem, many projects have launched programs designed to reduce erosion and
promote soil conservation. Of these programs we emphasize those that have been
adopted by farmers of Ribeira Seca which they demonstrated during the visit:
?
Construction of water and soil retention structure such as dams, bunds, retaining
walls and underground reservoirs done by public work fronts and the government.
?? Construction of contour vegetable barriers and contour rock wall terraces to
decrease runoff.
?
Conversion of large areas of rainfed com/beans intercropping to pigeonpea since as
a perennial, it protects better against erosion.
?
Use of trop residue as mulch to increase inbltration and reduce raindrop impact.
?
Use of a difFerent technique of cuhivating peanut on the slopes, for example, by
cutting ridges along the slope to act as water channels and decrease hill erosion on
peanut fields.
2. Researchers’ perspective
Soil erosion control measures used in Ribeira Seca watershed have significant
impact in reducing ru.noE, however, we believe that many farmers in that area are not
aware of its benefïts.
4. Conc.lusio~olnmendalion
In order to promote communication and transfer of experiences between
fàrmers, inserted in the interCRSP activiti.es, INIDA’s researchers organized a &-mer-
to-f%rmer vi& in tbe Ribeira Seca watershed. According to the farmers who
participated in this event, this activity was successful and very beneficial since they had
the opportunity to meet otber colleagues, learn and teach some techniques which will
help them reduce soil salinity, soil erosîon and loss of soil organic matter. They
suggested that this activity should be done as oRen as possible involving farmers tiom
other areas of Santiago island since they are sure there is a lot to learn. The consensus
was that the farmer-to-farmer visit was a great experience and worth repeating.

---

30
Adoption and Farm Level Impact of Improved Fertility
Maaagement Technologies in the Sudano Sahelian Zone
of The Gambia
Mohamed Kebbeh, National Agricuhural Research Institute (NARI), The Gambia
Abstract
L.-ow inherent soil fertility and decreasing Ievels of chemical fertïlizer use have partly
explained the significant decline in trop yields in Gambian uplands. The problem has
bec* aggravated by rapid expansions in cropped area and a reduction in the traditional
frxllow periods. Researchers are attempting to address this problem by evahzating
alternative soil fertility management strategies in eastem and western Gambia. Initial
results show increases in cereal yields resulting fiom the use of inorganic fertlfizer in
combination with chemical fertilizers.
Research efforts have so tifa, focused on agronomie parameters, with Little attention to
important socio-economic fàctors that determine potential adoption and the subsequent
impacts of these technologies. The introduction and adoption of new or improved
fextility management technologies have important socio-economic and farm resource
allocaltion implications, and an understanding of these fàctors would assist in the
generation of technologies that are consistent w=ith farmers’ production objectives and
adequately reflect the conditions and constraints fàcing fàrmers for whom these
technologies are being developed. Also, use levels for extemal inputs like chemical
fertilizer are determined largely by relative price levels. The sigr&ant decfine in
chemical fertilizer use in the country over the last decade has been attriiuted to sharp
price increases resultîng fiom the removal of ti subsidies.
The general objective of this study is to determine the adoption and farm level impacts
of different soil fertility management technologies under alternative pricing policies.
Whole fàrm models are developed to assess the adoption and tirm level impacts of
combinations of organic and inorganic fertiliser. The results show positive response to
chemical fertilizer price changes. Area under millet increases dramatically, at the
expense of groundnuts and ma&, with a 25% decrease in chemical fèrtilizer prices.
Also, levels of urea and compound fertiliser use and fart level profits increase with the
decline in chemical fertiliser prices. The results show a slight decrease in organic
fertilixr use under the lower input price scenario.
L Probiem Statement
With rapid population growths and limited access to production resources like
land and investment capital, Sahelians are under increasing pressure to address the
important factors that have constrained agricultural productivity in the region. In some
Sahelian countries, attempts have been made to address the problem of declinîng
:productivity through intensification of land use. Also, the introduction of draft animal
tractictn technology in the Sahel over the last twenty years has resulted in vast
expansions of area under trop cultivation and a near elimination of the traditional bush
fallow system that was the principal means of regenerating soil fertility.
Low trop yields and persistent declines in the production capacity of Gambian
uplanti soils are generally attrmuted to tow inherent soii iertiiity and low levels of
organic and inorganic fertilisation to replace removed nutrients fiom the soil.
S@ificant reductions in inorganic fertilizer use following the removal of farm subsidies
during structural adjustments have resulted in rapid soi1 degradation in Gambian

---

31
uplands and a faihu-e to replenish soil nutrients. In addition, access to organic fertilizer
is llimited for the major@ of farm households. Rapid expansions in area under trop
cuhivation and a near elimination of the traditional fallow system have compounded
the:se problems. It is not surprising, therefore, that a number of fàrmer constraint
studies have identifïed declining soil fertility as a primary constraint to increasing
productivity in the fàrm sector (Mills et al.; 1987; Boughton et aL, Torrence, 1989;
1987; Kebbeh et al., 1996).
Research efforts to address the problems of declining soil fertility in Gambia
uplands have intensified over the last few years. The Cropping Systems and Resource
Management Promam of the country’s National Agricultural Research Institute
continues to evaluate alternative fertility management options in a number of on-tirm
and station tri&. These research efforts have focussed on establishing trop yield
responses to alternative fertility management technologies. In eastern Gambia, d.ifEerent
combiuations of trop rotation and organic and inorganic fertilizer levels are being
examined in trials initiated last year. Initial results indicate that application of inorganic
fertilizer in combination with organic fertilizer significantly increased yields for millet,
maize and groundnuts. In the western half of the country, a long-term tria1 aimed at
assessing the long-term effects of soil amendments and the interaction of manure and
various levels of inorganic fertilizer was initiated last year. Initial results indicate that
organic manure in combination with chemical fertilizer gave higher mean yields than
inorganic fertilizer alone. Similar studies in Mali and Senegal have also reported
positive yield responses to inorganic fertilizer used in combination with organic
fertilizer.
As already pointed out, agricultural research efforts on soil fertility
management have SO far focussed on agronomie parameters, with little attention to
important socio-economic factors that would significantly determine potential adoption
and! the subsequent impacts of these technologies. The introduction and adoption of
new or improved fertility management technologies have important socio-economic
and. farm resource allocation implications, and an understanding of these factors would
as& in the generation of technologies that are consistent with farmers’ production
objectives and adequately reflect conditions and constraints facing farmers for whom
these technologies are being developed. This is critical for the sustainability of
introduced f&m technologies and potential payoffs to soil fertility maintenance
research in The Gambia. As efforts to generate sustainable soil fertility amendments
continue, the following questions are important.
--,
Given farm resource endowment and potential performance of the technologies
being tested, what is the potential of adoption (i.e., Who will adopt, what will
they adopt, how much will they adopt)?
mm
How Will adoption of technologies affect farm level resource allocation and
trop / livestock mix in both the short and long run?
- -
What Will be the farm/household level impact (farm output and inçome) of
adoption.
- -
How sustainable are these technologies?
- -
What are the effects of alternative farm policies on the sustainability of
potential technologies?
r’&li&ng these questibus \\til: be critical to tht lti,ig-rmi viability of poi&al
interventions to address the problems of soi1 degradation in the country. An important
consideration is that agriculture in the country be considered f?om a holistic approach.
At the level of the small farmer, soi1 fertility maintenance is viewed fi-om a whole farm

---

32
perspective, with farmers taking advantage of residual nutrients fi-om previous
applications through trop rotations, tethering of livestock and short-term fallow. The
adoption of improved soil fertility maintenance strategies would delïnitely have
re,source allocation implications at the household level, and it is important to
understand these. Also, farm households fiom difTerent resource categories may adopt
difEerent strategies, and it is, therefore, important to have an understanding of the
difiterent farm resource categories in the intervention areas.
2. Objectives
The general goal is to determine the adoption and fàrm level impacts of
potential soil fertility management technologies. The technologies evaluated are; use of
organic fertilizer at 2.5 tons per hectare, use of inorganic fertilizer at the recommended
rates, use of half the recommended rate of inorganic fertilizer in combination with
organic fertilizer, use of the recommended fertilizer rate in combination with organic
fertilizer, and a control of no organic or inorganic fertilizer. In addition, difFerent
fertilizer prices are used to evaluate the effect of alternative input pricing policies on
the adoption and farm level impact of improved fertility management technologies.
The specific objectives of the review are to:
1)
Determine the fàrm level adoption of di@erent combinations of
inorganic and organic fertilizers in upland fields,
2)
Determine the impact of new fertilizer management technologies on
cropping patterns
3)
Identti the fi-u-m level profit effects of improved fertility management
technologies; and
4)
Determine the effects of inorganic fertilizer price changes on the
adoption and farm level profits impact of improved fertilizer
technologies.
3. Methodology
Stndy Area: The Sudano Sahelian Zone
Agroecological Characteristics
One of three major agroecological zones in the country, the Sudano sahelian
zone covers all of Gambia except the Western Division and extreme north of the
Central River Division. The zone is characterised by savarma woodland vegetation with
an average atmual ratiall of 600 to 800 mm. Like other regions in the country, rainfall
in the sudano sahelian zone is very erratic, and the rainfall window has decreased
consistently over the last decade. The period of plant growth ranges fi-om 90 to 100
days, beginning in July and ending in September/October.
Prtduc tion Sys tems
The principal upland crops in this zone are groundnuts, millet, and maize.
Sorghum, sesame, cotton and cowpea are also produced but to a limited extent.
Agriculture in the sudano sahelian zone is characterised by both traditional rainfed and
semi-intensive production systems. The traditional rainfed system, referred to the ‘low
input extensive upland system”, is characterîsed by extensive cultivation made possible
Uq a relatively lob\\ r~&&~~ density. LOT,. :CI~& uf trop managcicltilri ;,, tlris system
are apparent in extremely low plant populations, use of traditional trop varieties and
limited use of extemal yield augmenting inputs like inorganic fertilizer. Production
increase is largely accounted for by expansion of area under cultivation. ‘I’he semi-

---

33
intensive system on the other hand is characterised by the use of improved production
techniques and the use of external inputs like inorganic fertilizer. In addition, a higher
trop-livestock interaction under this system facilitates greater access to organic
fertilizer.
Primary Constraints
Declining soil fertility has been identifïed as the principal constraint to
agriculture in the sudano sahelian zone. The problem is attriiuted to low levels of
inorganic fertilize use, decreasing Mlow and continuous cropping. In inland villages
away from the river, reduced access to organic fertilizer due to the out-migration of
livestock during the long dry season has compounded the problem. Soil erosion, lack
of access to f%rm credit and pest control chemicals, and inadequate farm labor supply
are other important constraints to the farm sector in this zone.
Land Use and Famz Mechanisafion
In the Sudano Sahelian zone, land use is categorised by both plot location and
structure of ownership. Inner fields are located close to village settlements and receive
most of organic fertiiizer used for trop production. These fields are usually tethered
during the dry season, and then put to maize the following rainy season. Outer fields
on the other hand receive very little organic fertilizer and are used predominantly for
the production of cash crops like groundnuts. In terms of ownership, fields cari be
either cultivated individually or communally. Individual fields are generahy put to
groundmtts, which is the principal cash trop. Household staples are cultivated on
communal fields.
The majority of farm households in the Sudano sahelian zone either own or
bave access to draft animal traction. Oxen, donkeys and horses are the predominant
sources of draft power in this zone. Crop production activities for which animal
fraction power is used include seedittg, weeding and harves&rg. Like in most parts of
the country, deep pfowing is uncommon in the sudano sahelian zone. Dry seeding just
prior to the first rains or direct seed immediately aRer is the most common practice.
Data Requirements and Sources
Data requirements include human and animal traction labor requirements and
supply, input and output prices, trop yields, farm resource endowments and seasonal
Tweather variation’. Human labor requirements are based on labor budgets for the key
crops, and supply is based on the structure of the typical farm household. Crop yield
data are derived fiom adjusted results of trials conducted at the research stations and
on farmers’ fields. Resource endowment covers land, liquidity and other fart inputs.
The Theoretical and Empirical Models
The adoption of soil fertihty maintenance strategies cari be seen as a problem of
portfolio allocation since adoption requires the allocation of farm resources given
production objectives and resource constraints. An ex-ante evaluation of the different
technologies (f?om a whole farm perspective) is done using mathematical programming
to simulate farm level adoption and profit efbects of soil fertility management
technologies. This paper adapts the mode1 developed by Coulibaly (1995) and uses the
Direct Expected Utility Maximization (DEMP) program2 which is generally formulated
as:
1 Data tilA LOI thL models represe.Il, o-b& &unates and are subjtbL cv LL1 udification followii~g
verifkation. The mode1 results presented later are therefore illustrative.
2 For a detailed presentation of the theoretical model, sec Coulibaly, 1995

---

34
Max E[u(W*)f = C 6,. U(W)
wbere a negative exponential utihty fimction is given as
U(W) = -e 73
Here
WS
= vector ofnet revenues tiom trop production activities
= states of nature
Er.1
= expectation operator
r:*
= coefficient of absolute ris%- aversion
Crop production activïties are undertaken subject to resource endowment
constraints to caver availability of land, human labor, draft animal traction and
liquidîty. In addition, non-negativity constraints are imposed on all decision variables.
For each resource category, the constraint is depicted as;
ZZAijXgIL;
where Ai = input requirement for trop activity j at period 1
X, = hectares of trop j at period I
Li = resource availability at period 1
The objective iùnction in the empirical mode1 is to maximize the expected
utility of net farm revenue subject to constraints on access to liquidity, land, lmman
labor and animal traction power. The farmer is assumed to be risk averse and
production activities caver three crops using five fertihty management options per
trop. The mode1 thus includes a total of fif&en activities. Three states of nature and
their probabilities of occurrence are used for çomputing expected yields and fàrm
.pro&s. Output prices also vary by state of nature. Liquidity and total land available for
production are restricted to reflect access to these resources by small and medium farm
households in the Sudano Sahelian zone. Production activities are broken into five
periods and constraints on human labor and animal traction availabihty reflect total
time available for identifïed activities and the active labor force and number of animals
per farm household in the zone.
4. Disctim ofRes&s
The mode13 is constructed to reflect production practices with different fertility
management technologies. Crop varieties and cultural practices reflect current farmer
practices. The results represent optimal trop choices under difherent organic and
inorganic fertilizer regimes. The control treatment reflects cropping with no organic or
inorganic fertilizer. Other treatments are; use of organic fertilizer alone at 2.5 tons per
hectare, use of chemical fertilizer alone at the recommended rates, use of organic plus
inorganic fertihzer at half the recommended rate, and use of organic plus inorganic
fertilizer at the recommended rate. The crops considered are groundnuts, millet and
maize. Sorghum and sesame are not inchtded in the models because they represent a
very smaII proportion of the total cereal production. The base mode1 represents
production choices under crurent inorganic fertilizer price levels. The mode1 is then re-
estimated assuming a reduction in the price of inorganic fertihzer. Ibis scenario depicts
3Cheffkients are based on estimates and may net lx accurate representations.
Al1 models
Will be te-estimated using more accurate cceffkients.

---

35
a pricing policy where the output prices increase relative to input costs. Estimation
results are discussed in this section.
Technology Choices and Total Farm Output
Optimal Soil fertility management technology choices and output under the two
price scenarios are presented in Table 1. At higher inorganic fertilizer CO~~S, the optimal
ph is to use lower levels of inorganic fertilizer. The results indicate cultivating 3.984
hectares of maize using organic fertilizer on@. Groundnut is also cultivated using a
combination of inorganic and organic fertilizers at half the recommended rates. These
results depict the trend in inorganic fertilizer use in Gambia over the last decade. With
the removal of fertilizer subsidies, a sharp increase rn the cost of inorganic fertilizer has
re,su.lted in signifkant declines in chemical fertibzer use. In the sudano sahelian zone,
maize is cultivated in inner fields close to the village, and these fields receive most of the
organic manure used for production. Groundnut on the other hand is cultivated primarily
in outer fields and is a priority trop for inorganic fertilizer application This could be
because groundnut is the principal cash trop, hence more likely to receive external inputs
like chemical fertihzer. Millet is not included in the optimal plan given the higher price
scenario although it is the staple cereal in the sudano sahelian zone. The results suggest
that at higher input costs, it is not profitable to invest in external inputs for millet
production.
The second scenario ikstrates the farm level effect of increasing output price
relative to input costs. Here we assume a policy of a decrease in the real cost of kputs.
The results show a signifkant decline in maize and groundnut production at lower
chemical fertilizer prices. Resources are diverted f%om maize and groundnut to millet
production under this price scenario. Use of organic fertilizer alone continues to be the
optimal technology choice for maize. For both groundnuts and millet, the optimal plan is
to use half the recommended rate of chemical fertilizer in combination with organic
fertilizer. The results suggest higher levels of cereal (millet) production. at affordable
fertilizer price levels. This is consistent with arguments that for cereal production to be
attractive to small-scale producers in the country, output prices must increase relative to
chemical fertihzer costs. This has important policy implications and will be a focus of
further analysis following re-estimation of coefficients and data modification.
--II-
---
-_-.
-

---

36
Table 1: Optimal Crop Choices and Output Levels under DifI-èrent Price Scenarios
*Scenario A
*Scenario B
Technology
Crop Area (ha)
Output (kg)
Crop Area
output
** Millet (C4)
-
0
0
3.027
4607.8
Maize (C7)
3.984
3854.376
1.407
1361.02
-
l Groundnuts(c14)
2.016
3268.379
1.5664
2539.088
* Scenario A. existing fertilizers prices, Scenario B: 25 % reduction in fertilizers prices
‘** C4 = NEllet with halfthe recommended rate of chemical fertilizer Phu manure
C7 = Maize with manure only
Cl4 = Groundnuts with halfthe tecommended rate of chemical fertilizer plus manure
Input Levels and Expected Fart Profits
Mode1 results for input use levels and expected farm profÏts under the two prix
scenarios are presented in Table 2. The results show that inorganic fertilizer use among
6u-m households continue to decline with higher chemical fertilizer price levels, resulting
in decreasing expected ~&II profits. At higher inorganic fertilizer CO&, the optimal plan
bxpports using chemical fertilizer on grouudnuts only, the primary cash trop. Under this
rxenario, compound (NPK) fertilizer is used for groundnut production, with no
investment in urea. Organic nxumre is used for both maize and groundnuts production.
The results explain the sharp decline in chemical fertilizer in The Gambîa followîng the
removal of înput subsidies and sharp increases in the price of inorganic fertilizer.
Table 2. Input Levels and Expected Farm Profits under Dif&rent Input FVices
Input Level/Farm Profit
- -
* Scenario A
Scenario B
Urea (kilograms)
0
151.349
Compound (kilograms)
100.807
305.338
- -
Organic Manure (tons)
12.48
9.26
- -
Expected Farm Profit (Dalasis)
D 15113.678
D 15557.254
1E Scenario A existing fertlizers prices; Scenario B 25 % less costly fertilizer.
Constraints to the Adoption of Fertility Amendment Technologies
Land, labor, animal traction and liquidity constraints are specified in bath
models. Binding constraints and their sbadow prices are shown in Table 3. The resuhs
show liquidity and land to be binding under both price scenarios. The lower shadov,
price for liquide under the higher price scenario provides fiu-ther evidence of decline
in chemical fertilizer use with input price increases. The decline in chemical fertilizer

---

3 7
use implies lower liquidity requirements for inorganic fertilizer purchases. Wth
reductions in the relative price of chemical fertilizers however, increases in the levels of
urea and compound fertilizer result in a higher liquidity requirement. The shadow
prices for land are extremely high under both scenarios, suggesting that the
specification of the land constraint in both models may not reflect farmer
cir~ances. This issue will be reviewed when the models are re-estimated using
more accurate coefficients.
Table 3. Farm Constraints under DifGrent Price Scenarios
Con.straint
Shadow Price
* Scenario A
Scenario B
Liquidity
101.921
111.993
Land
9341.651
8229.947
* Scenario A - existing fertîlizer prices; Scenario B - 25 % less costly fertilizer
4. Recommendations for Futnre Research
General Recammendatîons
The succes& introduction of new technology requires an understanding of the
important factors underlying potential adoption. An understanding of fhrme:r behavior
is critical to the adoption of new technology. It is important to understand the
socioeconomic characteristics and behavior of tiers for whom technologies are
being developed. Among these include the resource base of farm households and the
impact of farmer behavior on technology adoption. Technology adoption is very oRen
a problem of portfolio allocation, addressing questions of what will be adopted, wbat
will be the mix of new and old technologies, what proportions of farm resources will
be allocated to the Werent technologies, etc. There are d.ifIèrences in resource
endowments at the fart household level, and it may be useiùl to distinguish between
high, medium and low resource categories so that recommendations cari be made for a
broader range of producers based on levels of resource endowment. Very ofien, whole
tirm models are developed for the ‘average or representative farm’, and it may be
diflïcult to use results fiom such models to make recommendations Xlevels of resource
endowment are di&rent.
Given farm resource constraints, the levels of resource endowment Will
determine what technologies are feasible and what the mix of diEerent technologies
will be. DifEerent soil fertility management strategies may require Werent levels of
investment in inorganic fertilizers, other farm inputs, labor and implements. In addition,
there are implications in terms of ownership and control of Iivestock and access to
organic fertilizer. The Mali models for example use a representative farm with a land
endowment of 19 hectares. The result is that land does not corne out as a binding
constraint. However the initial argument is that high population pressure, together with
the introduction of animal traction and near elimination of traditional fallow have
increased pressure on farmlands, accelerating soil degradation. Ose would therefore

---

38
expect some form of land constraint. Land could be an important constraint for
hauseholds with low and medium resource endowments, and this may result in
dif&rent adoption behavior and d.i&rent fat-m level impacts. Whether or not land is an
important constraint could have significant implications for soil fertility management,
and this ne& to corne out in any analysis of fertiiity management technologies.
Sir arguments hold for labor or other inputs. Dalton (1996) for example
di%rentiated between highly capitalized and Eess capitalized farms in southern Mali
because ‘ less capitalized 43n-ms have only halfthe livestock un& per hectare of bighly
capitalized fm As such, their response to opportunities to intensify trop production
with organic fertilizers, derived f?om animal manure. is dramatically d.if&rent than
highly capitalized f5rms.’ It may therefore be useN: rcr develop fart models reflecting
at least three resource endowment categories, medium, low, and high. SensÏtivity
analysis mïght do it, but experience with ‘adjustments of cropping systems to producer
price changes in eastern Gambia’ is that better results are obtained by looking at
di.%nnt resource endowment groups. This will involve going back to the survey data
and looking at household resource levets to sec how they vary.
The response of fàrmers and farm househohis to different technologies will be
important in det ermining the relative merits of diflèrent fertility management
interventions. In addition, there is a need to be able to determine the potential
household or farm level impacts of introduced versus existing technologies. In this
regard, it is important to develop an understanding of who the potential adopters are,
what the constraînts to adoption are, tiat potential technologies will be adopted, how
much Will be adopted (i.e., mix between new and oldj, and what the implications are in
terms of resource allocation and meeting fàrm production objectives. It is important to
be able to determine the value of mitigating constraints to the adoption of specific farm
technologies and evaluate the effects of alternative fat-m policies on adoption and fat-m
income. The Mali mode1 evaluates technology impacts under devaluation, and the
results suggest that policies that affect the relative prices of inputs and fàrm output wih
determine adoption and farm level impact. ùritial mode1 results for Gambia support
these results and need to be fùrther anafysed.
Also, the models estimated in this paper consider only one level of ri.&
aversion, aad incorporation of different levels of risk aversion will give an indication of
how farmers will respond to technologies under di&rent levels of risk aversion.
Soil fertility management bas important long-term considerations that are central to thc
sustainability of interventions. Crop rotations (like cereal-legume rotations) and short
ter-m fallot35 are used to make effective use of residual nutrients fiom inorganic/organic
fertihzer application. The general abjective of the soi1 fertility maintenance research
project in Gambia is to build up soi1 organic matter and nutrient content over time to
support sustainab’fe production. It is therefore important that soc&economic analyses
of these trials take account of the long term sustainability of the interventions.
Following the development of farm models to evaluate potential adoption and farm
level impact of fertility management te&nologies, sustainability cari be evaluated C&I~
dynamic/muhi-period programming.
S’ec~jk Recomnendafions
For Gambia we intend to build on the experiences of this paper and work in
!Va!; !n develop fart prfrdur+ion models for thrree principal zones; Sshc?ian, Sudano
Sahelian and Sudano Guinean. These zones are the focus of current national
agricuhura’l research efforts. Unlike Mali and Senegal, more work is needed in Gambia
on putting together data for programming work. Options include supporting modest

---

39
farm surveys in intervention areas or adapting data fi-om similar zones in other
countries of the region where this Us applicable. Although Gambia is not in the CFA
zone, its physical location and volume of trade with Senegal warrants a look at the
impact of some CFA zone policies hke devaluation. This would allow good
comparative analysis.
Literature Cited
Ahmed, M.A. 1994. Tntroducing New Technologies On The Vertisols of Eastern
Sudan: A Dynamic Programming Approach. Ph.D. Thesis, Denartment of Agricuhural
Economie, Purdue University.
Coulibally, O.N.1995.Devaluatioq New Technologies, and Agricultural Policies in The
Sudanian and Sudano-Guinean Zones of Mali. Ph.D. Thesis, Department of
Agricultural Economie, Pur-due University.
Decosse, P. J. 1992. Structural Change in Gambian Agriculture: Stagnation or Silent
Transformation? Research Paper, USAJD, Banjul.
Kebbeh, M. 1990.
Adjustments of Cropping Systems to Producer Price Changes in
Eastern Gambia. MS Thesis, University of Wisconsin-Madison.
Shiferow, B and S. T. Holden. 1998. A Farm Household Analysis of Land Use and Soi1
Conservation Decisions of Smallholder Farmers in The Ethiopian Highlands. Paper
presented at the fourth Afiican Farm Management Conference, Stellenbosch, South
Africa. 26-30 .

---

40
Improving and Sust~ Food and Raw Material
Production in West Africa: A Participative Rapid Rural
Appr&al in Fansirakoro
A. H. Berthe, O.Doumbiaz, M. L Sylla’, MX. Djitteye’, II. A. Taïga’, S.F. Traoré’, RA
Kablan3, RS.. Yost’
1 ESPGRNSotuba, BP 9030, Bamako, Mali
2 Laboratoire Sol-plante-eau, Institut d’Economie Rurale, BP 438, Bamako, Mali
ity ofHawaiii 1910 East West Road, HonohtIu HI 96822
1. Introduction
The context qf th Participat-trve R@d Rural Appraisal (PRl&f)
Farmers like researchers, extension workers and firms are involved in generating,
tran$dg and utilizing technologies. In this respect the fàrmers’common knowledge
should be part and parce1 of the dominant paradigm of the development process. We cari
assume that &rmers have a crucial contribution to make, or rather, are already making a
crucial contribution by their continuous effort to adapt their fàrming to the changing
enviromnent.
Farmers’ effort should be integrated in our effort of officia1 system of
technology development process. Knowledge development and utilisation is the key
ingredients for continuity and growth in a society. Farmers and their knowledge are
potential components of the institutional knowledge system which cari or cannot be utilised
to advantage for rural development. In this process fat-mer cari be considered as a
knodedge manager.
This Participative Rapid Rural Appraisal (PRRA) was done as an attempt of
capturing firmer’s research effort as a basis for InterCRSP research and/or technology
development process. This paper ms the methodology and preliminary results of a
participative rapid rural appraisal conducted in the village of Fansirakoro in Southem Mali
Fansirakoro is located in the Zone of the Oflice de la Vallée du Niger (OHVN). The thirty-
eight rural agricultural production units @PU) of Fansirakoro were surveyed and in-depth,
open-ended interviews were conducted with the heads of PUS.
2. Objective
The main objective of this PRR4 was to provide baseline mformation on the
huming systems found in the OI-IV zone of Kati in order to establish research priorities for
the InterCRSP NRM project.
3. Sites presentation
The OHW Zone
The Office d’e Developpement de la Haute Vallée du Niger (OHVN) is a government
agricultural development parastatal institute created in July 1983. Its zone is divided
into six (6) administrative sectors: Kangaba, Ouelessebougou, Bancoumana,
Koulikoro, Kati and Banamba. These sectors were subdivided into thirty ti~re.~
d’Ex~xm.sion Rurale (ZERs) and the ZERs were :Cwrther subdivided into one hundred
and sixty Secteurs de Base (SB). The OHVN provides technical and extension services
to approximately nine hundred and thirty nine villages and settlements in the area.
The population of the OHV zone (533,687 persons in the last census in 1998

---

41
comprising :!65,029 males and 268,658 females) is predominantly rural, deriving the
majority of revenues tiom agricultural activities. The population density is 16
inhabitants per Kilometer Square. The ammal population growth rate varies from 3 to
3.5%). The ,main ethnie groups found in the area are the Malinké, the Bambara, the
Sarakhollé, fhe Peulh and the Bozo.
The QHVN zone encompasses an area of 34,500 km2 which represents 2.5% of the
.Malian territory and 11% of the cultivated rainfed area. It covers the area immediately
surrounding the capital City, Bamako,, and hence has access to the largest market in
Mali. The climate varies fiom Sahelian (600 mm rainfàll) in the northern sectors of
Banamba and North Koulikoro, to Sudano-Sahelian (1,200 mm ramfàll) in the south of
Ouelessebougou, Bancoumana and Kangaba. From 1981 to 1985 rainfall was low, but
in 1986 the drought appeared to dimini& and t,he pattern approached the 50 year
average. The mean temperature varies seasonally from 26 to 36OC.
Arable lands in the OI-IVN zoqe represent 60% (20 820 km2) of the total land
area and only 7% (145,700 ha) of the total arable land are cultivated. The plains
classified as ‘Q-pic Ustropepts are SO& of tic silt, clay materials, or sand, laying over
lateritic hardpans. The topography of tbe region is dominated by the Mandingue
mountains in the sectors of Kati, Banco-a, Kangaba and Koulikoro, and the Niger
river, which crosses the region fi-om Southwest to Northeast. Lands of variable
topography lay between the Niger and the mountain chains. Mountains and cliff areas
have .predominantly rock outcroppings and wash areas.
The broad alhrvial plains of the Niger are the agricultural heartland of the region.
Rice, millet, F;orghum, vegetables, and cash crops grow in this area. Rice is also grown in
Eowlands around the villages. In the sectors of Kangaba, Bancoumana, Ouelessebougou,
Kati and South Koulikoro, millet and sorghum are grown in gravelly sandy soils (typic
CuirustalfS), generally &y-silt or silt-clay mixtures. In the dominantly agropastoral sectors
of Banamba and North Koulikoro, souna millet is the principal trop. Tree crops (mangoes,
karité, etc..) are also important throughout the region.
The road system in the OHVN Zone comprises four main axes starting at Bamako
in the directions of Kati, Banamba, Kangaba and Ouelessebougou. In addition, the area is
served by a network of secondary roads, many of which become un& for motor vehicles
during the rainy season. Health and education centers are located mainly in the larger m-ban
centers and the main villages in some Sectors de Base. These centers are more developed in
the Sectors ofKati and Kangaba than in Banamba, where there are rare.
The Sector of Kati, the site of the project corresponds to the administrative
circle represents about 47% (16 100 km’) of the total area of the OHVN zone. It is
divided into three natural regions, the Plateau mandingue (PM), the Plateau of Koutiala
(52.38%) and the Haut Bani Niger which represent respectively 52.38%, 10.8 % and
37.45% of the area, respectively(PIRT, 1986).
The Plateau Mandingue comprises two agro-ecological zones (PM4, PM5) which
represents respectively 19.4 % (3210 km’) and 32,98% (5310 km’ of the total area of
the sector of Kati. PM 4 is located in the Southeast part of the Monts Madingue in the
south Soudanian zone. Arable lands caver 44% of the Monts Madingues and are located in
the valleys. Soils are quite deep and well drained. Non arable lands caver 53% of the area.
These soils are shallow and very susceptible to erosion. Rocky soils compose 3% of the
Mouts Mu:;il;ü,tii: area.
The Wenia agro-ecological zone (PM5) is limited in the East by the Plateau of
Koutiala (PK) and the Haut Bani Niger, in the West by the Gangaran, in the South by the
Monts Mandingue and in the North by the Bélédougou. This zone is located in the South

---

42
Soudanian zone. Arable lands caver 74% ofthe Wenia zone and non-arable lands 122%.
Fansirakoro belongs to this agro-ecological zone.
The second natural region, the Plateau de Ko&ala is limited to the North by the
Niier Delta, to the East by the Plateau of Banfora (Burkina Faso), to the west and South-
West by the Haut Bani Niger FN). Gnfàll varies fiom 650 mm to 1300 mm. The agro-
ecological zone of the Moyen Barri Occidental (PK3) which concems the sector of Kati is
located to the West of the Moyen Bani and Oriental @X2) runs through Baguiueda up to
the Bat&Bagoé river system Arable lands covers 53% of the total area sud are dominated
by plains (41%)
The third natural region found in the sector of Kati, tbe Haut Bani Niger (HBNl,
19.87% around 3200 km2, and HBN2.,17.58%
around La~0 km2) is limited by the
Plateau mandingue in theNorthwest, the Plateau de Ko&ala in the Northeast and the East,
the Plateau du Foniokourou in the South and by the border of the Country of Guinea in the
West. It is divided iu four agro-ecological zones fiom which two zones are part of the
sector of Kati
The Haut Barri-Niier occidental agro-ecological zone (HBNl) is located to the
South of ther Monts Mandingue and the Malian-Guinea border to the West, and the Moyen
Baoule to the East. The climate is of the Sot& Soudanian type. Arable lands represents
70% of the total area of the zone. The HBN2 corresponds to the agro-ecological zone of
Djitoumou .which is located to the Sauth of the district of Bamako. lt covers the extreme
North of HBNl and its climate is of the South Soudanian type. Arable lands represents
85% of the Qjitoumou zone.
The team before the survey conducted a literature review. It included documents
fi-om the difhèrent technical divisions of the OHVN, annual reports, follow-up reports,
specifk reports for assesrments and monitoring. D&ussions with resource persans
(extension StatTat OHVN headquarters and in the Sectors of Kati) were also conducted.
The survey was performed witlt the help of an open-ended mformal topic
guideline at two levels: village and production unit (PU). Its informa1 nature
contributed to the open and flexible interviews that the multidisciplinary teams were
able to conduct with the fàrmers at the time of survey implementation. The outline was
developed by team members using a format fiom the farming systems and natural
resource management team, and reviewed by OHVN personnel before going to the
field.. A sample topic guideline is presented in Appendix 2.
The methodological approach used for the survey comprised three steps by
which the choice of ZER, SB, villages and PUS was made in a logical sequence. The
first step, the choice of Fansirakoro was done during the InterCRSP preliminary survey
in the Sector of Kati in October 1997. OJ!IVN extension staff were part of the
preliminary survey team. The shallow soils and degraded land due to low soi1 fertility
and soil erosion characterized the situation of Fansirakoro compared to Torodo where
shallow, rocky and very degraded soi1 was predominant
The second step consisted of an interview with the population of the village
which involved mainly the village chief several counselors, a few other dignitaries, and
avdauk heads of PUS. ‘lhe SU;~ of the group during that interview was about 40
participants. The interview at that level was supported by one section of the open-
ended topic guideline.

---

43
The third step in&ded interviews with the PUS heads and were conducted using
the specific section of the open-ended topic guideline. The choice of the PUS heads to be
interviewed was based on their availability. The interviews were completed by a field
lmnsect of the village.
A multidisciplinary survey team conducted the PRRA. The team included six
researchers spe&hzed in the disciplines of agricuhural economics, animal science, ecology,
agronomy, soil science and gender issues. OHV extension agents and natural resource
management program officer assisted in conducting the interviews with fàrmers.
For the
interview process, beginning with Pus, the team was split in groups of two members of
each. The survey was conducted in 6 days.
Once the survey was completed, individual reports were prepalbti iollowing the
outline of the topic guideline. The data were also handled by Paradox Software and
analysis was completed using the SAS system. The farming systems identtied were
described, and hypotheses were formulated regarding the principal constraints of the
farming qstems. When possible, recommendations to alleviate or remove the
constraints were proposed. A list of these constraints and recommendations is
presented in the Results section of this report. The plan of research for the next years is
based on the identifïed constraints, modtied by aclditional information gathered by the
team through other activities.
5. Results
The village of Fansirakoro shares the same terroir with Fabougoula, Fansiratiani
and Fansiradj~èrobougou. The terroir of Fansirakoro is borde& by the village of Kobala at
3 km to the North-West, by the village of Djinidjala at 6 km to the West, the village of Sidjè
at 4 km to the South and the village of Yèkègougou at 8 km to tbe East. The total area of
the village is about 225 km’ (22,500 ha). Fansikoura and Fansirakoro are the two main
hamlets known under the name of Fansirakoro.
a. Population
J-.& pop-~at~a~
h fie
vmge
ofFm&&ora (570
pefmns)
is ~redo~~tly
mra&
and most revenues are derived Fom agricultural activities. The average labour force is
.lO persans per PU. The Barbara is the only ethnie group existing in the village of
Fansirakoro. The Bambara occupy tbe OHV zone, the central part of the zone of Kati,
the Ouelessebougou East and Koulikoro. The primary occupation of the population in the
village is agri&ture.
The village of Fansirakoro is organized in five extended familles (Sirimana,
blacksmiths, Tièkourabala, Dossorola and Dior&&). The thirty eight production uni& of
the village belong to these five families. The blacksmith Gtmily has 13 production units, the
Grimana tàmïly II, Dossorola and Dionkèla families 5 each and the family of the village
chief (Tièkourabal) 4 production units. The average population per PU is about 16 persans
comphsing of 7 fetîîales and 9 males.
The population of the village is living in two main hamlets (Fansikoro and
Fansirakoura). The history of the village started with three brothers who left the village
of Fabougoula to settle in Fansira. The village was established more than two centuries
ago. The founder of the village Samassékoro Traoré and his two brothers came fiom
Fabougou looking for good agricultural lands. The village used to be a cereal market
for nomadic population (Maures). The production of the village gave the population

---

44
some authority in the surrounding area. These conditions of authority (Fama) and good
production (finsoro) gave the name of Fansira to the village.
The population of Fansirakoro’ is practising Islam, Chr&Gsm and Traditional
religions. Fiieen chie& have ruled the village since its establishment, which was before the
arriva1 of colons The actual chiefis rulingthe village since 1983. Four counselors assist the
head of the village and by the council of eiders, which comprises the village chief and
representative of every extended fami&. The village of Fansira is organized into clans or
large families forming what is defmed ai a Production Unit (PU). Ail fàmilies live in large
firm fina@ units (Pus) around a patriarchical leader and work in communal Iields. The Pus
comprise on the average 2 to six households, with an active farm population of 16
. .
ll&alkmInM t4i27i~+ fiom 3 to 35 members.
Management decisions regarding allocation of land trop choice, use of agricultural
equipment, and distriibution of production are made by the patriarch of the PU. Individuals
responsible for specific fields make decisions on the daily operations in those fields, and are
accoumable to the head of the PU.
A typical farm includes both communal and indiviti fields. Women, heads of
households and Young active males, cultivate individual fie%. Women produce peanuts,
rice and millet for the local markets in these fields. In fàct, there are wide ranges of field
types used by the &~III f%unilies; there are village garden plots used to produce condiments,
near-village fields, bush fiekls, low-lying fiekls and flood recession fields, if available.
Management decisions on labor and other inputs to each of these type of fields derive fxom
a complex weighing of constraints, risks? and expected benefrts.
!kveral cmtmmtd organisations (Monchi, Gbtchi) are available in the village of
Fansirakoro, They are tmd&naLin the village and mainly age sets or peer groups. They are
mainly active in agr&ltural and cultural activities in labor management at the village level
Iabor migration, ahhough largely seasonal, is important enough in the village. Migrants are
ofkn males and females between tbe age of 14 and 30. They travel mostly to Bamako and
mi.
TheYèk&ougou-Fansirakaro road is the main road that serves the village. A
network of roads serves other amas, which might be inaccessible during the rainy season
hecause of waterways and moumains There is a lack of formal education and the rate of
illiteracy is extremely high. The impact of C~ernment effort of alphabetization is not high
in Fansirakoro.
The Lsurvey shows a widespread illiteracy in the village, which poses a problem for
agticultural and social development. l”he school and market are located in Yèkèbougou.
TheOHVkk~gQv
ermmtaL&~~elo~ agency operating solely in
the viige of Fansirakoro. Several rural development institutions have operated in
Fansirakoro.
The ‘Wakr development department established an equipped well in 1987. A
rehgious NC& (Seccurs Catholique) also established two large Wells and a milling machine.
The same N%O helped the village estabhsh a cereal bat& a rural mater& clinic. a
community pharmacy but none of these is fùnctionning at present.
The village of Fansirakoro bas good relationships with neighbouring villages
(Djinidjan, Fabougoula) including mariages and source of agricultural labor. The village of
Fansirakoro is located on the lattds of t.he village of Fabougoula.
The chief of the village of
Fabougcu!,1 is t’x owner of the lx~?r.. l%xver, there is no probl-?? with regards tp +%
land use ri&& even though land rights belong to Fabougoula. Soils in Fansirakoro
comprise low-land soils (very low potential), very degraded, rocky soils (upland soils)
which are dominant and plain soils.

---

45
Upland soils cari be cultivated for 6 years. Mer that period striga appears and
yields of crops may decline. Fallow cari take 15 to 20 years before the fertility cari be
restored. The growth of trees (Termin&a SP., Combretum SP., etc..) and the presence of
some gramineous species (A&omd sp. ) are good indicators for fe replenishment
for fàrmers.
Gravellous soils (lateritic, Typic Cuir&a&) cari be cultivated for 4 to 5 years. They
requùe 15 to 20 years for fzrt8t-y r$&nishment. Farmers report that the production
potential of these soils is lower than tlx+t of upland rocky soils (Aerîc Tropaquepts). Run-
off is a major constraint of production fbr these soils in addition to striga. The advantage of
hming these soils is the possibility of Fg equipment, which cari be dif6cu.k when those
are many sumes are .
In contrast plain soils (Typic pstropepts) cari be cultivated continuously when
fertilization is available. They also allo* the use of equipment (plough, catts, etc.). The
length of ctiltivation for these soils is at least ten years. The time required for fertility
restoration is also lower (3 to 4 years). Weeds cari be a problem but planting earlier in the
cropping season helps to alleviate this u$nstraint. The major constGnt with respect to plain
soils ,is the veq limited area they compr$e in Fansirakoro.
Some lands cannot be f&xd $ere animal are herded, grazed or watered. Fallow
lads and areas not cuhivated are u&I as rangelands. Planting of trees required an
agreement fiom the land owner’s. Womm get land fiom their husbands. The women utilize
trees like Néré (Parka biglobma), shea butter (Vitelluria padczxz) and baobab
(Adhsonia digitatn). Youngsters use d,ady the fhits of 2jzypIzu.s mauritimul.
The use of wild fhits is dso iegulated. Néré (Parkia biglobasa) and Tamarin
(Tkzwrindus indica) belong to owner of the cultivated field. But when these trees are
located in the f&llow land, their harve+g is not regulated and population fiom the three
villages cari ‘use them The harvesting of shea butter (yitdkria pandoxa) is fiee. The
cutting ofthese trees is forbidden because ofthe vahrable multipurpose uses.
b. F$rming systems
Land tenure system m Fansirakdro is typically complex. On the one hand there is a
lùnd of egalitarian access to land for ail f&nilies; on the other hand there is the historical
reality of founding fàmilies main-g a hold on more land over generations. The
introduction of cash trop (cotton, tomaio and tobacco), the use of animal traction (which
requires complete clearing for best effici+ncy) and the degradation of land under population
pressure have aggravated this situation.
Land cannot be sold in Fansirakdro. It cari acquired through patrimonial linkages or
through loans. The village chief has the gower to settle all land tenure disputes. Ail villagers
have the right to a parce1 of land to ‘support their w, but the overall land use is
determined $y the village chief and by ~Gllage tradition. An outsider fi-om a neighboring
village cari gain temporary access to tid by a loan fi-om a farmer with surplus land or by a
grant fi-om the village Chie£ However, hd cannot plant Guit trees.
Land has important economic Id material vtie due to its capacity to sustain lifë.
It has also a sacred character due to it~ procreative powers; the ancestors are thought to
have made an alliance with the sacred s@rit:s to obtain permission for its use. Specifically,
clcafi?g laqd for agricuhure is r-cri îs a Sort of sactilege. ?arncrs are rehxtant to r.‘*i,nc! rn
poor agricultural lands. Ordy 5% of famers in Fansirakoro are food self&fEcient for the
last five years. However, they stated that they would not leave even ifgiven land in another
sector or zone. They’re felt that they shobd remain to guard their spints and tombs.

---

46
No individual cari claim to own land. The user is only a temporary trustee; land
belongs to the past, present and future kinship group, and forms part of the group
patrimony or heritage. In Fansirakoro, the extended family owns a piece of land as a
collective proprietor. The head of the production unit cari distribute parcels to individual
fàmily members. Departures fiorn tbis general mode1 cari exist. For example, land near
villages which is cleared and manured & too valuable to allot to a another fàmily; this is
equally true for flood recession land, laud near irrigation water, and other valuable land.
Once a &mily has obtained access to @ni, the rights cari be handed down to succeeding
generations. Farmers who want additional land cari request it fiom the village chief or
others Eunilies. Forcefùl appropriation ofland is strongly resisted.
Farmers in Fansirakoro are bringing more and more land into cultivation, maiuly
increase with shallow soils and in erosion sen&ive upper parts of the catena. From 1996 to
1998,2.4% more land was brought under cultivation However, shiIEng cultivation is being
replaced by continuous cultivation; Iàllob land that was traditionnaly used to build up soil
£&y is becoming rare. The duration of fàllow has shortened drastically fiom 20 years to
4 years. Many households cannot rely anymore on fallow to build soil fertility.
Farmers in Fansirakoro complair@ about the lack of fertile lands. As a result, they
prefer to cultivate the surroundings steep slopes (ranging fiom 10 to lS%), claimiug
additional advantages of dimimshed wefid invasion in these fields. The availability of land
for cultivation is a major constraint in the village of Fansirakoro and the practice of
fàllowing is not usual in Fansirakoro because of the scarcity of land.
Mixed fàrming systems involvin~ cropping and livestock tending associated with
off-fàrm activities are dominant in Fansirakoro. The village of Fansirakoro is essentially
agriculturak filod and cash trop productions constitute the major activity of people. Cereal
grains, sorghum, millet, maize, rice are me main food crops. Cotton, tobacco and tomatoes
and to a lesser extent peanut are the main cash crops. Cowpeas, land peas (Vhzwkea
subtemmea), cucumber, orrions, potatoeff cabbage and lettuce are also cultivated.
Food crops concem 72.3 % of &tivated land in Fansirakoro and the average food
trop area per PU is 3.70 ha. Sorghum is me main staple trop because of the cultivated area
and production. It is mainly cultivated in’ association .with cowpeas or with millet. The area
cultivated in sorghum and millet repre+nted 47.9 % of cultivated area in 1998. Local
varieties such as Keniké, folaba, drong4 and Kedé are mahly used, Only 20% of farmers
are using improved seeds of sorghum. ~Fertihzers (organic or mineral) are not used on
sorgho Few fàrmers practice animal traction (main& ploughing). The low use of animal
traction might be explained by the predominance of rocky, shallow soils. Line seeding is
used by 40% of humers and stock management and conservation are not practiced at alL
Maize, the main food trop after sorghum, is cultivated on compound fields around the
village. Mai~: represented 14.33% of +tivated land in 1998. The improved variety
Tièmantiè of Lkmblara is mostly used. Maize in compound fields receives fàrmyard manure
(household waste and animal feces, etc.). Composting and/or manure production are not
usual.
IVlillet is the third food trop.. Rice~ and fonio (Digitaria sp) are secondary crops used
to support the main crops. The variety &gonu with recl seeds is the main variety of fonio
found in Fansirakoro. Rice is cultivated $r lowlands and mainly by women. Cowpea is the
most important legume trop because of the cultivated area and production level It is
cultivated w?h !n?d l-x 3s a single crcy or: 7,c-y smsll fields. Peacut :nd Gozrbo (Hibistnn
esculentus) are the main crops cultivated by women.
Only men eultivate cotton. Cash crops represent 27.6 % of the cultivated land of
Fansira and the average cash trop area per :PU is 1.70 ha. They include cotton, tomatoes

---

and tobam. Chtton represented 27.6 ‘X~ of cultivated land in 1998. Improved technologies
like improved sedq ploughm& hne *adcasting fspreadiug seed by ha@, weediug
minerai and organic fèrtihmtiou are apphed tminly on cotton.
Tobacco croppiug by both
men and womm is done duriug the dry season around the Stijèèko river looated 15 ktn fîom
Fansùakoro. Women help also in the f&ily crops fbr pkmting, hatvesting and trausportiug.
Cropyieadsintheareaarev~l~fordfyland~crops.Theyareestitlieted
at6to8oo~forsorghum,9ookgmafmmaizein~~~fieEds,200kg/h;afor
rd& (maidy associated with sorghum),~ 50 to 200 kg/ha fbr cowpeasj 100 kg/ha for land
peas (voandmu).
Fat-mers are awm ofthe declining yîekls of their rainfhd crops They
estimated the rate of decline of yidds at 100% per year (ie. 2 tons, 1 tonne, 500 kg per
year). En c.ontm they advocate tbat y&k?s of tomatoes bave been increased because of
improved &hn@ues Crop rotations in~ the area are sorghum - mi[let, cotton-som
~~sorglnnn-pe*
The use of inputs (tnineral f&t&+x, fnanure) is vety hmited because of the price of
fertitizetsandthelow~~of~~ofoTganicfèrtitizersMineral~areused
for the cottou, ma&, tomate and tabac+ Manure is used tnaialy on the compound fie&
fàrmed in IIX&.E. Only a third ofthese areas might receive orgauic fertiiiz;ers. The practice of
making of compost does not exist in @nsirakoro. Hmsehold wastes are stacked and
brought in the fields Mauure is mainly d.tIy Gxes of mminants and the fîee roaming of these
animalsihthedryseasonreducesitsman
availability.
Practices of water managemmt %
e ‘sting in Fa~&rakoro iuclude stones lines, physical
barriers in wauerway~, fàscines, and plat+ting trees, etc.. Stme lines are apphed by many
farmers in Fausirakoro even ifthey are net property desigml and plot size are very small
Several species of trees have been @xi in Fansirakoro for planting. They imlude
AzBdirackr innia-(netm), Elscafyptus dp., A mia aibi&, fbh), fios~iis ju11jloru,
khya .eraegaims~~(ca&xdrat) and J&O~!UI mrcm (‘ourghère). Poughère is mostly used
forphmting on stone lines
tiestock is quite important in pansirakoro. 67% of Production Unit (PU) are
involved in livestock teding activities R+mimnts (cattle, sheep sud goats) are dominant in
the hvestock system CattIe, sheq and goats represehts reqmtivefy 38.6,36,6 and 28.6 %
of domesticated anirnals, respective@ o+eh are mtne~ous and are 34.88% of the total
number ofbeefcattle. Goats represent 5718 % of srnall ruminants The mmbe~ of sheep sud
goats are increasiug in the village but cattle are decreasing because of robbety and sales
Farmers tend to invest ~II animal (cattle, ~ sheep sud goats). They hire outs&r herders to
tend these animak. The herder is paid a~fee and given rights to some of the milk of the
CO~S. Herd management and heahh tare is the respomibihty of the herder. Donkeys sud
chickm are also raised in the vihage. Interactions between crops and hvestock are of several
form.
Complementary relationships are esse&&@ related to forage (trop residues) for
animal.~ and manure for crops. Animds c&stitute an asset, which cari be used to reduce the
impact of trop &lure and iuvestment ri$ks. Ih-aught animais are used for cropping and
transportation. Important by-products, sueh as mamre, are also considered. Cmflicts have
developped as population pressure increa#es causing competitioh for land use to, cropping
and hvestock.
Agricukural credit in the village of Fansirakoro is provided by OHVN, the III&
extension ir+!~hx~ o~crati3c; in t!t- drn. OHVN credit s,~~.rm i s re’bed to cottu.,
production. At village level, the credit is ~handled by the ton villageois. It is thought that
social pressures would be used on dehncpxnt xmmbers who defaulted. Farmers said that
social pressures would force these individual to pay. They reasoned that village eiders knew

---

each individual ‘s capacity for repaymen better than any outsider, and that credit would in
t
this way be extended only to individuals iwho have demonstrated capacity to manage loans.
Credit is given in the form of làrm equipment (plough, seeders, etc) and inputs such
as f-s and pesticides, which are usually promised at the beginning of the cropping
season. It is associated usually with cash /zrop production. In 1997-98 Fansirakoro received
sixty bags of minera1 f-s (20 of urea, 40 of mixed f-s) for cotton production.
Agricultural credit was higher for the year 1998-1999 and concemed 316 bags of mixed
f-s, 85 bags of urea and 320 liters~ of insecticide. The village also received 8 oxen, a
cart and 5 ULV. The amount of credh fi-om OHVN has been increasing since 1994.
According to the population the village uses mineral feailizers for vegetable production (3
to 5 bags per agricuhural produL,~‘unit).
Natural resource management t&hnological packages recommended by OHVN
concem water management (stone lines bd tree planting), improved Gllow, living faces,
improved corrals and improved organic r$atter production. The adoption of these improved
technological packages by fàrmers is n’ t very high. For example only 6.4% of PU in
Fansirakoro used living fentes. These Les are S%, 8% and 13.3% respectively for
improved faîlow, composting and improved corrals Physical methods for water
management or soil conservation are gejting some interest among fiumers in Fansirakoro.
Stone lines are used by about 32% of fàr@s.
Diets in Fansirakoro are based on sorghum and/or millet with peanut sauce and
when availabb: meat and Gh as source ofproteins. Com and rice are inchtded to a limited
extent in the diet of fàrm fàmihes. C!or@ is considered a graden trop and is fiequently
consumed fiesh during the rainy season because of it is harvested earlier than other crops.
The population of Fansirakoro assumed ~that in the last decade very few of &rm fàmilies
were food sel&uflicient (only 5,6%). Du.ring years of good rainfàll hke in 1997-98, that
figure cari mach 50%. This meant that even in goods years 50 % of the population cannot
provide their own food considering aIl sources.
c. Evolutiod of natural resources
Farmers recoguized that the state of natural resources in Fansirakoro is changing.
For example yields of crops on gravellous soils (shallow soils over laterite -Typic
clwustalfs) have decreased drastically. This decreasing of yields are related to:
. . . .
-witfak
~
l
- lack of moisture;
- reduction of the density of trees;~
- reduction of organic matter in * soil;
- and increased mn-OK
Rice was cultivated on the plains, but today farming of millet on these soils is a
probletn Ma&x was cultivated on rocky upland soils; this is not the case today and maix is
cropped only on compound plots. Many t@es and grasses have disapearred or have become
less in number on the landscape. Trees $at are disapearing are Oqnanthera abyssinica,
Prmopis a)icana, Cordila pinnata, Bqkea afiicaq Kqya senegalertriis, Pterocaqzti
erinaceus, Amwna senegale&s, and ~
A@elia ajkicana. Crasses like Cjvnbomotz
gi@zntew, Loudetia sp. and Hyp@henia pp. have also disappeared..
Farmers noticed that sc)me gras$es mainly weed~ like Digifnricr .y., qtrign qn.,
Permisetum pedicellatum, Laudetia SP., ~
Andropogon pseudapricq etc.) are becoming
dominant. Trees species that increased k @mber are Bauhinia sp., Terminafia sp.., Acacia

---

l
49
nzacrosta&ia-s, Dicrmtachys &vmera’a, and De,tarium n~icricaqnm. Causes for the
evoh&ion of natural resources as related by &rmers are:
f
- aridity of climate;
-&&+J&lgramfall;
- uncontrolled cutting of trees b herders;
- cutting oftrees by fàrmers for Eharcoal and hrewood;
- controlled cutting of trees for c+hivation;
- and demographic pressure.
~
The causes of the degraded state of the natural resources are interrelated and cari be
illustrated by a problem tree. Farmers $I Fansirakoro are resource poor and they have a
deficit in cereal production. The intera&on oetween population growth and the need of
cultivated lands to meet subsistence requirements leads to the reduction of fàJlow periods in
Fansira below levels necessary to ma@& soil quahty. Cultivation is expanding onto
marginal soils that more fiagile Andy degraded rapidly. Growing markets for fium
ammodities as well as for consumer goods are putting more pressure on the villages’
natural resources.
. ~
Most of the total fàrm household mrne cornes fiom non-trop sources. Research on
adoption and dif%.sion of fàrm technolo ‘es has revealed that criteria like cash flow, food
secmity, labor burden and social a 4tance play a role in the acceptabihty of new
technologies, in addition to immediate e nomic and/or net present value of future benefits.
Tbe lack of capital limit the potential o fàrmers to participate in yield stabilizing or yield
r
improving technologies. Tree cutting f r flrewood and charcoal is also common among
villager to generate some cash. Tomato
d tobacco are the main cash crops, in addition to
cotton. However, the lack of water SO !fces for tobacco and tomato and the know-how
needed for CottOn croppinghmit the pot ’ tial of these crops for income generation. Natural
en
resource degradation hke decreased soil ~fertility, soil erosion and excessive runoff of water
are likely to exacerbate this situation ($gure 1). Manuring cari contribute to sustainable
production increases on the alluvial and ~acidic soils, which have low bufhering capacity, in
that it increases the soil buE&ng ca acity, and thereby leads to better nutrient use
efficiency. However, the lack of adeq $te livestock husbandry hmits the manure supply.
Likewise, the lack of carts for transportation of manure to distant fields in the bush land is
ah a constraint.
l
The main agticulmral production c0nstraints according to fàrmers are related to soil
and water conservation and manag
t. Rainfill has diminhhed drasticahy and ît is trot
regular in space and time. Run-off has
3 become important in the landscape, so has the
leaching of minerai f+om organic matter Water inflhration in the soil is very low. Soil loss
through erosion is alarming with gullies. Vegetation cuver is a.lso very poor in the area.
Farmers mainta.ined also that the kèrtihty of soil is very low. They were inhormed on
improved teclmolo@es for organic ma er production (compost and manure). but these
packages are seldom applied. Reasons f,or the low level of adopting extension packages
stated by farmers are $ven for food cro s (Table 1) and cash crops (Table 2). The lack of
tiormation and insu.fJicient equipment ‘d re important as constraints. Roaming of animais
during the dry season is tbe common pra#ice in Fansirakoro because of lack of forages and
drinking water. Also, Young males are no\\ available for tending herds because they move to
other villages for tobacco cropping.
e lack of equipment for compost or manure
handling (carts, doîLe; ;, etc) is also a

---

5 0
Table 1: Constraints fat Using
d-
Some , ood Crop Technologies
lackdameanaftransportatior
to fields on mountains is
-
pgt aaxs to mineral fertihefs
- 15% earlines allow varieti~ ta close their
cy:le.
pronebecauseafthehck
-10%Higllj3lwwion
-ne seeding 4~-40?hl qzitimum density for
higll flrdncborl
.-
@ck preservl-
micd products unavailable
stacks, panicles and spikes
-
$&ck mawwnent
Organic fertilizers like ruanu , household wastes are not used in Fansirakoro
+
either for food or cash crops. CO straints according to farmers are the lack of
P
information, the lack of equipment ~for manure transportation and the relief of the
landscape in and around the village. IFarmers are aware of advantages of using yields
increasing or yield stabihzing techno gies. For exanrple, many are not using chemical
k
fertilizers because they are not cotton growers. Cotton growing is becoming a current
strategy in Fansirakoro to get access to iuorganic fertilizer and other inputs.

---

R-CIfZldOptiOll
Omnic fertiliiziz
h4heral fertilizg
Using minerat fertilizem
-4O%jlOCKSOilS
- rxquuidfor higher prochhon
Line seedinp,
-4o%liltionaluseofcultivatedarea
- increasing plant &nsity
-dlOWthGUS3QfOthHtimaltiOIl
equipment
Herbicide
Fwsanddiseaseamtr0190%
-90% Hïgh frequency dplant pest and
6. Recommendation Domains ~
A recommendation domain is’i&ormation that f-s cari use to improve or to
maintain the! productivity of their re urces and or their huming systems. Depending on
constraint potemi& and constraints 4
for adapting specific technologies, and fat-mers’ needs
(recychng of organic matter by compo
* g or manure production, etc..), PUS in Fansirakoro
cari be categorixd in two recommen t-ion domains based on Gu-mers circumstances and
problems and practices. Recomm
tien domains are not necessarily the same for ah
4
technologies that are developed. DiKerent recommendation domains may be desirable for
different problems and related
‘OIE.
In the process of on-hum research
recommendations cari get redefined T
afte the prehminary diagnosis.
The needs for interventions id
?
improving water and soil manag
(stone lines, small dams, etc..);
?
intensifying the production of o
er (compost, improved corrals);
?
reinibrcing complementary relation
s between cropping and livestock systems
(forage rehitionship, manure relatio
?? developing and implementing
,sed natural resource management
strategy at village tevek
l
Stone hnes have been introduced in Fansirakoro in 1994 on a community basis. The
lack of labor resources at farm level dtig the dry season because of labor migration for
survival has been a handicap for the co4tinuity of this action. Uncontrolled use of vegetation
by the population of Fansirakoro and eighbouring villages (Fantiakoro and Fabougoula)
remains a big; concem for the preservati
of natural resources because of poverty.

---

l
52
Recommendation domsin 1
of iàrmers (PUS) in Fansirakoro. Resources
of equipment for animal traction and that
of of cattle (> 5 head) in addition to xen, goats and sheep. Because of îts resources, the
group cari adopt yield improving and
*
diversification of
Recommendation domain 2
This group comprises 70 % o fàrmers. Resources are very limited for this g~oup.
However, this group cari use yield sta ilizing and even yield imprwing technologies. Very
evident results might require reversin the aversion of this group to adopting improved
technologies. The labor force availab e for this group is very low, and most of the time,
equipment is lacking or incomplete. L’ estock available are smaIi ruminants and courtyard
i
c.hicken. Potentîal technologies the gro p cari use are the following:
.
Food crops diversification;
*
Ulilizing improved seeds;
Y
?? Recychg organic
?
hnproved cropping systems
0
Improved crop/livestock
0
Intensfjring vegetable trop produ
Farmers in Fansirakoro are re urce poor and fin-mers’ actual practices are lïkely to
increase costs per unit of production
of the continued degradation of the natural
resout-ce base (land) and by the
of cultivation ont0 margjnal SOI&S. The itnmediate
priority should placed on
management systems that are well adapted to the
situation of Fansirakoro. Participative t hnology adaptation should concem conserving the
Land, increasing yield
l
-m---

---

l
53
THEME 2 : S IL AN-D WATER
N
CO SERVATION

---

l
54
Impro-ving Food Crop
oduction in Subsistence Farming
and N’tétoukoro
A. H Berthe’, S.F.
B. Guinda’, RA. Kablan2 and
RS. Yostz
1 ESPGRN/S
a, BP 9030., Bamako, Mali
2 University ofHawaii, ”
19 1 East West Road, H o n o l u l u , H I 96822
1. Jid.Xcaiion
This consists of
of mixed fertilizers as extender of manure
with the tirmers practices
ilization or only manure. The assumption was
that the physical combination of mixe
manure will take advantage of the
physical praperties of the
an environment conducive for root
gTOWth.
2. Objectives
-
to compare effects of
and withaut f-s (Tl) on
-
to evaluaite the economic
The test was conducted 93
villages (Fansirakoro and N’tétoukoro) with 10
fia-mers in each site.
It cons& of the following:
” control ( n o f-lion)
-
3 tons of &.rmyard manure
3 tons + NPK mixed together (25% eq
alent applied manure).
4. Remlfs
l
Data used in the analysis c.o
one village (N’tétoukoro) and f?ve (5) out of
seven (7) who performed the test
. Two timers (m number 2 and 6) bave
been eliminated because of inaccurat
The variables used in the analysis
- Variable 1: (NPOQ) number of
- Variable 2: (NPLT) number ofp
- Variable 3: (NEPI)
number ofp
- Variable 4: (PEPI)
panic
- Variable 5: (PGRA) grain
- Variable 6: (GREP) Ratio grain/p
_~_-I-_._-.-~__-__---_-...I.
- “ e .

_ - -
-
--

---

5 6
Annex
List of observations
No Farmers Treat NPOQ
NEPI
PEPI
PGRA GREP
1
1
13600
21733
1 5 8 1
1247
0.79
1
2
12133
22000
2011
1468
0.73
1
3
12133
25467
2373
1627
0.69
* 2
1
19467
39067
2789
2245
0.80
* 2
2
26133
65733
2761
2024
0.73
* 2
3
19333
38800
2547
2027
0.80
3
1
20667
50133
2365
1 8 2 1
0.77
I
3
2
21733
45200
2355
1824
0.77
3
3
24133
47600
2645
2075
0.78
4
1
23467
47600
1 2 1 1
900
0.74
4
2
22400
56000
SS4
6 3 1
0.71
4
3
24667
52533
170s
1272
0.74
5
1
18533
29067
912
784
0.86
5
2
22133
47600
987
853
0.86
5
3
18667
41467
1187
819
0.69
*
6
1
13467
23333
1 4 2 5
995
0 70
* 6
2
13333
25867
1223
787
0.64
* 6
3
13067
23200
1019
38
0.75
7
1
20533
42933
1312
1003
0.76
7
2
21600
35600
987
736
0..75
7
3
22933
50400
2373
1755
0.74
Initial analysis fiom the seven fàrmers fi$lds (fkrms were replications)
NPLT
NEPI
P E P T PGRA
SREP
Control (Tl )
average
18533 42000
36266
11656
12S4
0.78
S T D
3736 15292
11571
672
543
0.05
c v %
20 36
3 1
40
42
6.52
ma-
23466 68133
50133
2789
2245
0.86
minimum
13466 24266
21733
912
784
0.70
Manure (T2)
moyenne
19923 48400
42571
1600
1188
0.74
écart-type
5 1 6 0 II3468
5810
7 6 3
572
0.07
C V %
25 38
37
47
48
9.01
maximum
26133 67066
55733
2761
2024
0.86
min*
12133 22533
22000
SS4
630
0.64
Ivlanure + fertilizers (T3)
moyexkne
1 9 2 7 6 45405i
39923
197s
1477
0.74
écart-type
5 0 9 8 1 6 1 3 5
11690
670
538
0.04
C V %
26
32
29
??
36
5.71
nLl&mum
2 4 6 6 6 6986(
52533
2645
2074
0.80
minimum
12133 23333
23200
1018
768
0.69

---

***** ANALYSE DE VARIANCE
DE 5 PAYSANS
ANALYSE DE 1
IN’IERACTION
-
-
SCE! test & TUKEW =%4747698.00
ANALYSEDEVARIANCE
-
-
S.C.E. DDL CARRES MO
JBA E.T. C.V.
VARTOTALE
%244717072.00 1 4
7479740.00
t
VAR.FACTEUR 1
3301952.00 2
165 76.00
0.72 0.5172
VAR-BLOCS
%223182256.00
r
4 5 795564.00
24.48 0.0002
VAR.RFSlDUELLE 1 %18232864.00 8 ~
2279108.00
Oh1509.67 7 . 6 %
TABLEAU DE!S MOYENNJZS
-
-
MOYENNEGENE&UE= 19955.56
MOYENNES FACTEUR 1
= JKERTILISATIO
19360.00 20000.00 20506.67
N
F I : 1(-I-l) 2!J2) 3Q-3)
MOYENNE!S BLOCS
= PAYSAN
-
-
F2: l(F1) 2(P2) 3(P3) 4(P4) 5( IL )
12622.22 22177.78 23511.11 19777.718 21688.89
ANALYSE DE L<A 2e VARIAEKE : Nbre pltsha ‘(NPLT)
~~
=z
IhfTERAcTION’-*BLOCS
-
-
SCEt&deTUKEY=%578518.31 PROBA 1
ANALYSE DE VARIANCE
1 o.8851
-
-
S.C.E. DDL CARRES MOY
TEST F
PROBA E.T. C.V.
VAItTOTALE
%3193808600.00 14 28129184.00
VAR-FACTEUR 1
%53079552.00
h
2 2 539776.00
1.07 0.3886
VAR.BLoCS
%2942585100.00 4 7fWl6270.00
29.70 0.0001
VARXESIDUELLE 1 %198144000.00 8 1 24768000.00
%4976.75 10.2%
TABLEAU DE!S MOYENNFS
cc-
MoyEINNE GENERALE = 48604.44
MOYENNESFACTEUR
= FERTILISATIONI
-
--
-
-
-
FI: l(T1) 2(T2) 3(T3)
46133.34 48986.66 50693.33
MOYJZNNES BLOCS
= PAYSAN
P-P
E‘2:
1 pl) 202) 3(P3) 4(w;l 5 ~

(w)
24977.78 54888.89 68088.88 48311.1) 46755.56
-~
-
ANALYSE DE LA 3e VARIABLE : Nbre épidha ~(NEPI)
-.--
-
-~ .--
.
-- .----
INIERACTION TRAITEMENIS*BLOCS
-~
SCEtestdeTUKEY=?/01057751.00 PROBA= ~
0.8686
ANALYSE DE VARIANCE

---

l

58
--~
S.C.E. DDL CARRES MO
PROBA E.T. C.V.
VAR.TOTALE
VARFACTEUR 1
0.99 0.4161
VAR.BLOCS
10.74 0.0030
VAR-RESIDUELLE 1 %276198020.00
%5875.78 14.3%
TABLEAUDES MOYEIWES
_-~
-~
MOYENNE GENERALE = 41022.22
MOYENNES FACIEUR 1
= FERTlLISATION
FI:
1c-w 2tT2) 3v3)
38293.33 41280.00 43493.; s
MOYENNESBLQCS
= PAYSAN
-
F
2
:
l(P1) Z(l’2) 3(P3) 4(P4) S(P5)
==--
=
ANALYSE DE LA 4e VARIABLE : Pds épis k&a (PEPI)
z-
=
INIERACDONTRAI1[EMENIs*BLOCS
z==-
SCE test de TUREY = 8143.95 PROBA == 0.7642
ANALYSE DE VARIANCE
_-
-
-~-
S.C.E. DDL CARRES MOYENS TESTF
PROBA E.T. C.V.
VARTOTALE
5715564.00 14 408254.56
VAR-FACTEUR 1
1190281.00 2
59514( .50
7.74 0.0137
VAILBLOCS
3910471.50 4 977617.88
12.72 0.0018
VARRESIDUELLE 1 614811.50 8
76851.44
277.22 16.7%
TABLEAU DES MOYENNES
z-p-
MOYENNE GENERALE = 1659.38
MOYENNES FACTEUR 1
= FERTILLSATION
Fl: l(T1) 2(T2) 3(T3)
1476.27 1444.53 2057.33
MOYENNESBLOCS
= PAYSAN
-
F2:
l(m) 2W) 3v3) 4(P4) 5(?5)
1988.44 2455.11 1267.56 1028.44
b
1557. 3
testdéNEWMAN-KEULS-s=5%
--_
FACIEUR 1
: FERTILISATION
-
-
-
-
NOMBRE DE MOYENNES
VALEURS DES PPAS
404.70 500.47
Fl LIBELLES MOYENNES
ENES
3 T3
2057.33 A
1 Tl
1476.27 B
2T2
1444.53 B
ANALYSE DE LA 5e VARIABLE : Pds grain kgh 1 (PGR4)
- - - -
!
IFJTERACTION Tm*BLGCS
==---
ANALYSE DE VARIANCE
- - -
S.C.E. DDL CARRES MO
PROBA E.T. C.V.
VARTGTALE
VARFACTEUR 1
494001.00 2
5.08 0.0376
VARBLOCS
2289175.80 4 572293.9
11.77 0.0023

---

59
VARWSIDUIXLE 1 389106.50 8
486 38.3
220.54 17.6Oh
TABLJZAUDESMOYENNES
w-
MOYlZNNE m = 1254.22
MOYENNEISFACIEURl
= FlxrlL1sATI01
Fl: I(Tl) 2çIz) 3(X3)
1150.93 110240 1509.33
MOYEW@SByxxIs
= PAYSAN
--L
F
2
:
l(P1) 2(P2) 3(F3) 4(P4) 5(P5)
1447.11 19fK.67 934.22 818.67 1164 44
test&NEwMAN-KEmLs-saIil=5%
-
-
=
F-1
: FERTTLISATION
NoMBREDEMoylzNNEs
2
3
VALEURS DES PPAS
321.% 398.15
Fl LIBELLES MOYENNES GROUPES HOh
3T3
15091.33 A
1 TI
1150.93 B
2T2
1102.40 B
BP
- --
-
ANALYSE DE LA 6e VARIABLE : Rapport grd4 i (GREP)
w-
-
- =--a
llVIERA~ON~*BIxx-S
-
-
SCEtezitdeTWEX= 0 . 0 0 PROBA= 0.1% i
ANALYSEDEVARIANCE
-
-
-
S.C.E.
DDL CARRES MOM NS TESTF
PROBA E . T . C.V
VARTOTALE
0.04 14
0.00
VARFACTEUR 1
0.01 2
0.00
1.83 0.2217
VAR.BLOCS
0.01 4
0.00
1.2.3 0.3695
vAR.REsTDuELL.E 1
0.02 8
0.M
0.05 6.2%
TABLEAUDESMOYENNES
-Y
MoYENNEl~=
0.76
-
-
MOW+iNES FACIEUR 1
= FERTlLïSATIO?
-
-
-
F l :
lcrl) 2cIz) 3m
0.79 0.77 0.73
MOYENNESBLOCS
=PAYSAN
11-u
F
2
:
l(P1) 2(R) 3(P3) 4(P4) S(P5)
0.73 0.78 0.73 0.80 0.7C

---

n 60
Eflixt of IVlanure and P-S urce Fertilizer on the Oplimization
of Soi-l Water and N trient IJse For the Min Cropping
System
Senegal Peanut Basin
?
M. S&e,
. Dia& and A. Badiane
slq senegal
AbStl-ZWt
?
A fi-yew national programme
underway t o boost the fbod and cash trop
production. P and Ca soil a
natural resources based materials is one of
the strategies dehed. A n
mized complete block design experiments
was installed within each of the
main cropping systems in the Senegal Peanut
Basin. The treatments consist of
ogypsum and phosphate rock mix, and mamue-
used separately or combined in o r d
to investigate the optimization tbe trop water
and nutrient use. The
il fertility change, trop water and nutrient
water balance, and y-i
For the experiment installed at Nioro Research
Station in 1997 within the peanutlc
cropping system, the important Ca movement
observed within the profile is stro
related to the amount of Ca added for each
treatment. Deep water percolation
so observed despite the moderate amont of the
annual rasa& and regarclless of t
eatment. Corn and peanut y-ields obtained are
low. The treatments
1. IIdroduction
I
In the Senegal peanut basin, fa
practices have almost disappeared from the
farmer’s land use system. This
tion is strongly related to the introduction of
peanut as a cash trop, but also re
fkom an increased demand for food crops by an
increasing population. The higb pr
e on the naturally fiagile soils combined with
the drought problem observed durin
e last 30 years is detrimental to the annual and
perennial vegetation caver. The
re, through soil organic matter loss and
acidification due to continuous cro
g and/or grazing, the food production system
has lest its resilience. In most
rmer’s fielcl, the degradation of soil water
characteristics favors an important
water percolation beyond the rooting depth.
even under moderate rainfàll CO
This also increases nutrient leaching risks.
Manure applications and plowing
very efficient in reducing the water and nutrients
Ioss through deep percolation by pr
oting a rapid trop root growth (Cissé, 1986).
Many studies have confirmed the effi
ncy of natural rock phosphate (RP) amendment
at an application rate of 400 to 5
kg/ha every 4 to 5 years to correct soil P
deficiencies. On soil with Iow pH,
agronomie efficiency of the rock phosphate
ranges fi-on1 82 to 91 % compared
triple super phosphate (Bationo et a1 1990).
This value depends on the chemi
characteristics of the rock phosphate mines for
which comparison results of the
dy are available and soils to which it is applied
(Ndiaye, 1978 ; Cissé, 1980).
However, for phosphogypsum (PG) r the combination of RP and PG now being used
in Senegal in the national 4-year rogram, there is little information in terms of
agronomie value or soil P and Ca am dment efficiency. The on-going experimentation
comparing those two minera1 camp unds is set to focus on that aspect. Assuming a
positive effect of the combination f RP and PG, the objective of mis study is to
--.--------__----.. ~- - __._ -
l

---

l
61
analyze the main cropping systems
r the efficiency of apply-ing combined P source
material and mamue to a degraded wil to optimii water and nutrient plant uptake in
order to attain a sustainable trop production increase.
2. MateriaIs and metbds
i
a. E$perimental sites
The sites were selected according to the existing main cropping systems. Within the
Peanut Basin, the improvement of ood secmity
f
cari be achieved in three cropping
systems. In the northern part of this groecological zone, peanut followed by millet is
the predominant if not the unique cr p rotation, whereas in the southem part, peanut
8
followed by com is a common practi ‘e. In the low lands of the Latter zone, continuous
rice is practiced.
One site for a long tertn experiment 1 as selected in each of these representative trop
systems: one site for peanut/millet rotation at Ouadior (Gossas Department) in the
north, two sites for peanut/com rot tion in Nioro area (one at the ISRA Research
a
Station and one on-&-m field near the station), and one site for continuous rice in the
Koutango valley (west part of the Nio’ o Department).
All the selected sites under the uplan 4 conditions are continuously cropped fields with
degraded soil fertility status, as ihustmted by Nioro soil analyses data (Table 1).
Table 1: Soil physical and chemical ch’ racteristics (O-10 cm). Nioro
f
+ Sand Total C
T S / T *lOO TotalP
P Olsen
%
%
%
meq/l OOg
h
@pm> _
92.2 2.5 9.2
0.6 1.6
64.5
0.2
28.5
blTreatmenks
The five treatments compared are sho
in Table 2.For cropping systems where com
or rice are involved plowing is perfo
when implementing the treatment, including
the control In fact, these two
deep tillage in order to express their
potential. For the peanut/millet
only a shallow hoeing is applied to prevent
fertilizer loss from wind blow.
Table 2: Treatment description in the different cropping systems
Treatments
T 3
T4 -
T5E---
O.M.
R P + P G
O.M. + RP+PG
P+O.M.. P+RP+PG
P+O.M. +
RP+PG
P + O.M. P !RI’ ! PG
P 1 OA:. +
RP+PG
N and K for Tl, and N, P and K fo T2 are applied annually as urea , ammonium

---

I
62
phosphate or KCl accor
are applied for the diffiient
crops. T3 consists of
tons/ha every two years for
the 2 fùst cropping systems,
&a for the continuous rice system, whereas
T4 refers to the application at, the
te of 700 kg/ba every four years of RP and PG
mix. Last, T5 is the combination
For all treatments fi-om T3 to T5,
nd K bave been added on an annual basis prior to
sowing. The experiment was inst
at Nioro station in 1997 does not include the T2
treatment, therefore only four tr
The experimental design is a ran
ck design with five treatments
repeated four times
The varieties used for the
for peanut, variety 55-43
duration (matutity at 90 days
after sowihg), variety 73-
for corn, var. Synthetic C (90 DAS
for rice, upland rice variety DJ-68
(matutity at about 90 DAS). Besides this rice
variety, the: fàrmers were provided
h two other salt tolerant varieties (Rock 5, and
Varl). The agronomie behavior
those varieties cuhivated
according t h e
farmers’practices in the bottom of
valley in submerged conditions have also been
evahtated for general assessment
the Koutango lowlands with respect to rice
production. In the fàrmer tiee field,
microplots were randomly harvested for yield
determinations.
c. Measur menti and monitor@
Sites characterization
Soil samples have been c
cterizuion of the sites; Ouadiour, on-
farm site at Nioro, and
ation experiment at Nioro installed in
1997, site charact
igetip, 1995) are used. Soit sampling
was also done a
1997. Ah the experimental plots
O-20 and 20-40 cm). Some of the
sented in this report.
Soil water balance
uptake but also for the purpose of
nutrîent balance, Soil water
il profile is measured once a week at
Nioro station site, and once eve
ten days elsewhere. Three methods are used:
neutron probe, and TDR probe at
station site, and the auger method elsewhere.
Access tubes for neutron probe
re installed at the depth of 265 cm; i.e.
deeper than the maximum trop
s . As for the TDR method, the probe
installation depth is limited by
h provided (120 cm). Concerning the soil
water content measurement me
e auger, the depth of augering is limited by
the wetting fiont as it advances in
ofile during the rainy season. Tensiometers are
used to determine the soil water
on at the depth supposed to be the maximum
rooting depth which is about
in sandy soils for peanut (Chopart, 1980). The
longest tensiometers tube ava
e 100 cm long. This limited the actual field
installation depth for aIl exp
plots to 100 cm for plots having an installed
accp~ tube, and 85 cm for
lots. Ddy y--asurements readicgs eZurted in
1998, but 1at.e (on septetnber 23
ember 1 Il) due to equipment availabihty.
The water balance monitoring wa
ot conducted for the on-fat-m experiment at
Diamaguene site (near Nioro Resea

---

l
63
below allows the determination of the soil plant
increment during the cropping season.
R - D - r f A S = E T R
R = rainMl, D = drainage, r
runo& AS = variation of stock, and ETR =
evapotranspiration. All these
1
camp nents are expressed in mm of water.
For drainage estimation,
was used. Soil hydraulic conductivity was
calculated using existing soil wate characteristic studies for Nioro site (Cissé et al.
1990).
- classe A; A + L = 20 %
- classe B; A f L= 21.5 % K(8, .-
- classe A; A + L = 20 %
where A + L = clay + silt
hydraulic conductivity; 8 = volumetrie water
content.
Nutrient balance
This is closely retated to soi1 w
ce in general, and to soi1 water stock and
drainage c,omponents in particula
t requires the use of the tensiometers to sampfe
soi1 solution at specified depths. W
soiI solution samples have been taken, starting
in late September. The nutrients
erest analyzed are nitrates, ammouim and
calcium. The results are not yet
The nutriertts lost through drain
of the nutrient balance components, wiIl be
calcuhtted by multiplying the wat
drainage and the chemical concentration of the
diffèrent elements analyzed.
Soil samples with depth at the be
each cropping season plant
samples at harvest wilI be collecte
allow the determination of
nutrient dynamics in root zone
Plant sampling
Plant samples were collected at flo
ring/pegging stage, around 60 days after sowing
for foliar diagnosis. The
curred alter noticeable feaf
chlorosis was observed on few tre
ent plots. Analysis of the plant samples is done
for the following elements:
Ca The plant analysis resuhs are presented.
Field operations
The multi-location experiment star-t
1997 on two sites; i.e. Koutango (contimous
rice) and Nioro (peanut/corn rotat
at the research station). The other sites were
implemented in 1998. The cropping
erations are presented (Table 3)

---

64
Table 3 : Field operation scheduling in the three selected sites.
Rainfàll itlput
Rainfall patterns are difkrent for the 2 years ( 1997 and 1998) in Figure 1.
In fact, we have experienced one early rainy season in 1997 as opposed to a late rainy
season in 1998. At Nioro, the total annual rainfall is about the same for the 2 cropping
seasons (580 mm). However, while the rainy season started early june 19977 the first
important rain was recorded late July 1998. Although characterized by a rather short
rainy season, the 1998 cropping season has a much better rainfill distribution. A long
drought period (over 30 days) occurred early during the 1997 cropping season,
causing a severe plant water stress, while in 1998 there was no major water stress
problem, except at the trop maturity phase.
At Koutango unhke at Nioro or Ouadiour, heavy storms were recorded (160 mm on
august 23 1997, and 100 mm on September 1998). Even though the soil vegetation
caver at that time was well established, a important part of the water fiom this rainfall
event is lost through runo@, the soil profile being near saturation at this time of the
rainy season. The positive aspect is the important input water to refïll the Koutango
river valley.
Data interpre tation
This concerns the y-ield data, and the nutrients and water data. For most data, ANOVA
methods will be implemented to compare treatment effects, whenever it is possible.
Otherwise, comparison of mean Will be used.
Most of the experiments have started in 1998. Therefore, only partial data are
available.
3. Results and Discussion
a. On-Station experiment at Nioro
The y-ield components data (cor-n in 1997, and peanut in 1998) are presented in Table
4. For corn, the ANOVA indicates no significant effect on grain yield or stalk, despite
the difberence in mean between the control and the other treatments. This is largely due

---

65
Figure 1: Nioro cumulative rainfall - 1997 and 1998
140
EL
n
El997
01998
Months

---

66
to important variability within treatments. The cor-n grain yields obtained are low but
higher than those obtained for phosphogypsum efficiency study (also presented in this
workshop). It is assumed that the rnanure application was the key fàctor to that
difberence. The soils samples collected fiom a11 plots down to the depth of 40 cm
indicates an emichment of the nutrient content (Table 5) but also a downward
movement of elements such as Ca (Figure 2).
Table 5: Soil chemical analysis results afier harvest in 1997. Nioro
Ca
Mg S T
VI.
%
0.7
0.9
73.3
0.7
0.9
76.5
1.0
1.3
78.3
0.8
0.8
98.0
/
0.8
0.8
84.5
1.0
1.2
82.7
For peanut in 1998, the plant population was close to the optimum in all plots. The
overall mean yields are good compared to the average yield observed this year.
However, no significant treatment effect for any yield component was obtained
through the ANOVA. Around the middle of the rainy season, plant chlorosis was
observed. Plant samples collected on each plot were analyzed. The foliar diagnosis
results (data not presented) show no significant difference between treatments.
Table 6: Yield Components at Nioro Station in 1997 and 1998
col-n 1997
Treatment
Stand 1 Stalk 1 Grain 1
Pl.ant
Pod+1
Population
T l
16800
1800
712
83330
T2
33230
3030
1740
82240
T 3
34690
3460
1820
91980
- c i - -
36420
3140
1880
90620
-
-
Mean
30290
2860
1540
87040
Sign. L e v e l N S
N S
N S
c v (%)
32
35
45
_-~~.
From the soil water monitoring done during 1998 growing season, changes in water
content in the profile are shown (Figure 3a et 3 b). Two major results cari be drawn
ffi-om these figures.

---

67
Exchangeable Ca (meq/l 00 g soil)
0 . 4
0 . 6
0.8
1.0
- 1 0
- 1 5
- 2 0
- 2 5
- 3 0 -_
-35
- 4 0
PG+PN
-45 -J
Figure 2. Ca dynamics in the soi1 profile at Nioro Station. Corn 1997

---

68
.
I
Water content (%)
2
4
6 8 10
12
14
0
i
-50
Dl : 14/07/98
D2 :20/07/1998
D3:24/08/1998
-100
&
D4 :28/09/1998
42
D5:15/1m998
“n
-150
a"
D6: 11/17/1998
-200
-250
-300
Fig 3a: Change in soi1 water content in the profile
Treatment: Control

---

69
Water content j%)
2
6
10
14
0
-50
-1100
3
c” -2 50
Y
-dr;l2 G!O0
-250
-300
/
Figure 3b: Change in soii water content in the profile
Treatment : Manure + (PG + RP) mix

---

70
First of all, regardless of the treatment, the water content values are well below the
water saturation for this type of soil (Cissé, 1990; Sène, 1995). This could mean that
water infïltrated in the soil is being used as long as the rains are falling.
Secondly, the wetting fiont has gone deep in the soil profile (below 2 m), so that water
drainage has occurred. Soi1 water balance requires then an estimation of the drained
water.
b. Ch-farm experiments
?
On-Farm Experiment in the Corn/Peanut Cropping System near Nioro Station
Yields components are shown in Table 6. The plant population lower than that
mentioned above for the on-station experiment is common in farm cropping situation.
The ANOVA performed shows no Sign&ant treatment effects on peanut yield
components.
ear Nioro. 1998
Hay
P o d
1380
1050
1260
1040
1080
750
1340
1090
1310
950
1270
974
N S
N S
20.3
22.3
No major water stress has occurred during the rainy season, due to the good
rainfall distriiution. In this degraded soil fertility situation of the continuously cropped
field (Diack et al, 1998), the lack of direct effect of P amendment application on peanut
could indicate that nutrients added are not readily available for trop. As for the manure
application, the lack of direct effect confirms research findings indicating that peanut
respond better to residual effect.
?
On-Fax-m Experiment at Ouadiour for thePeanut/Millet Cropping System
The effect of PG and RP mix and/or manure on yields for the peanut/mZlet
cropping system is shown (Table 7). Despite the .fairly good plant population obtained,
the yield components namely hay and peg yields are low. There is no significant effect
of the treatments. From the water balance measurements (data not shown), a very deep
water percolation is observed. This indicates a low water use efficiency which also
ocmrs in the sandy soil.

---

71
Table 7: Pe
1
C V ( % )
5 . 7
8,4
1
9,l
11,2
’
On-farm Experiment at Koutango for the Continuous Rice Cropping System in the
Valley Bottom.
The yield components obtained using an upland rice variety are shown in Table 8. The
ANOVA shows no significant effect of the treatments. The paddy yields values with a
mean of about 5 tons/ha show a real rice production potential irr the valley.
The soil water profile monitoring indicates a fairly good water availability to trop,
(data not shown). The water table during the cropping season has remained shallow (<
1 m ) (data not shown). Therefore, chances are that the groundwater through capillary
rise participates in plant water uptake when a drought period occurs.
Table 8: Effect of Treatments on Rice Yield (var. DJ-684D) at Koutango. 1998
Treatment
Plant population
Paddy
Stalk
Par&le
WT of 100 gr
TI
325000
4663
4989
5264
2 . 3 2
T2
316666
4692
5,320
5254
2.35
T-3
350625
4102
4267
4616
2.35
T4
304375
4440
5083
4983
2 . 4
l-5
32479 1
4006
4,286
4481
2.3
Mean
324292
4381
4789
4920
2 . 3 4
Sign. Level
N S
c v (%)
12.5
4. Conclusion
Long ter-m experiments were iastalled in tbree different cropping systems in the
peanut basin to address the sustainability of the food production increase, by means of
the enhancement of soil fertility. It is assumed that the key factor there is the
optimization of water and nutrient use. The results obtained tiom the first year of
implementation must be considered as tbe basis for a necessary continuation of the work
undenvay. For the next cropping season, focus till be put on the determination of the
different components of water and nutrieut balance.. Soi1 water percolation (drainage) for
upland. cropping systems and upward water flow in the rice root zone in the lowland
system must be determined as accurately as possible. This poses the problem of required
Pquipment. m‘0 +!itiornl needs in thzt regard concern the tOnc;nrncterT for soi1 water
pressure monitoring.

---

72
Literature Cited
Agetip, 1995. Programme de Réhabilitation des stations ISRA. Informations
pédologiques et étude cartographique des sols de la station de Nioro.
Bationo A., Sédogo M.P., Buerkert E., E. Ayuk, 1995. Recent achievment on
agronomie evaluation of phophorus fertilizer sources and management in the West
Afiican semi-arid Tropics. In : Ganry F. and Campbell B. (Eds), 1995. Sustainable land
management in Afiican semi-arid and sub-humid regions. Proceedings of the SCOPE
workshop, 15-19 november 1993, Dakar, Senegal Montpellier. France, CIRAD, 406~.
Chopart J.L., 1980. Etude au champ des systèmes racinaires des principales cultures
pluviales au Sénégal (arachide, mil, sorgho, riz pluvial). Thèse de Doctorat en
Production Végétale et qualité des produits. Institut National Polytechnique de
Toulouse, 16Op.
Cissé L. 1980. Suivi des facteurs physico-chimiques de la fertilité des sols sous culture
continue dans l’unité expérimentale de Thyssé-Kaymor. Bambey, Dot ronéo, C?%A-
ISRA, 50p.
Ci& L., 1986. Etude des effets d’apports de matière organique sur les bilans
hydriques et minéraux et la production du mil et de l’arachide sur un sol dégradé du
Centre-Nord du Sénégal. Thèse de Doctorat, Institut National Polytechnique de
Lorraine, Nancy, 184~.
Diack M., Badiane A.N., Sène M., Diatta M., Dick R, 1998. Cordyla pinnate en
association avec Piliostigma reticulatum: ‘Impact sur la régénération des sols dégradés
au Sénégal. Rapport final du projet ISRA-NRBAR,
LlRO2, ISRA- Sénégal, 25~.
Ndiaye J.P., 1978. Enquête fertilité en milieu paysan dans la région du Sine saloum,
ISRA, 15~. + annexes.
Sène M., 1995. Influence de l’état hydrique et du comportement mécanique du sol sur
l’implantation et la fructification de l’arachide.Thèse de Doctorat, Montpellier, France.

---

73
THEm 3 : NUTRIENT MANAGEMENT

---

74
Affect of organic and inorganic fertikers on the nutrient
status and yield of dryland crops
Eiaptîsta, 1. and 1. Anahory, Inida, Cape Verde
Capeverdean subsistence fàrmers cannot afford to invest in expensive fertilizers
to supply nutrients to their dryland crops; therefore, application of animal matmre
alone or mixed with moderate amounts of chemïcal fertihzers could be a possible
alternative. Animal manure, besides being vahtable sources of nutrients, increas
water and cations retention of the soil and, when applied alone or in combinatîon with
moderate amounts of chemical fertilizers, increase yields of crops and available nutrient
content of soils.
This field work was conducted at the Riieira Seca watershed to
determine the Hect of animal manure and NPK fertilizer in the yield and growth of
dryland crops, h paxtîcular, com (.%a may.s) and beans (Vigna unguicuhta and
LabZ& niger). The experiment was set up in a randomized complete block design
w3.h .four replication in which the treatments were: 1. Control, with no chemical or
organic fertilizers; 2. 60 kg N/ha irt which 50% was applied preplant as NPK ( 16 1 O-
20) and 50% as ammonium sulfate (50% N) as sidedress, four weeks after planting; 3.
20 tlha of cow mure incorporated into the soil; and 4. combinatîon of treatments 2
and 3. Sirme it did trot rain enough, the crops did not reach maturity. The parameters
evahtated were: com plant height, com dry matter yield, nutrient content of bean plant
tissue, bean grain yield and soil nutrient content. ‘Treatments did not affect com plant
heîght, soil nutrient content or bean grain yield. The application of NPK and manure,
alone or combined with each other increased corn dry matter yield. The highest dry
matter yield (3.92 Mg/ha) was obtained wîth the combination treatment, and the lowes&
(2.27 Mg/ha) with the control Ahhough there were some inconsistencies, the study
suggests a beneficial effèct of manure combined wîth NPK chemical fertilizer.
1. Iutroduclion
In Cape Verde, dryland crops, mainly cor-n sud beans, are grown under poor
management wîth no chemical or organic fertilizers as nutrient sources.
Since
Capeverdean subsistence fàrmers cannot afford to invest in expensive fertilizers to
supph, nutrients to their dryland crops, applicatim of animal manure alone or mixed
with moderate amounts of chemical fertilizers could be a possible alternative.
Animal manure, besides being vsluable sources of nutrients, increases water and
cations retention and, when applied alone or in combination with chemical fertilizers.
increases yields of crops and available nutrient content of SO&. Organic nutrient
sources such as animal manure may reduce the amount of fertilizers required For
optimum com growth and y-ield. However, the amount of animal manure is hmited and
E-trmexs usually do not have enough to apply on the slopes where dryland agriculture is
practiced. On the other hand, the animal manure used is generally low in nutrient
content, particularly N, because of the animal diet and handhng of the manure during
collection, storage and application.
The objectives of this field tria1 were:
?
1 .To extend the animal manure both chemically, biologically and physically

---

75
by combining a moderate amount of NPK fertilizer with the manure and
applying it to dryland soils where com and beans are grown;
?
2. TO evaluate the effect of both chemical fertihzer and animal manure on
trop growth and yield.
2. Materials and methods
This field experiment was conducted at a farmer’s field near S. Jorge, in the
Ribeira Seca watershed during the 1998 growing season. The soil at the experimental
site was a loam and laboratory analysis sk-,sd that initial soil pH was 6.8, OM
content 1.04% (dry combustion), P,O, (Olsen) and K20 (ammonium acetate) in the top
20 cm were 109 (very high) and 636 (very high) ppm, respectively. Nitrogen content
was extremely low. The cow manure used contained 1.1% N (Kjeldahl digestion);
0.8% P (colorimetry); 1.5% K (flame photometry); 19% ‘OM (dry combustion) and
26.2% moisture.
The experimental design was a randomized complete block with four
treatments replicated four times. Experimental units were plots w-ith 28 m’. The
treatments were:
?
l- Control, with no chemical or organic fertilizers;
?
2 - 60 kg Niha in which 50% was applied preplant as NPK ( 16- 10-20) and
50% as ammomum sulfate (50% N) as sidedress, four weeks alter planting;
?
3 - 20 t/ha of cow manure incorporated into the soil with a hoe before
seeding;
?
4 - and a combination of treatments 2 and 3.
Com and beans were sown on July 27 (after what later proved to be an insignXcant
rainfall) in bills at rate of 28 hills per plot and four maize seeds and 2 different bean
seed.s per hill. Com seeds used were of a local variety and the beans were cowpea
(Vignu unguiculata)
and feijao pedra (L,ablab niger).
Com plant height was measured fiom soil level to tip of first fully developed
leaf two months alter seeding, by randomly selected 6 plants in each plot. Bean tissue
samples (Young fùlly developed leaves) were collected f?om 20 randomly selected
plants in each plot for analysis of dry matter N, P and K concentrations. Total N, P
and K were determined by Kjeldahl analysis, colorimetry, and flame photometry,
respectively. Com ear leaf samples were also supposed to be collected for dry matter
nutrient analysis but, at that time, the trop was already sufTering fiom moisture stress.
There was no harvest of grain yield because lack of rain prevented the trop
fiom reaching maturky. Only the Vip unguiculata
produced some grains, which
were harvested on November 10.
Both grains and pods were weighed. TO
compensate for the lack of com grain yield, the straw in each plot was tut at soil level
and .weighed to determine the effect of treatments on dry matter yield. The straw was
not completely dry at cutting time; therefore, moisture content was determined by
drying straw samples at 65 OC for 48 hours. Plant population was also determined
because it was not uniform for all plots. Soil samples were collected at 20 cm depth in
each plot at the end of the experiment for laborat,ory analysis of pH (1:2 soikwater); N
(Kjelldahl); PzOs (Olsen), KzO (ammonium acetate) and OM (dry combustion).
Data were analyzed using analysis of variante and covariance. Fisher’s LSD
test and multiple range analysis were employed to separate treatment means.

---

76
3.Results and Discussion
Table 1 shows the rainfàll amount and distribution during the period of the
experiment. It cari be seen that the rains started very late in the season and ended very
early. In fact, the first significant rainfall only fell in late August and the last at the end
of September. RainGll was insuf3icient and poorly distributed during the tria1 period
which caused the crops to IX. Soil moisture was insufbcient at seeding time, causing
poor emergence and early growth of both corn and beans, even though seeding was
done twice. AIter tasseling it did not min anymore; therefore, there was not enough
soil moisture to promote grain filling and, consequently, there was no grain yield.
The analysis of varianee and çovariance (Table 2) indicated a highly significant
effect (pxO.01) of the treatments on corn dry matter yield (52% moisture). The use of
the number of plants as covariate turued out to be very significant (pcO.01) in reducing
the experimental error and detecting real difbxences due to treatments. A more
detailed analysis showed that NPK fertilizer and manure increased cor-n dry matter
yield both when applied separately and in combination with each other. However, a
positive interaction between NPK fatilizer and animal manure was not verified. It is
important to keep in mind that this soil was higb in P and K and, therefore the results
in this situation of high P&IC fertility no positive interaction occurred.
Table 3 shows the effect of treatments on mean dry matter yield and height of
corn plants. The mean dry matter yield of the NPK treatment was not signifïcantly
difl!erent fiom that of the animal manure treatment. The combination of the chemical
fertilizer and the manure resulted in an increase of42 and 23% over the control and the
manure alone, respectively. Dry matter yield appears to be additive between
application of NPK and animal mamue. There was no significant difference between
the combination and the NPK fertilizer alone. Mean dry matter yield was higher for
the NPK treatment than the manure treatment. Multiple range analysis showed a mean
corn dry matter yield of 2.4 and 4.0 Mg/ha for without and with NPK fertilizer,
respectively, while for the without and w-ith manure treatment it was 3.0 and 3.4
Mg/ha, respectively. Given the high soil P & K one is left to speculate that the yield
improvement might be due to N. It could be interesting to calculate the N contribution
fïom manure in comparison with that in the NPK fertilizers perhaps foliar N levels
would help test this hypothesis of the N effect.
In spite of the positive effect of the treatments on dry matter yield, the
treatments did not have any significant e.ffect on early cor-n plant height. In the same
way, the treatments did not have a significant effect on bean grain yield (Table 3).
Bean grain yield was very low, with a mean yield of 22 kg/ha due to lack of rain afier
flowering, causing poor grain filling. Grains were also very small.
Soi1 analysis at the end of the growing :season did not indicate a sign&xnt
effect of the treatments on soil N, P and K content (data not shown). This could be
due to the very high initial P& and K20 content of the soil.
In fact, according to
INIDA’s soils laboratory recommendation, a response to application of P and K
should not be expected in such condition. However, the application of mamre resulted
in a highly significant ($0.01) increase in soil organic matter content (Table 4). But,
in spite of the increase, the organic matter level is still in the low range for all
treatments (’ qn’
. &, 0). This could be due tci thL low OM content of the manure applied to
the soil and also because a large amount of OM is required to build up in the soil.
With the application of manure and NPK fertilizer the nutrient concentration in
the bean plant tissue increased significantly (Table 4). However, according to INIDA’s
--

-
*-.---Y

---

77
sd fertihty and trop fertilization manual (1997), in spite of the increase, the
concentration of N and P in the bean plants are &Il below sufhciency range. The low
nutrient concentration in the tissue samples could be due to water stress which limited
nutrient uptake. The combination of NPK fertilizer with aruma manure resulted in the
highest tissue concentration of N, P and K which means that the ‘combination
promoted more nutrient uptake than the individual application of NPK or mamue.
4. conclusion
It is obvious that the experiment did not give the expected resuhs because of
the delay, poor distribution and shortage of rain. Dryland experiments in Cape Verde
are always risky because of the uncertainty of the rains. With this tria1 we could not
evahutte the effect of the treatments on corn grain yield as expected, but we tried to
evaluate treatment effects on other parameters such as com plant height, soil nutrient
content, com dry matter yield and concentration of nutrients on bean plant tissue.
The results suggested that, although the application of the manure alone
increased com dry matter yield signifïcantly, the combination with manure is more
desirable because of the effect on nutrient uptake and other soil properties. However,
the tria1 should be repeated in favorable soil mois$tue condition to reach the conceived
objectives and to consolidate the resuhs of this study.

---

78
Table 1. ( -owhg season rainfiall data (cm) at S. Jorge, 1998
Months
1
m-
DZ&?!3
July
August
Septembe
October
lovemk
1
0
0
0
0. l
0
2
0
0
0.2
0.3
0
3
0
0
0.1
0.3
0
4
0
0.3
0
0. l
0.1
5
0
0
0
0.2
0.2
6
0
0
0
0. B
0.1
7
0
0
1.3
0.11
0
8
0
0
0
0
0
9
0
0
3.1
0
0
10
0
0
9.2
0.5
0.1
11
0
0.2
17.7
0.2
0
12
0
10
36.7
0.7
0
13
0
0
0.1
0.1
0
14
0
0.2
0
0.1
0
15
0
0
0
0
0
16
0
2.1
16.5
0
0
17
0
0
16.8
0.1
0
18
0.2
0
25.2
0
0
19
0
0
2.2
0
0
20
0
0
0
0
0
21
0
2.4
0
0
0
22
0
40.2
15
0.2.
0
23
0
0
0.1
0.1
0
24
8.5
0
0
0
0
25
5.5
0
0
0
0
26
0
0
0
0
0
27
0
10.5
1
0.2
0
28
0.2
11
32.6
0
0
29
0.8
0
7.5
0
0
30
1.6
0
0.1
0.1
0
31
0.4
0
0
0
0
Ota1
17.2
76.9
183.4
3.5
0.5

---

79
Table 2. Analysis ofvariance and covariance relaGve to the effect oftreatments on
corn dry matter yield
Sonrce of variation
ctf.
S S
M S
F
sig. level
plant #
1
-l3.982
13.982
11.02
**
Blocks
3
9.406
3.135
2.47
NS
Treatments
3
41.378
13.793
10.87
**
NPK
1
30.768
30.768
205.63
**
Manure
1
9.572
9.572
63.97
*
NPK * Manure
1
0.629
0.629
4.20
N S
El-IW
8
10.153
1.269
Total
1 5
74.920
* -p<o.o5;** - ~~0.01; NS. - net significant
Table 3. Effect of treatments on corn pla,nt height, corn dry matter yield and bean grain
yield
~
Treatments
Corn plant
Corn dry matter yield
Bean grain yield
l
height (cm)
kg/28 m2
M&a
kg/28m’
kg/ha
ContTol
1.24
6.37
2.27
0.052
18.58
NFK fertilizer
1.26
9.86
3.52
0.064
22.86
Manure
1.32
8.50
3.11
0.063
22.50
NPK + mamue
1.32
10.97
3.92
0.070
25.00
LSD (0.05)
0.15
1.83
0.65
0.029
10.26
Table 4. Effect of treatments on bean plant nutrient concentration and soil OM content
Treatments
N
l
P
K
Soil OM

---

80
An htegrated Fertilizatim Study of the Groundnut-Muet
Rotation Cropphg System
Babou Jobe, Soil Scientist and Program leader for Cropping Systems and Resource
Management program National Agricultural Rerxxuch Institute (NARI), PMB 526,
Serrekunda, the Gambia
Abstract
Like elsewhere in the Sahel, agriculture is the most important economic sector
in The Gambia. This sector provides employment for more than 75% of the rural
population, contriiutes between 20-25% to the countq+ Gross Domestic Product
(GDP) and generates ahnost 85% of foreign exchange earning and 40 % of total
export eamiugs. This agriculture is, however, at .a crossroad w-ith declining per capita
production and a deteriorating resource base. The region most affected is the North
Bank Division of The Gambia. The North bank division is one the most important
agricultural ,regions in The Gambia. This region covers an area of 221- 000 ha, has a
population of 156,000 people (15% of national population), and a total of 45,000
cattle, 64,000 small ruminants, 11,000 horses and donkeys (draft animals). These
resources put together account for 33% of the total agricultural output of The Gambia.
Currently, this region is experiencing a steady loss in trop production capacity,
which is on the one hand on account of the weather (low rainfXl), but most oRen. a
limitation posed by low soil fertility. Low trop yields in this region are due to the
combined effects of low inherent soi1 fertility and agricultural expansion characterized
by resource based (natural soil fertility) farming systems.
Resource-poor fàrmers, fùrther constrained by removal of subsidies, often
cannot apply recommended fertilizer rates, hence have sharply reduced fertilizer use.
Fertilizer use in The Gambia dropped to less than 5 kg/ha in 1996 fiom the application
rate of 16 kg/ha in 1985. These fàrming systems are nutrient mining. resuhing in the
negative nutrient slows and nutrient balances. Resource-based farming is unsustainable
and ctfIen goes hand-in-hand with land degradation.
Reduced dependence on chemical inputs and increased dependence on available
local resources is most appropriate for a fragile and stressed ecosystem under the
management of a resource-poor farming community. For environmental degradation to
be reduced and the production performance of the existing croplands increased, there
is the need to improve the nutrient budgets of these soils through organic or inorganic
sources. However, socioeconomic conditions, availability of inputs and hmited soi1
resïlience favor the integration of the two strategies.
The study (in a randomized complete block design with four replications)
utilized a groundnut-millet rotation system to aszess: i) the long-term changes in soi1
chemical properties due to soi1 amendments and ii) the long-term effect on the
production of the rotation system. The interaction eEects of different manure (stable
ar;i: cA>tV diüng) at 3 t/hc; vV;tE fi+* Icvels (0, 15, 30, 35 Ü..J CO kg/ha) PtOs CU N b+c;re
evaluated for groundnut and millet respectively at Njaba Kunda.
Two-year results showed that manure + fertihzer was superior to fertifizer
applied alone especially at low rates on trop biological and economic yields. The

---

8 1
mtqpted fertilization stmtegy inmased the sutput of the rotation system by over
twice the control (no fertilization) and 7% more than fe&zer applied alone. This
comptementary effect of manure and fertilizer was manif&& across rates and a sure
strategy for increasiug fertifizer use eflïciency.
In addition, the combined (integrated) fertilisation strategy impacted on
physical and physiochemical properties of the nutrient-poor and poor-buEering soil of
Njaba Kunda. The loss in soil productivity is oRen related to loss iu soil organic
matter. Soi! organic matter the key attriiute of soil fertility was enhanced significantly
over minerai fertihzer alone. The inerease in soil organic matter in manure and manure
+ f&ihzer apphed plots was due to the reduction in the rate of soil organic matter ioss.
The manure treatment acted as a strong L+k against decomposition. Also, bases
such as calcium and potassium were enhanced that cari bring about increased plant
nutrition and reduced acidification.
From the indications of this experiment, under current fàrmers’ practice there is
the threat of decline in productivity overtime with the likelihood of irreversiile
environmental degradation. Therefore, in order to maintain productivity, there is the
need to adopt the use of soi1 amendment particularly from organic sources.
Such an experiment needs to be long-term in order identi@ key indicators of
land degradation as soi1 organic matter and eistablish critical values for them for
difherent agroecologies. Also, in order to facilitate large-scale manure use especially on
faraway fields, tbere is the need to look înto ‘more efficient manure tmnsportation
means and proper field storage methods.
The maintenance of soïl ftity, given an increasing intensity of land use
associated with population growth and mechanizatîon is crucial to the long-run
productivity of Gambii farming systems. Yields of crops, particularly food crops, are
either declining or stagnant and the demand for food is increasing more rapidiy than at
anytïme in history. The steady loss in trop produ&on capacity in The Gambia is on the
one hand due to recent years of low raiefall, but most significantly, is a limitation posed
by low soil fMlity.
Resourmpoor fàrmers, fiuther constrained by the removal of subsidies on
fertihis, often cannot afford to apply the recommended fertilizer rates, hence they
have reduced fertilizer use. Fertilizer use declined fiom an aheady low of16 kgIha in
the mid-1980s to less than 5 kg/ha N+OS-K20 in 1996 (FAO, 1996). According to
Rhodes et. al. (1996), The Gamhia is a moderately nutient deficient coutttry tih a
negative nutrient slow of 14-3- 15 kg/ha (N-P205-&O). These farming systems are
nutrient mining, since output exceeds input, resuhing in a net negative nutrient flow.
Farmers’ reaction to the decreasing yield situation is to increase the planted area
more by using more marginal lands. No where is the problem is more evïdent than in
the North Bank JXvision (NBD), a region being among the most important in the
country agriculturaliy.
The NEHI covers an area of about 221,000 ha, of which 42% is suitabie for
agriculture and another 21% is considered marginal. Cropland covers shghtly more
than 45% i;? ;he area, with ; tgct,::on çover of less th&, iO% of the landsapt
(Dunsmore et al., 1976). It has a population of 156,000 people (15% of the national
population), a total of 45,000 cattle, 64,000 small ruminants, 11,000 horses and

---

82
donkeys (Population and livestock census, 1’993). These resources put together
accouru for 33% of total agricultural output of Ihe Garni&.
The cropping systems on the uplands are largely a groundnut-cereal (early
millet) rotation. The North Bank bas the largest area of groundnut in the country,
varying between 20,000 and 30,000 ha, with an average yield of littb more than 1
ton/ha. The early millet area has been growing rapidly and yiekis average about 900
kg/ha (Posner et al., 1989).
Most fàrmers have for a long time known the use and importance ofboth manure
and: inorganic fertilizer in increasing trop yields through soil fertility enhancement. The
use of manure is the major soil fertihty maintenance strategy in the inner fields (close to
compoLLIlL , on food crops. The effect of fertihzer, on the other hand, has been
demonstrated to these farmers on all major crops and adoption rates have been high for
cash trop unta subsidies were removed on fertilizer and it became unafhordable to the
majority of these resource-poor fàrmers.
Manure cari offer multiple bene@ but it is not available in the quantities
required. E:xclusive dependence on chemical fertilizers on the other hand even where
affordable, may have detrimental effect on the long-term productivity of the natwal
resource base. Reduction of dependençe on chemical inputs and increased dependence
on available local resources such as manure, compost and trop residues would be the
most appropriate technology for a fragile and stressed ecosystem under the management
of a resource-poor farming community.
This therefore, calls for an integrated plant nutrient system approach. The main
goal ofthis strategy will be to integrate the use of organic and inorganic sources of plant
nutrients for a sustainable increase in agri&ural productivity in an efficient and
enviromnentally fiiendly mariner.
SpciJic hypotheses
1. Stable manure is of higher quality in plant nutrient content and organic matter and
of more value than cow manure in improving trop response to applied fertihzer.
2 . Soils in tha study area (Njaba Kunda) are nutrient-deficit (negative flow) and are
likely respond to soil amendments and improvements in the nutrient budget.
SpeciJic objectives
i) to assess the changes in trop responses (yields) to difEerent soil amendment materials.
ii) to assess changes in soi1 properties due to the effect of difberent soil amendment
materials.
2. MateriaIs and Medmis
Experimental site
This two-year (1997 and 1998) experiment was conducted on an upland site in
Njaba Kunda in the central region of NBD, t.aking into consideration the dominant
cropping systems (groundnut-millet rotation system). Njaba Kunda is in the Sudano-
sahalian agroclimatic zone with a mean rainfall of 700 mm. The seasonal rainfall at the
site were 648 and 805 mm, being 93 and 115% normal, for the fïrst and second year
respectîvely. The fitst cycle of the experiment (1997) with groundnut, was replicated at
three sites other than Niaka Kunda, which differed in soil and annual mean rainfall:
Yundum (soil associationl); Jenoi (soi1 association 2); and Sapu (soil association 6).
Rainfall for these cite< in 1997 was; 900, 65 1 and 865 mm respectiv4y.
Soils
Generally, the soils of the uplands belong to the soil order of Alfisols, with low
fertility for reasons of geoiogy and human activity. Soil associations 1 and 2 have similar

---

83
charaoteristics; they are deep, weil-drained, moderately permeable soils of the uplands
with loamy sand or sandy loam surface layers and a sandy clay loam subsurface.
How~er, soil association 2 (at the point of Njaba Kunda) has an eolian deposited parent
material, hence coarser in texture because of Iower silt + clay (< 10%) content
(,Dunsmore et. al., 1976).
Association 6 soils on tbe other hand are deep, somewhat poorly drained and
moderately slowly permeable. For both years, preplant and postharvest soil samples
were collected at a depth of O-20 cm, air dried and ground to pass a 2 mm-sieve.
Measurements were made on soil pH (water and &Xl) in a soil to water/soltiion ratio of
1:2.5 using a glass electrode. Organic matter content was determined using the loss-on-
ignition method. Total N was determined by microXjeldah1 method and macrocations
(Ca, Mg, and Na) determined by flame photometry and available P was determined by
the colorimetric method using the spectrometer 21.
Field experiment
A randomized complete block experimental design was employed to evaluate the
integrated fertilization effect of organic (manure:) and inorganic (fertilizer) sources of
phtnt nutrients on trop yields and soil chemical properties in a groundnut-millet rotation
çropping system. Three types of soil amendments (no manure, stable manure and cow
dungb were applied at a rate of 3 tons/ha, in four replications, and were evaluated under
five levels (0,15,30,45, and 60 kg) of applied fertilizer.
Planting was done on flat land (no tihage), but mechanized for both crops
(groundnuts and millet) using animal traction. The groundnut variety 73-33 was planted
at a spacing of 50 cm x 12 cm and the millet variety (local) was planted at a spacing of
75 cm x 75 cm, inter-row vs intra-row. For both crops the experimental unit was 15 rn’.
Phosphorus fertilizer was applied to the groundnut trop at the specified rates as
triple superphosphate. In the second year, nitrogen in the form 15-l 5- 15 and urea was
applied to the millet trop. For groundnut, botb the manure and the phosphorus fertilizer
were applied and incorporated at tinte of plantîng. For millet trop, manure was basally
applicd and incorporated while N was applied in two splits, two-third the amount (2/3) at
3 weeks a&er planting and one-third (lJ3) at 6 weeks afier planting.
For each year, data on stand Count at harvest, dry pod yidd and biomass were
collected. The data were statistically ana’tyzed using the MSTAT and Minitab software
packages.
3.l.zesults aItd Ditx~om
Gromdnut cxop-year /
Table 1 presents mean groundnut pod yidd data of the 1997 experiment across
the four locations evahtated (Yundum, Njaba Kunda, Jenoi and Sapu). Yields of
groundnut pod weight ranged fiom a low of 700 kg/ha to a high of 2000 kgha in
Yundum. Differences in mean pod yield a.mong locations may in part be due to variation
in seasonal rainfall and soi2 environment. Yundum and Sapu have comparatively more
moisture (940 and 856 mm of rainfàll compared to 648 and. 651 for Njaba Kunda and
Jenoi), Heavier soils (> 15% silt + clay) at Yudum and Sapu than at Njaba Kunda and
Jenoi (< 10% silt + &y). Jenoi and Sapu also had lower plant populations due to 2 l-day
dry spell following planting and due to erosion respectively.
Anal>,& of\\:ariance on individtil sites (annex) indic~;;~ L;t there was a difL~&.I
response to soil amendments across locations. Yundum had significant manure x
fertilizer interaction effect, while Njaba Kunda and Jenoi indicated signifïcant main
efXect,s in added fertilizer and Sapu showed no treatment effect.
-.
-.-.-.-.--
..-UI_c
-1(1--

---

84
Combined analysis was only possible for Yundum and Njaba Kunda md
suggested sigttificant interactions among manure~ + fertilizer treatments. Table 3 shows
higher mean yields for manure + fertilizer, particularly with stable manure, than for
fertihzer added alone. Stable manure gave the highest yield at 15 kg PzOs/ha but stih
produced higher yields at each corresponding fertihzer level than cow dung or fertilizer
alon e.
Cow dung, however, produced higher mean yields than fertilizer alone and
showed a negatively linear response with incremental rates of phosphorus fertilizer.
Average fertihzer effect (Table 4) was higher in Njaba Kunda than at Yundum
with an mean average increase in yield of 896 and 394 kg/ha or 75 and 26% more than
their controk, respectively. The application of 3 t/ha of stable manure or cow dung
produced an average yield difherence of 410 and 960 kg/ha, respectively over the control.
Average petiormance of stable manure + fertihzer and cow dung + fertîhzer were
comparable, with 50 and 52% yield increase respectively over the control mean yield.
The better groundnut response to fertihzer at Njaba Kunda may be attributed to
the lower fertihty level of Njaba Kunda soils, thus increasing the likelihood of response.
W~GXI Njaba Kunda was analyzed separately the only significant difherent was between
soi1 amendment against no soil amendment. Even though non-significant, yields were
higher with manure plus fertihzer than for f-r alone, at the corresponding level of
applied fertihzers. The yield advantage ofthe manure + fertihzer plot was most likely due
to the additional nutrients and soil moisture enhancement provided by the manure to this
nutrient-poor and Jow-water retaining soil.
Table 1: Mea Pod yield (kg/ha) of Groundnut across Locations as Affected by Soil
Amendment m 1997.
Manure
PzOs-Level
Yundum
Njaba
Jenoî
Sapu
Mean
@3tiha
kg/ha
Kunda
“_______-_____________________________I_------------------.------------------------------~---
Nomanure
0
1514
1192
510
1125
1085
C‘
15
1799
2133
608
1285
1455
‘L
30
1847
2025
745
1344
1490
L‘
45
2076
2026
845
1177
1541
LL
6 0
1910
2125
608
1750
1598
Stable manure 0
1875
1650
700
1396
1405
L‘
1 5
2118
2333
777
1281
1627
“
3 0
2070
2000
883
1583
1634
“
45
1882
1708
790
1396
1444
“
60
2069
2033
608
1365
1519
Cow dung
0
2792
1833
753
1302
1670
‘L
15
2333
2200
567
1167
1567
‘L
30
1910
1917
748
1271
1 4 6 1
L‘
45
1972
2042
770
1052
1459
i‘
60
1958
2075
708
1177
1480
---~ ..-
Location mean
2008
1956
708
1311
1. Manure;
1= No manure,
2 = Stable manure,
3 = Cow dung

---

85
2. P-Level;
1 =o, 2=15,
3 = 30,
4=45,
5 = 60 kg P205
Table 2: Mean pod yield (kg/ba) fiom manure x fertilizer interaction averaged over
Yundum and Njaba Kunda
Added kg/ha
No manme
Stable
Cow dung
P205 (fertilizer)
manne
--_-,-__---_--_--___-____________________------~-------------------
_______--___-____--_---------------
0
1353
1763
2313
1 5
1966
2222
2267
30
1936
2035
1913
45
2072
1795
2007
60
2017
2035
2017
Regression Equations
1. Fertihzer:
Y = 1415+31.8P-0.371 P”, r=0.948X, = 43kgP?O&a
2 . Stable Manure Y = 1877 + 9.74 P - 0.146 P2, r = 0.342X, = 33 kgP?O&a
3 . Cow dung Y = 2353 - 16.3 P - 0.178 P2, r = 0.876 X, = 45 kgPz05/ha
X, = Amount of fertilizer nutrient required for maximum yield.
Table 3: Average site-specîfïc performance in groundnut pod yield as affected by
the difherent soil amendments.
-~
Yundum
% increases
Njaba
% increases
in yield
Kunda
in yield
--_____------_---------------------------------~-----------------------------------------------------
Fertilizer
1908
26
2088
75
Stable manure
1875
24
1650
38
Cow dung
2791
84
1833
5 4
Stable + fert.
2035
34
2019
6 9
Cow dung + fer-t.
2043
35
2058
73
Control
1514
1192
In order to answer some meaningtùl practic,al questions, preplanned
comparisons were performed. Orthogonal contrasts (Table 5) on groundnut pod yield
fiom Njaba Kunda revealed that the only significant comparison was between
amendment and no amendment. The application of soil amendments whether fertilizer.
manure or the combination of the two significantly increased yield over no application
of soil amendment. This indicated that the natural soil fertility at the site was low,
hence it could not provide adequate plant nourishment and, therefore, requires
supl!errnntation.
From the fitted orthogonal polynomials (Tabl’e 6), responses in pod yield to the
difberent amendments were curvilinear, an indication that the change in pod yield was
not constant for every incremental change in applied soi1 amendment. The response to
_
.-_-
p---1
‘-‘-“A
II-----

---

X6
fertiliser apphed alone was quadratic, cubic for stable manure + fertilizer and not
defined for cow dung + fertilizer. According to the fertihzer response Onction,
groumhmt pod yield will increase by 43-kg pod/kg of P,Os applied as fertilizer alone
and maximum yield was obtained at 41 kg PzOJla.
For stable manure + fertilizer, the change in pod yield at low rates was 84 kg
pod/kg P,Os.
However, yield declined faster than the decrease in fertihzer at
intermediate rates (> 30 kg/ha), but peaked up again at higher rates. Especially
notjceable on tropical sites was the effect of manure as a P fertibzer and the improved
effectiveness of minera1 P fertilizers when combined with manure. Initial available P in
these soils was low (cl0 ppm), and are largely stored in the soil organic matter which
is readily depleted. Fertihzing with uurinure cari counteract P deficiency or its reduced
availability on cultivated soils. How manure brings about an increase in available P
may be chemical, physical or biological in nature :
1) organic colloids prevent dissolved phosphate Fom coming into contact with hee
ahuninum and iron,
2 ) when organic matter decays, the organic acid that forms dissolves phosphate,
3 ) organic phosphorus is less strongly fïxed by soil and
4 ) microorganisms minerahze organic phosphate compounds.
Mil/et cropyear 2
Mean grain yield of 557 kg is below the average yield in the area (Table 5) and
was attributable to late planting and disease infestation (Downy mildew). Farmers in
this area ‘dry seed” their early millet in order to gain time for planting groundnuts
whereas, the experiment was planted when the soi1 was sufliciently soaked by rains.
Grain yield ranged Fom a high of 866 kg/ha for cow dung + 60 kg N/ha to a low 125
kg/ba for the control (no fertilization).
Grain yield was afbected also by Downy mildew infestation, especially in the
low fertility plots. The effect of the disease on grain y-ield was manihested through
depression of plant vigor affecting parameters such as tiller production for panicle
Count and plant height. Plots with disease scores of 7 and 9 (high inGestation) had
signifïcantly fewer panicles and shorter plants. Tbe occurrence of heavy downy mildew
disease infestation corresponding to low fertihty plots was an indication of the
importance of fertility in pest and disease management.
The main effects of the experiment (Table 5) manure and fertilizer, were
significant for grain yield, panicle Count and plant height with least signiftcant
difberences at the 5% level of 135 kg/ha., 11 panicles and 1 I centimeters respectively.
Manure signifïcantly outperformed no manure m all of the stated parameters. The
average effects of manure on grain yield was 30% greater than that of no manure and
among manure, cow dung produced the highest mean grain yield of 623 kg/ha. The
response to fertilizer was cubic with extrema at 15 and 60-kg PzOs/ha rates.
Orthogonal contrasts of treatment means (Table 5) revealed significant
difference on millet grain yield in year two of the rotation experiment among the
difberent soïl amendments and also a manure x fertilizer interaction. In the fïrst year of
the rotation experiment with groundnuts, the only signifrcant différence detected was
between some soi1 amendment and no soil amendment on pod yield. Soil amendment
was superior to no soil amendment and between the soil amendments, organic
amerl-‘meEt combined with fertihzer was superior to i*-xganic amendment app!ied as
fertihzer alone. Between the manure types, the interaction effect of cow dung +
fertihier was higher than stable + fertilizer.

---

87
The positive effect of manure on fertilizer use efficiency might be attributed to the
delayed effects of manure. After repeated applications (two), both the immediate effect
and the delayed effects of earlier applications coincide, and mamue starts to have its
maximum impact on yields. The impact of manure on yields depends strongly on the site
that is, on primary effixt on soils (as fertilizer, biological or physical) and on the state of
the soil The soit at the experhuental site is both at nutrient-poor and poor-bufkring
capacity status. Therefore, the likely impact of manure is through nutrient supply and
enhanced moisture retention and availability to tbe plant.
Table 4: Effect of difberent soil amendments on the grain yield, yield components
and agronomie parameters of millet (year 2 rotation)
Treatnmt
Grain Yleld
Stand
Panicle
Plant Ht
DowTlymildew
c0lm.t
CQm
(cm)
h4ainc.d%ct
Mamrre
1
1.463b
1 4
35b
139b
6
2. 583 ab
1 5
54 a
151 a
5
3. 623 a
1 5
47 a
151 a
6
SY
47
1
4
4
0.03
Id .05
135
?Js
11
11
Fertilizer
1. 414 c 14 35 b 135 b
9
2. 606 ab 14 46 ab
154ab
4
3. 539
abc 15 44 b 139 b 6
4. 522
bc 15 45 ab 152 ab 7
5. 702 a 16 59 a 154a
6
2 ~_-_-----I_--------____________l_____l_---------------------------------------______________I_____--
3
s, 61 1 5 5
4 Lsd
45
174 ns 14 14
5
~-_-_-----~-_----_-------------------------------“--------____---_--_-___-_-_------------------------

---

88
hteradion &ect
7
Manure
Fertiliser
1
1 125 d
13
10d
114d
9
1
2 642
abc
14
47 abc
154 abc
4
1
3 467 bc
16
34cd
132 c d
6
1
4
442bc
13
38bc
149 abc
7
1
5
642
abc
17
49 abc
149 abc
6
2
1 542 bc
16
51 abc
138 abc
6
2
2 517bc
15
51 abc
154 abc
5
2
3 542
bc
15
49 abc
148 abc
6
2
4 717 ab
16
59 ab
160 ab
5
2
5 600 abc
16
61 ab
154 abc
5
3
1 575 abc
15
44abc
154 abc
7
3
2 658 abc
13
42 abc
155 abc
6
3
3 608 abc
14
48 abc
137 bcd
7
3
4 408 cd
17
38bc
149 abc
7
5 866 a
15
66 a
161 a
4
.--~
-~-. -- -~-
Meân
557
15-
46
147
6
s.
106
1.5
9
8
hd. 05
302
ns
24
24
Marure: 1~10 mmure, 2=Stable manm, 3=Cow dmg.
Fertilizer: I=O, 2=15, 3=30, 4==45, 5=60 kgN/ha

---

89
Table 5: Orthogonal Contrasts for 1997 Groundnuts and 1998 Millet Soil
Amendment Treatments.
-
-
Gromldnut
Millet
----“----
--__“-__-__-_-_--------
_--____-_____-__---------
-----/“m- ---
Contrast
D F
F value
Prob.
F value
Frob.
-
-
1. Amt. Vs Noue
1
12.68** 0.001
17.83
o.ooo**
2. Org. Vs Inorg.
1
<l
5.84
0.020*
3. F2 Vs SF2
1.30
0.260
<1
4. F3 Vs SF3
<1
<l
5. F4 Vs SF4
2.23
0.143
<l
6. F5 Vs SF5
<l
1.06
0.309
7. F2 Vs CF2
<l
<l
8. F3 Vs CF3
<l
<l
9. F4 Vs CF4
1.337
<l
10.F5 V s C F 5
<I
6.62
0.014”
ll.SF2 V s CF2
1.30
Cl
12.SF3 V s C F 3
<l
<l
13.SF4 V s C F 4
<l
6.29
0.016*
14.SF4 V s C F 5
1
<1
3 . 3 6
0.07
--
F2 == fertiliser at level2;
SF2 = Stable manure + fertilizer at level2; CF2 = cow
dung + fertihzer at level; Amt = amendment; Org. = organic; Inorg. = inorganic
--mm----
.._----
8,

---

90
Table 6: Response Curves of Groundnut pod ykld and Millet grain yield due to
DBerent Soil Amendments.
--
GroundnLlt
Millet
---------_---------3____)____I_
____--___------_----____________
SS
F value
SS
F value
DF
Fertilizer E f f e c t 4
Linear
1294560.4
6.56*
277555.6
6.19*
Q u a d r a t i c :
747054.0
3.79*
66654.0
1.49 ns
Cubic
1
455395.6
2.31 ns
337824.7
7.54**
Deviation 1
<l
<l
Stable manure Effkct 4
Linear
1
17305.6
<1
39942.4
4
Quadratic 1
106314.3
<l
390.3
<l
Cubic
1
898800.4
4.56*
46785.6
1.04 ns
Deviation 1
<l
-4
Cow
4
dung Effect
Linear
1
1 4 7 4 5 . 6 <l
44622.4
<l
Quadratic 1
4536.0
<l
103544.0
2.31
Cubic
1
9 6 4 3 2 . 4 <l
250905.6
5.60*
D e v i a t i o n 1
-
<l
Regression equations
Groundmt
Fertilizer :
Y= 1318 + 42.8X - 0.513X2, R” = 0.340, P=O.O3
Stable manure:
Y = 1653 + 84.6 X - 3.52 X’ + 0.004 X3, R” = 0.383, P=O.O5
Millet
Fert ilizer :
Y = 136.0 + 58.6 X- 2.19 X2 + 0.023 X3, R’= 0.518,
P-o.007
Cow dung
Y= 563.0 + 28.6X- 1.57X’+ 0.020X3+2=0.310, P--l10

---

91
Table 7: Crop production index and fertilizer use efficiency (in parentheses ) of the
groundnut-rotation as affected by soil amendment.
Nutrients added
Fertilizer
Stable manure
Cow dung
kg/ha (N+P20+KzO)
OL
1.002 (-)
1.66 (-)’
1.83 (-)
9 0
2.10 (16.1)
2.16 (17.0)
2.17 (17.1)
120
1.89 (9.8)
1.93 (10.2)
1.92 (10.1)
150
1.90 (7.9)
1.84 (7.4)
1.86 (7.6)
180
2.10 (8.0)
2.00 (7.3)
2.23 (9.0)
1. N-o nutrient added as minera1 fertiliser.
2 . Crop production index (1 .OO = 1192 kg/ha groundnut pods + 125 kg/ha millet
grain Yields over the rotation perîod 1997- 1998.
3 . Fertilizer use efficiency, kg produce kg ml of nutrient applied.
Prductivity of the gmmutit-millet ruttztion
The output (pod +grain yields) of the groundnut-millet rotation system (Table 7)
is shown to be enhanced more by the integrated fertilisation strategy than the use of
minera1 fertilizer alone when compared to the control Manure and fertilizer when
applied alone increased output by 75 and 99% respectively. On average, manure +
fertilizer increased output by 102% and at lower levels of applied fertilizer, combined
fertilization was 7% more efficient than fertilizr alone.
A common assumption is that a groundnut-millet rotation system is the
maintenance of soil fertility through nitrogen fixation by the groundnut trop. Analyses
of long-term experiments form Semiarid Francophone Afiica indicated that nutrient
balances are negatîve particularly in nitrogen, are especially negative given the typically
low, less than 10 kg/ha (N + P205 + K20), fertilizer usage. Studies in Bambey,
Senegal (Pieri 1995) using ‘?J concluded that a groundnut trop actually causes
nitrogen impoverishment of the soil under the current condition of farming (i.e. no
return of residue to the field).
B. SoiI conditions
Soil amendments (Table 8) tend to enhance the resource base through a positive
influence on most of the measured soit chemical properties. Resource-base farming
(control) or nutrient balance negative farming in comparison is depleting of soi1 nutrient
reserves as measured by soil chemical properties..
Manure or manure + fertilizer tends to increase soil chemical measures of
nutrients than fertilizer applied alone, especially on soi1 organic matter content (SOM),
exchangeable bases (K and Ca) and total nitrogen. Alter one application, manure and
manure + fertilizer plots showed a SOM content of 13 and 17% respectively, more than
plots that received fertilizer alone. After the second application in year two, SOM

---

92
content resulting fiom manure was higher than that of the fertilizer alone by 22 and 2 6
% respectively.
The wide gap in SOM content. between fertilizer alone plots and manure +
fertilizer plots was a result of higher annual SOM loss (k%) f?om the fertilizer alone
plots. Soil organic matter loss cari be due to both mineralization and erosion. Fertilizer
was applied as nitrogen in year two and nitrogen fertilizers are known to hasten SOM
mineralization through increased microbial activity.
The loss of soil productivity is ofien related to loss in organic matter. Soil
analysis f?o:m long-term experiments in the West Afiican Sahel has indicated that decline
in soil productivity cari be ascr&ed to two processes: the decline in soil organic matter
and soil acidification (Pieri, 1995.) For example, in the very sandy sols of Bambey
(Senegal), ammal plowing combined with fertilizer leads to an ammal k% of 5% or more.
Also, on b&ter-textured soils, (silt + clay > 101, of Saria (Burkina Faso) and BebedJe
(Chad), the higher the N fertiliser used the higher the k%. The lower k% for manure +
fertilizer may be as result of addition of organic matter, directly or indirectly acting as
strong check on decomposition of the soil humus.
Benef3s of manure cari be through the physical and physio-chemical effect, and
provision of plant nutrient notably N, P and K and also some micronutrients. Soil organic
matter is increased and this usually leads to improved moisture infiltration and increased
water-hohiing capacity.
Complementary effects of manure and minera1 fertilizers have been confirmed in
the tropics; Richards 1967, Roche 1970, Lamarc 1972 and Ganry et a1 1974 (all cited in
Mokwunye 1980). Higher yields are achieved when the same amount of nutrient is
applied in crombined form than applied as fertilizer alone. This is said to be true in the
long-teIm and when the level of minera1 fertilizer is relatively low.
.-.--
X.*--l-
-IMII-
---

---

93
Table 8: Effèct of Dif&rent !&il Amendments on Measured Soil Cbemical
properties.
PH
Avail. P
Exch. K
Exch. Ca.
OM
k
(water)
PPm
PP*
w*
Yo
Yo
Year l= Grouminut and PhosDhorus
Control
6.2 8
2 5
178
0.92
-
Fertiliser
6.5 75
2 3
162
0.64
-
Stable manure
6.2
15
2 5
123
0.79
-
cow dung
6.1 13
19
129
0.64
-
Stable + Fert.
6.1 14
2 8
168
0.80
-
Cow dung+ Fert. 6.1 23
2 6
1 8 1
0.70
-
Year 2 = MiIlet and Nitrogen
Control
5.8 9
0.68 14
FertiIizer
5.8 5
0.49 12
Stable mure
5.7 5
0.63
10
Cow dung
5.7 7
0 . 5 7
6
Stable + Fert.
5.5 6
0.63
11
Cowdung-t-Fert. 5 . 7 6
0 . 6 0
7
3. conc.lusions
The results of the integrated fertilisation experiment lead to the concbsion that
there is a threat of soil productivity decliue over time under the current cropping
practices, no fertilization or minerai fertilizer alone. In order to maintain trop yields
under continuous cultivation there is the need for additional soil amendments. Fertilizers
are essential under such stressed soil conditions, but are expensive inputs, and tnust be
well managed to maximize efficiency.
Though there was variation in the two years between the two manures,
regardless of the type, manure and fertilizer combined always outperformed the
fertiliir alone.
The exchtsive use of fertilizer on tiagile ecosystems as the coarse-textured
uplands of The Gambia, may lead to physical land degradation and ultimately to nutrient
fatigue. The sohttion is then an integrated plant nutrient system approach, by combining
minera1 fertilizers with organic ones such as manure. The addition of manure cari
improve the properties of these fragile soils as soit organic matter content for sustainabie
agricultural productivity.
The integrated fertilization strategy employed in this study has increased the
productivity of groundnut-millet rotation by increasing fertilizer use efficiency over
mine., A ftxtilkr especially a~ low rates of applied ft, Gxr, thus reduciug y1 uduc;Gon
cost and increasing profit.
--.--II--
--

-

---

94
Recommendations
There is need for the study to continue over the long-ter-m in order to identifl key
indicators of soil fertility decline relating to soil organic matter content. This may help us
to understand the soil causes of land degradation and establisb critical vames of soil
organic matter for different agroecologies.
For a large-scale adoption of mamue use on field crops, there is the need to
facilitate the transportation of mamue fiom the source to field as this is the principal
constraint especially for the fields far away frorn dwellings.
TO increase the value of manure, proper field storage methods need to be
recommended. Manure is transportad to the field before the rains and at this time cari
only be surface applied which exposes it to the elements (heat, v&~ and rain) of tbe
weather.
References
1.
Central Statistic Department. 1993. Population and Housing Census, Banjul,
The Gambia.
2.
Dunsmore, J. R, A. B., Moffet, 1). J., Andersen, I.P., Williams, J.B., 1976.
Agricultural Development of The Gambia: An Agricultural, Environmental and
Socioeconomic Analysis. Land Resource Study, No. 22, vol.: 2.
3.
FAO year book, Fertilizer, vol. 46. 1996.
4.
Jallow, B. P. 1989.
Rainfall analysis for simple climate application in
agricultural planning.
Department of Water Resources (DWR) Technical
Paper.
5- .
Mokwunye, U. 1980. Interactions between farmyard manure and fertilizer in
Savamta soils. FAO Soils Bull. No. 43 : ‘192-200.
6.
Pie* C. 1995. Long-Term Soil management Experiments in Semiarid
Francophone Afiica. IK RLal. and 8. A. Stewart (ed.) Advances in Soi1
Science:Soil Management: Experiment basis for sustainability and
environmental quality.
7.
Posner, J. L. and Jallow, T. 1989. District Agricultural Profile of Central
Badibou, North Bar& Division. Gambia Agricultural Research Working Paper
# 2.
8.
Republic of The Gambia. 1993 Livestock Census. Department of
Livestock Services/ITC, Banjul, The Gambia.
9.
Rhodes, E., A. Bationo, E. M. A. Smaling, and C. Visker. 1996. Nutrient
Stocks and Flows in West Af?ican Soils: .rrz: A. Uzo Moh-wunye, A. de Jager
and
E. M. A. Smaling (eds.) Restoring and maintaining the productivity of
West Aftican Soils: Yey to Sustainable Deve! npmcnt.

---

95
Annex 1 : hlividual Analysis of variante of Groundnut Pod yield per Site.
-
-
Source
Degr=
Sums of Squares
o f
o f
Variation
Freedom
Yundum
Njaba lhnda
Jenoi
sspu
----“--------_-----------------------------*---~--------------~------____________1___“_______I__
Repkatiun
3
1220433.625
792740.477 375356.279 3402314.704
mure(M)
2
1325016.346*
113592.655
78072.589
460474.455
Fertiliser (F) 4
17>,;ci.313 2936851.740* 317219.280* 4 6 3 1 3 6 . 5 6 8
MxF
8
2961125.678* 1206037.055
240260.75 1
864467.625
Errer
42
333370.672
8052259.578
868654.830 428692 1.398
c v (%)
20.81
22.39
20.13
24.37
~~_-
---~_----- .---

---

96
Pilustigma reticulatum used for Soil Oganic Mi&ter Build up:
Effkcts on the Soil Quality and Crop Yield in the Peanut
Basin of Senegal
M. Dia& M. Sène and A. Badiane
ISRA, senegal
Abstract
Given the hi& demand for trop residues and the insuf&iency of animal mame,
the use of naLtve vegetation could be an alternative solution for improving soil quahty
and trop productivity. This study was conducted to determine the effect of PiZiostigma
reticulatum (PR), a native shrub, in improving the soil properties of a degraded soil and
to develop management systems that effectively utilize the species to maximize input
efficiency and trop productivity. The experiment was conducted at Nioro, Kaolack, on
a Deck Dior loamy sand (fine sandy, mixed Haplic Ferric Lixisol), leached ferrugeneous
soil tropical soil, where peanut and millet were grown. The above-ground biomass of
PR, collected afier each cutting fiom the field was applied at the soi1 surface, at a rate
of 2 t/ha on the dry matter basis. The study compared a control (TO), PR biomass fiom
weeding (Tl), Tl + minera1 fertihzer (T2), at recommended rate for the growing trop,
‘f2 ,f biomass i?om cutting aRer harvest (T3), PR biomass fi-om clear cutting during soil
preparation (T4), and T4 + minerai fertihzer at recommended rate for the growing trop
(T5). As treatment application was sequential partial results showed slight increases of
21 and 26% for Tl and T2 respectively in peanut plant population, 6 and 17% for T 1
and T2 respectively in pods. As for millet, increases in stalk of 8 and 188% for T 1 and
T2 respective& were recorded. For both crops, these increases in yield components
were more likely due to the chemical fertilizer that is more available to plants so far.
Key words: Pibstigma reticdatum, biomass, soil organic matter, soi1 qualii, trop
productivity.
1. htrdwtion
With 1/3 of the country’s area, the Peanut Basin provides 75% of the peanut and
80% of the millet production of the country. However, whîle the soils are intensively
cuhivated with a high degree of land utilization (land in fallow is less than 3%), trop
production is still low. Soils are degraded.
Degradation of the soil resource, in Senegab has resuhed fiom the combination of
decreases in ralnfall, inappropriate land management practices (less land in shorter
fallow periods, removal of nearly all trop residues fi-om fields), and declining numbers
of trees fiom traditional parkland agroforestry ~systems. Consequently, there has been
an intense extraction of nutrient N and P, and a decrease in overall soit organic matter
(decreased tiom the naturally occurring 2% down to 0.3%). The changes to the soi1
resource h.ave caused a decrease in food production and reduced rural income.
Therefore, efforts must be redoubled to replace nutrients fhat have been loti fi-om the
farms through the use of fertilizers and organic matter. Given the high demand for trop
xsidues ir~d the iüsticiency of ,,,iu~l n13nure, the use ofi,,tive iegetation çouX L,
an alternative sotition for improving soi1 yuahty and thereby increasing trop
productivity.

---

97
fiiiostigma reticulatum, an endophitic legume, has no nodules, does trot fix N,
belongs to the Cesalpiniaceae
fhilyy and is usually a shrub but it cari occasionally be a
tree. It grows on sandy, clayey and lateritic soils. During the dry season, it cari grow up
to 90 cm with a canopy diameter of 100 to 17.5 cm Ammally, 1268 g of dry weight
biomass per shrub are produced, for an average density of 317 shrubs per ha (Diack et
al., 1998). In some areas, these shrubs nearly caver the landscape but in others, there is a
less dense distribution. This difherence in density is due to differences in soi1 types. If lefi
uncut, these shrubs continue to grow, but in fàrmers’ fields they are tut at the soil surface
and burned just prior to the rainy season.
A survey was conducted (Diack et al, 1998) to determine the farmer’s level of
appreciation and the uses of P. reticulatum within the parkland system The survey
showed that P. reticulatum was the third most important species that tir-mers would like
to conserve in the fields after two tree species: Cordjda pinnata and Acacia albida. The
reasons for the choice of these three species were fruit production, nutient value of their
biomass and forage for livestock. For P. reticuiatum, farmers thought that it cari play an
important role in soil fertilization and help protect soil against erosion. The population of
P. reticulatum is still acceptable despite a decrease during the past ten years, due to
mecbanization. The management sytiem of P. reficulatum consists of cutting the shrub
(april-june) fo r soil preparation and cutting 2 to 3 times more, depending on the trop,
during the growing season. Ash as minera1 failizer is the main product that farmers get
f?om the management system. Aceording to fmers, the effect of P. reticulatum on trop
productivity is showed by the good response of crops such as peanut, millet and to some
extend cowpea, under the shrub canopy. These information collected f+om the survey
show that Pihstigma reticulatum plays an important role in the farming system and has
a potential of improving the soil productivity.
A decomposition study of Pilicx@yma retkulatum biomass (Diack et al., 1998)
showed a greater mass loss under field conditions than under controlled conditions. This
was probably due to the role of soil Iàuna whi& may have been involved under field
conditions but would have been excluded under laboratory conditions. With such a fast
decomposition rate under field conditions, accumulating residue biomass at the soil
surfàice would certainly avoid a discontinuous distribution of organic resource for the
microbial activity. The specific surface area-to-mass ratio, obtained for P. reticulatum,
should allow such a continuous application of the biomass to the soi1 for both caver and
organic matter build up.
The objectives of this study are to determine the influence of P. reticulatum on
the soi1 physical, chemical and biological properties and the trop yield.
2. MateriaIs and Me&&s
.sGtJ : The experiment is being carried. out in Paoskoto (Kaolack) in two
farmers’fields which are under a two-year rotation of peanut (Arachis &ogaea)
and millet (Pennisetum glaucum L.). The soil is a Deck Dior loamy sand (fine-
sandy, mixed Haplic Ferric Liisol), leached fem@rous tropical soi1 (probably
an Ultisol). The soil has a pH varying fiom 5.7 to 6.7 and a low fertility status
(OC: 0.47%, N: 0.45%).
mater&:
Plant
The above- ground biomass (lcz:~ and $SKIS) of PiIiostiArilu r.dicuIatunz wert:
collected alter each cutting from the field.
Exnerimental design
.^_-
-----
_II-

---

98
The design was a randomized complete block design with 6 treatments replicated
4 times.
The treatments are the following:
TO = Contro~
Tl = Biomass fiom cutting (weeding) and application between rows and at the soil
surface, at a rate of 2 ton/ha of dry matter;
T2 := Tl + minera1 fertilizer at the recommended rate for the growing trop;
T3 = T2 + biomass f+om cutting afier harvest and application at the soif surface, at a rate
of 2 tonIha of dry matter;
T4 == Biomass f?om clear cutting (soir preparation) and application at the soil dace, at
a rate of 2 ton/ha of dry matter;
T5 == T4 + minera1 fertilizer at the recommended rate for the growing trop.
Peanut, a 73-33 variety, will be planted wîth inter-rows of 0.50 m while millet (Souna
3) Will be planted with inter-rows of 0.90 m, in an experimental unit area of 45 m’.
Measuremen&s
On plants: N, P, K, Ca, Mg and S contents Will be determined at the maximum
vegetative stage and before harvesting for each trop.
On soils: samples Will be collected hefore and aiter each growing season to monitor
soil moisture in relation to soil physical, chemicak biological properties and C
dynamics. The following properties Will be measured:
4
mal characteristics: moisture content, bulk density, infiltration rate, soif
resistance to penetration and seahng index as a measure of aggregate stabiliry;
W
amical characteristics: total C and N, NO;, NI&+, P205, K’, Ca2+ and Mg”+ ;
cl
Biological characteristics: microbial biomass C, particulate organic C (POC)
and enzyme activity (j3glucosidase).
3. Partial ResuU.s md Dkussioc
Since the application of Piliostigma reticulatum biomass is sequential, treatment
T3 has been applied right aRer harvest while T4 and T5 will be apphed early next season.
Therefore, the yield components presented in Tables 1 and 2 are only for treatments TO,

---

99
Tl and T2. The yield components recorded for T3, T4 and TS treatments could be
considered as control for the Iirst year.
For peanut (Table l), even though yields were relatively low, the P. reticula&n
application has resulted in a slight increase in yield components. For plant population at
harvest, there is an increase of 21 and 26% for ‘treatments Tl and T2, respective& For
pods, there has been an increase in yield of 6 and 17% for Tl and T2, respectively
whereas 6 and 29% were the respective increases in peanut hay yield for Tl and T2.
As for millet, yields in general were low (Table 2). However, while treatment Tl
has not yet shown any increase, T2 in the other hand has resulted in a strong yield
responsee. Plant population has increased fi-om 1 to 6% and stalk fiom 8 to 188% for Tl
and T2, respectively. For both crops, increases in yield componenrs were more likely due
to the chemical fertilizer that was added to A? reticzdatum biomass for its greater
availability.
Perspectives
Since the experimentation is a 4-year program we need to complete first the
treatment applications and then monitor the soil property changes over time.
Table 1. Yield components of the peatmt as alZected by Piliustigma reticuiatum
-~
Treattnent
Plant poputatim
Pod+Hay
Hay
POd
-
-
u@w
(kg/ha)
(E;g/ha)
TO (Chtrol)
44480
1250
600
650
-~
- T 1 (2 tons
- biomass)
54060
1320
640
690
T2 (2 tons biomass +
55980
1530
770
760
ntieral fcdilizers)
-
-
-
Mean
-
53300
1460
720
750
Table 2. Yield components of the millet as affected by Piliastigma reticdatum
Plant popdation
NUllibW
!Mk ww
Grain &Id=)
of
spikes
lba
8240
11140
1240
210
8320
8170
1340
150
8700
27170
3560
660
8380
18650
2040
410
Literatme Cited
Diack M., 13adiane A.N., Sène M., Diatta, M. Dick R, 1998. Cordyla Pinnata en
association avec Piliostigma Reticulatum: Impact sur la régenération des sols dégradés
au Sénégal. Rapport ISRWNRI3AR LIlU02.-USAID, 25 p.

---

Phosphogypsum Efficiency to correct Soil P Deficiency and/or
Soil Acidity
M. Sène, M. Diack and A Badiane
ISRA, Sénégal
Abstract
TO Jircrease agricuhural production in Senegal, the national program uses a
phosphogypsun (PG) and Taiba phosphate rock (PR) mix as a soil P amendment to
correct soil P deficiency and/or acidity. However, there is no previous study
demonstrating the efficiency of PG to corrç
:hese soil constraints. This long ter-m
experiment started in 1997 at Nioro agrucultural research station to study the
efficiency of PG as compared with PR and lime to increase trop yields for the
corn/peanut rotation and improve soil P and pH status. For the degraded soil fertility
site selected, a randomized complete block design composed of 8 treatments and 4
replications is used.
In 1997, no significant direct effect was observed on cor-n yield components. This cari
be explained partly by rainfall shortage. The analysis of soil samples taken aRer harvest
has shown a important Ca movement within the soil profile. In 1998, the residual effect
of treatments is significant only on pod yield. The positive effect is more important for
the lime treatment. As compared to the çontrol, ail the treatments (except the 75 % PR
and 25 % PG comhination) signifïcantly improved the pod yield. This confirmed the
positive effect of Ca application on the pod filling. The foliar analysis has not shown
any significant treatment effect, despite the plant chlorosis observed.
1. htroduction
The production of inorganic fertilizer phosphate f?om the local industrial produces
phosphoric acid and the natural rock phosphate deposits resuh in a huge amount of
phosphogypsum by-product piled up near the fàctories. This materials for which
chemical characteristics are given (see annex ) is now being used on a large scale
(nationwide) for the national program aiming at increasing the agricultural production.
The newly processed P-source amendment material bags distributed to farmers consists
of the mix of of 50 % rock phosphate (RP) and 50 % phosphogypsum. There is,
however, no previous study in Senegal that documents the efficiency of PG to correct
soil P deficiency for trop uses, or to reduce soil acidity given the large Ca content.
The objective of this experiment was to study the efficiency of PG as compared with
rock phosphate and lime.
2 .Materials and Methods
Site selecticwz
This long ‘ter-m experiment, started in 1997, was located in the NIORO ISRA
agronomie research station in a ferruginous, leached soil. For the put-pose of this study,
a highly chemically degraded soil site (pH < 5.5) but with available P < 30 ppm was
selected. This had been enabled by the soil fertihty assessment undertaken throughout
the r search. station three ye-
UL3_ ag, (Agetip, 1995). T!ic 2C m by 20 m grîd stinnplii~g
used allowed a spatial variability analysis of the various plots within the station.
Cropping system
The Nioro area receives relatively reliable rainfall. Peanut (variety 73-33) is oRen
-----.---“--‘IIIyu
..--------
-
----UI----

---

101
grown as a cash trop as is cor-n (variety Synthetic C) which is very sensitive to soil
fmility. The corn/peanut rotation was started in 1997 with cor-n. The potential yield of
the region is 4 t/ha for corn, and 25 t/ha for peanut.
TreatientY
Treatments under comparison consisted of the combined use of phosphogypsum and
rock phosphate as indicated (Table 1) ; the lime treatment acted as a reference for soi1
acidity control
75%Pf?omRP+25%PfromPG
lOO%PtiomRP+O%Pf?omPG
100 % Ca fiom lime (CaO)
The rate of application was different for the two P sources : 100 % P from RP
corresponds with an application rate of 400 kg/ha of RP, and 100 % P from PG defines
an application rate of 700 kg/ha of PG. For treatment 8, 100 % Ca fiom Ca0 refers to
a lime application of 400 kg/ha. From the chemical analysis data of these fertilizersY
the P and/or Ca quantity added on each plot cari be determined. These rates Will be
applied once every 4 years after 2 complete rotation.
Experimental design
The experimental design was a randomized complete block design with 8 treatments
and 4 repetitions. The size for each of the 32 plots is 84 m2 (15 m x 5.6 m). The
number of rows to be sown each year depend on the trop : 7 for cor-n sown at a
spacing of 80 cm and 11 for peanut sown at a spacing of 50 cm
A4eawremenfs
a) SO~LS
AfIer the initial soil physical and chemical characterization, as described above, soil
samples are taken once a year alter harvest to monitor soil pH, P and Ca contents
within the profile. For the first sampling performed in December 1997 alter the com
harvest, 108 samples have been collected according the following scheme :
?
For treatment 1, 3 and 7, ah the plots were sampled at 4 depths (O-10, 1 O-
:20, 20-40, and 40-60 cm The treatments have been chosen to allow a
analysis of Ca movement in the profile ;
?
For the remaining treatments, a11 the plots have been sampled at only 3
depths (O-10, 1 O-20,20-40 cm)
The soil analyses r-un on those samples were : particle size analysis, pH (water and
KCl), carbon, nitrogen (total N and nitrates), bases, CEC, aluminum and sulfùr.
b) Plant
This sampling was not done in 1997 for com.
Peanut plant samples are taken fi-om each of the 32 plots at flowering/pegging stage

---

102
for foliar diagnosis. For each plot, four replicate samples were analyzed for N, P, K,
Ca, and Mg.
FieU oprations
AfIer applying the phosphocalcic amendment, the plots were plowed to mix the
fertilizer in the top 20 cm soil layer. The timing of the difberent operations performed in
1997 for corn and in 1998 for peanut is presented (Table 2).
Nitrogen and Potassium were applied on each plot at the following rates :
12 kg/ha of N was applied at sowing, 22 kg/ha N was applied for both applications to
com ; 12 kgIha of N was applied at sowing for peanut
40 kg/ha of K was applied at sowing for both peanut and corn.
Rainfall conditions
Rainfàll patterns were different for the 2 years. In fact, the rainy season of 1997 was
ended early while rainy season in 1998 was late. The total ammal rainfàll was about the
same for tbe 2 cropping seasons (580 mm). However, while the rainy season started
early June 1997, the first important rain was recorded late July 1998. Although
characterized by a rather short rainy season, the 1998 cropping season had a much
better rainfall distribution. A long drought period occurred early during the 1998
cropping season, causing a severe plant water stress, while in 1998 there were no
major water stress problem, except at the trop maturity phase.
3. Results and Discussion
As mentioned above, the selected site is part of the most degraded block of Nioro
Station. The comparison of the effects of P and/or Ca amendments on corn in 1997
and on peanut in 1998 is shown in table 3.
For the first year of the experîment 6orresponding with corn production, the ANOVA
indicated no significant treatment effect. An important variability has been observed
among treatment. Com grain yields obtained were low compared to the variety yield
potcntial -hi& is around 4 tons/hq Two severe drc-ght p--iodr have occurrcd du?*:g
the cropping season, early in the vegetative phase and late at the maturity phase, which
partly explains the low yield obtained this year ~JI the Peanut Basin.
_-~
.-.--mm
1111
---

---

103
Peanut (1998)
I \\--o---,
13930
2580
1350 d
2670
1480 c
100 %PG (T3) 23610
1312
1110
82760
3990
2800
1590 b
25 -75 (T4)
21670
1240
970
81350
4260
2370
1590 b
50-50 (T5)
17920
1000
740
86510
4530
2900
1630 ab
z “5 (T6)
19760
910
780
75100
3760
2370
1390 d
100 %asRP
17470
990
730
- 86040
4330 12800 / 1540 bc
tT71
Lime alone (T8) 18890
980
750
86350
14380
12720
1 1660a
Mean
19580
1150
880
- 83740
4170
2640
1530
Level of sign
N S
NS NS
- N S
N S
N S
S
CV%
2 5 . 2
39.2
44.2
- 7.9
11.5
16.6
7.8
In 1998, no significant effect is observed except for pod yields. Highest pod yields
were obtained with the treatment where lime was applied and by the treatments where
phosphogysum was added; the best among these was the treatment receiving the tnix
of SO % PG + 50 % RP. This indicates the relative importance of Ca in pod
production. The positive action of Ca to the degraded soil could arise fkom two
aspects. First of ah, there is the improvement of pod filhng, and secondly the soil pH
increase and/or aluminum toxicity decrease could occur. This cari be confirmed by the
soil profile Ca emichment resuhing fi-om the difk-ent P or Ca source amendments.
This hypothesis is supported by the fact that the plant analysis done following the
chlorosis observed at mid-season does not reveal any significant treatment diBerence.
4. Conclusion
Alter a second year study, the trials show the possible effects that cari result fi-om the P
and/or Ca source of soil amendments. The remaining 2 to 4 years of the experiment
should give a better idea on the real agronomie value of phosphogypsum as an
amendment material.
----“--
. ..----“I~
---

---

Annexe: Chemical characteristics of phosohogypswn (PG), PG + Taiba Rock
Phosphate (RP) mix, and Taiba RP
Total Elements
PG
50% PG + 50% RP
Taiba RP
CaO%
32,3
40,6
49,4
MgQ%
0,Ol
0,04
0,06
K20 %
< 0,5
< 0,5
0,02
Na20 %
0,07
08
0,09
Fe& %
0,15
0,49
1,59
Al203 %
0,27
0,77
0,98
P,Os %
0,99
19,12
37,2
S%
14,64
6,79
*
m PPm
2,3
108
404
CU PPm
335
27,7
72
zn P P ”
< 1.0
290
522
Pb PPm
5,3
5,3
5,4
cr PPm
24,2
96,7
198
Ni ppm
23
29,8
86,4
C d ‘ppm
15,8
51,6
70,8
Soluble elements
Samples
C a 0 %
10,48
10,9
Mg0 %
V,Ol
0,03
KZO %
0,44
0,32
Na20 %
0,07
0,07
P*O5 %
0,39
0,47
s %
4,86
4,81
pH eau %
4,67
3,91
----“II-

---

105
Phosphorus Buffer Coefficients of Selected Soils of West
Africa
A. Sidii, M. K&a, O.B. Coumaré, MD. Doumbia, A. Bationo, R k Kablan et R S.
Yost
Abstract
Labcrratory incubations were conducted to validate P bufEer coefficients predicted by
the Phosphorus Decision Support System (PDSS) modeL Soil samples were analyzec’ “T
general soil properties.
Collected soil samples were first analyzed for clay and plant available P. These
parameters were fed to the mode1 to generate buflèr coefficients referred to as h. These soils
samples were then incubated to estimate buffer coefficients referred to as ai. Then, a, and ai
were compared.
Steeper slopes (indicating higher buffer coefficients, but lower buffeting capacity) were
obtained for sandy soils (0.87 for the sandy, Seno soil), in contrast to the less steep slopes for
clayey soils (0.21 for the clayey, Moursi 1 soil). Buffer coefficients derived fi-om the mode1
(a,) were significantly higher than those obtained fiom laboratory incubation (ai). Differential
responses of soils to methods of bufIer coefficient determination were obtained: the buffer
coefficient derived from the mode1 (am).was significantly higher for the clayey Mowsi 1 soi1
(under irrigated rice), but not for the clayey Longorola BF soil (bottom-land, waterlogged
SOil).
Future work on PDSS will focus on (i) field testing of the predictions and (ii) including
phosphate rocks into the predictions. Phosphate rocks will require defining both specific P
buffer coefficients and soil P depletion rates in relation with phosphate rocks.
1. Backgrmnd
It has long been believed that rainfall was the most limiting factor for trop production
in Mali as well as in many regions of West Af?ica. Recently, many research programs have
prov-ided evidence that soil fertility, not ratiall is the number one factor limiting trop
production in these regions.(Stroosnijder, 1981; Doumbia et al., 1998; Breman et al., 1998).
Phosphorus is the most deficient and is the plant growth limiting nutrient in these soils
(Poulain, 1976; Manu et al., 1991; Daumbia et al., 1993). Jones and Wild (1975) documented
that P deficiency could be SO acute tbat plant growth stopped once the seed reserve of P had
been depleted.
Fertilizer recommendations in Mali were made according to Chaminade (1965). For
economical reasons deficiency corrective rates were recommended (PG-i, 1973). For the case
of P, rates of about 20 kg ha- 1 are recommended for most crops in most soils of Mali (Poulain.
1977). When applied, these rates lead to soil mining (Van der Pol, 1992; KieR et al., 1994).
For increased trop production and sustained soil :fertility, there is a need for recommendations
based on fi,xtors such soil properties, yield goal, etc. These site specifïc recommendations are
better handled by decision aids. Phosphorus decision support system (PDSS) offers a strong
potential in making these recommendations. It gives P recommendation on the basis of
primatily soi1 clay and plant available P contents (Yost et al., 1992). However, the mode1 needs
to be adapted to soil and socio-economic conditions of West Ahica. Phosphorus requirement

---

106
predictions are made according to th ‘!: following formula:
P = (Bc - Bo) * al/aZ * DAO, wher’ e
P = P requirement prediction,
Bc - P mitical level in the SO&
BO = P initial level in the soil,
at =:long term retention rate of P
a2 - P buffer CoefIicient,
D =: depth of application.
3.Ob,jectives
The general objective of this research was to test the PDSS model. Tbe specific
objective was to test P-buffer coeffic&nts determined by PDSS.
4, MIatensrls and Methods
Forty five SO~I samples representing major agricuhural regions of Mali and Niger were
selected on the basis of texture class. These samples were analyzed for clay content, Bray- 1 P.
and pH (Spark et al., 1996).
Laboratory incubations were petiormed to determine P buffer coefbcients in the
laboratory. These coefficient were compared to those predicted by PDSS. Known rates of P
solutions were added to SO~I (1: l), and the mixture were allowed to dry at room temperature
during 6-7 days. The slope ofthe regression obtained byplotting bray P1 extractable P against
the amount of P added was considered as P buEer coefficient (ai). This coefficient was
compared with the coefficient (a,) obtained by the model, according to the following formula
(Yost at al., 1992.
a, = 0.6665 * exp(-0.02739
* clay).
5. Results and lllischns
SO& were selected to represegt cropping systems md of Mali and Niger including
rainfed agricuhure, bottom land (waterlogged), and irrigated systems These soil samples were
classified into two textural and two Bray-1 P groups. The textural groups included: (i) the
clays such as soils fiom Kollo, Niger (56.4% &y), Moursi, Mali (43% clay) and Longorola bf,
Mali (35% clay) and (ii) the sands su& as Seno-S, Mali (3.9% clay) and Dounga S. (2% clay).
The Bray- 1 P levels included: (i) soil witb “sufbcient” P such as Kollo and Longorola bf (with
>lO mg P kg-’ ) and (ii) soils very deficient in P such as Moursi and Tanda, Niger (with < 1 mg
P kg”‘). The soils Kollo and Longorola bf had each high contents of both clay and Bray-1 P.
The difberent buffer coefficients are shown in Figures 1 through 7. In general buffer
coefftcients estimated by the mode1 are higher than those determined by incubation. In
addition, sandy soils have higher btier coefficients.
Figure 1 shows buffer coefficients for a sandy soil (Seno), a loamy soi1 (Kita), and a
clayey soil (Moursi 1). The steeper slope (higher buEer coefficient of 0.87) for the sandy Seno
has a lower buffering capacity. In contra& the less steep slope (lower buffer coefficient of
0.21) for the clayey Moursi 1 has a higher btiering capacity. The Moursi 1 is a soil under
irrigated rice cropping system. Similarly, Figure 2 shows the same pattern between the sandy
soil of Dougouba and the clayey soil of Longorola BF. The Longorola BF is a bottorn-land
soi1 (waterlogged for about 3-4 months) under rice production. Figure 3 shows the above

---

107
pattems for soil fiom difherent cropping .systems of Niger.
Figure 5 compares buffer coefficients obtained fiom laboratory incubation (ai) to those
derived fiom the mode1 (a,). In general the mode1 predicted signifrcantly higher coefficients, as
indicated by Duncan test. Figure 6 shows the comparison buffer coefficients a, and ai of 10
samples collected fiom major cropping systems of Mali. The mode1 predicted a significantly
higher coefficient for the clayey Moursi 1 soil (under irrigated rice cropping system), but not
for the clayey Longorola BF soil (a bottom-land soil, waterlogged for about 3-4 months, under
rice production).
6. Conclusions and Perspetkives
For soil containing a wide range of clay content (2 to 56%) steeper slopes (indicating
higher buffer coefficients, but lower btiering capacity) were obtained for sandy soils (0.87 for
the sandy, Seno soil), in contrast to the tess steep slopes for clayey soils (0.21 for the clayey,
Moursi 1 soil).
Buffer coefficients derived fi-om the mode1 (a,).were siguificantly higher than those
obtained fi-om laboratory incubation’ (a;). This would lead to slight under prediction of P
fertilizers requirements by PDSS in addition there were difTerentia1 responses of soils to
metbods of bufher coefficient determination. The buffer coefficient derived from the mode1 (a,,)
was signifïcantly higher for the clayey Moursi 1. soi1 (under irrigated rice), but not for the
clayey Longorola BF soil (bottom-land, waterlogged soil).
In the future, predictions fiom PDSS will be tested in field plots against ‘%onventional”
fertiliser recommendations. These tests will be implemented with sorghum and maize.
One of the limitations of PDSS is that it does not make recommendations using
phosphate rocks PR). Attempts will be made to include phosphate rocks, especially Tilemsi
PR into the recommendations. Phosphate rocks will require defining both speci& P bu&er
coefficients and soil P depletion rates in relation with the use of PR’s.

---

108
PBuffer 1
?? ??????? ????????
0
Kita
0
Seno
100
1!50
200
250
300
Added P (mghil)
Figure 7. Buffer coeffkient of selected sols (Incubation met/-@)
- --.
--II---
--
-.--
--(M

---

109
50
40
30
20
?? LongordaBF
10
0
Tierouala
v Cinzana
added P (ct/ml)
Figure 2. Buffer coefficient of setected soils (Incubation method)
-
-,-.--
-,
---uL-
--

---

110
Pbuffer3

---

111
PBuffer4
4.8% clay
56.4 clay
li
t
46.8% cAay
0 Kollo P I
0
LongorolaBF
v MoursiGl
v Dougouba
0
60
1 0 0
150
200
250
300
added P(pg/ml)
Figure 4. Buffer coefficient of selected soils (Incubation method)

---

112
PBuffer5
a
0.6
l
0.4
0.2
l
0.0
Méthode de détermination du pouvoir tampon
Figure 5. Comparaison de méthodes de defefminafhn de coefficients de pouvoir tampon en F
(a’ and ‘6’ indiquent les classes du fest de Duncan)S

---

113
PBuffet-6
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
I’
0.0 l--
0
1
2
3 4 5 6
7
8
Numéros des sots
Figure 6. Com~atison de méthodes de determination de coefficients
de pouvoir tampon en P de queicyes sols du Mali..
---

---

PBuffer 7
0.8
0
f
0.6
2?
-LP0cl.-8 0.4
œ PDSS
I
Incubation
-
1
26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46
Selr*ted soils
:ie
Figure 7. P Buffer coeffic Nnt of selected soils of Niger

---

115
Literature Cited
BREMAN,, H., A. COULJBALY, G.H. DIJKSTERHUIS, and P.W. J. UITHOL. 1998.
Approche systémique, acquis de la modelisation et risques de la fertilisation lies a la
phviometrie. pp. 285-302. In Breman H. and K Sissoko (eds). L’itensification agricole au
Sahel. Editions Karthala. The Nethe&ands
CHAMINADE, R 1965. Recherche sur la f&é et la fèrtil&tion des sols tropicaux:
Principes de base et tecniques. Agronomie Tropicale 20:1014-1017.
DOUMBlA M.D., HOSSNER L.R, and ONKEN A.B., 1993. Variable sorghum growth
in acid soils of subhumid West Aliica. Arid Soil Research and Rehabilitation 7:335-346.
DOUMBIA M.D., HOSSNER L. R, and ONKEN A.B., 199. Sorghum growth in acid
soils of West Af%ca. Variations in soil chemical properties. Arid Soi1 Research and
Rehabilitation 12:179-190.
JONES M. and WILD A., 1975. Soi& of the West Afï-ican Savanna: the maintenance and
improvement of their fertility. Techn.. Comm 55. Commonwealth Agric. Bureaux.
Famham Royal, U.K.
KIEFT H., KEITA N. et VAN DER HEIDE A., 1994. Engrais fertiles? Vers une fertilité
durable des terres agricoles au Mali. 99 p. ETC, Leusden, The Netherlands.
MANU, A.., A. BATIONO, and S.C. GEIGER 1991. Fertility status of selected
millet producing soils of West Ahica’with ernphasis on phosphorus. Soil Science
152~315-320.
PIERI, C. 1973. La firmure des céréales de culture sèche en République du Mali.
Agronomie Tropicale 2875 l-766.
POULAIN J.F., 1976. Amélioration de la fertilité des sols agricoles du Mali Bilan de
treize années de travaux (1962- 1974). Agronomie Tropicale 3 1:402-4 16.
SPARKS, D.L., A.L. PAGE, P.A. HELMKE, RH. LOEPPERT, P.N. SOLTANPOUR.
M.A. TABATABAI, CT. JOHNSTON, and M.E. SUMNER. 1996. Methods of soil
analysis. Part 3 - Chemical methods. pp. 1390.
STR.OOSNHDER
L., 1981. A soil water balance study of the Malian Sahel. Agronomy
Abstracts. 1981. p. 144.
VAN DER POL F., 1992. Soil mining. An unseen contributor to farm income in southern
Mali. Bulletin 325. Royal Tropical lnstitute. Amsterdam The Netherlands.
YOST, RS., A.B. ONKEN, F. COX, and S. REID. 1992. The diagnosis of phosphorus
deiicdency and predicting phosphorus requirement. in proceedings ot the TropSoils
Phosphorus Decision Support System Workshop. pp. l-20. March 11 -12. Texas A&M
University. College Station, TX.

---

116
Nutrient Balances Under Contrasting Millet Cropphg
Systems
M. D. Doumbia’, A. Sidiié, M. Koné, B. S. Coulibaly, N. Coulibaly, A COI.&~&, R A
Kablan ‘, and R S. Yost 2
1.
Wnditut dEconomir Rude, Bamako, Mali (IER)
2. Uniwmily of Hawaii, Honolulu, USA.
Abstract
On-farm research was condqcted to estimate nutrient budgets under two cropping
systems of Mali These cropping systems included the sandy soils of Central Mali under
permanent miUet cropping system and the cotton cropping system of Southern Mali.
One tirm was selected in one village of each of the selected areas. Dougouba, near
Segou, was selected as the village for low input, millet cropping system while Dampela, near
Koutiala, represented the %igh’ input, cotton based cropping system Partial nutrient balances
were cond.ucted in each farm. In eac$ of these cases, soil samples were collected to estimate
the initial nutrient status. Additions of nutrients were estimated by measuring application rates
of soil amendments and performing laboratory analysis on samples of applied soil amendments.
Nutrient experts were estimated by measuring biomass production and performing laboratory
analysis on samples of the produced biomass.
Nutrient balances were negat&e under each cropping system: Dampéla: -21 kg N ha.‘.
-3,2 kg P ha-‘-9,l kg K ha-‘; Dougouba: -105.8 kg N ha-‘, -17 kg kg P ha-‘, -54 kg K ha-‘. The
very low inherent fertility of sandy soils of Dougouba coupled to permanent millet cropping
without any addition of minera1 fertilizer is leading to a strong nutrient mining.
An implication of this study tiould be an integrated soil fertility management progmm
aimed at soil fertility restoration and iricreasing farmers’ income.
1. Backgrmnd
Several research programs have provided evidence t.hat soil fertility is indeed
the number one factor limiting trop production in Central and Southem Mali and West
Africa in general.(Stroosnijder, 1981; Doumbia et al., 1998; Breman et al., 1998). In
fact, several constraints lead to soil fertility decline under cropping systems in Mali:
?
The low inherent fertility and fiagility of Malian soils (Poulain, 1976; Piéri, 1989; Wilting
and Hossner, 1989; Traoré, 1974; Doumbia et al., 1993 and 1998),
*
Application of very low rates of minera1 and organic sources of plant nutrients (Pién,
1989; Berckmoes et al., 1989; Van der Pol, 1992; KieR et al., 1994),
0
continuous removal of nutrients under traditional cropping systems (Van der Pol, 1992 ??
e
cmstraints related to t5.e nse of the local, Tilemsi phosphate rock, PNT, (Kamara et al.:
1994),
?
dïsappearance or shortening of the fallow period (Hoefsloot et al., 1993),
?
lack of trop residues management in traditional systems (Van der Pol, 1992),

---

117
?
cuhivation of marginal lands (Van der Pol et Giraudy, 1993),
?
minimum or lack of SO~I conservation practices in traditional systems (KieR et al., 1994).
?
the need for Iirewood (KiefI et al., 1994)
?
traditional pastoral systems (Van der Pol 1992; KIeft et al., 1994)
e
socio-economic conditions not favoring the use of chemical fertilizers (Piéri, 1989; Kieft et
al., 1994).
Nutrients balances have been estimated for cropping systems of Mali as well as other
countries of West Afi-ica. (Ro~se, 1981; Pi&, 1983; Duivenbooden, 1990; Veldkamp et al.,
199 1; Van der Pol, 1992; Smaling et al., 1993 et 1996; Stoorvogel et al., 1993; Cretenet et al.,
1994; Traoré, 1995). These nutrient balances are negative for N (-40 kg ha- 1 ), P (-2 kg ha- l ),
K (-33 kg ha-*), Ca (-8 kg ha-l), Mg (-10 kg ha-l), and CaC03 (-16 kg ha-l) for the cotton
cropping systems of Southem Mali (Van der Pol, 1992). A negative balance in N (-8 kg ha-I),
P (-1 kg ha-l), et K (-7 \\g ha-l) was made for agricultural soils of Mali. Tbis nutrient balance
w-il1 be more negative in N (-11 kg ha-l), P (-2 kg ha-l), et K (-10 kg ha-*) in year 2000
(Stoorvogel et al., 1993).
2. Objectives
The general objective of this study is to improve food security and environmental
rehabilitation in sandy areas under permanent millet cropping systems.
The specific objective was to estimate partial nutrient balances under two contrasting
cropping systems (on-farm) of Mali.
3. Materials and Methods
a. Sites Selection
Dougouba, near Segou, was selected as ,the village for ‘low” input, millet cropping
system whiIe Dampela, near Koutiala, represented the ‘%igh” input, cotton based cropping
system. These villages were selected on the basis of their involvement in on-going IER
(Institut D’ESconomie Rurale) activities. Selected characteristics of these villages are given in
Table 1.
m---

---

118
Tableau 1: Selected characteristics of the villages of Dougouba and Dampela.
Dampela
4 4
Population
386
642
Farms
32
4 4
Average fàrm size (ha)
8
1 8
Area under cronning. (ha)
261
792
Area under fallow (ha)
0
0
Communal pasture land (1- ’
1 2
6 4
CoIlaborating NGO’s or extension services
None
cotton company
b. Partial Nutrient Balances
One farmer was selected in each village. Mr. Lassine Djiré was selected from
Dougouba and Mr. Bougoussama bembélé fi-om Dampéla. Selected characteristics of these
farms are g&n in Table 2.
A very simple approach, easy to explain to farmers, was adopted. Only N, P, and K
nutrient additions (minera1 and organic sources) and exports of nutrients (biomass of crops)
were considered and estimated. Any other addition/export of nutrient was ignored.
Nutrient additions were estir$ated by first measuring applied quantities. Then samples
were collected to determine their composition in N, P, and K (Sparks et al. 1996). Similarly,
nutrient e-ports were estimated by first measuring the biomass produced (grain plus straw).
Then samples were collected to determine their composition in N, P, and K..
Partial nutrient balances werei made according to Van der Pol (1992), Van der Pol and
Traoré (1993). However, as stated earlier the following simple formula was use: B = A - E;
where B = partial nutrient balance, A = nutrient additions, and E = nutrient exports.

---

119
Table 2: Selected characteristics of the farts of Mr. Djiré (Dougouba) and Mi. Dembélé
(Dampela).
Characteristics
Mi. L. Djiré
Mr- B. Dembélé
-
-
(Dougouba)
(Dampela)
-
Farmer class
-
-
C
A
Land ownership
Owns the land
Ownstheland
-
Familysize
16
2 7
-
2 3
- -
3
2
-
loamy: 25
clayey: 3
bottom-Iand: 4
--
Crop rotation
Millet-millet
cotton-maize
-
-
-
Crops: No. of fields
Millet../cowpea: 2
cotton: 3
Groundnut:2
Rice: 7
Fonio: 1
Sorghum: 3
vouiuldzou: 1
Millet: 9
-
-
_ Cassava:l
Maize/cowpea: 1
-
-
Ca#le
-
-
0
33
-
-
-
4
1 5
-
-
Goats S&E-..-
2
4
Donkeylhorse
2
1
-
Chicken
7
3 0
Cart
-
-
-
-
-
Plow
Weeder
-~
-
-
Other inputs (purchased)
* Cco is a Cotton fertilizer blend (14-22-l 2-SS-1.5B), Cte is a cereal fertilizer blend
(15-15-15)

---

120
4. Re&ts and Discussion
Selected characteristics of the villages and farms were given in Tables 1 and 2.
Dougouba receives a mean annual rairtf%ll of 600 mm (Sivakumar et al., 1984). This village is
located in the ‘Albida Parkland” (Ayacia albida), characterized by deep sandy soils. Dampela
receives a mean ammaf rainfàll of 800 mm (Sivakumar et al., 1984). This village is located in
the ‘Qld Cotton Belt”, where cotton cultivation is the oldest in the country. in addition,
Dampela has a strong comparative adsantage over Dougouba: the cotton company provides
on a ban basis, ail the inputs required.
Additions and exports of nutrients (N, P et K) to/fiom the f&rms Mr. Dembélé
(Dampela) and Mr. Djiré (Dougouba) are indicated in the tables, 3 through 8.
Table 3. Mr. Dembélé (Dampela) farm: nutrient additions by minera] and organic sources.
Additions
N
(Crop: fertilizer)
Millet Bl: Urea
23
Millet B5: Urea
2 0 . 7
Cotton B6: Manure
77.9
10.8
58.1
cotton B6: cc0
2 1
3 3
1 8
Cotton B6: Urea
23
-
-
Maize B7: Urea
23
Maize B7: Cte
1 5
6.6
12.5
cotton B8: cc0
21
33
1 8
Cotton B8: Urea
23
-
-
Sorghum B9: Urea
23
Sorghum B9: Cte
7.5
3.3
6.2
Sorghum B12: Urea 46
SorahumB12: CCe
7.5
3.3
6.2
Sorghum B13: Urea 23
Sorghum B13: CCe
7.5
3.3
6.2
Cowpea B15: Urea 4.6
VoandzouB16: Urea 2.3
Sorghum B23: Urea 46
Sorghum B23: CCe
1 5
6.6
12.5
TO’TAL
734
169.1
2 9 7 . 2

---

1 2 1
Table 4. Mr. Djiré (Dougouba) farm: &rient additions by organic sources.
Crop
N
P
K
(Crop: fer-r)
0%)
ck)
23
IHillet Dl: Manure
26.0
3 . 6
19.4
-
-
MiIlet D3: Mamue
3 4 . 6
.4.K
2 5 . 8
-
-
Millet D5: Manure
8 . 7
1.2
6.5
-
-
TOTAL
6 9 . 3
9.6
5 1 . 7
Table 5. Mr. Dembélé (Dampela) farm: nutrient exports by crops.
(Crop: fertilizer)
1 (kg)
cottm BS
121.9
BSorghum 9
123.7
--j-s--
Grou.ndIlut B10
1.4
0.1
0.5
Rise
7 . 7
1.7
9 . 0
-
-
Sorghum B12
4 0 . 0
6.8
8 . 4
Sorghum B13
1.5
3.1
3 . 9
Groundnut B14
5.1
0.4
2 . 0
CowDea B15
0 . 6
0.2
1.0
Voanclmu B16
1 . 5
-
-
Rice B17
9 . 9
-
-
Rice B18
8.6
Rice B19
17.8
Rice B20
12.3
2.7
14.4
-~
Rice B21
-~
9.9
2.2
11.6
Rice B22
3 6 . 6
8.1
43.2
-
-
Sorgbum B23
251.1
-12.5
52.4
TOTAL
1 1407.2
272.7
1 589.0

---

122
Table 6. Mr. Djiré (Dougouba) fàrm: nutrient exports by crops.
~
Crop
N
P
(Crop: Code)
tw
1
MiIlet Dl
120
2 6 . 4
5 7 . 6
Cowpea Dl
12
3.6
2 7 . 8
3 8 . 4
7 6 . 8
4 . 8
3 8 . 4
19.2
7 3 . 6
voandzou D7
I 8 . 2
3 8 . 8
TOTAL
1 704
1111.3
3 7 5 . 4
Table 7. Mr. Dembélé (Dampela) Farm: partial nutrient balance.
--
Addition/Export 1 N
Additions
734
169.1
-
-
2 9 7 . 2
Experts
1407.2
272.7
5 8 9 . 0
Balance
-673.2
~
-103.6
- 2 9 1 . 8
Table 8. Mr. Djiré (Dougouba) Farm: partial nutrient balance.
1 AdditionlExpwt 1 N
--
@a)
1
Additions
69.3
9 . 6
5 1 . 7
--
Experts
704.0
111.3
3 7 5 . 4
Balance
- 634.9
- 101.7
- 323.7
Tables 7 and 8 show that nutrient balances are negative for N, P and K, at bot.h
locations, Dampéla and Dougouba. In addition, tbese nutrient balances indicate soi1 mining in
the following rates:
Dampéla:
-21 kg N/ha
Dougouba
- 105.8 kg N/ha
-3.2 kg P/ha
-17 kg P/ha
-9.1 kg Wha
-54 kg Kfha
By comparison, Van der Pol (1992) and Van der Pol et Traoré (1993) reported the foflowing
mean values for nutrient balances for cotton based cropping systems of Southern Mali:
- 25 kg Nlha
0 kg P/ha
- 20 kg Klha

---

1 2 3
5. Condnsions and Perspectives
a. Conclusions
Complete nutrient balances are difbcuh to measure. The nutrient balances were
negative under each cropping system: Dampéla: -21 kg N ha-‘, -3,2 kg P ha-r-9,1 kg K ha-‘;
Dougouba: -105.8 kg N ha-‘, -17 kg kg P ha-‘, -54 kg K ha-‘. One of the main constraints in
Dampela is the lack of arable land, thus resulting in a higher input cropping system, especially
using chemical fertihzer on cereal crops despite very low ratios of product prices to input
prices (Kelly et al-, 1998). This high input cropping system is supported by the cotton
company, by making bath crédit and inputs available to f-s of Dampela. Eve9 with these
‘high” inputs the soils are being mined. Nevertheless, the lower inherent soil fertility at
Dougouba and permanent millet cropping with application of Iower rate of nutrient input
resulted in an even more negative nutrient balance.
b. Perspectives
One of the objectives of this research was to develop an integrated soil fertihty
management program aimed at improving the negative nutrient balances. The foïlowing
activities are suggested to reduce the negative nutrient balance at each site.
0
Addition of Tilemsi phosphate rock during composting. This would not only improve the P
balance, but also alleviate several of the constraints reported by Kamara et al. (1994).
0
Phosphorus recapitalization (usirig Tilemsi phosphate rock) to alleviate P deficiencies and
improve P balance of these SO& (Poulain, 1976; Piéri, 1989; Doumbia et al., 1998;
Sanchez et Izac, 1995).
?
Early planting to capture N flushes at the on-set of the rainy season (Birch, 1964).
?
More integration of legumes in the cropping system.
In addition, at Dampela, there is a need to increase the efficiency of use of chemical
fertihzers, especia& when applied tu cereal. An efficiency of 20 kg mihet/sorghum grains per
kg offertihzer is required to achieve a valuefcost ratio of 4 (Kelly et al., 1998).
Literatwe Cited
BERCKMOES W. E., JAGER et KONE Y., 1988. L’intensi6cation agricole au Mali-sud.
Souhait ou
réalité. University of Arkansas, Fayetteville, FSRE Symposium DRSPR
Sikasso, Mali/KIT Amsterdam The Netherlands
BIRCH, ELF. 1964. Mineralization
of plant nitrogen fohowing alternate wet and dry
conditions. Plant and soil20:43-49.
BISHOP J.. and ALLEN. J., 1989. The on-site cost of soil erosion in Mali. The World Bank
Policy and Research Staff.Environment Working Paper 21. pp. 71.
CRETENET M., DU&AU D., TRAORE B. et BALLO D., 199-l. Ftl&t& et fertilisation
dans Ia région sud du Mali: du diagnostic au pronostic. Agricuhure et Développement~ 3:4- 12.

---

124
DAY J.C. and AILLERY P., 1989. Economie effects of soil and water management
ethnologies: Preliminary results Erom a case study analysis in Mali. In ICRISAT. 1989. Soil,
trop, and water management systems for rain&d agriculture in the Sudano-Sahelian zone:
proceedings
of an international workshop, 11-16, Jan. 1987. p. 363-371. ICRISAT Sahelian
Center, Niamey, Niger. Patancheru, A.P. 502324, India: ICRISAT.
DOUMBIA M.D., COULIBALY A., KOUYATE Z., SIDlBE A,, ET SMTI’H F. J. 1998.
Baseline study of socio-economic conditions and decision making processes with a focus on
small farmexs, nutrient management and millet production in the Segou region of Mali.
LaboSEP, IER 11 pp.
DOUMBIA M.D., HOSSNER L.R, and ONKEN A.B., 1998. Sorghum growth in acid soils
of West Af?ica: variations in soit chemical properties. Arid Soil Research and Rehabihtation
12:179-190.
DOUMBIA. M-D., HOSSNER LR, and ONKEN A.B., 1993. Variable sorghum growth in
acid soils of subhumid West Atica. &id Soi1 Research and Rehabihtation 7:335-346.
DUIVENBOODEN N. VAN., 1990. Sustainability of West African cropping system in tenus
of nutrient elements. A study for the 5th region of Mali. CABO, Wageningen, The
Netherlands.
GIGOU J., COULIBALY L, WENNINCK B, et TRAORE KB. 1997. Aménagements des
champs pour la culture en coufbes de niveaux au sud du Mali. Agriculture et Développement
14: 47-57.
GUIBERT, H 1997. Appui 8 la réalisation d’une étude au Mali-sud sur les stratégies des
agriculteurs en matière d’installation des cultures et de gestion des intrants. Rapport de
mission au Mali. CIRAD-CA. Montpellier. France. 101 pp.
HOEFSLOOT B., VAN DER POL F. et ROELEVELD L., 1993. Jachères améliorées.
Opti.ons pour le développement des systèmes de production rurale au Mali. Vol 2. lER/KlT.
Bamako/Amesterdam The N etherlands.
JOLY C., 1989. Réponse des cultures aux engrais et influence des facteurs agroécologiques
sur les rendements dans la zone soudano-sahélienne. In ICRISAT. 1989. Soil, trop, and water
management systems for raimèd ag&&ure in the Sudano-Sahelian zone: proceedings of an
international workshop, 1 1 - 16, Jan. 1987. p. 201-2 13. ICRISAT Sahelian Cerner, Niamey.
Niger. Patancheru, A.P. 502324, India: ICRISAT.
KAMARA .A, RITA J., DEFOER T. et FOMBA B., 1994. L’utilisation du PNT. Une analyse
des problèmes et des possibilités d‘adoption. Document 94/14 ESPGRN - Sikasso.
KELLY V.:, REARDO T., YANGGEN D., ET NASEEM A. 1998. Fertilizer in Sub-Saharan
Airica: Breaking the vicious circle of high prices and low demand. Policy Synthesis 32.
!?sATP. Washington D.C.: USA.
KlEET H., KElTA N. et VAN DER HEIDE A., 1994. Engrais fertiles? Vers une feailité
durable des
terres agricoles au Mali. 99 p. ETC, Leusden, The Netherlands.
-
-
-llll-Il--

---

1 2 5
KOUNKANDJI B., KANTE S., KEÎTA B., VLOT J.E., DIARRA Z., DIONI L., DEFOER
T., DIARRA S., KAMARA
A., COULIBALY L., BENGALY A. and VELDKAMP W.J.,
1995. Méthodologie de l’étude toposéquentielle au Mali-Sud. Document LaboSEP Serie TO 3.
LaboSEP Sotuba.
NEDELEC G., 1994. Problématique de la production agricole en zone CMDT. Présentation
suctinte. CMDT. DTDR/LRD.
ONKEN A.B. and WENDT C. W., 1989 Soil fertility management and water relationships. ln
ICRISAT. 1989. Soil, trop, and watel management systems for rainfed agriculture in the
Sudano-Sahelian zone: proceedings of an international workshop, 1 l- I6, Jan. 1987. p. 99- 116.
ICRISAT Sahelian Cerner, Niamey, Niger. Patancheru, A.P. 502324, India: ICRISAT.
PAYNE W. A.: LASCANO RJ., HOlSSNER L.R, WENDT C.W. and ONKEN A-B., 1991.
Pearl millet growth as inlluenced by phosphorus and water. Agron. J. 83:942-948.
PIERI C., 1983. Nutrient balances in rainfed farming systems in arid and semiarid regions.
Proceedings of the 17th Colloquium of the International Potash Institute. Bem, Germany.
1983:181-209.
PIERI C., 1989. Fertilite des terres ‘de Savannes. Bilan de trente ans de recherche et de
développement agricoles au sud du Sahara. Ministere de la Cooperation et CIRAD-IRAT.
France.
POULAIN J.F., 1976. Amélioration de la fertilité des sols agricoles du Mali. Bilan de treize
années de tra,vaux ( 1962- 1974). Agronomie Tropicale 3 1:402-4 16.
ROOSE E.J., 1981. Dynamique actuelle des sols ferralitiques et ferrugineux tropicaux
d’Afrique occidentale. Paris, ORSTOM. 569 p. Travaux rt documents ORSTOM No. 130.
SANCHEZ P.A. and IZAC A. M., 1995. Soil fertility recapitalization in Afiica: A concept
paper. ICRAF, Nairobi, Kenya.
SIVAKUMAR, M.V.K., KONATE M., ET VIRMANI, S.M. 1984. Agroclimatology de
l’Afrique de l’Ouest: le Mali. Bulletin d’information no19. Patancheru. A.P. 502 324, Inde:
International Crop Research Institute Iir the SEmi-Arid Trop&.
SMALING E.M.A., STOORVOGEL J. J. et WINDMEIJER P. N., 1993. Calculating soil
nutrient balances in Afiica at different scales. II. District scale. Fertilizer Research 35:237-250.
SMALING E. M. A., FRESCO L. et DE JAGER A, 1996. Classifying, monitoring and
improving soil nutrient stocks and flows in Ahican agriculture. Royal Swedish Academy of
Sciences: Ambio Vol. 25, December 1996.
SPARKS, D.L., A.L. PAGE, P.A. HELMKE, RH. LOEPPERT, P.N. SOLTANPOUR,
M.A.
TABATABAI, C.T. JOHNSTON, AND M.E. SUMNER 1996. Methods of soi1 analysis. Part
3 - Chemical methods. pp. 1390.

---

126
STOORVOGEL J. J., SMALING E. M. A. et JANSSEN B. H., 1993. Calculating soil
nutrient balances in Aca at clifferent scales. 1. Supra-national scale. Fertilizer Research
35:227-235.
STROOSNIJDER L., 1981. A soil water balance study of the Malian Sahel. Agronomy
Abstracts. 1981. p. 144.
TAKOW S.A., DOUMBIA M. D.’ and HOSSNER L. R, 1991. Acid soil profiles of the
semiaxid and subhumid tropics in C&tral and West Ahica. In RJ. Wright et al. (Eds.) Plant -
soir interactions at low pH. p. 3 23-320. fiwer Academic Publishers. The l+&erlands
TRAORE M.F., 1974. Etude de la fimnue minérale intensive et du role spécifique de la
matière organique dans la fertilité des sols du Mali. Agron. Trop. 29:567-586.
TRAORE B., 1995. La fertilisation dans les systèmes de culture à base de coton, Une
évolution souhaitable. Atelier Toposéquence DER-KIT. Sikasso, 21 - 25 Août 1995.
VAN DER POL F., 1992. Soi1 m$.ing. An unseen contributor to farm income in southern
Mali. Bulletin 325. Royal Tropical @titute. Amsterdam The Netherlands.
VAN DER POL F. et GIRAUDY F., 1993. Etude des relations entre pratiques d’amélioration
des sols et variablessocio-économiques dans la zone Mali-sud. IER/KIT/CMDT. Bamako.
Mali
VAN DER POL, F and TRAORE B., 1993. Soil nutrient depletion by agricultural production
in Southern Mali. Fertilizer Research 36:79-90.
VELDKAMP W. J., TRAORE A., N’DIAYE M. K., KEITA M. K., KEITA B. et
BAGAYOKO M., 1991. Fertilite des sols du Mali: Mali-sutioffice du Niger. Interpretation
des données analytiques des sols et des plantes. Cellule Agro-Pédologie. SRCVO - BZR
Sotuba
WILDING L. P. and HOSSNER L, R, 1989. Causes and effects of acidity in Sahelian soils. In
ICRISAT. 1989. SO& trop, and ;Yater management systems for rainfed agriculture in the
Sudano-Sahelian zone: proceedings of an international workshop, 1 l-16, Jan. 1987. p. 2 1 S-
227. ICRISAT Sahelian Center, Niamey, Niger. Patancheru, A.P. 502324, India: ICRISAT.

---

127
THEME 4 :ENVII ONMENTAL IMPACT
-----

---

128
Drinking Water Qualit# as Influemed by the Cropping
System
M. D. Doumbia”, A Sidïbé’, M. Ko&, A Couhbalyl, R A Kablan2, and R S. Yost”
1 Laboratoire sol-plante-eau, Institut d’Economie Appliquée, BP 438, Bamako, Mali
2 University of Hawaii, 1910 East West Road, Honolti, HI 96822
Water samples were collected fiom Wells in two villages located in two contrasted
agricultural zones of Mali to assess the impact of cropping systems (inputs) on the quality of
drînking water. The first village, Dougouba, is located on sandy soils under permanent millet
cropping system The second village, ‘Dampela, is located in Southern Mali, on loamy soils
under very intensive cotton cropping $stem Collected water samples were analyzed for pH.
N, P, Ca, Mg, K and Na.
Concentrations of P, Ca and Mg were low as 0.04 mg P L-‘, 0.02 mg Ca L-‘, and
0.01 mg Mg L-‘. HIighs variations were found in the concentration of minera1 N (0.25 to 1 .OO
mg L-l), K (12.5 to 200 mg L-l), Na ~(5 to 40 mg L“), and the values of pH (5.5 to 7.5).
Higer concentrations of N were ~found in water samples fi-om Dampela. Higher
concentrations of K and Na were fond in the sample fi-om the sandy soils of Dougouba.
The data presented here are just preliminary assessments. More literature review and
research are needed to validate these data.
1. Background
The low inherent fertility and fiagility of Malîan soils (Wildîng and Hossner 1989:
Doumbia et al, 1993, 1998) coupled with factors such as erratic rainfall, continuous removal
of nutrients under traditional cropphrg systems, lack of trop residue management in
traditional systems, cultivation of marginal lands, minimum or lack of soil conservation
practices in traditional systems, the needs for firewood, and traditional pastoral systems ha1.e
resulted in soil degradation and yield reductions (Van der Pol, 1992; Kiefi et al., 1994).
Despite the above constraints cotton yield .in the cotton cropping system of S0uthe.m
Mali has gone fi-om 200 Kg ha-’ (early 1960’s) to 1300 Kg ha-’ (Traore, 1995). Yield
projections of 1500 - 2500 Kg ha-’ have been made. These yield increases are due to the use
of variable rates of chemical fertihzers (200 Kg ha”‘), organic sources of nutrients ( 1 - 5 Mg
ha-‘), herbicides (1 - 3 L ha-‘) and insecticides (6 - 12 L ha-l). Alter more than 30 years of
cropping practices involving the above inputs, very limited information is available on their
impact on th.e environment. This research attempts to initiate an environnemental impact
n<~~P~ment study of these rropping systems on grollndlvater quality.

---

2. Objectives
The general objective of this research was to assess the impact of cropping systems
on the environment. The specitïc objective was to assess the impact of inputs on the quality
of drinking water (fiom Wells).
3. Materials and Methods
l
Villages were selected fiom ~ contrasting agricuhural zones. The first village,
Dougouba, is located in Central Mali, in sandy soils under permanent millet cropping systeru
Farmers of this village use several sourices of manure, but use neither chemical fertilizers nor
pesticides. The second village, DampeLa, i.s located in Southern Mali, on loamy soils under
very intensive cotton cropping system. Farmers of this village use not only high rates of
manure and chemical fertilizers, but ah+ several treatments of insecticides and herbicides.
Families selected (25) at Dougouba for the baseline evaluation of the Soil
Management CRSP were retained for this study. That evaluation colects socio-economic and
croppmg strategies data on these farmers. At Dampela, 25 families were also selected on the
basis of their participation in on-going $urvey of cropping systems.
Water samples were collected from Wells (in September, when most of the chemical
inputs have assumedly reacted), using local materials traditionally used.
Collected water were analyzed for pH, N, P, Ca, Mg, K, and Na according to
procedures suggested by Sparks et al. (~1996).
4. Results and Discussions
Concentrations of P, Ca and Mg were very low, within and across villages. These
concentrations were as low as 0.04 mg IL-’ for P, 0.02 mg L“ for Ca and 0.01 mg C’ for Mg.
These values .need to be compared to threshold concentrations.
Concentrations of minera1 N are presented in Figure 1. These fi-equency histograms
show high variations in the concentratbon of minera1 N in drînking water samples (0.25 to
1 .OO mg L-‘). Higher concentrations of minera1 N are shown for the samples from Dampela,
the site under intense cotton productio n. None of the nitrate levels are dangerously high at
the moment.
Concentrations of K are pref;ented in Figure 2. Here again, these fiequency
histograms show high variations in the oncentration of K in drinking water samples (12.5 to
200 mg L-l). Hïgh
k
er concentrations of are shown for the samples from Dougouba, the site
with sandy soils under permanent millet cropping systems, These higher rates may be due to
leaching of K from manure application,s (the only input used in the cropping system beside
seeds).
Concentrations of Na are presented in Figure ? Here alsn. these fiP?nenc~
histograms show high variations in the ~ concentration of Na in drinking water samples (5 to
40 mg, L-l). Higher concentrations of Na are shown for the samples from Dougouba. These
higher concentrations may be due to the impacts of househ’old activities carried out near

---

130
wells.
The pH of collected water samples are presented in Figure 4. Here also, these
fiequency histograms show high variations in the pH of drinking water samples (5.5 to 7.5).
‘Higher pH values are shown for the sa$tples fi-om Dougouba. These pH values cor-relate weh
with the higher concentrations of Na r+orted for tbese samples (Figure 3).
5. Conclusions ad Perspectives ~
c&lchlsions
Concentrations of P, Ca and Mg were low as 0.04 mg P L-‘, 0.02 mg Ca L-‘, and
0.01 mg Mg Cl. Hïgh variations were found in the concentration of minerai N (0.25 to 1 .OO
mg L-l), K (12.5 to 200 mg L-l), Na (5 ~to 40 mg C’), and the values of pH (5.5 to 7.5).
Higher concentrations of N were found in water samples from the loamy soils of
Dampela under intensive cotton cropping system Higher concentrations of K and Na were
fond in the sample fiom the sandy SO& of Dougouba where the use of mamue is probably
higher.
The data presented here are just preliminary assessments. Data have yet to be found
En order to compare the concentrations iof the above elements to threshold values.
Perspectives
There is a need for firrther anale sis of the data or collection of more data to explain
the high variability found not only wit ’
k samples fiom the same village, but also between
samples fiom the two villages.
Future works on this activity will be implemented in the same villages, but thorough
sampling in each household may be de+able for better understanding. The use of a tool such
GPS may help in both the data collection process and the trends in concentrations.
Thorough analysis of the water isample is also desirable. In fact, components such as
heavy metals and biological activities are important in evahtating the impacts of cropping
systems on the enviromnent, especially the quality of drinking water.
There is also a need for methods for laboratory analysis of the samples. A literature
review will be requested fiom our collaborators at the University of Hawaii on this issue.
_-.~-
---1111

---

131
WQuall
16
I
0 Dougouba 1
2
--
-
0
r-
0.00
0.50
Nitrogen concentration (mg/l)
Figure 1. Freq ?ncy histogram of N concentration in wells

---

132
WQual2
14
12
10
2
0
100.0
Potassium concentration (mg/l)
Figure 2. Frequel ~cy histogram of K concentration in Wells

---

WOual3
1 0
$
D Dampela /
a
0 Dougouba 1
E
8
3
3
i
8
6
E
i
i
rl
2
0 L
i -
5
10
20
>dium concentttaion (mg/L)
Figure 3. Frequen y histogram of Na concentration in Wells
---

---

I Dampela ~
-
0 Dougouba ~
20II
5.5
6.5
7.5
Figure 4. Frequedcy histogram of the pH of Wells waters
-----.-

---

135
DGUMBIA M.D., HOSSNER L.R., sud ONKEN A.B., 1993. Variable sorghum growth
in acid soils of subhumid West Af%ca. Arid Soil Research and Rehabilitation 7:335-346.
.DOUMBJA M.D., HOSSNER L.R., and ONKEN AB., 199. Variable sorghum growth in
acid soils of West AfXca. Variations $ soil chemical properties. Arid Soil Research and
Rehabilitation 12: 179- 190.
KrcFT HI., REITA N. et VAN DER FIEfDE A., 1994. Engrais fertiles? Vers une fertilité
durable des terres agricoles au Mali. 99 p. ETC, Leusden, The Netherlands.
SPARRS, D.L., A.L. PAGE, P.A. L(ELMKE, RH. LOEPPERT, P.N. SOLTANPOUR
M.A. TABATABAI, C.T. JOHNSTGN, AND M.E. SUMNER 1996. Methods of soi1
analysis. Part 3 - Chen&al methods. pP. 1390.
TRAORE B., 1995. La fertilisation dans les systèmes de culture à base de coton. Une
évolution souhaitable. Atelier Toposéq&nce IER-JKIT. Sikasso, 21 - 25 Août 1995.
VAN DER POL F., 1992. Soil min&) An unseen contributor to farm income in southern
Mali. Bulletin 325. Royal Tropical Inst/itute. Amsterdam. The Netherlands.
WILDING L. P. and HOSSNER L. R, 1989. Causes and effects of acidity in Sahel&
soils. In ICRISAT. 1989. Soil, trop, and water management systems for rainfed
agricuhure in the Sudano-Sahel& zone proceedings of an international workshop, 1 l- 16.
Jan. 1987. p. 215-227. ICRISAT Shhelian Center, Niamey, Niger. Patancheru, A.P.
502324, Mia: ICRISAT.
--

---

---

1 3 7
Report of the Field trip~ to the K.outango and Nioro Sites
M. Diack, ISRA Senegal
As part of tbe workshop program, the field trip gave an opportunity to the participants
to vi& the Koutango and Nioro sites~where the EnterCRSP experimental plots were
located.
l
7H3Oam
The visitors lefl Kaolack for the Koutango site, 69 km, South-west fiom
KaOhCk.
9H 05 am
l
They arrived at the site and were introduced to Mr. Mansour Dème, field worker and
owner of the field by Dr. Modou Sène. A nutrient management progam is being carried
out on rice under rainfed. Rock phosphate and phosphogypsum as P and Ca sources
and ~manure are the fertilizer forms used in the experiment (see papers).
Several questions about the perception that farmers have on the phosphogypsum
progam, initiated by the Government of Senegal, were asked.
Was the phosphogypsum efficient as fertilizer?
Yes, and I am also using it to control ~salinity, reptied Mr. D&ue.
What ifthe Govermuent stopped providing fertihzer freely?
1 Will purchase it, he said.
Beside rice, Mr. Dème grows vegetables because water is available.
The site is within an interesting valley for which ISRA intends to run a research
and development program under lowland conditions. Dr. Dia&, coordinator of that
program, presented preliminary data obtained fiom both socio-economic and
hydrologie surveys conducted in the valley. The data showed a great agricultural
potential of the valley. However, soil ~constraints such as salinization, acidii and water
management need to be addressed.i From the socio-economic standpoint, villages
established arround the valley are organized into 57 groups based on economic
interest. The R and D program will be carried out by a multidisciplinary team
composed of scientists f?om ISRA, th’ University of Dakar and an NGO.
1
9H45am
l
The participants left Koutango for the Nioro Research Station, second site
llH45 am.
As they arrived at the Research Station, the participants were introduced to the local
staff by Dr. Modou Sène. The visit
arted with experimental plots amongwhich the
Piliostigma reticdatum tria1 under %eanut/millet rotation, a second tria1 where rock
phosphate and phosphogypsum as P I”d Ca sources and manure were compared under
a peanut/miIlet rotation and a third. experiment combining nutrient management and
water balance (see papers). The partrcrpants also visited the small dam implemented to
control runofT/erosion in the watershed.
For each trial, several questions were asked, leading t’o important discussion.
From the discussion, one could appreciate the amount of work done SO far?
meaning the InterCRSP program in Senegal is going well.
Iii 20 pm
The participants IeR Nioro for Kaolack.
2H 15 pm. Arriva1 at Kaolack.

---

138
Tbe f&wing participants took part 1o tbe field trip:
1 . RmselYOST
2. Gaoussou TRAORE
3 . Samuel BRUCE-OLIVER
4. Aminata SIDIBE
5. Isuarina BAPTISTA
6. AminataBADIANE
7. Abou BERTHE
8 . Mamadou DOUMBIA
9. KevinBRANNAN
10. Richard KABUL ,
11. Mouhamed KEBBEH
12. Babou JOBE
13. Cheikb NDIAYJZ
14. Modou SENE
15. Mateugue DIACK

---

139
REPORI S ON SESSIONS

---

140
Report for themes 1 and 2
Monday, Jamtary 11,99
Chairman: Dr. Jean Pierre N’DIA$E
Secretary: Ms. Aminata SLDIBE
~
The themes discusxd below were preceded by keynotes addresses tiom Dr. Gaoussou
Traore, Head of Division at INSAH, +rd Dr. Russell S. Yost, Coordinator of the West
InterCRSP Croup.
Dr. Traore gave an overview of the activities and strategic workplan of MSAH. He
indicated that InterCRSP is a program falling into the natural resource management
program of LNSAH. He insisted that
terCRSP should work under the umbrella, and
should therefore keep JNSAH up dbted on anything regarding this collaborative
research program Discussions related this speech were on information procedures
between the West InterCRSP Croup and INSAH, the ROSELT program, and
interactions with other CRSP.
Dr. Yost gave an overview of activities of the West JInterCRSP Croup. He recalled the
overall objective of the program: nahual resources management for food security and
income. As a justification of this objective, he pointed out that food supply decreased
by 18% in the target area of the West InterCRSP. There is an urgent need to reverse
the fàctors contriiuting to this decline in production. Dr. Yost fistrated by a slide
show some of these factors in each of the 4 countries of the West InterCRSP group.
These tictors include: loss of soil organic matter., nutrient depletion, soil acidification,
soil loss by erosion, erratic rainfàll, degradation of natural tesources, etc... These
factors were grouped into 4 major themes for the workplan: Ci) Farm programming and
policy, (ii) Soil and water conservation, (iii) nutrient management, and (iv)
environmental impact. Discussions related to this presentation were about localized
manure application at Dougouba (Mali), spatial variability in a sorghum field at
Fansirakoro (Mali), Salt intrusion in Senegal and The Gambia, and erosion control on
hill slopes of Cape Verde.
l
Theme 1: Farm Programming and kountry Policy
Country presentations under this theme were made for Cape Verde, The Gambia, and
Mali.
Cape Verde: Farmer to Farmer Vi$t by Isaurinda Baptista and Isabel Anahory
In Cape Verde, soil characteristics, toPography conditions and climate constraints have
produced a unique agricuhural system Major constraints to this agricuhural system are
salinity, loss of soi1 organic matter , and soi1 erosion. Inter farmer visits were organized
to exchange views on techniques used by farmers to alleviate these constraints. These
visits in diZerent areas were suç~~&iiI and farmers suggesrcd that this type of activiry
is worth repeating.

---

Discussions related this presentation were on soil types, drip irrigation, soi1 erosion
measurements, and the benefit fàrm s get fiom interacting amcmg themselves. It was
recommended to increase the
of these “làrmer-to-fat-mer” visits.
The Gambia: Adoption and Parm Level Impact of Improved Fertility
Management Technologies in the Sudano-Sahelian Zone of The Gambia by
Mohamed Kebbeh
Low inherent soil fertility and decreasing levels of chemical fertilizer use have partly
explained the signiCcant decline in {rop yields in Gambian uplands. Researchers are
attempting to address this problem by evamating alternative soil fertility management
strategies in eastern and western Gambia. Whole fàrm models are developed to assess
the adoption and fàrm level impa,cts of combinations of organic and inorganic
fertilizers. T’he results of the simulations show positive response to chemical fertilizer
price changes. For example, area under millet would increase drastically, at the expense
of groundm~t and maize, w-ith a 25% decrease in ehemical fertilizer prices.
Discussions related this presentation were on The use of value/cost ratio in compatison
with the mode1 described in the prese@ation and the relevance of changes in the prices
of fertilizers. Such a change of 25% decrease was thought unlikely as chances for
fertihzer subsidy or tax tut on fertilizers are very slim. It was suggested to address
issues such that: (i) who will adopt the tedmology? (ii) what will be adopted? (iii) what
proportion of the technology? (iv) hbw sustainable is the technology? In addition, it
was suggested to conduct this type of evahution in each country of the West
InterCRSP group.
Mali: Improving and Sustaining Food and Raw Material Productiion in West
Aftica: A Participative Rapid R~I-@ Appraisal in Fansirakoro by Russell S. Yost
and Abou Berthe
l
A participatory rapid rural appraisal (PRRA) was conducted to provide a baseline
information on the farming systems found in the area of Kati (Mali) in order to
establish research priorities for the bterCRSP project. According to farmers, yield
decline was due to dimG&ing ra’
R lack of moisture, reduction of the density of
trees, rectuction of organic matter 7
in e soil, and increased mn-off. It was concluded
that the main agricultural production constraints are related to soi1 and water
conservation and management. Re&mmendations to alleviate the above constraints
were divided into 2 recommendation domains, one representing 30% of the farmers
and the other representing 70% of the farmers (very resource-poor farmers.
Discussions related this presentation were on the lack of information on organic
fertilizer production, the method used. to conduct the PRRA, the definition of
recommendation domains, and steep billside farming in the area.

---

142
Theme 2: Soil and Water Conseru
ion
Country presentations under this the : were made for Mali and SenegaL
Mali: Improving Food Produt o n i n Subsistence F a r m i n g Systems i n
Fansirakoro and N,Tentoukoro b A. H. Berthe, S. F. Traore, B. Traore, C. D.
Kamissoko, B. Guinda, R A. Kab n, and R S. Yost.
An on-farm tria1 involving 10
rmers in both villages of Fansirakoro and
N,Tentoukoro was conducted to e\\ uate the effect of di%xent fertilizer management
systems on the productivity of sor um. Ear and grain yields were not statistically
dif%rent. Fax-mers ranked the treatm ts, according to visual’ hùbessment, the combined
manu.re/fertïlizer first. the manure se nd, and the check la&
Discussions related this presentatiol vere on a yield increase of 400 kg of sorghum
grain per ha, the treatments of the
periment, .the variability and the sign%cance of
some of the data collected.
Senegal: Effect of Manure and P ource Fertilizer on Water and Nutrient Use
for Main Crops in Senegal Peanut :asin by M. Sene, M. Diack, and A. Badiane.
A long term experiment was initiatec
o evaluate the effects of several combinations of
phosphate rock and phospho-gypse
In water and nutrient use for several trop in the
Peanut Basïn of Senegai. This expe rient is going on for the second year. The data
generated so fàr show little impact
’ the treatments due to the low sohxbility of the
Taiba phosphate rock and the lim d nutrive value of tbe phospho-gypsum (it is
however reach in caicium and sulfixr’
Discussions related this presentation
‘ere on the relevance of long term experiment on
thîs InterCRSP project which “quicl *esults-to -the-farmer” oriented, soil properties,
and phospho-gypsum issues.
Tuesday, January 12,99
Presentations of Session 3: Nutrient
anagement
Cape Verde Isaurinda Baptista
Effect of Organic and Inorganic 1 -tilizers on: the nutrient status and yield of
dry land crops (maize and beans)
Objectives : 1) to evaluate the effect f bot11 chemical fertilizer and animal manure on
trop growth and yield ;
2) to improve soi1 tond ons both physically and chemically.
CQrdi4sion :
The tria1 was not followed to
e end because of trop failure due to drought.
-uI---

---

However, the manure plus fertilizer out-performed other treatments with the highest
dry matter yield of 3.92 Mg/ha compared to 2.27 Mg/ha for the control.
It was suggested to continue the study~ under favorable moisture conditions so as to
meet the objectives.
Discussion
Question 1 : Why was it that you apply organic fertilizer ? 1s the soil degraded ?
Answer : We were interested in the long-term effect, but the soil at the site is not
degraded.
Question 2 : Why did YOU apply 20 tia of manure ?
Answer : It is the rate that is recommended.
Question 3 : How was the manure handled ?
.Answer : There is no special management. It was just collected fiom pens and
transported to the field for application. )
Question 4 : 1s there any marketing on manure ?
Answer : No, but we brought the mate+1 for our tria1 from a Fann.
Question 5 : There appears to be inconsistency in the measured values from organic
matter for the control ?
Answer : Yes, I’ve realized the incons&ency. Hope to look at it again.
Question 6 : Why cow manure per se ?#
Answer : Cow is most commonly used,: but poultry manure is also available but not
used on large because it bums the crop$.
2. Mali : Aminata Sidibé
Phosphorus Buffer Coeffxcients of Selected Soils of West Africa
Objective : TO validate P buffer coeffic+ents predicted by the Phosphorus Decision
Support System (PDSS) model.
~
Conclusion : P buffer coefficients determined fi-om .the mode1 were not significantly
different fi-om those of the incubation studies. How~~er, the clay soils tend to have
lower coefficients with the incubation method. Further study is needed to investigate
this efhect.
Discussion
Question 1: What type of clay mineralogy are your soils ?
Answer : Mostly kaolinite (1: 1) clays.
)
Supplement : Upland soils in Mali have fixation of 12 g/kg and the Smectites of the
lowlands fix at the rate of 150 g/kg.
~
Question 2 : Did you analyze for clay mineralogy ?
Answer : No, we do not have the capacity. However, we cari always send it to the
other labs for X-ray diffraction.
Mali : Mamadou Doumbia
Nutrient Balances under Contrasting’Millet Cropping Systems
C?Aject,‘+c : to estitiiate nutnent Ldgets u.nJ.cr thrcL milL cropping bystern, ;ti MA.
Cunchion : All three cropping systemsregistered negative nutrients balance but it
was highest under the cotton based croppinl systems.

---

144
Discussion :
Question 1 : What is your present st$ategy for soi1 fertility restoration ?
Answer : No particular one as yet b&ause we are just starting. However, we are
currently looking into composting dth rock phosphate, chemical fertiliz,er and
improved manure storage.
Question 2 : How do you account fol soil input ?
Answer : There are d.ifKerent approaqhes :
1) Input as amendment - output (crob removal)
2) Stial level + addition - output.
~
Question3 ,: Do you have a methodofpgy 1 cari use ? 1 want do a similar study in Cape
Verde ?
Answer : As 1 said, there are differeni approaches. You use .the method most
appropriate to your conditions. We +ill however talk more.
Question 4 : What would you reco
end ?
Answer : We are not yet ready
T
with ti e data. We have to look at it again and iffünds
are available, we will continue the wcirk.
Senegal : Mateugue Diack
Use uf Piliostigma reticulatum (locz+l shrub) to improve upland soil conditions in a
milletigroundnut rotation system. ~
Objective : TO evaluate the tiuence bf Piliostigtna reticulatunz on the soil quality and
trop yield.
Conclusion : Tria1 implementation in process. ResUtts obtained SO far suggested
decomposition of plant parts to be fa$er under field than laboratory incubation
conditions. It was either that soi1 mic+bes were excluded from soil during sample
preparation for laboratory incubation br the soils are biologically degraded.
Discussion
l
Question 1 : Why was millet yields too low ?
Answer : Tbe millet trop was lately planted coinciding with dry spell.
Question 2 : What was the methodolqgy for N analysis during laboratory incubation ?
Answer : We used the classical methofi, but one has to do it fast under optimum
temperature.
Question 3 : How cari animal traction ~be adapted in managing the residues of this
shrub ?
Answer : Animal traction is not an obtiacle in the management of shrub residue
especially at weeding when the re-gro*h is tender and gets incorporated readily for
faster decomposition.
Senegal : Modou Sène
Use of Phosphogypsum (PG) to corqect P deficiency
Objective : to evaluate PG against Rock Phosphate
Conclusion : No rtzponse by com in tbe firu: year.. For secon< ycar, ,ïoundnut kud
yield was most significantly increased by lime alone treatment followed by treatment
combinations containing more by ratio: of PG than RP.
-111---
.--
-.----

---

145
Discussion
Question 1 : What liming material did You use ?
Answer : Calcium oxide (CaO) at 400~ kg/ha.
Question 2 : Are there severai forms of PG ?
Answer : There are different types of RP but PG is a by-product of RP.
Question 3 : What isthe variability in PG ?
Answer : 1 cannot tell you now. We are still working on the data.
Comment : PG sent to Mali contains 1% P205 and 38% CaO.
Question 4 : Did you assess PG on biological N &ation.
Answer : No.
Comment : A work with RP which had positive effect on tir, was recalled.
Question 5 : What was your initial soil exchangeable calcium test ?
Answer : 0.6 meq/lOO g of soil and robe to 0.8 aRer amendment.
Comment : That may explain the supei-iority of lime and PG in increasing pod yield.
Groundnuts will likely respond to calcfum in soit with less than 1 .O meq/lOO g of
exchangeable calcium
Mali : Mamadou Doumbia
:
Environmental Impact Assessment ktudy
Objective : to assess the impact diBerent cropping practices have on the quality of well
water.
Cumlusim :: Minimal levels of nitrate Jand phosphorus detected but elevated levels of
potassium and sodium were tested in water.
Discmsion :
Suggestion on methodology for such assessment were invited.
comments :
1. You need to measure electrical conductivity (EC)
2. There is a threshold value for P forilake eutrifiçation but not for drinking water
3 . Identify proximity of locations to wells where human activity are possible source for
contamination
4. C?roup wells by locations.
The Gambia : Babou Jobe
An Integrated Fertilization Study of the Groundnutknillet rotation system of
North Bank Division
Objective : 1) to assess changes in crob response {biological and economic yields) due
to difherent soil amendments
2) to assess changes in soi1 properties due to difherent soil amandements.
co;,,:lusi”ii : Regardlcss of t>c tJi ; of manu;;, manure plu, fertilizer Llfluencb trop
yield and soi1 properties pore positively than fertilizer atone. TO manage fertilizer well
for reasons of cost and environmental integrity, it’s best to explore the complementary
effects of manwe plus fertilizer. ‘Ihere is also a need to improve on the transportation

---

and field handling of manure in promoting large-scale adoption of manure use for field
crops
Discussion :
l
Question 1 : Was this a station or on-f$rm experiment ?
Answer : This was a research-managed on-farm trial.
Question 2 : What was the initial soil P level because the control yielded more than 1
t/ha?
Answer : The initial P test is not shown here but still the significant increase in yield
due to soil amendment may indicate that the initial P supply may be below optimum
Question 3 : How did you get to this regression equation ?
Answer : These were derived fkom contrasts and regression fiom 20 points and not
fkom 5 point-mean data.
Question 4 : Why was combined analykis on performed on two site ?
Answer : Only these two sites were compatible for combined analysis following test on
homogeneity of variante.
~
Question 5 : Why was not the Split-plot design used instead of the RCB factorial ?
Answer : Split-plot measures interactions better than main efIects and is best applicable
when large differences exist among main factors. This certainly cannot say among the
main fàctors here. Also, the added a+tage of fàctorial RCH is that it measures both
main effect attd interaction effect with ~equal precision.
Question 6 What do you mean by Max ?
Answer : By Xmax, 1 mean the amount of input needed to obtain maximum response.
It only indicate treatment effect not optimum or economic level.
Question 7 : In what form was the cow manure ?
Answer : The cow manure was in the dung form not with bedding/litter as farmyard
manure. It was reduced into smaller fragments to -increase surface area for faster
decomposition.
Project Activity Planning WW-$000
Virginia State University : Kevin Bran+kran
Runoff Modeling with the KINEROS Iode1
Use : 1) Compare practices
2) Assist in design and locating practices in watershed
3) Assist in monitoring
Discussion
Question 1 : Can the mode1 be adopted under the structures we have III Cape Verde !
Answer : Yes, the structure 1 saw in Cape Verde like stone wall dams across channels
because they have geometry.

---

147
Question 2 : How cari the Gambia benefit fiom the mode1 in terms of information on
topography such as slope for cost/be$efit analysis of erosion control practices.
Answer : 1 was involved in lots of modeling with the U. S. Soil and Water Conservation
services and. the information cari be usied by economists for use in cost/benefit analysis.
The Gambia : Mohamed Kebbeh ~
Farm Level Economie Programming/Policy
Applications : 1) evaluate the perfornntnce of improved technology
2) effect of risk
~
3) Questions of adaptation (who, what and how much).
Discussion
Question 1 : From your analysis in Gambia, fertilizer price is an important constraint to
fertilizer use. This is important in the issue of soil nutrient mining. What we do to
improve policy on fertilizer ?
~
Answer : We should try to make policy makers involve in the analysis and use of
results. But how do we get the results to policy makers ? May be they should be
exposed to fora like this.
Comments : It is a key issue and at th@ levek has been discussed at various cycles and
at various levels even up to presidential. But, still the issue of making fertihier
affordable. It may also be usefùl to bonsider farmgate prices. For example, Gambia
increased producer price for groundnut last year by 21% and fertilizer use is picking
up. Farmer organizations cari be very effective in pressuring governments to improve
policy on inputs.
Recommendations
1. TO harmonize methodologies on thematic InterCRSP field studies ;
2. TO shorten research cycle as most of the topics being looked at now have been
previously worked on and also if negative impact is not envisaged of technology
being developed ;
3. Researchers argued in favor of long-term studies in order to be sure of what effect
the developed technology would bave on the environment ;
4. Mamadou Doumbia should use the results of nutrient balance studies to advice
farmers of the implications of their practices on their soil quality ;
5. Need to explore agroforestry to for& stoneline for runoff control ;
6. Need to explore biologically oriented cropping systems such as improved fallow and
imercropping with high N fixing legume species for both soil fertility maintenance
and SO~I and water conservation.

---

148
Minutes of the plenary session 0: ‘InterCRSP
Thursday, January 14 th, 1999
Chairman: Dr. Samuel Bruce-Oliver
Secretary: Dr. Abou Berthé
Dr. Yost in his introduction asked ifcc luntries have specific problems to discuss before
he cari move on project activities for th : 3 years. He noticed that he has ben- involved
in meeting with each country represent Itives except Mali to discuss budget. He will
discuss this matter with Mali represent; tives in Bamako.
Representatives of country asked more informations about the following topics.
1. Baptista (Cap Vert) : When we cari s :nd receipts about expenses ?
Dr. Yost :
You cari send receipts a expenses are going on. Bank transfer cari take
at least 2 weeks.
Dr. Bruce :
1 think it will be better 1 3 indicate deadline to u-se fùnds by the time Dr.
Yost gets to hawaii. Th : budget has been tut to 60% in Year 2. It is
possible to use the fùnd! of the project to caver administrative and
management costs (e-mi 4 fax, etc.,:)
Dr. Yost :
of course you cari. For 1ixample , nutrient balance cari be funded on on-
farm trials. 1 advice you to combine activities across tasks in a very
efficient ways to fit your situation.
Dr. Traoré Gaoussou :
Do you h tve any mechanism to talk to the East group.
You migl t have a combine workshop in Bamako.
Dr. Yost :
1 talked to the East group, but in an Informa1 way. We need to work
with INSAH for identifling the group. We might need some
translation. The workshop çannot be in english next time.
l
Dr. Badiane : The main financial problem of the project is overhead costs. We are
using all facilities at ISRA to make the project running.
Dr. Kebbeh : 1 suggest that 10% of F+nds cari be used for overhead expenses.
Dr. Yost:
If there is some consensus, we might consider that situation.
Dr. Sene:
1s it AID rules not have overhead.
Dr. Ycw
1 think it is allowed. Perhaps, 1 Will be very happy to look for that. You
are suggesting 10%.
-------
--

---

149
Dr. Badiane: The other problem is the bill. We need fimds to run. The labor cannot
wait. They should net tut the fimds. This year for example we will be
doing our PRRA.
Dr. Yost:
Receipts in our financial department are not too restrictive. 1 have
nothing to do with thi$. You should be consistent with your accounting
system.
Dr. Bruce:
At least you give initial money to start with the activities. After that
initial fùnd the system of prefinancing will go on.
Dr. Sene:
We are in year 2 now?, When year 1 did start.
Dr. Yost:
First year was frorn may 1’ 1997 to April30 th 1998. We had our trip
in October 1997. All activities for year 2 should end by September
1999. 1 don? have a straight answer. You should spend all year 1 and
year 2 fùnds by MarchI to avoid cutting. We got tut before, we
didn’t spend all the money.
Review of project tasks, present Statu$, factors causing present status and revised
completion date
Objective
T a s k s
Expected Output
1. Evaluate farmer PRRA
Report, Workshop presentation
practices
Annual workshop in Year 2000, January 10 - 14
Training task 33: cross visits consultancy reports, improved skills and programs
50% costs to be coverd by receiving country
50% cost by InterCRSP Headquarters
Modou Sene : 1 would like to see cross visits as all groups getting together to visit
project activities in a country
Dr. Gaoussou Traoré :
1 like cross visits. Even you have local expertise you
need to ring in somebo#y. You cari try also E-mail conferences among
the group.
Annual work.shop pappers : Due date December 20*
Workshop proceedings :
March 1; 1999
Annual reports
July 15, 1999
Dr. Gaoussou Traoré :If you want your results to be largely disseminated, they should
!;r in frcr;eh.
Dr. Sene :
Papers for Gambia should be in English
Mali and Senegal :
French
-----

---

150
Cap vert :
+wtugese
Dr. Brannan : Extension fact sheets shouki be in national langages, but scientifïc
papers mighî be in enghsh
Abstracts : Enghsh and french for paper executive summaries
Cut-off date for report/s Febl , 1999
Forword for proceeedings Will be made by Dr. Yost.
Comme& uti Workplan needed:
(
February 1,1999
Workshop proceedittgs:
April, 1999 (due date)
Annual reports (progess report) July 115. 1999 . Will have to submit for September 1”.
Annual workshop papers, due date for workshop papers , Jan 31, year 2000 in
Bamako?
Comtry Workpïan presentaîi*
Senegal
Kineros model, PRFG$
Crossvisit t o u r ~
Yo:st:
Farm level economic programming
Badiane:
Kebbeh Will give us the formuIa and data requirements. 0tt.r economist
cari nm the modet
I
BTUCXX
Are you doing fat-mer to fi-amer tisit
Badiane:
Yes, we do.
I
M a l i
- Farming system researcb and natural resowce
management
1. Task 2.6
Estimates of runoff a$d soil loss in Mali OHVN zone (use of Kinero’s
model)
2. task 2.14
Farm level economic programming
3. Task 3.23 On fàrm tria1
~
(manure extender)
w
contour Iine cuitivation
vegetating stone lines

---

151
4. T’ask 4.26 Farmer to farmer visit (contour line farming, organic matter
production, agroforestry)
5. task 4.29
Field days diflùsion (1 und table discussion with NGO’s ,Extension and
commodity research, etc. .)
- Nutrient Balance
1. Soil fertilii restoration
PR&PDSS (SM activities)
2. Environmental impact assessment
3. GIS data development, componen of other activities
Gambia
-
Nutrient balance study (Njaba Ku dal
-
Farm level economic programmir
(incorporate other crops into the model)
- Acidîfïcation study
-
Runof%oil erosion study: benefit :ost analysis for different erosion control
strategies
- On-fiam trails
-
Pre-extension fiam trials of the m rient management study at Njaba Kunda
-
Assess the stability of the techno18 3 across different farm conditions.
-
F?amer to farmer visits.
- Lmpact analysis
-
Fnviromnental impact study
Dr. Gaoussou Traoré (INSAH)
1 don’t see the relationship of your rorkplan and farmers’ needs.
Dia&: What are the norms for water tuality?
Kevin: Maximum contaminated level 1 10 ppm for Nitrogen
Tbey are well established for pesticidc
Dr. Yost: some pesticides are known ) be leachers. Pesticides are difficult and costly
to analyze. You cari make a list ofpe :icides used in CMDT zones and we wil rank
them as leaching or as no leaching.
Minimum concentration level (mcl) is ;et for a lot of pesticides.
Cap Vert
-
Nutrient balance study (com, pea lt, beans and banana) at two locations
1 need some references on that w( .k.

---

152
Available references are f?om: Gigou, 1980’s and Pieri 1978 in Senegal
(,Agronomie tropicale)
~
smaling
Van der pool and Nick in Mali
(
-
Runoff studies (Kinero’s model)
-
Farm level econmoic programming & adoption studies
-
Environmental inpact study (Salt i@rusion and pesticides)
- On-f?am triais (two locations)
I
m
Integrate trop and forage system$ (biomass yield and build-up of SOM)
-
Framer to fi-amer vi&
-
Regional impact (GIS caver of the watershed)
Badiane: meet deadline
Dr. Gaoussou Traoré (INSAH):
-
send a report to INSAH
-
Bring others groups for next yeari workshop
-
Interna1 discussion of reports in each conntry before coming to the group.
Dr. Bruce: Take an opportunity to thbnk all participants, Badiane, people to Kaolack.
Thank you very much again

---

153
IList bf Participants
Aminata Niane Badiane
ISRA-DG, Route des Hydrocarbures
Bd Air, BP 3120, Dakar, Senegal
e-m& an&ian*s. arc. sn
TeL: (221) 832 24 31/ 832 24 28
Isaurinda Uaptista
rNrD&CPs4
Praia, CAPE VERDE
e-mail: inidacv@mailtelecoacv
Tel.: (238) 71 1127
Fax: (238) 71 11 33
Abou Berthé
ESPGRNISotuba, BP 9030
Bamako, Mali
e-mail: abert&@$Malinet.ml
Kevin Brannan
Biological System Engineering
Virginia Tech.
Blaeksburg, VA 24061- 0303
e-mail: kbramw@vt.edu
Tel.: (540) 231 2145
Fax: (540) 231 3199
Samuel Bruce-Oliver
National Agricukural Research ht te
P M B 5 2 6
Serrekmda, The Gambia
e=mail: sbs@qanet.gm
TeL: (220) 48 49 25 / 48 49 31
Fax: (220) 48 49 21
Mateugue Diack
ISRA/Saint-Louis, BP 240
Saint-Louis, Senegal
Tel.: (221) 961 17 5 1
Fax: (221) 961 18 91
Mamadou D. Doumbia
LabosepmR, BP 438
Bamako, Mali
e-mail: madu@labosep.ier.ml
Tel.: (223) 24 61 66
Fax: (223) 22 37 75

---

1 5 4
Babou Jobe
National Agricultural Research Institu e
P M B 5 2 6
Serrekmda, The Gamhia
e-ma% babjobe@qanet.gm
Tel.:(220)484931/4831 63
Fax: (220) 48 49 21
Richard Kablan
University of Hawaii
1910 East West Road,
Honom HI 96822
e-mail: rak@hawaii.edu
TeL: (808) 956 68 83 / 946 94 21
Mohamed Kebbeh
National Agricultural Research Institu e
P M B 5 2 6
Serrekunda, The Gambia
e-mail: mbkebbeh@qanet.gm
Tel.:(220)484931 /4831 67
Fax: (220) 48 49 21
Cheikh Ndiaye
CARITASKaolack
BP 482
Kaolack, Senegal
e-mail: caritaskl@s. arc. su
TeL: (221) 941 20 30 / 941 27 30
Fax: (221) 941 35 34
Jean Pierre Ndiaye
ISRA/DG, Route des Hydrocarbures
BP 3 120, Bel-Air
e-mail: jpndiaye@sra.refer. sn
Tel.: (221) 832 24 28
Modou Sène
ISRAKNRA, BP 53
Bambey, Senegal
e-mail: isracrak@syfed.refer. sn
Tel.: (221) 973 60 50
Fax: (221) 9’73 60 52
Aminata Sidibé
LabosepER, BP ?38
Bamako, Ma.li
e-mail: ami@~abosep.ier.ml
Tel.: (223) 24 61 66

---

155
Gaoussou Traoré
Institut du Sahel, BP 1530
e-mail: gaoussou@padres.imah.ml
Tel.: (223) 23 4 0 6 7
Fax: (223) 2 2 5 9 80
Russe1 Yost
University of Hawaii
1910 East West Road,
HonoM% HI 96822
e-mail: rsyost@hawaii. edu
Tel.: (SOS) 956 70 66
---1I*II-

---

Programme of the III xcrsp West group workshop
Kaolack, Sene: d January ll-14,1999
Sunday januray 10 *, Arriva1 of the articipants and departure for Kaolack
Monday, january 11,
9h30 - lOhl5 : Opening ceremony
. Dr. Russe1 YOST
?? Dr. Gaoussou TRAORE
??
DGIISRA
lOh15 - lOh30 : Coffee break
lOh30 - 11 h30 : Keynotes adresses
Chairman : Dr. J.P. NDIAYE
Secretary : Ms. SIDIBE
lOh30 - 1 lh : Dr. Gaoussou TRAOR
1 lh -- 1 lh30 : Dr. Russell YOST
Theme 1 : Farm Programming and Jountry policv
1 lh30 - 1 lh50 : Cape Verde
1 lh50 - 12hlO : The Gambia
12hlO - 12h30 : Mali
12h 30 - 12b50 : Senegal
12h50 - 13h30 : General Discussions
13h30 - 14h30 : Lunch Break
Theme 2 : Soil / Water Conservatio
14h30 - 14hSO : Mali
14h50 - 15h’LO : Senegal
15h 10 - 16h30 : Discussions
Tuesday, january 12 th
Chairtnan : Dr. Gaoussou TRAORE
Secretary : Mr. Babou JOBE
Theme 3 : Nutrient management
9h - 9h20 : Cape Verde
9h20 - 9h40 : The Gambia
9h40 - lOh20 : Mali (2 présentations)
lOh20 - 1 lhO0 : Senegal(2 presentati 1s)

---

157
11 hO0 - 1 lh20 : Break
1 lh20 - 13 h 00 : Discussion
13hOO - 14hOO : Lunch break
Theme 4 :Environmental Impact ~
14hO0 - 14h20 : Mali
14h20 - 14h40 : Discussion
Theme 5 : Proiect Activities Plannif-
14h40 - 15h15 : RunofVErosion Cor rol (M. SENE and K BRANNAN)
I5hl5 - 15h45 : Nutrient managerne t (M. DOUMBIA)
15h45 - 16h15 : Farm level economi programming/policy (M. KEBBEH)
16hl5 - 17h30 : Discussion (GIS)
Wednesday, janusry 13 *
7h30 - 13h30
Fidd trip : Nioro and Koutingo siites
Thursdayjanuary 14 th
Plenary session
Chairman : Dr. Samuel BRUCE-OLItrER
Secretary : Dr. Abou Berthe
9hOO - lOh30 : Project management 1RYOST)
1 Oh30 - lOh45 : Break
10h45 - 123245 : Countries presentati ms (Plans and Taks)
12h45 - 13h30 : Wrap up meeting
13h30 - 14h30 : Break
15hOO : DEPARTURE fiom KAOL/ C K

---