MALAWI DIAGNeSTIC S(JURVEY WORK. NTClIU NRDP AR1A FEBRUARY 1980
Dcetaild _Guidelino nort on nthe r_-mi 'tin 1 fa.:ni 1sste .
Discussions hpld initially with ,LD1P staff and local Agriculture staff,
together with evidence from the available AES publications, suggested that
EPA's 2,4,7,8 could be taken ;s. following a similar type of farming.
On the whole this was verified during the Exploratory survey though some
North South and local differences were noted.
1. In EPA 2 many more farmers had adopted improved maize varieties,
with hybrids also common fertiliser use probably more widespread than
in the South. EPA 7 where maize still predominant as a food crop,
and VCA had some adoption as a food maize.
'2. Cattle concentrations higher in the North c. 40-50% of households
compared to perhaps 10-15% in the South.
3. Dimba areas were more common in the North And allowed some off
season activity by a higher proportion of farmers there.
4. Cotton was more widely accepted in the South c 20-25% farmers tailing
off to none in EPA 2 in the North..
5. Land in general less densely cultivated in the South, consistent with
a lower density of" population.
6. The production of beans, the most favoured relish crop, was a
problem throughout. the area, but much more serious in the South
where production was very difficult.
7. Cassava and sweet potatoes were widely seen as insurances against the
failure of starch cereal staples. However in areas where the
keeping of pigs had become popular their feeding on roots was
beginning to sqeoze out both crops. The problem was emphasized
more strongly in the South of the area.
In spite of these variations, of.which the importance of cotton in the
Southern p.rers is perhaps the most significant, it is felt justified to
treat the whole of EPA's 2,4,7,8 as a uniform type of farming area.
Variatiiurn. were noted and verified durif.g fieldwork
a) More hybrid adoption in the North EPA.2.
b) More cotton grown in the South EPA 7. 20-25% far3rsra
c) Cattle concentrations higher in North 40-50% 10-15%
d) More dimba land available in the North.
The underlying food consumption patterns and food production problems were,
however, very much the same throughout the area.
Security of food supply was dominant in farmers priorities.
After fieldwork it was considered valid on these grounds to treat the EPA's
2,4,7,8 as a type of farming area, despite the variations noted.
3. Background information
AES survey data was used to gain an idea of dominance in the cropping pattern
(90% of area cultivated in maize and maize mixtures) and the importance of
livestock for some farmers, local market price'data... Dry maize, beans,
Finger Millet and cassava are virtually always available on the,local markets.
Low prices for maize are in June and July 4.4..tambala/kg, high.prices in
January March 6.?. tambala/kg. Bean prices are generally high, reaching
22 tambala/kg in one market in January and ranging around 14.or 15 tambala/kg.
with lowest prices In May at 11 t/kg.
Finger millet prices are usually higher than maize prices reflecting the
profitability of the brewing business. Groundnut prices are slightly
lower than bean prices (contrasting with Admare prices) and groundnuts are
not as consistently available.
Rainfall and evaporation n data was examined to pin down the growing season,
and to look at 3o' probabilities creating uncertainty for farmers and
requiring insurance management'
Rainfall figures fr Balaka show.
1) For planting maize assume 2, 31" needed (giving soil water balances).
October gives I in 10 years
November give 5 in 10 years
Decumb-."r .rives 9 in 10 years
2) For maize tasselling assume 5" needed in area (low evaporation 120mm p.m)
December gives this 6 in 10 years .
January gives this 9 in 10 years
February gives this 7 in 10 years.
March gives this 6 in 3:0 years.
It was fascinating to be in the area in a 1 in 10 January. Farmers.:
immediately reformed for comparison to 1940 and 1922. Rainfall 1949 January
was .0" i.e. perhaps li3O Jan.
Major point it' will have biased the interpretation'
it will 4aso have biased farmer attitude to food secu-rity.
Management towards this end is likely to be reinforced.
population data, some 17,000 farm families and 75,q00 people. o Density
varies: higher in North than South., <'
4. The arming System
In their local circumstances farmers allocate. their land, labour and cash
to meet their priorities
how theydo this makes up their farming system look at this for Ntcheu.
1) The background information told ussomething about local cir4@bW4g;ces,
other important aspects forming a context within which farmers have
to operate include:
a) Relatively long distances to official markets and limited trans-
b) Land becoming scarce, more in North than South, but an apparent in-
flexibility in the tenure system perhaps aggravated by the matri-
lineal system most farmers accepting they have to work with what
c) A sharpening division between cattle owners and nonaownqes as the
increasing population density also increases grazing pressure.
d) A s-carcity of dimba land especially in the South, which forms a
valuable management asset for o$f season production of food.
e) Limited opportunities for off farm employment.
2) Farmers priorities were clear and general throughout the area; a secure
supply of staple foods and relishes with maize for Nsima and beans for
relish as dominant preferences. Production for cash as a secondary
consideration. Remember food is needed every day this 'flow' idea is
important in ma4geaentt by small farmers,..
3) Their farming system
So how did Ntcheu farmers, in their local circumstances, manage their
land, labour and cash to. achieve theirfood supply priorities, and
their cash needs. Best single reflection of the farming system is to
see the way resources are allocated and. consumption noeds met are the
agricultural year. Remember .
a.) Each planting presents a, pattern of resource, commitments land
preparation;I planting,. weeding, harvest.
b) Each harvest presents a new supply of .food or cash,"
.1 1. ,
NTGHEU: CROP PRODUCTION AND FOOD SUPPLY CALENDAR
THS OF SEASON A* AUG
jpprox. rain inches .
p2land Crop: "
Locai maize and : -
int .planted beans (most)' ----
interplanted millet (some
(North: and se
(most - -
(many on in;
crisis and us
- Use --
e .... ....
OCT NOV Dro S
.6 2.8 6.3
- O4 1!' '' -o '
---- Plant Use le
sel lPlant _
* Plant use le
' . I
P- lat .
use le waves harvest and use
- --- - plant
a fo on
Food deficit and cash
problems extent +
severity depending on year
in eer4es har
t and ug
MAR APR !MAY JUN JUL
5.2 .1.6 7
mai ze useO Harve~t SSi and.
Harvest" [a -r I .4dw..&V .e&a*
s t o r e - - - - -
ai t e :H r v e al a n
1 or42tvears 'karietits
Two sets of comments on the picture presented by the crop production and
food supply calendar above.
(4) Food security strategies
Food supply priorities are reflected 4n the set of strategies for
guaranteeing mopth to month food supply
a) Local food Maize -takes pride of pJce for resources once rain falls.
b) Maize used green allows 'early bite' ,
c) The remains of finger millet probably the traditional starch
staple. offer a dual purpose role
i) For sale for beer brewing the price information showed it to
be available through the year on local markets, a valuable
source of contingency cash.
ii) For use as Nsima during the Oqtpber-February period in years
when maize sto es are empty.
d) Sweet potato", harvested in May will be stored to October if the
maize situatcn lopks difficult
e) Cassava; giyes a flexibility to good supplies due to its ability
to wait in the ground. Two varieties'in evidence: one giving edible
roots after a year with a further years storage possible
: the second giving roots after two years with 2-3 years
further storage possible
These five strategies of management are important to meeting food
supply priorities. Two other contingency strategies were noted as
f) Paying cash for maize and beans from September to February or
shorter appending on the severity 'pf the year. Often buying maize
(hybrid)> pack from Admarc at the higher retail price.
g) Working off the farm for cash during this period.
(5) Resource Allocation Compromises
a) Preparation of lands during the dry season, very vital to the system.
The intensive labour required for peeparation can be spread over a
b) The fact that pretty well all lands are ready for planting by the
rains creates a high labour peak during November December January
for planting and for weeding and ranking All farmers reported
December January as their busiest time of the year, and most hirings
were reported at this time.
c) Farmers in need of food af thr.i critical period are forced to:work for
it thus compromiFing thUt reoxt prantirngs on their own holdings.
d) In years when adeqi.ate rvin for planting are delayed to December
(suggested about 4/10), wvork. miuo. be hectic with non-priority'crops,
particularly cotton and hybrid mti.zo, pushed into January plantings.
Husbandry standards will probably fall and weeding be delayed with an
effect on populations and fertilizer responses..
e) Significant cash eums are spent op maize and beans instead of
consumer or producer goods -- incentives are squashed and the system
may tend to stagnate due to the vicious circle..
6) General principles behind a choice of strategy for R + D
a) Development strategies and technical packages which conflict with
food security priorities and management strategies which farmers have
developed for meeting them.
i) Are difficult for the community. to accept and lose project.
ii) May exacerbate problems in poor years, deepening the crisis and
the need for contingency action from government.
Machakos Kenya Sorghuxr/Birds, Maize and food aid? every
-. Ntcbeu pigs and root crops
On the otherhand packages which reinforce reliability in food supplies
are highly acceptable, generate project goodwill and reduce problems
in poor years.
b) Development strategies and technical packages which fail to
acknowledge farmers resource limitations. In Ntcheu, labour
in December January, land for food supplies, and limited cash
sums for input purchases.
i) Are difficult to accept for the community ,' .
and loose project goodwill .
ii) May so exacerbate resource constraints as to actually
contract the system and farmers incomes.
extreme example 5 crops to be planted ,,
in the samv week 1/Sth of .present incomes.
On the other hand packages whichwseek to ,fit into the resource capacities
of local farmers are early absorbed and generate project goodwill as well
as expanding incomes.
7) Application to Ntcheu
a) Basic Strategy: Raise food crop productivity and enhance 'the major
insurance crop cassava.
Accept thattthe local maize bean intercrop should have priority in
resource allocation immediately after the start of the rains.
b) Follow-up strategy: Promote one or two.cash crops with complementary
resource requirements at establishment weeding and perhaps even harvest
where labour required is intensive for picking;, groundnuts, cotton
Details of research and development emphasis n' .... .
*, . i .
8) Food supplies -. -
a) Local knowledge and research resources
For project goodwill and rapid impact-'beans and their improved per-
formance is a key area in farmer priorities.
Particularly severe problems, with complete loss of flowers and
.** . . . . ,-.. i ,
fruit's were reported in Central and South of the Ntcheu area."
Key characteristics are the climbing nature of the existing varieties
favoured for their supply of leaves, and a red, brown, or colour
preferred for taste.
I am not a bean agronomist but a wealth of boan technology exists
with Dr..Edge at Bunda, acknowledged internationally as an authority
He believes thqt low density planting causes the local climbing
V1 variety to build up too many fruiting points and sources in the
plant cannot develop them all and flowers and fruits fall. The lack
of sources is exacerbated by leaf picking.
It is possible perhaps probable that low densities are a result
of a vicious circle in scarcity of beans and therefore the high cost
of seed. Local market prices consistently show beans the highest
priced item reaching 22 tambala per kg.,,.with a usual level 14-15
tambala per kg.
A detailed R+D proposal' requires further work,
gramme is sketched out again I am not a bean
violating some simple principleaf
.But a possible pro-
agronomist. I may be
J-. ',. .
Selection has been done for the ability to carry a. low
density (from recommcnrided 22,$4 p,0 m) 8 p m has beer.found
possible.,withopL loss of yield-but lower yields would be acceptable'-;
to farers. Teh tpportant point is to improve on'present performance
in firmers fields wtth seed productivity as a key criterion.
Available climbing selections with-red brownSpec k .4'olour should
be supplied on credit, ) .
These should be recq=oended for'interplanting with lo4j' maize at
densities adequate to give sufficient pressure to bean e source and
fruit.development. ; '
Possibly the climbing crop should be taken green and,,a second crop
Sof dwarfs planted on the sides of ridges carries through to matu~ ty
in the dry weather in April May.
an improvement in bean storage methods should be sought with useful
receptacles offered on credit by tho project.
c) A second major arm of the project aimed at securing food supplies,.
should Ie to improve the productivity of local maize push up planting
densities of local maize and induce low levels of fertiliser.use:
C(w 5-10 per acre), Pushing up mgize densities will allow. incsased
cdensitiep of climber! and enhance both ma.ze;, and bean supplies.'.',
A sho-rt.term improved paize should be sought with anti rot, pounding
and storage characteriptics of the local variety as major selection.'
The shooter term will allow an ebriier crop
both frop upland 'rnd Dimba plantings, reducing the hunger period, .at
will also pull back harvesting time opening the period for potential
cash crop harvest, but rot. resistance will be important.
Maize storage management should be tim.ruoved with Inputp supplied by
the project. .
d) A third element in the food reliability ,rog amme
Would be the supply o: improved cassava varieties and disease free
planting material (A pathologist should look at the areA)with the:
desired keeping characteristics. Populations should be increased.
Cassava should be Incated in the Je3ds well away from the village
to minimize the interaction with pig,
'- Further work should bo!.started on planting cassava late in the season -
into the maiae crop with a view to undersowing the cassava with
groundnuts the following year.
9) Cash Crops
The development of cash props should be seen as second phas in area
development ano' a guiding principle in the choice variety .
and management components should be complementar1i~with the resource
requirements of the local maize, beans and cassavagrown for food.
a) Overall strategy would be a gradual reduction;, the food crop
area as efficiency and reliability in production improved dhe to the
Urst phase of the project'. This wo4d release bqth land, and labour
for growing marketable cash crops. The beginning'pf December to the
-middie of April (including some 20 dais of repidua moisture) would
give 4 40 days of rain. However, the redi fall probabilities indicate
a delay of some 20-30 dys in planting rains in 4 o'ut of 10 years, with
food crop in activity delayed into December. Similarly,. rainfall will cut
off early i# March, in 4/1 .years and a 110-120 water requirement
profile for selqctdd a..iph crops would give a more reliable result -
particularly <4ortant in reducing the risk in outl of cash on purchased
b) The crop possibilities
i) Groundnuts, as a rel.tivoly well oatabllahed cash crop through the
area should be offered with ,Banagemeint components which remain
profitable under late December planting.
ii) A veecond cash crop; sunflower or cotton with management .
profitable-under every January planting, and complementary
to group puts for harvesting labour requirements, may be
feasible. The profitability of low 'lev0.s f purchased iputs,
particularly fertltiser, should be established by on farm trials
using achievable planting times when the season is late.
iii) For the majority of farmers cash crop input levels should be
held to Kw 15-20. Higher outlays will be too risky for them,
even with credit facilities.
It is worth re-emphasising the need for complementarity with the local
maize/bean enterprise in demand for labour for establishment, weeding and
harvesting if new techniques of m.uagement are to be widely and rapidly
S10. Other possible programme content
a) more flexibility in land tenure, especially in the re-alloction
needed to balance changes- n family needs and labour capacy.-
f b) organization of access to water and grass for stock owners
c) organized marketing for pig? ,
C) General points, no1,.specific to; tcheu .., ..::
1. I am extremely impressed with project framework, institutions
organisation and enthusiasm. If you can't do it with this oet
up yqu can't do it!
2. The coneept of BPA's is good for management and implementation but
for selecting programme content type of farming areas are aiperstive.
3."" I would see former participation as important on both. idesaot the
farmer/research/extension circle. Farmers groups .and group:.fields
after a'g excellent vehicle; for evaluating farmers problems and
priorities on one side of circle and fto testing out adopted.i '
treatments in area trials on the other. The Interaction between
farmers, researchers and extension staff at the pit of.n'f-a trials
is an extremely powerful participatory device. ..
4. Costs of this type of udigaptoa.s exercise carried out in Ntcheu are
very low and the whole approachh on the far er/reseatch side of
the circle can be complete within a two or three month period.
* '-'' '" ''*