TECHNICAL REPORT NO. 2
SOCIAL SCIENCE PILOT STUDY IN NORTHERN MALAWI--
Kaveh Abani, M.S.1
Pat Barnes-McConnell, Ph.D.2
Michigan State University
East Lansing, Michigan 48824
Bean/Cowpea CRSP.Malawi Project
"Genetic, Agronomic and Socio-Cultural
Analysis of Diversity among Bean Land-Races in Malawi"
IAbani is a Ph.D.
for the past two
computer science at MSU assigned to this project
2Barnes-McConnell is Director of the Bean/Cowpea
professor at Michigan State University. She is
science collaborator on this project.
CRSP and an associate
the principal U. S. social
SOCIAL SCIENCE PILOT STUDY IN NORTHERN MALAWI--
Kaveh Abani, M.S.
Pat Barnes-McConnell, Ph.D.
This is an initial report of the outcome of the social science pilot study
conducted among limited resource farm families in northern Malawi in September
of 1982 as part of a major international collaborative research project. The
project entitled Genetic, Agronomic and Socio-cultural Analysis of Diversity
Among Bean Land-Races in Malawi is directed by Dr. M. W. Adams (MSU) and Dr.
Todo Edje (Bunda College). It is one of eighteen projects which presently make
up the Bean/-Cowpea Collaborative Research Support Program (CRSP) funded under
Title XII of the U.S. Foreign Assistance Act through the Agency for Interna-
tional Development (AID). The legislation, entitled "Famine Prevention and
Freedom from Hunger," encourages U.S. universities to work with AID in support
of research carried out jointly with developing country research institutions.
Such efforts address specific constraints to increased food production and
consumption. Beans, when available, are frequently the major source of protein
and some vitamins in the diets of many persons in the developing world and thus
bean production/consumption research is a significant activity under this
The Malawi project concentrates on evolving a methodology which allows both
production agriculture and social science perspectives, firmly anchored in the
Host Country setting, to contribute to an understanding of small farm family
subsistence. These two perspectives are developed in the research design by
the appropriate professionals with specific attention to complementarity and
comparability. The social science component has evolved to give attention to
the family farming systems context in understanding the maintenance of the wide
bean diversity which exists in the country. Of particular interest is under-
standing the relationship of such bean diversity to family health and welfare.
In March of 1982 during the harvest season the principal agricultural and
social science researchers spent three weeks in northern Malawi on a site
selection trip. In a Land Rover or on foot, the team of scientists travelled
through the mountains and across the plains of this area stopping at both
remote and accessible homesteads wnere beans were seen to be drying, or where
there were other indications that the family likely had beans (i.e., appro-
priate elevation for bean growing, mature bean fields visible in the area,
etc.). Samples were taken from each of the bean collections available within
each family. Every collection was assigned a code number, the first one of
which became the dominant label (Code #) by which that family was known. This
code number was written at the designated spot on all documents and forms
related to that family in the subsequent research. The scientists took
extensive field notes, describing each family and farming system, noting grid
references on the topological road maps of the area and recording all other
geographical, and political designations which would help in finding that
family again later. The very helpful extension agents, who were familiar with
the often obscure terrain, frequently travelled with the team and assisted in
making the appropriate notations.
The next stage was a more in-depth pilot socio-cultural study of the chosen
farm families in that area which would contribute to an understanding of the
constraints to bean production and consumption and an understanding of the
great bean diversity that exists there. This stage took place in August and
September of 1982.
The methodology for this pilot study was reported in Technical Report No. 1
which included the research training procedures, the pre-pilot training study,
and the research instruments used. Following a training period, the team of
nine female students from Bunda College of Agriculture, University of Malawi,
travelled to the north to carry out the pilot study. A tenth, who was trained,
had to drop out early in this study; the case loads of the others were adjusted
accordingly. These young women, whose homes were in the north, spoke the
language of that area and were quite adept at making their way around the
This pilot study was carried out in homes selected from among those where
bean and soil samples had been taken by the primary research team on the
earlier germplasm collecting and site selection trip some months before.
Findings presented in this report are the preliminary results of that pilot
work and come both from family reports to questionnaires administered directly
by the nine enumerators and simultaneously from timed observations recorded by
them during the day over a five-day period.
Twenty-five families, made up of 168 people, comprised the research popu-
lation of this study. These families consisted of 82 males and 86 females
including the children living at the home site at the time of the study, 45
male children and 45 female children. Some of the children were adults. The
age distribution of the family members by gender is reported below (Table 1).
*Dr. Julia Miller from Virginia State University was the field supervisor for
this effort. The young women from Malawi participating as recorders in this
project were Wanangwa Banda, Kupingani Gondwe, Lonely Gondwe, Filly Kamanga,
Caroline Mhango, Ndindasi Mkandawire, Estere Moyo, Emmie Nyirenda, Flora
Zulu. Augustine Mwamukamgama, a former Bunda student who had accompanied the
team in March and was a technical officer on this trip, was the interpreter
for the field supervisor and the liaison with local extension agents.
Table 1. Age and Family Member by Gender
(Years) Male Female Total
0-9 28 24 52
10-19 17 21 38
20-34 22 24 46
35 + 15 17 32
Total 82 86 168
Bean/Cowpea CRSP, Malawi Project, 1982
Family, for the purposes of this study, was defined as all persons living
in the same dwelling and, in some cases, included hired laborers who lived in
but were not a part of the immediate family. However, since they contributed
to both production and consumption, they were counted as part of the dwelling's
Although most families could be described as limited resource families,
there were clear status differences among them. Status of a family was deter-
mined based on family resources:
1. One or more persons working off the farm.
2. Possession of income-generating animals above medium number.
3. Existence of a corrugated iron roof and/or cement floor.
4. Average adult years of schooling 8 years or above.
There were 9 least-limited resource families (richest) having two or more
of the above resources, 8 average-limited resource families having one of the
above resources, and 8 most-limited (poorest) resource families having none of
In this study beans were frequently discussed in terms of pounds. These
are converted'on the basis of 4 cups to the pound and 600 cups or 150 pounds
to tne acre (amount observed from harvested acres intercropped with maize).
Beans were found in an assortment of states and when observed were recorded by
lots. A lot was any relatively stable grouping of beans--for example, stored
in a container such as a basin, drying loose, stacked in a pile, or growing
together in a field. Sub-groupings of these lots were observed over the
five-day observation period carrying the primary lot number--for example,
twenty cups of a primary lot in a clay pot soaking, cooking or being eaten, or
wrapped in a scarf being taken to market to be sold. A lot could be made up
of one seed type or many.
Among the families in tnis study, all had beans at the time of this
research. Most of the families, given the average rate of consumption, had
produced and stored enough beans to last until the next harvest. The seed type
reported to have out-performed the others was the large red kidney seed type
(Saaba 23% and Marikala 7%). The families indicated a preference for these
large (72%) and red (53%) beans although they were able to name a total of 32
others they grew at the same time. One respondent indicated that the large red
ones usually produce even when others do not. Green (17%) and yellow (10%)
were also mentioned as preferred colors.
Ninety-two percent of the families indicated their intent to plant from
their own seed stock. However, when they have had to eat their planting seed,
40% said they would buy from the government (ADMARC) while 30% would acquire
their seed from friends, neighbors, or relatives. Thirty percent indicated
they never eat their planting seed.
At the next planting season, 21% indicated they would first plant Saaba
(large, red kidney) and 12% would first plant Nyauzembe (round, green).
Twenty-one other varieties were listed by the remaining families as bean types
they would plant first at the next planting season.
Eighty-eight percent of the families plant other crops with beans (maize
67%). Of those that intercrop maize and beans, 68% plant beans and maize in
the same planting hole. Forty-eight percent of the families use some kind of
fertilizer (of these 46% used 20-20-0, 36% used sulfate of ammonia, 13% used
calcium ammonimum nitrate, and 5% used urea).
Crops per year ranged from one to three with 80% having two crops per year.
The last total crop harvested reportedly yielded for the family, on the
average, approximately 190 pounds of beans. At the homesites, however, 5 of
the families had less than 50 pounds of beans in stock, 13 families had 50-200
pounds, while 6 families had 200-600 pounds. One family, defined as an
average-limited resource family, was unique with an estimated 1800 pounds of
beans at the time of the study. This family reported having had approximately
11 acres (which could mean 11 different plots) in beans growing at that time
plus additional beans in storage from the last harvest. While this family
reported never buying beans, they do sell them. Among the families of the
study the average stock was 217 pounds of beans. The average storehouse of
beans was just over 142 pounds when this one family was removed from the
Of the 25 families, 18 had 4 or fewer lots (groups, batches) of beans and
7 families had 5 or more lots. In addition, among their lots 5 families were
observed to have a total of 5 or fewer seed types, 5 had a total of 6-14 seed
types, and 15 had a total number of seed types in excess of 15. Because it is
likely that all seed types in a lot were not seen (dark storeroom, not every
pod opened), it is likely that there was even greater variability in the beans
in all the homes than indicated here.
Within a given household, lots were found stored as only single seed types
(in 7 homes), mixed seed types (in 12 homes), and mixed and single seed types
(in 6 homes). On the average, those with the smaller numbers of lots tended
to have them as mixed lots and those lots tended to have greater numbers of
seed types in their mixture than those from homes with a larger number of lots.
Those with the greater number of lots, on the other hand, tended to have them
as single seed types. This suggested the families might be separating the
mixed seed types into lots of single seed types after harvest, an economically
advantageous move for the families, the vast majority of whom do sell beans at
some time. At this time lots of monocolored beans brought a higher return
than mixed in the marketplace and from the government.
In addition, the households with the greatest amount of beans tended to
have them in single seed type lots, although most of them also had a few mixed
lots with a range of seed types.
For the 80% of the families without a third crop from a dimba garden
(residual moisture or bottom land), the average of 142 pounds would have to
last until the next harvest (approximately late February). Assuming 22 weeks
to go, with average family size of 7 eating beans 3 times a week at the rate
of 1/2 cup per person per meal, the amount required would be approximately 231
cups or 58 pounds. Nine of the 25 families were observed to have less than 58
pounds of beans, 90% of those reporting having only two crops per year
(apparently no dimba garden).
Although those families with a dimba crop were just harvesting or still
had beans in tne field, one-fourth of all the families reported, in answer to
a direct question, that they did not harvest enough beans at their last
harvest to last them until the next crop was in. These twenty-five percent of
the families reporting too limited a yield did not overlap completely with or
were not completely within the 36% appearing to have too few beans on the
oasis of observed stored amounts. Nevertheless, 100% of both groups had only
two crops a year, the extent of insufficiency likely being influenced by
family size. The major problem of the last crop was reported to have been too
much rain (30% of the families) which affected fruiting and made weeding,
harvesting, and drying difficult. Additionally, insect problems (20%) were
cited as well as too little or too late rains (20%). Ninety-six percent of
the families indicated their desire to have grown more beans, but 85% of these
reported they did not have enough land to do so, which undoubtedly, for most,
meant dimba garden land. Time and labor were also given as important
constraints to growing more beans.
Bean production is a family affair although some of the families hire
laborers to help when needed. Of the 25 families, 9 reported that this year
they have hired laborers for weeding, 6 hired them for plowing, and 4 hired
them for making ridges. Another 4 hired for harvesting maize. At this point
it is unclear about the ages of the laborers or how much in-kind payment may
be involved. For example, in the case of weeding, older children may be
involved or, in the case of.ridge making, neighbors may exchange services.
Among the family members, gender and age roles supported the production
cycle. Bean production among the families of this study was the major
responsibility of the women although all family members, including the male
and female children and live-in laborers, had a role to play (Table 2). They
report that, for the most part, the land was cleared and the ridging was done
by the mother and father together. The planting and weeding were reported to
be the work of the mother who also took care of watering, harvesting,
threshing, putting to dry, and storing. Children were reported to assist in
most of the work.
When there is no family male around to do the production work usually
contributed to by males, most of the women indicated they would hire laborers,
perhaps paying them in local beer which is commonly brewed by women. A small
percentage of the women (8%) indicated they, or they with their children,
would do all the work themselves. Even though 84% of the women indicated
agricultural field work was the hardest work for them to do, it involves
factors other than strength. Of the women' who cited the difficulty of field
work, when asked why this work was such a problem, 33% responded because it is
a lot of work (a combination of time and energy--"many processes are done"),
26% responded time was the factor ("must spend the whole day"), 26% responded
strength was the factor ("soil is hard, needs a lot of power"), and 15%
responded health was the factor ("causes backache afterwards"). While
strength is commonly considered the main constraint to women's assuming
agricultural chores commonly thought to be performed by males, time is clearly
a companion factor.
The time pressure on women is discussed in an earlier unpublished paper
from these data (Barnes-McConnell and Goduka, 1983). The analysis of the
timed observations of the family members suggested that women do up to six
things at a time (requiring an average of 48 minutes each) while men do only
one (requiring an average of 4.4 hours each, such as rethatching a roof,
Table 2. Percent of Families in Northern Malawi Reporting
Member Contributions to Bean Production (N = 25)
Family Planting Clearing into
Member Site Land Plowing Planting Watering Weeding Harvesting Drying Threshing Storage
Mother 96 72 72 88 28 96 100 96 92 100
Father 20 68 68 48 12 64 24 8 8 4
Girl 20 24 24 40 16 40 36 28 44 8
Boy 12 32 36 32 16 36 20 8 28 8
Laborer 0 12 36 4 4 44 16 0 0 0
Communal Work 0 0 4 8 0 4 1 0 0 0
Bean/Cowpea CRSP, Malawi Project, 1982
furniture making, or plowing). The phenomenon of required time is likely a
significant factor in the distribution of production responsibilities by
gender and one explanation for the resistance in the exchange of gender roles.
That is to say, with many critical but shorter chores to perform
simultaneously, women have difficulty fitting in another work activity which
demands more time than her normal complex of responsibilities. This appeared
true because agricultural work time was less than 15% of the total observed
time among adult females during this dimba harvest period.
The women's roles and constraints such as time and energy hold important
implications for production research. The bean varieties preferred for
growing under residual moisture or for use in bottom lands have reported
problems which research in plant architecture may address ("the gardens are
near streams where there are reeds"--"it's so difficult to remove them from
tne reeds"). Further, among the drought-resistant plants, lines bred with
long, strong root systems may be unacceptable because of the time and energy
required to pull them at harvest. Over 90% of the women interviewed reported
that when harvesting they pull the entire plant, especially when the beans are
dry (which is also when the soil is likely to be dry). Thirty-one percent
indicated they pull the whole plant because of the ease in transporting the
harvest when it is tied up in bundles. Additional reasons cited were
(1) reduced labor, (2) to get the plant to burn for the ash which will be used
in bean seed storage, (3) easier to harvest whole plant than removing pods
when dry--saves time.
During this relatively slack period in the agricultural production cycle
at the time of this study, percent time spent in agricultural tasks for women
and men 20 and over were 12% and 9% respectively (Table 3). The interaction
of time, labor, production, and ecology demands in the maintenance of genetic
diversity will be explored more systematically in the subsequent longer-term
Table 3. Percent Time Spent in Categories
of Activities by Age and Gender
._ .. AAges 2
Tasks F M F M
Agriculture 0.9 3.2 1.7 7.1
Childcare 0.0 2.3 1.0 .4
Domestic 1.7 3.0 9.6 4.0
Economic 0.0 0.5 0.1 1.3
Personal 48.3 48.0 35.4 27.8
Social 49.1 43.0 52.3 59.4
100.0 100.0 100.1 100.0
Bean/Cowpea CRSP, Malawi Project, 1982
20 and Over
Within the last year, seventeen of the twenty-five families sold beans but
only four bought them. Only two reported they never sell beans, while nine
indicated they never buy them. The beans sold included twenty different seed
types including some mixed lots, the predominant beans being sugar beans
(16%), Saaba--large, red kidney (13%), and mixtures (10%). Those buying
purchased varying lots of nine single seed types and one mixture made up of
sixty seed types. The predominant ones bought were Saaba (24%), sugar beans
(24%), and Nyauzembe (12%). The average number of beans purchased was
approximately 25 pounds, although tne range was from less than 1 pound to 135
pounds. Of the 25 families, 48% reported buying beans for seed, 8% buying them
to eat, and 4% buying them for both purposes. Forty-four percent indicated
that beans were expensive at the time of:the study, costing on the average lOt
per cup (1/4 of a pound; 10t = 84 US).
The women interviewed indicated that they earn money by selling maize
(24%), selling vegetables other than beans (20%), selling beans (15%), and
selling fruit (11%). Other income-generating activities include brewing beer,
operating a grinding mill, and, one woman indicated, selling cattle and sheep.
Sixty percent of the families had at least one person working off the farm,
thirty-six percent had at least two people, and twelve percent had at least
three people doing so. Of the people employed off the farm, nine were female
and nineteen were male. In one family there were five people (three daughters
and two sons) engaged in off-farm employment.
Beans require very little in the way of financial investment. The most
frequent item purchased for bean production was fertilizer (16%), although
several indicated the fertilizer was really put on the maize and the inter-
cropped beans use it as well. Other items mentioned were seed, hoes and hoe
handles, and laborers. The most prevalent item purchased for bean storage was
chemicals (40%) with DDT, sevin, and actellic mentioned. Other purchases
include clay pots, tins, and baskets, all for storage. Of the 25 families, 12
reported purchasing salt and oil for cooking beans with vegetables also
mentioned, such as onions (8), tomatoes (8), and groundnuts (2) as well.
Cooking pots, mostly clay but one person mentioned metal, were also purchased
While this study did not include a systematic assessment of general family
finances, it was clear that bean production and utilization made limited
financial demands on the farm family system at the time of this investigation.
Research which seeks to increase the availability of beans through introduction
of new inputs will have to seriously consider this issue in the context of the
limited resource family and the buy/sell/consumption pattern.
The families in this study reported an average of approximately ten pounds
of beans lost in the field and another ten pounds lost in post-harvest
handling. In the field the most serious problem reported by the families (60%)
was beans' spoiling through germination and rotting from heavy rains or morning
dew. Also mentioned were problems of insects (9%) and animals (rats, baboons,
chickens, birds) eating the seed. Before, during, and after harvest, pod
shattering was mentioned as a serious source of loss (56%). Responses also
included "not all collected," "lost in carrying home," and "some left from
threshing." Some noted that seeds broke during handling and threshing which
was another source of loss.
Over 80% of the families spread the beans on the ground to dry in the sun
although several also mentioned hanging the tied plants with pods at one stage
in the drying process. During the dry season harvest (July/August) the beans
were reported to be completely dry when harvested or they will take only a day
or two more to complete drying. During the raining season harvest (February/-
March) they are reported to take longer to dry, up to as much as a week
depending on the amount of sun. While still in the pod, the beans may be hung
in the shade or in mud barns. After shelling, most of the families stored the
beans in clay pots and used fertilizer bags or sacks. They also mentioned
gourds and baskets as likely receptacles for the beans.
Sixty-eight percent of the families reported using chemicals in storing the
beans (DDT, actellic, sevin). Nearly as many reported using ash from the
burned bean plants or from the cooking area. A few indicated preferring this
latter method for their eating beans and the former for their planting seed.
Also mentioned as additives for storage were whole groundnuts and sand.
On the average, it was reported that beans were stored up to six to seven
months before they were completely consumed. Fifty-four percent of the
families reported insect damage occurred during this time when some also rot
or are eaten by rats. The beans reported to be the most susceptible to insects
by these families were sugar beans (22%). The beans reported to be the most
susceptible to becoming hard in storage were the small round whites and
Nyauzembe--the small, round greens.
These problems of hard seed coat and susceptibility to insects in storage
are widespread and have important implications for production science
research. This project will be alert to varieties which suggest superior
genetic potential for addressing these problems.
The families in this study reported cooking beans an average of twice a
week, eating them an average of three times. When questioned, the women
indicated they prepared an average of four cups of dry beans when they cooked
them for their families (average size of seven), making an average of 1/2 cup
of dry beans per person. The observation data concurred with this finding,
demonstrating an average number of beans observed cooking at 4.4 cups with a
range observed of 2 to 8 cups. From these observation data, an average of .35
cups of dry beans per person per meal is calculated with a range of .21 to 1.6
cups (adjusted for family size). The families rinse their beans (96%) one to
five times (average twice) before cooKing depending on whether or not they had
been treated with a chemical. Only 38% reported soaking their beans before
The nearest source of the water necessary for bean preparation comes from
a well (44%), underground spring (20%), piped system (16%), river or stream
(16%), or bore hole (4%). Females are responsible for transporting this water
approximately five times a day which, on the average, was 238 feet from the
door of the homestead (range 21-639 feet). Nothing was done to purify this
water for 60% of the respondents beyond its boiling in normal use during
Traditionally served as a companion to nsima (dough-like food from maize
flour), the respondents indicated vegetables--such as tomatoes, onions,
groundnuts--were sometimes added to the beans during cooking. Cooking oil was
another frequent addition. Other items mentioned included sweet potatoes,
Irisn potatoes, bananas, flying ants (termites), cassava, pumpkins, pepper,
and pigeon peas. Most of the bean dishes were variations on boiled beans,
with some persons reporting liking mashed boiled beans and others liking their
boiled beans to remain intact. Several indicated they remove the seed coat.
Other bean foods eaten were fresh beans cooked in the pod (cooked pods
sometimes eaten, sometimes discarded) or boiled with the leaf as vegetable.
Twenty percent of the respondents reported that the large red kidney bean
kept the best for more than a day or so after cooking without becoming sour.
On the other hand, over 26% agreed that the small round whites did not keep at
all well after cooking.
When the women were asked about cooking time, the responses suggested that
the beans requiring the longest time were sugar beans (13%). They also men-
tioned the small round whites and the black beans.
The major reasons given for producing the difficult beans were to sell
(28%), to get a high yield (14%), they taste good (17%), and sometimes they
just turn up in the mixture and are hard to separate (10%).
The observation data were analyzed to see if, during cooking time, there
was any interaction among temperature (average cooking temperature 190.3"F),
time (average cooking time four hours), and elevation (average elevation 4753
feet). These data showed that there was no significant interaction among
temperature, elevation, and cooking time. The highest correlation was .44
between elevation and temperature. Since most of the cooking was done outside
over a three-stone wood fire, the variation throughout the cooking time was
considerable as the fire would sometimes go out before more wood was added.
Temperatures were recorded as low as 166.250F and as high as 200.660F. For
this reason, it was felt that cooking time and temperature studies would have
to be done some place other than in the field where assured constant
temperatures as well as controlled variability in temperatures can document
CONSUMPTION AND HEALTH
Families in this study reported they ate beans three times a week at the
time of this study, more often as mixed beans (52%) rather than separated
(32%). The favorite beans for eating were Nyauzembe--the round greens (18%),
the large red kidney beans (12%), and sugar beans (12%). The two
characteristics mentioned as most important in choosing these beans were good
flavor (38%) and cooking fast (19%). When asked why they generally preferred
to have beans which were red (green second most popular color), the
respondents cited two things as important. They liked the thick, red (or
perhaps brown) gravy wnich made an attractive relish and pleasing soup. They
also indicated that these red beans usually produced even at times wnen, for
various reasons, the other varieties did not.
Saaba (large red kidneys) and sugar beans were the beans most often given
to children whose diet included nsima as the staple and sweet potato and
banana. There were reported to be no differences in the diet of male and
female children. On the average, children began eating beans (or at least bean
soup) at thirteen months old (range 2-24 months). Twenty-four percent indi-
cated that there were various forms of digestive problems when their babies
were first given beans. Of this group, the age range when their children were
first fed beans or bean soup was 5-24 months.
Among the adults, it was reported that 12% have stomach problems from
eating beans. Some of these bean-related problems, for both adults and
children, may be exaggerated by malabsorption of other origins such as
parasites and various infections. These latter problems were reported at all
ages although the most numerous and most serious were associated with children.
General health of the family members was observed on each day of the 5-day
observation period. Of the persons whose general health was observed by the
researchers, the females (32%) demonstrated generally twice as many days of
health problems as males (17%), with the dominant symptoms being runny nose
and cough. The age group distribution is shown in Table 4.
Table 4. Percent Person Days of Health Problems by Gender and Age
(168 Persons in Observation Periods of 5 Days Each)
Gender 0-4 5-9 10-19 20 and above
Male 50% 19% 5% 10%
Female 67% 43% 25% 21%
Bean/Cowpea CRSP, Malawi Project, 1982
In light of the other data, it was somewhat difficult to interpret the
very low figure recorded for males 10-19 years old. It may be that this
particular category was confounded by the fact that the surveyors were all
female around the same age as the older boys in this age group. The
relatively small number of persons also contributes to the percent
variability. Nonetheless it is quite clear that females of all ages generally
had more health problems than their male peers. In addition, among the 25
families the women reported having had a total of 166 children, 26% of whom
were deceased--all before the end of their fifth year (miscarriages included).
Although the data reported here are from a pilot study carried out in only
25 homes (168 people) surveyed for a total of 5 days each, they are most
suggestive, especially when the agricultural reports from this project are
All of the families reported that women have the major responsibility for
bean production which together with their other agricultural work was computed
from the observations to make up only 12% of their total time allocations.
Observed to be engaged in up to six things at a time and demonstrating twice
as many health problems as males, the women report that agricultural field work
is among the hardest work they have to do. The women's tenuous health status
is further demonstrated by the high rate of miscarriages and infant deaths
reported among these limited resource families. Further, the observation data
suggest that frequently work overload unmanageable by the women is passed to
the older female children, a group whose nutritional needs for growth and for
preparation for upcoming reproductive years are particularly acute. Thus, the
labor force on whom bean production is highly dependent is heavily constrained
by questionable health and limited time and energy. A vicious cycle
perpetuates itself as the bean protein, vitamins and calories which would be
so helpful in raising the level of health and energy of this labor force is in
limited supply because of the limited health, time, and energy available to
put into their production.
What is the relevance of this tight production system to the maintenance
of the great diversity of beans, the major focus of this project? Concurrent
studies being conducted by the agricultural team on this project have demon-
strated the genetic diversity that exists in the favored beans of this region:
the whites, the red kidneys, the round greens, the mottled sugar beans, and the
yellows or tans. Several experiments have been run to date in which the first
generation hybrid from a cross of two of the above resulted in second and sub-
sequent generations producing a plethora of different seed types, in some cases
more than eighty. In the 1983 Annual Report, this project reports:
From several standpoints one gets the impression that particular kinds
of beans (Saabas, sugar beans, Nyauzembes, etc.), not mixtures, are
preferred. Many families maintain two or more of these particular
kinds. When planted in adjacent fields, natural crosses will occur
between them, resulting in the development of 'hybrid' types within
the initially homogeneous stocks. Unless these 'hybrid' types are
continually removed by hand sorting, the homogeneous character of the
basic stock will be lost and in time, a complex mixture will replace
There is economic advantage to separating the seed types because the
preferred varieties usually bring a better price when sold separately.
However, this is yet another task which the female producer may not feel is
worth the effort on any real systematic basis. While the data suggest that
there may be considerable sorting after harvest, we have not yet determined
what the farmer does with her non-preferred types and thus how much they are
around to re-establish themselves in the system.
This issue poses a serious dilemma for the breeders. If they were to
concentrate on increasing the yield potential of the 3 to 5 preferred seed
types, they would simultaneously have to address the family farming system
components which would support the integrity of those varieties. Otherwise in
a few short seasons the new seed might well be lost and its components
reclaimed back into the vast diverse genetic pool, an occurrence easily
facilitated through the use by these limited resource farm women of their own
saved seed stock.
Conversely, even though there appears to be more of a preference for
identified single types over the mixtures, all of the families had an array of
seed types in their possession. One cannot discount the potential importance
of maintaining this complex genetic pool. Interventions will have to address
this issue which is as crucial over the long term for the breeders as for the
These ideas suggested by the pilot data are yet to be confirmed by the
longer-term study which should also identify critical points in the family
farming system where there may be a few degrees of freedom. A more thorough
analysis must await those data to be gathered in this same region of Malawi
across several growing cycles over the course of a full year.
As pointed out by the External Reviewer of this project, the bridge between
agricultural and social sciences must be painstakingly evolved in a way that
the latter can articulate significant and important insights useful to the
S. the project is in fact breaking ground which is new in detail;
and to develop multidisciplinary cooperation into interdisciplinary
thought and action for all participants takes time and is seldom
The exceptional cooperation and esprit among the multidisciplinary U. S.
and Host Country participants of this project and the respective Governments
promise the success of the team in analyzing the research problem and utilizing
this understanding to make a positive impact on bean production and consumption
in Malawi, and perhaps, beyond.
Barnes-McConnell, P., and Goduka, I., "The Acquisition of Sex Role Behavior in
Subsistence Farm Families in Northern Malawi." Research Display Presenta-
tion. Annual Meeting of Society for Research in Child Development, 1983.