SECTION V. Economics & Post-Harvest

SECTION V

Economics & Post-Harvest

364

365

A study on the adoption of improved maize technologies in northern Ghana K.O. Gyasi1 , L.N. Abatania1 , T. Paulinus 1 , M.S. Abdulai1 and A.S. Langyintuo1,2 1

Savanna Agricultural Research Institute (SARI), Box 52, Tamale, Ghana 1,2 Presently: Purdue University, Lafayette, Indiana, USA. Abstract A Tobit regression model was used to study the adoption of improved maize (Zea mays L.) technologies among 264 maize farmers in northern Ghana. The objectives of the study were to (1) analyze the factors influencing farmers’ decisions to adopt improved maize technologies, (2) measure the rate and intensity of adoption, and (3) examine the pattern and spread of adoption. In 1999, an estimated 39% of the total maize area in northern Ghana was devoted to improved varieties. Okomasa, Abeleehe, and Obatanpa are the popular maize varieties cultivated in the study area. The proportion of farmers growing an improved maize variety increased from 19.6% in 1988 to 75.4% in 1999. Important factors influencing adoption among farmers include farmer’s knowledge, extension education and farmer-to-farmer exchange of seed and new knowledge. The regression results indicated that adoption of improved maize technologies was significantly influenced by farm size, family labor availability, farming experience, yield potential of the variety, tolerance/resistance to diseases, maturity period, and grain quality. The results suggest that the socioeconomic conditions of farmers (farmer-specific characteristics) and technology specific characteristics of the varieties are important in adoption of improved maize technologies. The results underscore the importance of education of producers (adult males, adult females and young farmers) in facilitating rapid adoption of improved technologies. Résumé Le modèle Tobit a été utilisé pour analyser des données obtenues de 264 cultivateurs de maïs dans le nord du Ghana. Les objectifs de l’étude sont : (1) analyser les facteurs influençant la décision d’adopter les technologies améliorées du maïs, (2) la mesure du taux et de l’intensité d'adoption, et (3) examiner le modèle et l’étendu de diffusion de l'adoption de variétés de maïs améliorées dans le Nord du Ghana. Quelque 39% de la superficie totale de maïs dans le nord du Ghana était consacré aux variétés améliorées en 1999. Okomasa et Obatanpa sont les variétés populaires de maïs cultivées dans la

366

zone d’étude. La proportion de cultivateurs produisant n’importe quelle variété améliorée de maïs a significativement augmenté de 19,6% en 1988 à 75,4% en 1999. Les facteurs importants qui influencent l’adoption entre producteurs incluent : le niveau de connaissance du producteur, l’éducation sur la vulgarisation et les échanges de semence et de connaissance d’agriculteur à agriculteur. Les résultats de la régression indiquent que l'adoption de technologies améliorées du maïs était significativement influencée par la taille de la ferme, la main d’œuvre familiale, l'expérience en matière de méthodes de culture, le potentiel de rendement de la variété, la tolérance / résistance aux maladies, la période de maturité et la qualité des grains. Les résultats suggèrent l'importance des conditions socio-économiques (spécifique à l’agriculteur) et la technologie spécifique requise par la variété dans la prise de décision d'adoption. Les résultats de l'étude ont aussi montré l'importance du niveau d'instruction du producteur (qui inclut équitablement femmes et jeunes agriculteurs) pour faciliter l'adoption rapide de technologies améliorées.

Introduction Increasing agricultural productivity and production with improved agricultural technologies is a precondition for achieving food security in Ghana without food aid. As long as farmers continue to use traditional low yielding crop varieties, the vision to achieve agricultural growth rate of 4% as stipulated in the country’s Vision 2020 (Republic of Ghana 1995) will be a mere illusion. Efforts have therefore been made by the Savanna Agricultural Research Institute (SARI) and the Crops Research Institute (CRI), in collaboration with the International Institute of tropical Agriculture (IITA), and other international research institutes to develop improved technologies for use by farmers. Maize is one of the crops that has received priority research attention (Morris et al. 1999) possibly because it is the most important cereal grown on more than 500 000 ha (MoFA 1998) and consumed nationwide (Gyasi 2001; Alderman and Hingis, 1992; Boateng et al. 1990). Several improved maize te chnologies have been developed and extended to farmers. Among them are the following improved varieties: Dobidi, Aburotia, Okomasa, Abeleehe, Obatanpa, Mamaba, and Dorke. The superiority of these varieties over the traditional ones in terms of grain yield were confirmed in on-farm tests before extending them to farmers together with the complementary crop and soil management practices. There has not been any follow up in northern Ghana to investigate the extent to which farmers have adopted the varieties. There are non -quantified indications that farmers have not adopted the varieties to any appreciable extent, at least in some parts of northern Ghana. In the Upper East Region, for example, farmers

367

still prefer to plant early pearl millet that gives them low yields, rather than maize that would produce reasonably higher yield with minimum soil management practices. It is unclear the extent to which these varieties have been adopted by farmers in northern Ghana in general. This suggests that despite the superiority of improved varieties over traditional ones, widespread adoption and efficient use cannot be guaranteed. There is the need to look beyond the yield advantage of these improved varieties to identify what other factors are important in motivating farmers to adopt a particular crop variety. Various institutional, economic, psychological and social factors are known to be important in determining the adoption of improved technologies (Adesina and Zinnah 1993). This study was therefore motivated by the apparent lack of information on the level of adoption or non -adoption of improved maize varieties extended to farmers in northern Ghana. The primary objective of the research was to examine and evaluate the factors influencing the adoption of maize technologies in smallholder farming systems in northern Ghana. More specifically, the study was designed to: i. ii iii. iv

determine the factors influencing the adoption or nonadoption of improved maize varieties in northern Ghana, measure the rate and intensity of adoption of improved maize varieties in the study area, examine the pattern and spread of adoption of the improved maize varieties, and identify the constraints to the adoption of the improved maize varieties.

Materials and Methods Sources of data Field survey data were collected from a total of 261 farmers in 12 districts in northern Ghana between October and December 1999. Northern Ghana comprises of Northern, Upper East, and Upper West regions. Since maize production is equally important in all regions, four districts were randomly selected in each region. In each district, five villages were randomly selected from among villages that previously had experience with improved maize varieties through the Ministry of Food and Agriculture or CRI/SARI. The 261 farmers (maximum of five per village) randomly selected were interviewed using structured questionnaire covering the following issues: (i) socioeconomic and institutional variables, (ii) technology-specific variables includi ng a comparison of the best local varieties to improved varieties, in terms of yield, pest resistance, etc., and (iii) psychological factors such as social participation, market perception, and opinion. Vertical and horizontal effects of these factors on farmers’

368

decision to adopt or not to adopt an improved maize technology were assessed.

Data analysis Data were analysed using appropriate descriptive statistics to estimate the rates and dynamics of adoption. A Tobit model was used to investigate the dete rminants of adoption of the improved maize varieties in northern Ghana. Logit curves were used to analyse the spread of the new varieties. In technology adoption studies, limited dependent variable models such as Logit, Probit and Tobit continue to have extensive applications in obtaining information from the non-normal distribution of such data (see for instance , Adesina and Zinnah 1993). The ordinary least squares regression is inappropriate when the dependent variable is discontinuous (Goldberger 1964, Feder et al. 1982; Pindyck and Rubinfield 1998). Logit and Probit models are appropriate when the dependent variable is discrete, usually taking two values, 0 or 1. These models are useful if the question is whether to adopt or not, but are not appropriate when it is important to measure the intensity of adoption of a technology. The Tobit model which better handles censored dependent variables (continuous between some lower and possibly upper bound) (Pindyck and Rubinfield 1998, Shakya and Flinn 1985) is superior to the Logit and Probit. It measures both the probability of adoption and intensity of use. In this study, the Tobit model was used to achieve the stated objectives (See Appendix A for the empirical model).

Results and Discussion Socioeconomic households

characteristics

of

the

sampled

A household in the study area consisted of a group of people living in the same compound, storing products together and eating from the same pot. Consequently, a household in the study area encompasses the extended family system. It was not surprising to observe family sizes ranging from 1 to 57 members, with a mean of 13. Because a bulk of the farm labor force is provided by family members in the study area, large family sizes remain an asset to the farmer. The ages of the respondents ranged from 21 to 69 years, with a mean of 43 years. About 65% of the respondents were 41 years or above. The main characteristics of the sampled households show large gender disparity in maize cultivation in the study area (Table 1). Only 5% of the maize farmers interviewed were women. Low participation of women in maize cultivation in northern Ghana is probably linked to traditional gender roles. At the household level, men are primarily responsible for the production of basic staples.

369

Women play a supporting role. However, women who are single parents (mostly widowed) without male children produce the basic staples like the men to support the household. Table 1. Main characteristics of sampled households.

Characteristic Sex % Male (%) Female (%) Age (years) Family size (No.) Level of education Illiterate (%) Nonformal (%) Formal (%) Plot/farm size (acres) Household heads with sources of off-farm income (%) Off-farm income No off-farm income Households with other members having off-farm income Member with off-farm income N o member with off-farm-income Primary purpose for maize cultivation (%) Mainly for food Mainly for cash For both food and cash

Frequency or Mean

% or std dev.

248 13 43.4 13.1

95 5 9.7 8.4

174 25 62 4.4

66.7 9.6 23.8 3.6

128 133

49.0 51.0

207 54

79.5 20.5

63 6 172

24.1 2.2 73.6

Farmlands in the study area are communally owned and vested in the chief. However, families control the land they have been farming on which are bequeathed to generations of the family. The power to dispose of the land (permanently or by lease) is vested only in the chief. Male farmers have permanent custody to lands but females only have temporary custody of land for the cultivation of cash crops of their choice. However, females who are widowed and remain in their matrimonial home may have permanent custody to land, provided there is no male adult child that would assume headship of the family. Most of the respondents (81%) cultivate maize on family lands while 13% obtain land from other sources including friends and land owners who normally take a portion of the maize harvest as rent for the land. Another 4% farm on community lands. Maize farm sizes in the study area range from 0.2 ha to about 10 ha, with mean farm size of about 5.2 ha (Table 1). The mean farm size for women maize farmers is about 0.6 ha. Farming is still the major occupation of the people of northern Ghana. However, 49% of the respondents are engaged in nonfarm activities to occupy themselves during the lean farming season and to raise extra income to support the household (Table

370

1). For those without off-farm income, 25% have household members with off-farm income source. Regular remittances to this group can greatly influence adoption decisions. The major non -farm activities are artisanship, craftsmanship, and petty trading. Few are wage earners. The impact of such social characteristics on the farmers’ decision to use improved technology is examined in subsequent sections. Maize is produced mainly for household consumption but it is common to find farm households cultivating maize for cash income to buy consumer goods and meet other cash responsibilitie s (Table 1).

Maize cropping system in northern Ghana. Two main types of farms exist in the study area: the compound farm and the distant (bush) farm. The compound farm is located around the homestead and is usually cultivated permanently. Compound farms are usually enriched by farmyard manure and therefore are reserved for the production of staple cereals including maize. The distant (bush) farm, on the other hand, can lie between 5 and 25 km away from the village. Its fertility is sometimes improved by bush fallowing for a few years. It is usually cultivated to all crops including maize. In the study area, maize is grown in pure stands (monocrop) or in mixtures. Important intercrops are sorghum, cassava, millet, groundnut, cowpea and soybean. Many of the interviewed farmers intercrop maize. Intercropping is practiced to make maximum use of the available land and other resources as well as hedge against the risk of crop failure. The household is the most important source of labor for all farm operations in the study area. Hired labor is, however, important for land preparation and weeding. Table 2. Gender responsibility and household task division in maize production (%). Task Land preparation Planting Weeding Fertilizer application Harvesting Transporting Threshing

Male mainly 89.7 12.6 77.0 36.4 5.0 24.5 14.9

Female mainly 0.4 31.1 2.3 3.4 9.0 6.5 13.4

Both sexes 10.0 36.3 20.7 20.7 85.8 69.0 71.6

Hired labor is more commonly used by large -scale and women farmers. Communal and exchange labors are largely employed during harvesting. Labor type engaged for farm operations include both male and female. Land preparation and weeding are major male operations. Table 2 presents information on gender and division of labor in maize production in the study area. Although male participation is dominant for many of the tasks, the complementary role of women in maize production is prominent.

371

Land preparation is by mechanization, bullock or hand-hoe plowing. The availability of tractors for cultivation is one of the factors influencing the decision of farmers to increase their farm size and grow more of improved maize varieties. The use of bullocks in land preparation is very important in the study area. Few of the respondents use tractors in plowing their fields. Fiftyfive percent of the respondents use bullock while 27% use handhoes to till their fields (Table 3). Table 3. Farming practices in northern Ghana as indicated by farmers in the survey. Land preparation methods Land ploughed by hand hoe Land ploughed by bullock Land ploughed by tractor Planting Methods Row (line) planting Random planting Any Soil Fertility management practices Yes No Types fertility management practice (if Yes) Chemical fertilizers Manure Agroforestry Composting Planting method vs fertility management Row (line) planting + fertility management Random planting + fertility management Frequency of Extension visits Once every week Once every two weeks Once every month Once every 3 months Occasional Not at all

Frequency 68 144 69

Percent 26.8 55.2 18.6

200 61

76.6 23.4

223 48

85.4 14.6

85 108 14 16

38.1 48.4 6.3 7.2

200 23

89.7 10.3

0 64 61 6 86 44

0.0 24.4 23.4 2.3 33.0 16.9

To help conserve soil moisture, 79% of the farmers cultivate their maize on ridges prepared either with bullock or manually by hand hoe. The survey revealed that most of the farmers (51%) practice row planting with appropriate crop spacing to facilitate the performance of other farm operations such as weeding and fertilizer application. A major constraint to maize production in northern Ghana is soil fertility (NAES, 1982/83; NAES, 1983/84; NAES, 1984/85; NAES, 1985/86; NAES, 1988; NAES, 1989; Gerner et al. 1992). Eighty-five pe rcent of the farmers interviewed practice some form of soil fertility management. In some cases a combination of organic and inorganic fertilization is practiced. Of those who practice soil fertility management, 38% apply chemical fertilizers, 7% apply com post while 48% apply farmyard manure, usually applied to the compound farms (Table 3). The inorganic fertilizers commonly used are compound fertilizers (NPK) and sulphate of amonia or urea. Majority of the farmers (58%)

372

combine the two, usually not at the recommended rates. Twentyeight percent (28%) apply NPK alone while 4% apply only urea. The study revealed that fertilizer application is associated with planting methods; majority of those who apply fertilizer plant in rows (Table 3). Most of the farmers interviewed belong to groups and associations established along commodity lines to improve their chances of benefiting from technology and credit facilities. The association members hold meetings that are periodically attended by extension agents to give some advice on farming practices in general. Besides these meetings, the extension agents normally visit individual farmers. Only 17% of the farmers claimed that extension agents never visited them on individual basis (Table 3).

Adoption of improved varieties1 The improved maize varieties Abeleehe, Aburotia, Dobidi, Mamaba, Obatanpa, and Okomasa are widely cultivated in the three northern regions of Ghana. Dorke and Golden Crystal are also cultivated but to a smaller extent (Table 4). Table 4. Major mai ze cultivars grown in Northern Ghana (%)*. Variety Major variety grown Dobidi 16.0 Aburotia 17.6 Abeleehe 20.0 Obatanpa 21.4 Safita 10.7 Dorke 11.9 Okomasa 3.4 Mamaba 1.5 Golden crystal 12.4 Local 62.4 *Multiple answers were allowed in the survey.

Preferred variety 14.2 11.1 13.0 8.4 6.4 7.7 0.0 0.0 0.0 25.0

It was difficult to get specific names for the local varieties since various communities have different names for the same varieties. Some of the local varieties are thought to be old improved varieties. All the improved varieties of maize have been proven to perform better than the traditional ones on farmers’ fields under good crop management practices. But farmers in the study area frequently mix varieties in their fields 2 and often cultivate more than one variety, including local varieties, at the same time.

1Rate

of adoption can be measure in two ways: (1) in terms of number of farmers who adopt the technology or, (2) in terms of area under the improved technology. Both measures are accurate (Morris et al. 1999). Given the multiple objectives of this study we used both measures as appropriate. 2It is thus difficult to estimate the proportion of area under each variety when farmers crop more than one variety on their field.

373

In all, 75% preferred improved varieties. Some of the major constraints to the adoption of improved maize varieties included lack of improved seeds (11.3%), high cost of complementary chemical fertilizer (21.5%), lack of cash to purchase seed (23.7%), and preference for local varieties due to their better storage and cooking qualities (9.6%). Most of the farmers in the study area obtained their first improved maize seed from extension agents, Ghana Seed Company (now defunct), neighbors (other farmers), seed dealers or at the open market. Dissemination of information on the potential of innovations is still very much crucial in the diffusion process. While extension education has played an important role in creating awareness about improved varieties in the study area, farmer-to-farmer exchange of seed (horizontal diffusion) is an important mode of spread of improved maize varieties. Table 5 shows that the proportion of farmers growing any improved maize variety at all increased from less than 20% in 1988 to about 75% in 1999. The estimated adoption rate in northern Ghana, in terms of area cultivated to improved maize varieties, was 39% in 1999. Our data suggest that this was an underestimation. Table 5. The Rates of adoption of improved varieties of maize in northern Ghana. Year 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999

Abe lehe 3.4 3.1 4.6 7.3 9.6 11.9 14.2 17.2 19.9 20.3 20.3 22.2

Aburotia 4.2 5.0 5.7 9.2 9.2 12.3 13.8 15.3 16.9 18.4 18.4 17.6

Dobidi 8.4 8.8 10.3 10.3 11.1 12.3 11.9 14.9 16.1 16.1 15.4 15.3

Mamaba 0.0 0.0 0.0 0.0 0.8 0.8 0.8 0.8 1.1 1.5 1.1 1.5

Obatanpa 0.0 0.0 0.o 3.1 4.2 5.4 9.2 11.9 14.2 16.1 17.3 17.6

Okomasa 4.6 6.1 8.8 9.6 10.7 11.9 14.2 14.2 20.3 22.6 22.2 19.9

Any imp’d variety 19.6 23.4 27.9 34.7 36.2 44.2 48.7 60.4 69.4 73.2 76.2 75.4

A natural adoption process is observed in which farmers increasingly cultivate large proportions of new varieties as they gain more confidence. This phenomenon is clearly depicted by the logistic curve of cumulative adoption levels based on farmers’ recall of adoption years presented in Figure 1. The cumulative adoption tends to follow a typical "s-shaped" curve in which there is slow initial growth in the use of the new varieties, followed by rapid increase and then slowing down as the cumulative proportion approaches the maximum in 1999.

Cum. percent

374

90 80 70 60 50 40 30 20 10 0 88 89 90 91 92 93 94 95 96 97 98 99 year

Fig.1. Cumulative adoption curve for improved maize varieties in northern Ghana.

Adoption decision model Choice of variables and hypotheses As indicated earlier, some farm - and farmer-specific factors as well as technology-specific characteristics are hypothesized to influence the adoption of innovations. Farm and household specific factors included in the adoption decision model are farmer’s age, available family labor, farm size, farming experience, and extension contact. Technology and psychological characteristics included in the model were yield potential, resistance to diseases and drought, fertilizer requirement, maturity period, grain quality and storability. Taste was the only psychological factor included in the model. It is assumed, generally, that younger farmers are more inclined to accept innovations than older ones. However, studies by Panin (1988) suggest that decisions relating to adoption of innovations are influenced positively by age. The reason being that older people are more likely to be endowed with resources (e.g. land, household labor, etc.) than younger ones. Available family labor (number of persons) is one of the important factors of production in peasant agriculture. A large family is more likely to have a large number of working members and will be more likely to accept innovations (Panin 1988, Kebede 1989). In a region like Northern Ghana where labor market is virtually nonexistent, family labor availability may influence adoption decisions positively. Farming experience reflects the number of years in

375

maize cultivation. With increasing experience, farmers may be able to assess production constraints better, and for that matter opt for appropriate innovations to address them. However, household heads addicted to “safety first” subsistence agriculture may be risk averse and less likely to accept innovations that are not time tested. Farm size (in ha) is expected to positively influence the adoption of improved varieties, as those operating large farms tend to have greater financial resources, access to innovations and more land to allocate to improved technologies (Feder and Slade 1984). Similarly, extension visits are expected to positively affect technology adoption. Farmers’ perception of the superiority of improved varieties, in terms of yield performance, disease and drought resistance, and maturity period are expected to positively relate to adoption decisions. Similarly, farmers are not likely to accept varieties with poor grain and storage problems. High external input demand is likely to influence technology adoption negatively. Adoption and use intensity of improved varieties (dependent variable) is measured as the proportion of maize farm planted with improved varieties. The dependent variable (land under improved maize varieties) is censored with values which are positive or zero. Table 6 provides descriptive statistics of variables used in the empirical model.

Empirical results Table 7 presents the maximum likelihood estimates of coefficients in the Tobit3 regression equation for improved variety adoption decision model. The model correctly predicted 87% of the variance in adoption intensity. The age of the head of the household and farm size are found to have positive and significant impact on the probability of adoption. Older farmers may be the elders of the farming communities and for that matter, resource owners (access to land, family labor and, to some extent, informal credit due to their social status). As a result they may have preferential access to new technologies through extension services. This is particularly relevant in the situations where the success of any new innovation is seen to be dependent on its acceptance by opinion/community leaders. These results seem to affirm the important role of resource endowment in observed adoption behavior (Adesina and Zinnah 1993). Certainly, farmers with large farms are more likely to have more opportunities to learn about the new varieties.

3The

Tobit model was estimated using Time Series Processor (TSP) Version 4.3

376

Table 6. Descriptive statistics of variables used in the empirical model. Variable Age Hsize. Farm size Extension visits Farming experience Yield potential Disease resistance Drought tolerance Fertilizer req’t Market value Maturi ty period Grain storability Grain quality

Description Age of farmer (years) Household size (no.) Farm size (Ha) Number of extension visits

Mean 43.4 13.1 4.4 2.8

Min. 21 1 0.2 0

Max. 69 57 5 10

No. of yrs cropping maize

12.3

1

41

Measure as binary factor: 1 if farmer thought improved varieties were superior, zero otherwise Measure as binary factor: 1 if farmer thought improved varieties were superior, zero otherwise Measure as binary factor: 1 if farmer thought improved varieties were superior, zero otherwise Measure as binary factor: 1 if farmer thought improved varieties were superior, zero otherwise Measure as binary factor: 1 if farmer thought improved varieties were superior, zero otherwise Measure as binary factor: 1 if farmer thought improved varieties were superior, zero otherwise Measure as binary factor: 1 if farmer thought improved varieties were superior, zero otherwise Meas ure as binary factor: 1 if farmer thought improved varieties were superior, zero otherwise

.81

0

1

0.45

0

1

0.50

0

1

0.56

0

1

0.62

0

1

0.65

0

1

0.55

0

1

0.9

0

1

They are also likely to have more incentives to adopt new technologies and are more able to bear risks associated with early adoption of improved technology (Abdelmagid and Hassan 1996; Feder et al 1985; Feder and Slade 1984; Feder 1980) 4. Contrary to expectation, family size (proxy for available family labor) has a negative influence on the use intensity of improved maize varieties. This suggests that farmers with small family size use improved maize seeds more intensively. 4Risk

associated with improved (maize) variety is assumed to be low since farmers can easily abandon the technology without grave consequences.

377

Table 7. Tobit model estimate for the intensity of adoption of improved maize varieties in Northern Ghana. ________________________________________________________________________ Parameter Estimated coefficient Std error t-statistic Constant -0.983 0.573 -1.715* Age 0.024 0.012 1.961** Family labour -0.061 0.014 -4.252*** Farm size 0.828 0.031 26.565*** Extension visit 0.025 0.042 0.605 Farming experience -0.051 0.013 -3.972*** Yield potential 0.671 0.417 1.691* Disease resistance 0.724 0.253 2.857*** Drought resistance 0.054 0.279 0.192 Fert. Requirement -0.079 0.228 -0.349 Market value 0.251 0.340 0.739 Maturity period 0.222 0.122 1.823* Storability 0.104 0.267 0.388 Grain quality 0.538 0.254 2.116** SIGMA 1.672 0.077 21.701 Dependent variable: Proportion of maize farm under improved variety * Significant at 10%, two-tailed level, ** Significant at 5%, two-tailed level, *** Significant at 1%, two -tailed level. Percentage of correct predictions = 87. Log of likelihood function = -490.71

Farming experience has also been shown to have negative impact on adoption of improved maize varieties. Farmers’ experience with local varieties may limit land allocation to new varieties. This may be due to the fact that farmers who have traditionally been inclined to "safety-first" subsistence production may not easily give up their old and time-tested ways of doing things. Time and more education are needed to influence their choice decisions. This suggests that new maize farmers are more likely to adopt improved practices. Superior qualities of improved maize varieties, relating to yield potential, maturity period, grain quality, and resistance to diseases, are some technology-specific characteristics that play significant role in adoption decisions. Farmers therefore continue to cultivate improved varieties that satisfy these qualities. The farmers prefer improved, short duration maize varieties. Factors such as extension contact, storability of grain and drought resistance did not have significant effect on the intensity of adoption of improved maize varieties in Northern Ghana. The estimated model predicts that an “average farmer” with an average age about 44 years and the other mean characteristics presented in Table 6 would almost certainly (99% chance, Appendix B) adopt an improved maize variety. Such a farmer would put about 4 ha of his land under the improved maize variety. The results of this study, to a large extent, corroborate the findings of other empirical analysis of the impact of farmerspecific characteristics and technology-specific attributes of

378

innovation adoption in sub-Saharan Africa (Gyasi et al. 1999, 2000; Abatania et al. 2000; Adesina and Zinnah 1993; Hassan and Faki 1993; Kabebe et al. 1990; Gerhat 1975; Fulusi 1970).

Conclusion This study shows that improved maize technologies have been introduced to farmers in Northern Ghana and a large proportion of the farmers have adopted the technologies. Okomasa and Obatanpa are the most widely cultivated normal endosperm varieties and some farmers are also cultivating the quality protein maize (QPM), Mamaba. In general, the proportion of farmers cultivating any improved maize variety has increased from less than 20% in 1988, to about 76% in 1999. However, inspite of the high preference for improved maize varieties, the proportion of farmland cultivated to improved maize variety is still relatively low. The continued use of local varieties is blamed on the high input demand associated with improved varieties, non-availability of seed, lack of cash to purchase seed and lack of knowledge. The role of extension services in the diffusion of improved maize varieties has not been significant. The adoption rate and intensity of use of improved maize seed is influenced by characteristics of the farmer and some technologyspecific factors. Factors such as age of household head, availability of family labor, farm size, yield potential, grain qu ality as well as disease resistance significantly influence adoption decisions. Addiction to "safety first" subsistence farming and learning behavior of the farmer also affects the proportion of land allocated to new varieties. Extensive technology transfer strategy that equitably involves women and young farmers is needed, not only to promote and facilitate adoption, but also to encourage the extensive cultivation of improved maize varieties in Northern Ghana.

References Abatania, L.N., K.O.Gyasi, O. Coulibaly, R. Adeoti, and A.B. Salifu. 2000. Adoption and impact of cowpea technologies in Northern Ghana . PRONAF Socioeconomic Report. Abdelmagid, S.A. and F.K. Hassan. 1996. Factors affecting the adoption of wheat production technology in the Sudan. Journal of International Agriculture 35(4): 325–337). Adesina, A.A., and M. Zinnah. 1993. Technology characteristics, farmer perceptions, and adoption decisions: A Tobit Model Application in Sierra Leone. Agricultural Economics 9:297– 311.

379

Aikens, M.T., A.E. Havens, and W.L. Flinn. 1975. The adoption of innovations: the neglected role of institutional constraints. Mimeograph. Dept. of Rural Sociology, Ohio State University, Columbus, Ohio. Alderman, H., and P. Hingis. 1992. Food and nutrition adequacy in Ghana. Cornell Food and Nutrition Policy Paper no. 27. Ithaca, New York: Cornell Food and Nutrition Project. Boateng, E.O., K. Ewusi, K. Kanbur, and A. Mckay. 1990 . A poverty profile for Ghana, 1987–88.. Social Dimensions of Adjustment in Sub-Saharan Africa. The World Bank Working Paper No. 5. Washington, D.C. CIMMYT Economics Program. Agricultural Technology: A Mexico. D.F: CIMMYT.

1993. Guide

The adoption of for survey design.

Falusi, A.O. 1974. Multivariate probit: Analysis of selected factors influencing fertilizer adoption among farmers in Western Nigeria. Nigerian Journal of Economics and Social Studies 16: 3–6. Feder, G. 1980. Farm size, risk aversion and the adoption of new technology under uncertainty. Oxford Economic Papers 32: 263–295. Feder, G., and R. Slade. 1984. The acquisition of information and the adoption of new technology. American Journal of Agricultural Economics 66:312–320. Feder, G., R.E. Just, and D. Zilberman. 1985. Adoption of agricultural innovations in developing countries: A survey. Economic Development and Cultural Change 33:255–298. Fey, M. 1992. Household food consumption survey in Northern Ghana. MSc. Thesis.University of Gottingen. Germany. Gerhart, J. 1975. The diffusion of hybrid maize in Western Kenya. CIMMYT, Mexico. Gerner, H., E.O. Asante, E. Owusu-Bennoah, and K. Marfo. 1995. Ghana fertilizer privatization scheme: Private sector roles and public sector responsibilities in meeting needs of farmers. International Fertilizer Development Center-Africa. Goldbeger, A.S. 1964. Econometric theory. John Wiley and Sons, New York. Gyasi, K.O. 2001. Integration of maize markets in Northern Ghana. Pages 433–446 in B. Badu -Apraku, M.A.B. Fakorede, M. Ouedraogo, and R.J. Casky (eds.) Impact, challenges and prospects of maize research in West and Central Africa. Proceedings of a Regional Maize Workshop, 4–7 May, 1999. IITA-Cotonou, Benin Republic. WECAMAN/IITA. Gyasi, K.O., L.N. Abatania, and P. Terbobri. 1999. Evaluation of the adoption of improved rice technologies in Northern Ghana. SARI/EU-OFAR Technical Report.

380

Hassan, R.M, and H. Faki. 1993. Economic policy and technology determinants of the comparative advantage of wheat production in the Sudan . CIMMYT Economic Paper No. 6. Morris, M.L, R. Tripp, and A.A. Dankyi. 1999. Adoption and impact of improved maize production technology: A case study of Grains Development Project. CIMMYT Economics Program Paper 99–01. Kebede, Y., K. Gunjal, and G. Coffin. 1990. Adoption of new technologies in Ethiopian agriculture: The case of TeguletBulga District, Shoa Province. Agricultural Economics 4: 1. MoFA (Ministry of Food and Agriculture). 1997. Accelerated agricultural growth and development in support of vision 2020. Draft. MoFA. 1998. (Ministry of Food and Agriculture). Agriculture in Ghana: Facts and figures. PPMED, Accra, Ghana. NAES (Nyankpala Agricultural Experiment Station). 1983/84, 1984/85, 1985/86, 1988, 1989. Annual Reports, Nyankpla, Tamale, Ghana. Nowak, P.J. 1987. The adoption of agricultural conservation technologies: Economic and diffusion explanations. Rural Sociology 52(2): 208–220. Panin, A., 1988. Hoe and bullock farming systems in northern Ghana. A comparative socio-economic analysis. Nyankpala Agricultural Research Report No. 1. Pindyck, R.S. and D.L. Rubinfield. 1998. Econometric mo dels and economic forecasts. Irwin/McGraw-Hill, USA. Ralm, M.R., and W.E. Huffman. 1984. The adoption of reduced tillage: The role of human capital and other variables. American Journal of Agricultural Economics 66: 405–415. Shakya, P.B., and J.C. Flinn. 1985. Adoption of modern varieties and fertilizer use on rice in the Eastern Tarai of Nepal. Journal of Agricultural Economics 36(3).

381

Appendix A: Empirical model? ? Following Shakya and Flinn (1985), the empirical model for this study is specified as follows: Y* = ? + ? X + ? and Y = g(Y*) where Y=Y* for Y* > 0 Y = 0, otherwise

…..1 …..2

Y* is an index reflecting the combine effect of X (farmer specific and technology specific) factors that influence adoption decision. Y* i s not observable. What is, however, observed is whether the farmer adopts the technology (when Y* >0) or not ( when Y* =0). Where Y is the area (ha) under improved maize varieties, X is as defined above, ? , ? are parameters to be estimated and ? is a stochastic error term The probability of adoption and use intensity can be estimated using the following conditional expectation function: E(Y|Y*) = Y*.F(Y*/?) + ? .f(Y*/?

…..3

E(Y|Y*)= the expected area (ha) given that the variety is adopted ? = standard error of estimate (reported as Sigma in Table 7) Y*/? = standardardized index F(Y*/?)= Tobit probability of adoption, calculated from the cumulative normal distribution f(Y*/? ) = normal density function when Z = (Y*/ ? ).

Appendix B Predicting the proba bility of adoption for an ‘average’ farmer. Substituting the average values from Table 6 and the parameter estimates (Table 7) into equation 1, Y = 4.039 ? Y/? = 2.415 From the normal distribution Table, the probability of adoption is estimated as: F(Y/ ? ) = F(2,415) = 0.9922 Expected area to be put under improved maize varieties is given by: E(Y|Y*) = Y*.F(Y*/?) + ? .f(Y*/? ) Y* = 4.039, F(Y*/ ?) = 0.9922, ? = 1.672 (sigma in Table 7), f(Y*/? ) = 0.0216 ? E (Y|Y*) = 4.0387(0.9922) + 1.672(0.0216) = 4.043 ha. QED. ? ? The empirical model was adapted from Shakya and Flinn (1985.)

382

Pattern of spread of extra-early maize varieties in the Sudan savanna ecology of Nigeria J.E.Onyibe1 , C.K. Daudu1, J.G. Akpoko1, R.A. Gbadegesin1, and E.N.O. Iwuafor2 1

NAERLS, Ahmadu Bello University, P.M.B. 1067, Zaria, Nigeria 2 IAR, Ahmadu Bello University, P.M.B. 1044, Zaria, Nigeria.

Abstract Production of maize (Zea mays L.) in the Sudan savanna zone of Nigeria has been constrained by non-availability of varieties specifically adapted to the 2½-3 month erratic seasonal precipitation characteristic of the zone. Two new, high yielding, extra-early maize varieties (95 TZEE-W1 and 95 TZEE-Y1), developed by the West and Central Africa Collaborative Maize Research Network (WECAMAN) were introduced to 65 individual farmers at two locations (Kafinsoli and Ladanawa) within the Sudan savanna ecology of Nigeria at the beginning of the 1997 cropping season. The varieties were also introduced to other farmers, Government officials, input and seed companies during field days held at the end of the season in the same year at Ladanawa and at Kafinsoli. Late in the year 2000, a survey was conducted at the two locations and 20 other communities surrounding the two locations to assess the extent of technology adoption, to identify the reasons for acceptability or otherwise of the new maize varieties and to analyze the determinants of area expansion. The result shows that 95% of the original seed recipients were still growing the varieties. All 22 communities surveyed had at least six out of 15 respondents cultivating the maize. About 72% of the 464 surveyed farmers were not original recipients of seed but were growing either or both varieties in 2000. The result also revealed high (over 1000%) rate of expansion of area under cultivation of the varieties. The study identified short duration to maturity, higher yield relative to traditional staples such as sorghum and millet and effective media advisory support as factors that contributed to the wide spread of the new varieties. Résumé La production du maïs (Zea mays L.) dans la zone de la savane soudanaise du Nigeria a été entravée par la non disponibilité des variétés spécifiquement adaptées à la saison pluvieuse de 2,5 – 3 mois de durée caractéristique de la zone. Deux nouvelles variétés de maïs hautement productives, extra précoce (95 TZEE-W1 et 95 TZEE- Y1), créées par le réseau collaboratif de recherche sur le maïs en Afrique de l’Ouest et du Centre (WECAMAN) ont été introduites chez 65 paysans

383

dans deux localités (Kanfisoli et Ladanawa) dans la sava ne soudanaise du Nigéria au début de la campagne 1997. Les variétés ont été également introduites à d’autres paysans, chez des personnalités officielles du gouvernement, des compagnies semencières et d`intrants durant les journées agricoles organisées en fin de saison de la même année à Ladanawa et à Kafinsoli. Tard dans l’année 2000 une enquête a été réalisée dans les deux localités et dans 20 autres communautés autour des deux localités pour évaluer le taux d’adoption des technologies identifiées les raisons d’acceptabilité des nouvelles variétés et analyser les facteurs déterminants des zones d`expansion. Les résultats montrent que 95% de ceux qui ont reçu la semence continuent à cultiver ces variétés. Les 22 communautés enquêtées ont au moins six des 15 répondants qui cultivent le maïs. Près de 72% des 464 paysans enquêtes n’ont pas reçu les semences au départ mais sont entrain de cultiver une ou les deux variétés en 2000. Les résultats révèlent aussi un grand (plus de 1000%) taux d`augmentation des superficies emblavées par ces variétés. L`étude a identifié le cycle court de maturité, le haut rendement par rapport aux cultures de base comme le sorgho et le mil et un encadrement effectif par les médias comme des facteurs qui ont contribué à la grande diffusion des nouvelles variétés.

Introduction The international Food Policy Research Institute (IFPRI) estimated that by the year 2020 the population of the world is likely to approach a billion and the rate of growth will be most rapid in Africa (IFPRI 1995). It is also estimated that by that time extreme poverty and insufficient economic growth will stimulate unprecedented food and social crises in both rural and urban settlements in Africa. The poverty and food crisis in Africa reflects opportunities for technology innovation, generation and transfer that would elicit new levels of cropping system intensity and diversity to mitigate the crises. Development and deployment of such technologies will achieve optimal effect if areas that offer high potential for substantial economic benefits are targeted. The savanna ecology of West and Central Africa (WCA) holds enormous potential for the cultivation of maize. As a major food and industrial crop, maize holds considerable promise as a weapon against poverty and food crisis in the region. But its production until recently had been confined to the forest and the wetter savanna ecologies. The potential for its production in the marginal zones (Sudan savanna) though high (Onyibe et al. 1999) is limited to the use of extra-early maize varieties unless where irrigation facilities had been developed. Our efforts to accelerate the cultivation of maize in the Sudan savanna of Nigeria commenced in 1997 with the introduction and promotion

384

of appropriate extra-early varieties using a mix of extension advisory support services. Conventional extension philosophy assumes rightly, that with some promotion efforts, a good crop production technology will diffuse quickly between farmers and farming communities within areas of simi lar natural resource endowments. After four years of promotion of the extra-early maize varieties within farming communities, it is pertinent to document the pattern of spread as a basis for the development of technology deployment strategies for accelerated maize production in the Sudan savanna zone.

Methodology Study area The study was conducted within the Sudan ecology of Katsina State of Nigeria (Lat. 12.0-13.2 oN and Long. 6o 40’-8o 20’E; Fig. 1). Similar to the Sudan ecological zone of WCA, the rainy season is between July to September and the rainfall is monomodal averaging about 500mm per year. Detailed description of the ecology has been reported by Kowal and Knabe (1972). The soil is predominantly Entisol. The people are Hausa-Fulani and household sizes up to 30 persons are a common feature. Farming is a major occupation and most farmers till their land with traditional hand hoe. But the use of animal traction is gaining popularity due to the problem of availability of labor during the short farming season. Major crops produced in the area include sorghum, millet, cotton, cowpea and groundnut. The soils of the area are low in most nutrients required by crops and chemical fertilization of these crops is limited due to high cost and non-availabili ty. Maize production in the zone was limited to areas where irrigation facilities were available. In 1997, under a WECAMAN Maize Promotion Project, 97.5 kg of seeds of two extra early maize varieties, 95TZEE-W1(white) and 95 TZEE-Y1 (yellow) were distribu ted to 65 farmers in 61 households. These farmers expressed willingness to plant the seeds and were trained on maize production techniques. Each farmer received 0.5 kg of each variety along with 6 kg N, 3 kg P2 05 and 3 kg K2 0. The farmers were from two villages, Kafinsoli and Ladanawa (Fig. 1). A field day was held at Ladanawa in 1997, with participation of farmers from Kafinsoli and neighboring villages. A similar field day was held at Kafinsoli in 1999. Also radio and T.V broadcast of the potential of the varieties and the field day were made annually from 1997 to 2000 at the State broadcasting stations. Farmers were encouraged to give or sell seeds to interested farmers.

385

In 1999 six farmers’ groups were organized, registered and encouraged to produce the seeds of the extra early varieties. The groups produced a total of 3360 kg of 95 TZEE-Y1, and 3430 kg of 95 TZEE-W1. The seed was sold in local markets at premium prices that ranged from N45 to N75/kg ($1 = N90) at the beginning of the cropping season of 2000 compared with the price of maize grain that ranged between N22 and N30/kg and N16–24/kg for grain sorghum.

386

Survey At the end of the fourth cropping season (2000), an extensive survey was carried out at the two initial locations where the varieties were introduced and 20 other villages surrounding these two locations. During the survey, the lists of farmers, including relations and friends, that received or bought seed from the initial seed recipients were obtained. A similar procedure was employed to generate information on the number of farmers and quantity of seed sold or distributed by the farmers. Also 10 kg of seeds of each variety was given to the Katsina State Agriculture and Rural Development Authority (KTARDA) to multiply in their seed multiplication farms at Ajiwa and Kafinsoli. The list of beneficiaries of the seed produced by the KTARDA and their locations was secured through informants and used along with the information provided by informants on the first, second and third generations of seed recipients to determine the villages to be selected for the administration of questionnaires. Since rainfed maize cultivation was uncommon in the zone (Sudan ecology of Katsina State) before the WECAMAN Maize Promotion Project, it was assumed that all rainfed maize produced in the area were from the extra early varieties introduced. The same sampling procedure was applied to all categories of seed recipients and maize growers. Following the identification of recipients, 464 households were selected for the survey (Table 1). Four of the first generation seed recipients left the villages while three abandoned farming for the sale of petrol. These were not included in the analysis. Table 1. Categories of extra-early maize farmers identified and used in the analysis. Recipients (Respondents ) Categories of seed* 1997 1998 1999 2000 First generation 58 Second generation 51 Third generation 183 Fourth generation 172 Total 464 *First generation-received seed in 1997 directly from WECAMAN Project. Second generation-received seed in 1998 from first generation recipients or bought from the market. Third generation-received seed in 1999 from the WECAMAN Project/KATARDA relations or bought from market. Fourth generation-received seeds in Year 2000 from KTARDA, relations or bought from market.

The selection procedure had limitations in that beneficiaries outside the Sudan or Sahel savanna of Katsina State were not included. Also farmers within the zone that may have benefite d from the project but who were not introduced to the survey team could not be included in the study. The data collected included

387

farmers’ assessment of the new varieties (both positive and negative attributes), agricultural practices (planting pattern), primary source of seed, quantity of seed harvested, household method of disposal of harvest, socioeconomic characteristics of growers, crop area, yield estimates, and production constraints. Geopositioning system (GPS) units were used to collect georeferenced data for each measured field in order to analyze the spatial pattern of spread of the varieties.

Data analysis Descriptive statistical methods were employed to interpret the data. A correlation analysis of all 20 variables in the survey tool was conducted to identify the variables that were highly correlated. Non-correlated variables were fitted in a stepwise multiple regression model to identify determinants for the expansion of the new varieties of extra-early maize as used by Manyong et al. (1998) for new varieties of soybean in southern Kaduna of Nigeria. The empirical analysis assumed a linear statistical model for 20 non -correlated explanatory variables; that is: Y= ? o + ? 1 Sex + ? 2 Age + ? 3 Edu + ? 4FS + ? 5 KepLiv + ? 6 V1 loca + ? 7 S Seed + ? 8 Plt Patn + ? 9 Yr Cont Crop. + ? 10 Priss + ? 11 Qharv 00 + ? 12 Q sold 00, + ? 13 AdopC. + ? 14 Pest Prob + ? 15 Var + ? 16 Trend + ? 17 Fert Prob + ? 18 Exten Supt + ? 19 Mkt Priz + ? 20 Negat + ? In this model: Y = cultivated area of the two extra-early maize varieties in the 2000 and ? = the error term. The inclusion of these variables was based on the evidence of their significance as observed in previous studies by Onyibe et al. (1999) at these locations and that of Manyong et al. (1998). It is hypothesized that factors likely to explain the reasons for the spread of new varieties include socio-economic factors, natural resources endowment (soil and climate), market, extension advisory support, and technology specific attributes (such as yield, plant height) (Table 2).

Results and Discussion Sources of seed of extra-early maize to farmers The data in Table 3 show the sources of seed of extra-early maize varieties available to respondents in the study. As expected, all first generation farmers indicated that they received their seeds from the maize promotion project implemented under WECAMAN.

388

Table 2. List of explanatory variables used in stepwise multiple regression analysis. Sex Age Edu FS Kepliv Vloca SSeed Plt Patn Yr. Cont Crop Priss Qharv00 Q sold 00 Adop C Pest Prob Var Trend Fert Prob Ext. Supt

Mkt Priz Negat

Sex of respondent Age of respondent Education level of respondent Family size Keep livestock Location of village Sources of seed Planting Pattern Years of continuous cropping of the maize Primary source of seed Quantity of grain/seed harvested in year 2000 Quantity of grain/seed sold in year 2000 Adoption categories of respondent Pest problem (e.g., Striga, termite) Variety of extra-early used Trend in area expansion of the maize varieties Fertilizer Problem Received extension support in any of the following ways (publication, training, extension agent visit, TV and/or radio program on extra early maize production, attended field day). Price of maize at local market Negative attributes of the extra early maize varieties.

Table 3. Sources of seed of extra-early varieties (95 TZEE-W1 and 95 TZEE-Y1) available to farmers in the Sudan savanna of Katsina State, 1997–2000 cropping seasons. Seed source Project Market KATARDA Relations Others*

1997(N=58) 100 (58)** -

% of farmers using source 1998(N=51) 1999(N=183) 25 (46) 43 (22) 54 (98) 20 (10) 15(28) 35 (18) 6 (11) 2 (1) -

2000(N=172) 9 (16) 72 (123) 13 (23) 2 ( 3) 4 (7)

**Figures in parenthesis are the actual number of respondents. *Others – Those who obtained the seeds during field days.

The number of respondents that obtained their seeds from the project decreased progressively to 9% in the fourth generation (Year 2000). In contrast to this, sourcing seeds from local markets increasingly became important over time. The percent of respondents that obtained their seeds from the market was 43% in 1998 compared with 54% in 1999 and 72% in 2000. KTARDA provided the seeds used by 20, 15 and 13% of the respondents in 1998, 1999, and 2000, respectively. The results suggests that the promotional activities implemented by the project such as field days, radio and TV program broadcast and training of farmers stimulated the interest of a substantial number of the farmers to the extent that they were willing to pay

389

for the extra-early maize technology. The expansion in the contribution of the market as a source of seed was enhanced by the involvement of the State and the communities in the multiplication of the seed. The involvement of the Farmers Supply Company (FASCOT) of the State in the marketing of the seed through its various outlets was also a factor that contributed to the success. The roles played by the different State agencies in addition to buying the seed produced by the farmers greatly facilitated the spread of the two maize varieties. Also the premium price attracted by both maize seed (N45–75/kg) and grains (N22 000 to 30 000/ton) in the area during the period under review probably contributed to the adoption of the technology by farmers. Expansion in production would depend on the mechanism of distribution of the seed of the extra-early varieties. There is a great potential for adoption of these and similar technologies in areas where the crop is yet to be introduced within the Sudan and Sahel savannas and in areas where only a few farmers are currently involved in maize production.

Farmers assessment of extra-early maize Negative attributes of the varieties. Only 82 out of 172 farmers that cultivated the maize varieties in 2000 responded properly to questions on the attributes of the two maize varieties promoted. Majority of the farmers that responded (55%) mentioned no negative attributes in the two extra-early varieties (Table 4). Table 4. Farmers’ assessment of the negative attributes of two extraearly maize varieties in the Sudan and Sahel savanna of Katsina State, Nigeria in 2000.

Attributes No negative attribute Too short (relative to sorghum) Requires fertilization Susceptible to Striga Low yield

95 TZEE-W1 (N=36) 56 (20)* 11 (4) 28 (10) 0 6 (2)

% of respondents 95 TZEE-Y1 (N=46) 54 (25) 13 (6) 26 (12) 7 (3) 0

Total (N=82) 55 12 27 4 2

*Figures in parenthesis are the actual number of respondents.

A few however indicated that the short height of the crop and relatively high fertilizer requirement of the varieties as compared with sorghum and millet as well as susceptibility to Striga were negative attributes of the varieties. The requirement for tall plants among farmers is related to the various uses of stalks of cereals such as fencing, thatching of roof, livestock feed and domestic energy. Striga is a problem in the zone and has led to abandonment of fields (Lagoke et al. 1994). The concern of farmers regarding susceptibility to Striga is therefore

390

understandable. Fertilizer scarcity and high cost are common problems facing farmers in the region, and this has been particularly acute since the withdrawal of the fertilizer subsidy by the government (Ahmed et al. 1998). Unless the issue of availability of fertilizer is resolved, the expansion in the production of the varieties is likely to be constrained. Positive attributes of the varieties. The overriding positive attributes of the two varieties include high yield relative to those of sorghum and millet, early maturity and availability of a market for the seed and grain (Table 5). Table 5. Farmers’ assessment of the positive attributes of two extraearly maize varieties in the Sudan and Sahel savanna of Katsina State, Nigeria in 2000.

Attribute Relatively high yielding* Early maturity Good taste Tolerate inter cropping Easily sold

95 TZEE-W1 (n = 36) 58 (21) 19 (7) 6 (2) 6 (2) 11 (4)

% of respondents 95 TZEE-Y1 (n = 46) 54 (25) 13 (6) 11 (5) 7 (3) 15 (7)

Total (N = 82) 56 16 9 6 13

*Relative to traditional staple crops such as sorghum and millet.

Also compatibility with other crops in mixed cropping and good taste were considered good attributes of the varieties. The results apparently reflect farmers’ growing concern for improvement in income. The rating for both varieties did not differ remarkably suggesting similar levels of acceptance by the farmers.

Cropping system Farmers in the first generation of seed recipients were trained in maize production by the maize promotion project. A few of the third generation seed recipients who received seeds directly from the project were also trained. However most of the farmers adopted intercropping with cowpea, cotton or groundnut (Table 6). Table 6. Planting pattern of involving extra–early maize varieties use by the respondents.

Cropping pattern Sole* Intercropped with cowpea Intercropped with sorghum and millet Intercropped with cotton Intercropped with groundnut With other crops

% of respondents 1999 (N=183) 2000 (N=172) 12 14 33 28 7 5 27 34 18 17 3 2

* Comprised mostly of seed multiplication plots.

391

In contrast, intercropping with sorghum and millet was very low. Sole cropping appeared limited to seed multiplication, otherwise intercropping is a preferred system. From the crop combinations, it appears that the farmers prefer to mix crops of diverse morphotypes. This combination has the potential of permitting high levels of plant densities (Onyibe and Adeosun 1998). Adoption of maize production packages intended for this zone appears to be contingent on the level of compatibility of maize with these preferred cropping systems. Intercropping is used as insurance against crop failure. For the adoption of new technology, the need for this at farmer’s level seems compelling.

Area cropped to extra-early maize The area cultivated to the two extra-early maize varieties increased from 2.9 ha in 1997 to 155 ha in 2000 (Table 7). The area expansion was more remarkable with 95 TZEE-Y1 than 95 T2EE-W1. This is probably related to its higher yield (about 2.52.9 t/ha) compared with 2.2-2.5 t/ha obtainable with 95 TZEEW1(Onyibe et al. 1999), and the preference for yellow maize in some of the communities (Kafinsoli, Baud and Safana). Average farm holding also increased progressively but was less than 1 ha per farmer. In 1999 and 2000 average farm holding was 0.67 and 0.90 ha/farmer, respectively. This high level reflects inclusion of seed multiplication plots that were in units of 2–5 hectares. The overall rate of expansion was over 1000%. The expansion in cultivated area was however constrained by lack of seeds (41%), fertilizer (34%) and suitable land (14%). Uncertainty of the weather accounted for the remaining 11%. Table 7. Estimated area cropped to extra-early maize varieties by respondents between 1997 to 2000.

95 TZEE-W1 (ha) 95 TZEE-Y1 (ha) Total (ha) No of farmers Average holding (ha)

1997 1.5 1.4 2.9 58 0.05

1998 11 1.5 26 51 0.51

1999** 49 74 123 183 0.67

2000** 68 87 155 172 0.90

*Includes area derived from inter-cropped systems based on 1:1 ratio combination. **Area for these years include group and KATARDA seed multiplication plots.

Spread Figure 1 shows the spatial locations of the categories of the respondents and the area coverage. The first-generation seed recipients were distributed in only two locations Ladanawa and Kafinsoli in 1997. In 1998, a southward trend in spread at both locations was observed. An attempt was made to repos ition the technology farther north in 1999 by introducing it at Katsina and Rimi, and westward at Buade. The pattern of spread in 2000, though diffused, was essentially southward. Communities with

392

large markets such as Kafinsoli, Ladanawa Kankia, Dutse -ma, Karoji, and Maje adopted the technology and recorded larger number of respondents, especially among the third generation farmers perhaps due to better accessibility to the seeds sold in these markets. The rainfall distribution in the zone is such that there is a gradual reduction in quantity from the south to the north. Thus the uncertainty of the weather becomes a factor in the selection of crops in the higher latitudes. This factor may have accounted for the more southward spread of the technology. Arokoyo et al. (1996) noted that some of the maize farms located north of Ladanawa did not produce harvestable yields while fairly high yields were obtained at Ladanawa and Kafinsoli. This probably made the farmers in the areas north of Ladanawa less confident to try the varieties. In spite of the confirmed potential of the technology, the erratic nature of rainfall in the zone will continue to be a major factor in the spread of the varieties, especially northward, until reliable weather monitoring and forecasting services are available to the farmers.

Determinants of area expansion Differences in the level of contributions of the 20 explanatory variables were found. Five of the variables (Table 8) accounted for 58% of the variation in cultivated area. The quantity of harvested grains or seeds (Qharv00), planting pattern and extension support services (such as training and access to radio and TV programs about the technology) were highly significant positive factors that contributed to area expansion. Fertilizer supply had a highly significant negative effect on area expansion. The quantity of produce sold (Qsold) was significant only at 5% level of probability. It can be deduced from the results that farmers appreciated the high yield potential of the varieties in their environment relative to traditional cereals. They tried to maximize the benefit of the new crop by selling most of their maize harvest and to use the proceeds therefrom to buy other grains. This accounts for the high T-ratio for Qsold00 (Table 8). Table 8. Statistically significant parameter estimates from the stepwise multiple regression of land area cultivated to extra-early maize on several variables in the Sudan and Sahel savanna zones of Katsina State, Nigeria. Parameter Qharv 00 Plt Patn. Qsold 00 ExtSupt Fert Prob Overall

Estimated coefficient 0.493 0.389 0.413 0.331 -0.403

T-ratio 4.268** 3.466** 2.394* 3.273** -3.16**

*,**Significant at 0.05 and 0.01 levels of probability, respectively.

R2 0.563 0.553 0.439 0.531 0.522 0.581

393

Apparently the higher the level of harvest the higher the impetus to sell due to favorable price. The intensive extension support services stimulated the awareness and diffusion of the technology in the zone. Also the varieties blended with the traditional cropping practice and thus reduced the risk of crop failure. The availability of market for both grain and seed of the varieties provided opportunities for enhanced income from this farming enterprise. These factors were therefore certain to stimulate cultivation of the varieties. Fertilizer scarcity as a variable was a disincentive to cultivation of maize in the zone. Apparently, sustained expansion in cultivated area is likely to be achieved when fertilizer supply is assured. Conclusion The study has shown that the extra-early maize varieties (95 TZEE-W1 and 95 TZEE-Y1) have high yield potential in the dry savanna zone of Nigeria. The study revealed that the technology had diffused widely and this was due to the high yield potential, early maturity and compatibility of the varieties with traditional cropping practices. The quest for improvement in income was also an important factor contributing to the spread of the varieties. The study provides additional evidence that multilateral collaboration in research and technology transfer can elicit new possibilities in cropping system diversity, intensity and enhanced opportunities for improvements in farmers’ income in the dryer savanna zones. References Ahmed, B., A. O. Ogungbile, J. O. Olukosi, and M. I. Kolawole. 1998. The impact of fertilizer subsidy withdrawal on the sustainability of maize production in the farming system of Northern Guinea Savanna of Nigeria Pages 243–252 in T. Bezuneh, S. Ouedraogo, J. M Menyonga, J. D. Zongo, and M. Ouedraogo (eds.) Towards sustainable farming systems in sub-Saharan Africa. Publication of the African Association of Farming Systems Research-Extension and Training, Ouagadougou, Burkina Faso. Arokoyo, J.O., J.E. Onyibe, C.K. Daudu, J.G. Akpoko, E.N.O. Iwuafor, and K.A. Elemo. 1996. Promotion of maize technology transfer in savanna ecology of Nigeria. 1996 Annual Report of WECAMAN Maize Promotion Project. 11p. IFPRI. 1995. A 2020 vision for food, agriculture, and the environment. Washington D.C., International Food Policy Research Institute, USA. Kowal, J. and M. Knabe. 1972. An agro-climatological atlas of the northern states of Nigeria. ABU Press, Zaria, Nigeria. 111p.

394

Lagoke, S.T.O., J. Y. Shebayan, J. Adeosun, E.N.O. Iwuafor, J. Olukosi, J.K. Adu, O. Olufajo, A.M. Emechebe, M. Zaria, A. Adeoti, J. Onyibe, and S. Chobe. 1994. Survey of Striga problem, and on-farm testing of integrated Striga control packages and evaluation of various Striga control methods in maize, sorghum, and cowpea in the Nigerian Savanna. Pages 50–97 in R. Tranboulsi, S.T.O. Lagoke, R. Hoevers, and S.S. M’Boob (eds.) Towards an integrated control of Striga in Africa. Proceedings 1st General Workshop of the Pan-African Striga Control Network (PASCON), 11–14 March 1990, Ibadan Nigeria, Accra Ghana FAO. Manyong, V.M., K.E. Dashiell, B. Oyewole, and G. Blahut. 1998. Spread of new soybean varieties in traditional soybean growing areas of Nigeria Pages 151–161 in T. Bezuneh, S. Ouedraogo, J. M. Menyonga, J. D. Zongo, and M. Ouedraogo (eds.) Towards sustainable farming systems in sub-Saharan Africa. Publication of the African Association of Farming Systems Research-Extension and Training, Ouagadougou, Burkina Faso. Onyibe, J.E. and J.O. Adeosun. 1998. Inter-cropping soybean with maize or sorghum under farmer management in northern in T. Guinea savanna zone of Nigeria. Pages 143 –150 Bezuneh, S. Ouedraogo, J. M. Menyonga, J. D. Zongo, and M. Ouedraogo (eds.) Towards sustainable farming systems in sub-Saharan Africa. Publication of the African Association of Farming Systems Research-Extension and Training, Ouagadougou, Burkina Faso. Onyibe, J.E., C.K. Daudu, J.K. Akpoko, and E.N.O. Iwuafor. 1999. Challenges of maize technology transfer in marginal zones of Nigeria. Pages 383–393 in B. Badu-Apraku, M.A.B. Fakorede, M. Ouedraogo, and R.J. Casky (eds.) Impact, challenges and prospects of maize research and development in West and Central Africa. Proceedings of a Regional Maize Workshop, 4–7 May, 1999, IITA-Cotonou, Benin Republic. WECAMAN/IITA.

395

Potential impact of input policy on maize supply in Ghana A.S. Langyintuo, K. Foster, and J. Lowenberg-DeBoer Department of Agricultural Economics, Purdue University, 1145 Krannert Building, West Lafayette, IN 47907, USA Abstract As part of the conditions for the implementation of the International Monetary Fund and the World Bank structural adjustment program in the early 80s, the government of Ghana removed input subsidies. Subsequent calls on the government by stakeholders in the agricultural sector to reintroduce subsidies to boost agriculture failed to yield positive results. To widen its revenue base, the government levied a 10% Value Added Tax (VAT) on domestic sales and imports in 1998. There were a few exemptions including agricultural inputs. This study analyses the potential direction and magnitude of the impact of 10% VAT or subsidy on fertilizer demand, marginal cost of production and supply of maize in Ghana. The results suggest that a 10% VAT on fertilizer will decrease fertilizer demand by 15%, unambiguously increase marginal cost and depress maize production by 24%. The potential tax revenue would be inadequate to compensate for the revenue loss by maize farmers. In contrast, a 10% fertilizer subsidy will reduce marginal cost by 17%, increase fertilizer demand by 22.4% and expand maize supply by 27%. Maize farmers would experience a gain in net revenue of about 30% over the case without a subsidy. The corresponding subsidy cost would be 6% of the revenue gain by farmers. These estimates should be used in policy analysis to compare with the benefits of other programs and/or the costs of other tax revenue sources. Résumé Dans le cadre de la mise en oeuvre des programmes d’ajustement structurelle du FMI et de la Banque Mondiale, au début des années 80, le gouvernement dU Ghana a supprimé les subventions sur les intrants. Les appels subsequents des acteurs du secteur agricole au gouvernement pour la réinstauration de ces subventions pour booster l’agriculture sont restés sans suite. En 1998, le gouvernement du Ghana a imposé une taxe de 10% sur la valeur ajoutée (TVA) des importations et ventes domestiques. Les intrants agricoles et quelques autres produits étaient exemptés de cette TVA. Cette étude analyse la direction potentielle et l’amplitude de l'impact de 10% TVA ou subvention sur la demande d'engrais, les coûts marginaux de production et d’approvisionnement du maïs au Ghana. Les résultats suggèrent que les producteurs de maïs minimisent les coûts. Une TVA de 10% sur l'engrais diminuera

396

la demande d'engrais de 15%, augmentera clairement le coût marginal et découragera la production de maïs de 24%. Le revenu potentiel de la taxe ne pourra pas compenser la perte de revenue des producteurs de maïs. Au contraire, une subvention de 10% d'engrais réduira le coût marginal de 17%, augmentera la demande en engrais de 22,4%, permettra d'accroître l’approvisionnement en maïs de 27%. Les producteurs de maïs obtiendront un gain net additionnel de l’order de 30% comparé au cas sans la subvention. Le coût correspondant de la subvention serait de 6% le gain en revenu des producteurs. Ces estimations devraient être prises en compte dans les analyses politiques pour comparer avec les benefices d’autres programmes et/ou les coûts des autres taxes sur les sources de revenue.

Introduction Agricultural input subsidies have the potential to increase production, raise the net income gains from a given level of input use and promote input use. New seeds and fertilizer, for example, are complementary inputs. This means that the highest levels of yield are only achieved by simultaneous increase of both types of inputs in the correct proportions. Productivity gains of the new technology may be limited if one input is missing. Under such circumstances, the farmer may do better to stay with traditional varieties that are less sensitive to levels of chemical inputs. Subsidies motivate farmers to overcome the risk-averse behavior which causes them to underestimate the returns to using new inputs. This thus provides an ince ntive for the more rapid adoption of modern inputs than would occur in their absence (Miller and Tolley 1989; Ellis, 1992). However, input subsidies involve departures from social opportunity cost and unpredictable budgetary burdens (Ellis 1992; Roche 1994). They can also cause, for instance, resource inefficiency at the farm level including excessive input use, inefficient substitution of scarce for abundant resource inputs, inefficient substitution of crops that use much of the input for crops that use li ttle (Ellis 1992). In response to similar adverse impacts of input subsidies, the International Monetary Fund (IMF) and the World Bank (WB) advocate for the removal of subsidies in their structural adjustment programs (SAP) in developing countries. Following the IMF and WB SAP directive, the government of Ghana removed input subsidies and privatized fertilizer distribution between the mid- and late 80s. Since the implementation of this policy, widely condemned by farmers, there has been unceasing call by farmers and other stakeholders in agriculture for the re -introduction of the subsidies to boost agricultural productivity and production. This call unfortunately is unpopular with the government that is cash strapped and taking all steps to harness all sources of revenue. In the process, a Value Added Tax (VAT), seen as part of a broader program of tax

397

reform envisaged under the Economic Recovery Program, was introduced in March 1995 but was withdrawn after three months following nationwide demonstrations against its imposition (Assibey-Mensah 1999; Osei 2000). In December 1998, VAT was reintroduced at the standard rate of 10%, as against 17.5% initially. Most traded commodities in the country excluding a few essential goods and services including agricultural inputs were taxed (ISSER, 1998; Osei, 2000). Part of the food sector such as restaurants which were initially exempted was later covered. There are growing fears that further expansion of the tax net could eventually cover the agricultural inputs sector. It is argued here that such a move will have pervasive impact on the agroindustrial base of the economy, especially sectors intensive in the use of such inputs as fertilizer and insecticides. On the other hand, re-introducing the subsidy on inputs would decrease production cost, promote input use and increase grain supply. To provide an indication of the potential direction and magnitude of the impact of re-introduction of input subsidy or extending the VAT on inputs on the supply of grains and the demand for such inputs, it is thought expedient to undertake this study. This study uses a cost function analysis approach to estimate the potential effects of imposing a 10% VAT or 10% subsidy on chemical fertilizers on its demand, marginal cost of production and supply of maize in Ghana. Chemical fertilizers are by far the most important purchased variable inputs in terms of their yield impact in conjunction with new seeds, and in terms of the volume and gross value of their consumption compared to other inputs. The choice of maize as the test crop is because it is the most extensively grown crop and accounts for over an estimated 60% of all chemical fertilizers used in crop production (IFDC 1995). The Translog Cost Function (TCF) for maize production is estimated and the marginal cost (MC) function derived. Supply response for maize representing a tax or subsidy on fertilizer is calculated from the MC function. Output supply and fertilizer demand elasticities are also computed. The policy variable, tax or subsidy, is analyzed in terms of the quantitative effect on MC, output supply, and fertilizer demand. Relevant maize production and input supply data from 1970 to 1998 were obtained from various sources. The main source of maize production and prices of both maize and fertilizer was the Policy Planning Monitoring and Evaluation Division (PPMED) of the Ghana Ministry of Food and Agriculture (MoFA). Estimates of fertilizer (NPK and SA) consumption were obtained from FAO statistics and ISSER (1996, 1998). FAO gave overall consumption and not by crop. An estimate of quantity of the fertilizer used on maize was obtained from an IFDC 1995 report and Donhauser et al (1994). Costs of labor were estimated from Statistical Service reports (various years) agricultural wages. Capital data were estimated from various sources including Donhauser et al (1994)

398

and FAO (1996, 2000). All costs and prices were deflated using the consumer price index numbers.

Model specification and estimation procedure Duality in production economics, linking production and cost relationships, has permitted researchers to estimate parameters of a cost function and relate the information to the underlying production function (Chambers 1988; Cornes 1992; Silberberg 1990). Cost function estimati on has generally been limited to the computation of input demand elasticities and elasticities of substitution (Binswanger 1974; Akridge and Hertel 1986; Capalbo and Antle 1988). In other cost studies, MC is derived as a means to estimate economies of scale (Akridge and Hertel 1986). Marginal analysis has been applied to data on supply and demand elasticities, prices, quantities, and estimated cost and yield effects from input policy (Lichtenberg, Parker, and Zilberman, 1988). Cost function estimation permits a focus on changes in MC and supply responses as a result of input policy. This study estimates a cost function to analyze changes in costs at the margin, and to analyze supply response from fertilizer taxation or subsidization. The elasticity estimates of MC and supply are derived from the cost function to estimate the changes in fertilizer demand which are used for comparing the effects of the tax or subsidy policies. Various cost functions are available for use but the translog cost function (TCF) is used here because of its flexibility (Capalbo 1988; Pollak and Wales 1992). The TCF is a second order approximation to an arbitrary twice -continuously differentiable cost function (Heathfield and Wibe 1987; Pollak and Wales 1992; Capalbo 1988). The mathematical expression of the second order logarithmic Taylor series expansion of the cost function, C, around variable levels of output, Q, and input prices, Wi(i=1,…,n) is given as lnC=f(lnW 1 , …, lnWn ; lnQ), where lnC is cost of production in natural logs. The second-order Taylor series expansion of this function generates the TCF as: n

ln C ? ? 0 ? ? ? i ln Wi ? ? q ln Q ? i ?1

?

n

?

i ?1

1 n ? 2 i

n

? j

1 ? ij (ln Wi )(ln Wj ) ? ? qq (ln Q) 2 2

? qi (ln Q)(ln Wi )

where ? , ? and

… (1)

? are parameters to be estimated.

The main concerns here are the relevance of various theoretical production and cost properties and the reliability of the data, especially inputs use data. Three main inputs are considered: fertilizer, labor and capital. Cash cost of consumable inputs (e.g. seed, fertilizer, pesticides) are usually easy to deal with, but there are often controversies about capital inputs (e.g. useful life,

399

pattern of depreciation, discount rate). Fertilizer, an aggregate of nitrogenous fertilizer – nitrogen-phosphorus-potassium (NPK) and sulphate of ammonia (SA), often assumes over 50% of production costs, especially in developing agriculture. The input that attracts the most controversy in such analysis is labor. The main sources of labor are family and hired labor. At the aggregate level, hired labor predominates because about 90 % of maize is produced in southern Ghana (PPMED 1998) where farmers depend mainly on hired labor, some of which is migrant labor from northern Ghana with minimal family labor input. In this part of the country, use of machinery is limited by high vegetation of trees. It is often argued that family labor is conceptually more difficult for cost considerations since the farmer owns it, and therefore collects producer rents rather than an out-of-pocket cash payment that reflects an input’s purchase price. But given that at the aggregate level family labor constitutes a small proportion of all labor use, it is of less consequence in this study. Family labor is valued at the opportunity cost of labor in agriculture. It should be noted that before fertilizers and labor were aggregated, they were tested for weak separability using the Likelihood Ratio Test (LRT) (Blackory, 1978; and Moschini et al. 1994). Land is excluded because determining the value of the land input is problematic, especially that most farmland is under a “stool5” that prohibits its sale. As a result, land is either free as in the north or used under various tenancy arrangements such as share cropping which is of less consequence to the farm cost structure. The consequence of omitting land in the model is the inability to estimate a substitution elasticity between land and fertilizer. The fertilizer-reducing (increasing) policy may result in a substitution to greater (lesser) land area in cultivation. A three -input (fertilizer (F), labor (L) and capital (K)) TCF is estimated simultaneously with two inputs cost-share and one revenue share equations using seemingly unrelated regression (SUR). By construction, the share of factor cos ts sum up to unity hence the need to drop one share equation prior to estimation (Capalbo 1988). The following simultaneous system of equations were estimated, imposing homogeneity ( ? FL

5Stool

( ? i ? ij ? 0) and symmetry

? ? LF ):

land is a piece of land belonging to a clan and usually divided among clan members. The head of the clan still takes custody of the land and makes sure it is handed down from generation to generation. No sale or alienation of such a piece of land is permitted. Portions may, however, be given out to non-members on temporary basis.

400

ln C ? ? 0 ? ? F ln WF ? ? L lnWL ? ? K ln WK ? ? Q ln Q ? 0.5? QQ(ln Q)2 ? ? TT ? 0.5? KK(lnWK )2 ? ? QF ln QlnWF ? ? QL lnQlnWL ? ? QK ln QlnWK ? ? FT lnWFT

? ? LT ln W LT ? ? KT ln W K T ? ? QT LnQT ? ? TT T

… (2.1)

S F ? ? F ? ? FF ln WF ? ? FL ln WL ? ? FK ln WK ? ? F ln Q ? ? FTT

… (2.2)

S L ? ? L ? ? FL ln WF ? ? LL ln WL ? ? LK lnWK ? ? L ln Q ? ? LT T

… (2.3)

RQ ? ? Q ? ? QF ln WF ? ? QL ln WL ? ? QK ln WK ? ? QQ ln Q ? ? QTT

… (2.4)

where WF , WL and WK are prices for aggregate fertilizer, aggregate labor and capital, respectively; SF and SL are the cost shares of fertilizer and labor (capital cost share equation was dropped); RQ is the revenue cost share (ratio of revenue and total cost) equation. The specification ensured adding-up and homotheticity. Negativity condition is satisfied when the own -price Allen partial elasticities of substitution are less than zero for each input (Moschini et al. 1994):

?

A ii

?

? ii ? S i ( S i ? 1) ? 0. S 2i

… (3)

Marginal cost (MC) and input demand functions directly related to the system of estimated equations were derived after statistically testing the restrictions and checking of parameter estimates to ensure model consistency. The translog MC function is expressed as:

MC ? {?

Q

? ? QQ ln Q ?

??

Qi

ln Wi }AC

… (4)

i

where AC is average cost and Wi is the ith input price. Implications of any policy change such as an introduction of any tax or subsidy may be observed from changes in MC. The fertilizer price elasticity of MC, which represents the perce ntage change in MC due to a 1% change in the price of fertilizer, captures the effect of the tax or subsidy. The price elasticity of MC for the translog cost function is defined and computed as: PF ? MC ?

AC.? QF ? MC W F . ? ? WF MC MC

… (5)

The MC function is the supply function at all of the points above minimum average variable cost. In a competitive market settings producers supply a level of output where MC is greater than AC and often equal to the output price. Because the cost function so specified does not include all the variable costs, the elasticities of supply are short-run estimates.

401

The output supply and input demand equations may be specified as:

C ?? Q ? ? QF ln W F ? ? QL ln WL ? ? Q K ln WK ? ? QQ ln Q? … (6) P C F? ? F ? ? FF ln W F ? ? FL ln WL ? ? FK ln WK ? ? QF ln Q … (7) WF

Q?

?

?

By Shephard’s lemma, the partial derivative of the input demand function with respect to the input price produces the output compensated own -price input demand elasticity computed as:

? WFF ?

? F WF . ?W F F

? Q? 0

?? ? ?? FF ? SF

? ?? ? S F ? 1 ?

… (8)

The own -price compensated elasticity, or the substitution effect, represents the movement along an isoquant measuring the change in fertilizer use at a given output level (Heathfield and Wibe, 1987). If by introducing a policy such as a tax (subsidy) alters the price upwards (downwards), the higher (lower) price of the target input, fertilizer in this case, would result in two negative (positive) effects on the quantity demanded for fertilizer: the substitution effect and an output effect (Binger and Hoffman 1998, Nicholson 1998; Varian 1985). This substitution effect is given by :

?F ? WF

? ? Q? 0

F C C .S F ? 2 .? FF ? 2 .S F WF WF WF

… (9)

The output effect which is the change in fertilizer demanded as a result of the higher (lower) price of fertilizer causing costs to rise (decline) which, in turn, causes the supply to decrease (increase) is given by:

?F ? WF

? ? Q? 0

AC.? QF ?Q MC .S F .? . WF WF2 ? WF

… (10)

Multiplying both equations by the input price, for example WF, and dividing by the quantity of the input, F, expresses the total change in demand in elasticity form followed by a small manipulation to obtain the uncompensated input demand elasticity given as:

? WFF ?

?F WF . ? WF F

? ? Q? 0

? F WF . ? WF F

? ? FPF ? ? Q? 0

? WSF ? SF

… (11)

402

Consequently the uncompensated input demand elasticity depends on the magnitude of the elasticity of substitution of each input type by the other and labor, the input price elasticity of supply, and the input quantity elasticity of supply (Garcia and Randall 1994).

Cost function estimation results The descriptive statistics of the variables in the estimation and in the computation of the elasticities are presented in Table 1. The e stimated regression results for the base case and the tax and subsidy scenarios are reported in Table 2. All the estimated coefficients show expected signs. The three inputs are shown to be Allen complements implying that, if fertilizer price goes up, for instance, less of it is purchased inducing a decrease in the demand for both labor and capital ad vice versa. Table 1. Descriptive statistics of variables used in estimation of elasticities. Units Mean Std dev Minimum Maximum 140.80 1034.20 Maize output (1,000 mt) 554.61 264.68 16.54 8.80 64.64 NPK fertilizer (1,000 mt) 34.66 8.27 4.40 32.32 SA fertilizer (1,000 mt) 17.33 36.37 15.38 164.72 Labor (1,000 mandays) 67.02 0.12 845.67 Maize price (1,000 ¢/mt) 100.49 196.16 223.54 0.31 780.00 NPK price (1,000 ¢/mt) 126.56 268.64 0.23 860.00 SA price (1,000 ¢/mt) 137.48 0.66 0.01 2.05 Labor wage (1,000 ¢/manday) 0.43 Note: N=29Sources: FAO, 1998-2000; PPMED, 1998, 1999; ISSER, 1998. SA = sulphate of ammonia.

Elasticity estimates Tables 3 and 4 compare the price elasticities of input demand, the MC elasticities with respect to input price and quantity, and the supply elasticities with respect to own price, input price and quantity under the base case and the two policy regimes scenarios. Homogeneity and symmetry are statistically tested using Chi -square test by the likelihood ratio method (Lopez 1988; Moschini et al. 1994; Kmenta 1986) at the 5 % level and found to be statistically significant for the three models. This implies that the restrictions are necessary for the theoretical condition(s) to be satisfied. Adding-up and homotheticity conditions were effectively imposed with cross-equation symmetry and homogeneity restrictions. Negativity conditions are satisfied because the own -price Allen partial substitution elasticities are negative for all the models (Tables 3, 4). Monotonicity in input prices requires the cost shares to be greater than zero

( S i ? ? i ? 0)

at the point of approximation. This

403

condition im plies that the cost function is non-decreasing in input prices. Monotonicity in output requires that MC be greater than zero at the point of approximation (i.e.,

? Q ? 0 ), the cost

minimization hypothesis restriction on the first-order coefficient on quantity. Table 2. Maize cost of production estimation results for Ghana using seemingly unrelated regression, 1970–98. Coefficient

Base case

Coefficient Base case

0.726*** 0.110 ? QF ?F (0.117) (0.119) 0.002 0.052*** ? QL ?L (0.010) (0.012) 0.272** -0.162 ? QK ?K (0.115) (0.126) -0.077 -0.001 ? ?Q FT (0.209) (0.004) -0.015** 0.000 ? LT ?T (0.007) (0.000) 0.125* 0.001 ? KT ? FF (0.062) (0.004) -0.024*** 0.984 ? QQ ? FL (0.005) (0.722) -0.101 0.006 ? QT ? FK (0.061) (0.032) 0.050*** -0.002*** ? TT ? LL (0.002) (0.001) -0.026*** INT 2.203*** ? LK (0.006) (0.116) 0.128** ? KK (0.061) Note: SE are in parenthesis; n = 29. *,**,*** Significant at 0.10, 0.05, and 0.01 probability level, respectively. Table 3. Input demand, marginal cost, and supply elasticities for output price and fertilizer prices and quantities under the base case scenario.

Allen Partial Elasticity of Substitution Input cost share evaluated at the mean Compensated own price elasticity of demand for input Uncompensated own price elasticity of demand for input

Fertilizer –0.398 0.554

Labor Capital –2.740 –0.749 0.067 0.379

–0.220

–0.184

–0.284

–0.725 0.569

–0.278 0.074

–0.666

MC elasticity with respect to input quantity

0.855

0.793

0.935

Elasticity of supply with respect to input quantity Elasticity of supply with respect to input price

0.137

0.127

–0.069

–0.012

0.150 –0.057

MC elasticity with respect to input price

0.357

Elasticity of supply with respect to output price

0.16

Average cost evaluated at the sample mean(¢1,000/mt)

9.68

Marginal cost evaluated at the sample mean(¢1,000/mt)

68.89

404

Table 4. Input demand, marginal cost, and supply elasticities for output price and fertilizer prices and quantities. Fertilizer Labor Capital 10% 10% 10% 10% 10% 10% tax subsidy tax subsidy tax subsidy Allen Partial Elasticity of Substitution Input cost share at the mean Compensated own price elasticity of demand for input Uncompensated own price elasticity of demand for input MC elasticity with respect to input price MC elasticity with respect to input quantity Elasticity of supply with respect to input quantity Elasticity of supply with respect to input price

–0.344 –0.516 –2.847 –2.538 –0.798 –0.644 0.579 0.503 0.063

0.075 0.358 0.420

–0.199 –0.260 –0.179 –0.192 –0.286 –0.272 –0.727 –0.717 –0.266 –0.300 –0.645 –0.697 0.594 0.520 0.069

0.085 0.337 0.396

0.860 0.844 0.799

0.779 0.937 0.931

0.132 0.147 0.123

0.136 0.144 0.162

–0.070 –0.067 –0.011 –0.015 –0.052 –0.069

Elasticity of supply with respect to output price

0.154 0.174

Average cost evaluated at the sample mean (¢1,000/mt)

10.32

Marginal cost evaluated at the sample mean (¢1,000/mt)

76.07 56.68

8.60

This implies that the cost function is non-decreasing in output. Monotonicity of prices and output is satisfied in each case. Since there is only one output, convexity of the cost function requires that the second-order coefficient on quantity should be non-

?

?0

negative (i.e. QQ ). With respect to the curvature conditions of the cost functions, one must determine whether the model is concave in input prices and convex in output (Capalbo and Antle 1988) 6. The curvature conditions were checked locally and are satisfied. The compensated and uncompensated demand elasticities for the three inputs, fertilizer, labor and capital are all less than zero in each case, implying that the demand for the inputs are reduced with an increase in the respective input price. The uncompensated demand elasticities are larger in absolute values than the compensated elasticities reflecting that the output effect moves in the same di rection as the substitution effect from a ‘The concavity condition of the cost function with respect to P, is based on the second-order Hessian matrix of partials of input prices on cost, requiring the Hessian to be negative semi-definite. A necessary and sufficient condition requires that the principal minors of the Hessian matrix alternate in signs (between non-positive and non-negative values, starting with a no-positive value for the first principal minor). The convexity of the cost function is based on the second-order partials of cost with respect to output. The necessary and sufficient condition for convexity requires that the Hessian matrix with respect to output needs to be positive semi-definite. 6

405

change in input price. As indicated in Table 3 the uncompensated elasticity of fertilizer demand for Ghana is –0.725, higher than observed for labor or capital. When the policy scenarios are implemented, the uncompensated elasticities of substitution are similarly larger than the compensated elasticites (Table 4). The output price elasticities of supply are positive implying that producers would be willing to increase maize supply in expectation of higher profits. Under the policy scenarios, producers are more willing to increase production with the subsidy compared to a tax. This appears to be low but justified because of the high risk of agriculture in the area. As a result of this risk, farmers are less prepared to intensify production. The input price elasticities of supply are negative as expected implying that as the price of an input increases, the costs increase and producers reduce output. The positive inputs price elasticities of MC seem to suggest that none of the inputs is inferior. A unit increase in the price of input results in an increase in MC. The change is larger for fertilizer than for capital, the second largest cost component of maize production. Similarly, the MC elasticities with respect to input quantity are positive, suggesting that reductions in the quantity of an input used (holding the price of the input constant) reduces output, resulting in a decrease in costs. This elasticity is interpreted as reflecting changes along the MC curve. Potential implications of imposing a tax or a subsidy on fertilizer To provide an indication of the magnitude and direction of the changes in maize supply and fertilizer demand of introducing a tax or subsidy on fertilizer, we employ the elasticities of supply with respect to fertilizer price and quantity as well as the input demand elasticities in the analysis. The percentage changes in supply and fertilizer demand from a fertilizer tax or subsidy are reported in Table 5. At the current level of fertilizer consumption and prices in the country, the projected tax revenue by imposing a 10% tax on fertilizer would be 1.13 billion cedis, about 60% (676.9 million cedis) of which would be from the maize sector alone. Considering the maize sub-sector alone (since it is the scope of the study), the tax will reduce fertilizer consumption by 15%. This suggests that only 85% (580.23 million cedis) of the projected tax revenue would be obtained. As a result of the reduced demand for fertilizer, the fertilizer distribution sector will experience a reduction in fertilizer gross sales revenue of 1.04 billion cedis related to sales of fertilizer for maize. (It is reasonable to assume that the elasticity of demand for fertilizer for maize is lower than for other crops. Therefore, it is unlikely that the consumption of fertilizer in the other crop subsectors would increase to compensate for the reduction in the maize sector.)

406

Table 5. Partial analysis of the incidence of a tax or a subsidy on fertilizer on the maize subsector 1. Before policy (Quantity)

After tax % change Quantity

After subsidy2 % change Quantity

Quantities (‘000MT) NPK fertilizer

34.66

-15.4 (90.7)

29.46

22.4 (40.6)

42.29

Sulphate of Ammonia3

17.33

-15.4 (-90.7) -23.5 (-81.2)

14.73

22.4 (40.6) 27.0 (88.5)

21.14

Maize supply

554.61

Costs and revenues (Millions of Cedis) Value of fertilizer used 6,769.10 Value of other inputs 5,538.35 Tax revenue or subsidy cost4 0.00 Net revenue from 5 maize 43,425.31

421.50

704.35

5,802.34

8,258.30

4,747.37

6,756.79

580.23

825.83

31,807.18

56,591.34

Change in revenue of maize farmers6 -11,618.13 13,166.04 1In parentheses are the ranges. Note: 2Although the price is stated as was the case before the policy, government subsidy cost absorbs 10%. 3Also includes urea. 4Tax revenue is an income transfer from maize farmers to government while subsidy cost is an income transfer from the government to maize farmers. 5Values reflect only gross revenues from maize production. 6Maize farmers lose (11.6 billion cedis) with the tax but gain (13.2 billion cedis) when fertilizer is subsidized.

Hence the reduction in fertilizer gross sales revenue as a result of the tax would have far reaching consequences on the fertilizer distribution sector of the economy. Regarding the policy impacts on the maize production sub-sector, the results suggest that the tax will unambiguously increase marginal cost of production with corresponding worsening terms of trade effect on maize farmers. Maize supply would decrease by 24% with a corresponding decrease in net sales revenue loss of about 11.62 billion cedis compared with the case without the tax. This is more than 10 times the projected gains in government’s total tax revenue from fertilizer sales. Furthermore, the uncompetitive maize sector would lose labor to other sectors demanding less of fertilizer. This would have adverse effect on the maize subsector in particular and agriculture in general (since maize is one of the largest subsectors of the agricultural sector). Ignoring the impact of the policy on the fertilizer distribution and related sectors, the tax revenue to government will be inadequate

407

to compensate for the loss in revenue to maize farmers. Therefore such a policy will not be socially optimal and should be avoided. In contrast, the results suggest that a 10% subsidy on fertilizer has the potential of increasing fertilizer demand by 22%, more than twice the value of the subsidy. This will mean an increase in fertilizer gross sales revenue by 22% (1.5 billion cedis) over and above the situation without the subsidy. It is reasonable to suggest that such a change in gross revenue would expand the industry thus creating jobs in the fertilizer distribution sector. This change however has a budgetary implication of 825.8 million cedis, which has to be financed somehow by the government. The subsidy cost is an income transfer from the government to farmers. The policy measure will also lead to a decrease in marginal cost by 17%. The combined effect of the increased use of fertilizer and decreased marginal cost would be an increase in maize supply by 27% (nearly three times the rate of the subsidy), similar to the observations made by Roth and Abbott (1990) in their analysis of agricultural input subsidy reforms in Burkina Faso. Net revenue (including subsidy cost transferred) to farmers would increase by 30% (13.2 billion cedis) over and above the situation without the subsidy. This would attract more farmers and labor invariably increasing employment in the sector. One would also expect spillover effects from the expanded maize supply; for example, increase in employment in the food processing sector. Under perfectly competitive market conditions, maize prices will be set at marginal cost levels meaning that prices will be more affordable to the urban poor. It is worth noting that estimation of the optimal levels of the subsidy and fertilizer use in maize to guide policy decisions are relevant issues to consider but are beyond the scope of the present study. These results suggest that re-instating fertilizer subsidies would improve the performance of the maize production sector: reduce cost of production, increase maize output and improve maize farmers’ welfare. However, any such decision must take into consideration the budgetary implications and sources of funds to pay for the subsidies.

Concluding comments Within the limits of a crop specific cost function estimation, an appropriately specified and estimated translog cost function greatly enhances the examination of potential implications of input policy. The marginal cost function so derived from a traditional cost function play an important role in determining the effects of input policy on cost, supply, and input demand in the Ghanaian economy. This study examines the potential impact of imposing a 10% tax or subsidy on fertilizer in Ghana using a translog cost function (TCF). The estimated supply elasti cities generally satisfy a priori

408

assumptions. Subject to the limitations of the data, the results seem to suggest that a 10% VAT on fertilizer will decrease fertilizer demand by 15%, unambiguously increase marginal cost and depress maize production by 24%. The potential tax revenue would be inadequate to compensate for the revenue loss by maize farmers. In contrast, a 10% fertilizer subsidy will reduce marginal cost by 17%, increase fertilizer demand by 22.4% and expand maize supply by 27%. Maize farmers would experience a gain in net revenue of about 30% over the case without a subsidy. The corresponding subsidy cost would be 6% the revenue gain by farmers. These estimates should be used in policy analysis to compare with the benefits of other programs and/or the costs of other tax revenue sources.

References Akridge, J.T., and T.W. Hertel. 1986. Multiproduct Cost Relationships for retail fertilizer plants. American Journal of Agricultural Economics 68: 928–38. Assibey-Mensah, G. O. 1999. The value -added tax in Ghana. Public Budgeting and Finance 19: 76–89. Binger, B. R., and E. Hoffman. 1998. Microeconomics with calculus. Addison and Wesley. Binswanger, H.P. 1974. A cost cunction approach to the measurement of factor demand elasticities and elasticities of substitution. American Journal of Agricultural Economics 55: 377–86. Blackorby, C., D. Primont, and R.R. Russell, 1978. Duality, separability and functional structure: Theory and economic applications . Elsevier North-Holland, Inc. Capalbo, S.M., and J.M. Antle (eds.). 1988. Agricultural productivity: measurement and explanation. Washington, DC: Resources for the Future. Chamber, R.G. 1988. Applied production analysis: The dual approach . Cambridge University Press. Cones, R. 1992. Duality and modern economics. Cambridge University Press. Donhauser, F., H. Baur, and A.S. Langyintuo. 1994. Smallholder agriculture in Western Dagbon. A farming system in northern Ghana. Nyankpala Agricultural Research Report No. 10. FAO. 2000. Food and Agricultural Organization of the United Nations (FAO) Statistics. Garcia, R.J., and A., Randall. 1994. A cost function analysis to estimate the effects of fertilizer policy on the supply of wheat and corn. Review of Agricultural Economics 16(2): 215–230. Heathfield, D.F., and S. Wibe. 1987. An introduction to cost and production functions. Macmillan Education, New York.

409

IFDC (International Fertilizer Development Center). 1995. Ghana fertilizer privatization scheme. Private sector roles and public sector responsibilities in meeting needs of farmers. IFDC. ISSER (Institute of Statistical, Social and Economic Research. University of Ghana, Legon). 1999. The state of the economy in1998. ISSER, 1999. Kmenta, J. 1997. Elements MacMillan.

of

econometrics .

New

York:

Lichtenberg, E., D.D. Parker, and D. Zilberman. 1988. Marginal analysis of welfare costs of environmental policies: The case of pesticide regulation. American Journal of Agricultural Economics 70: 867–74. Lopez, R.E. 1998. The structure of production and derived demand for inputs in Canadian Agriculture. American Journal of Agricultural Economics 62: 38–45. Miller, T., and G. Tolley. 1989. Technology adoption and agricultural price policy. American Journal of Agricultural Economics. 71(4): 847–857. Moschini, G. D. moro and Green. 1994. Maintaining and testing separability in demand systems. American Journal of Agricultural Economics 76: 61–73. Nicholson, W. 1985. Microeconomic theory: Basic principles and extension . New York: CBS. Osei, P. D., 2000. Political liberalisation and the implementation of value added tax in Ghana. The Journal of Modern African Studies 38 (2): 255–278. Pollak, R.A., and T.J. Wales. 1992. Demand system specification and estimation. Oxford University Press. PPMED (Policy Planning Monitoring and Evaluation Division, Ministry of Food and Agriculture, Ghana). 1998. Annual sample survey of agriculture. Regional and district cropped area, yield and production estimates, 1998. Ray, S.C., 1982. A translog cost function analysis of U.S. Agriculture, 1939–77. Amer ican Journal of Agricultural Economics 64: 490–98. Roche, F. C. 1994. Technical and price efficiency of fertilizer use in irrigated rice production. Bulletin of Indonesian Economic Studies 30 (1): 59–83. Roth, M. J., and P.C. Abbott. 1990. Agricultural pri ce policy, food aid and input subsidy reforms in Burkina Faso. Journal of Agricultural Economics 41(3): 327–345. Silbergerg, E. 1990. The structure of economics: A mathematical analysis. Second Edition. McGraw-Hill, Inc.

410

Recent advances in the development and promotion of quality protein maize in Ghana P.Y.K. Sallah, K. Obeng-Antwi, E.A. Asiedu, M.B. Ewool, and B.D. Dzah Crops Research Institute, CSIR, P.O. Box 3785, Kumasi, Ghana. Sasakawa Global 2000, CRI, P.O. Box 3785, Kumasi, Ghana. Abstract Maize (Zea mays L.) is a very important staple crop in subSaharan Africa where half of the population consumes this cereal in various forms. The normal maize varieties commonly grown in the region contain 10% grain protein, but this protein is deficient in two essential amino acids, lysine and tryptophan. Consumption of normal maize-based foods without adequate protein supplementation, therefore, leads to widespread malnutrition, particularly among infants, pregnant women and lactating mothers. To help solve this problem, interinstitutional and multidisciplinary research was initiated in 1989 to (i) develop high and stable yielding quality protein maize (QPM) varieties which are high in these two essential amino acids, and (ii) promote the production and utilization of these varieties in Ghana and other countries. This effort resulted in the development and release of Obatanpa, a medium maturing QPM composite in 1992. In 1997, Obatanpa was planted to 21% of the 650,000 ha under maize in Ghana. Apart from Ghana, Obatanpa has been released officially in Guinea, Mali, Mozambique, and Uganda. Large-scale production of 3-way QPM hybrids, Mamaba, Dadaba and CIDA -ba, released in 1997 is currently being promoted in Ghana. Current emphasis is on developing QPM varieties of the three maturity cycles (extraearly, early and intermediate/late) that possess tolerance to the major biotic and abiotic stresses that limit maize production in the sub-region. Two new extra-early white and one yellow intermediate maturing QPM experimental varieties have been developed. Preliminary yield data across six sites in year 2000 showed that one extra-early variety designated GH99- 80DWDQPM-S yielded 5.22 t/ha compared to 4.50 t/ha for the improved normal extra-early variety. GH99- 80DWDQPM-S has been submitted for testing in the WECAMAN-coordinated regional uniform variety trials in 2001. Résumé Le maïs (Zea mays L.) est un aliment de base très important en Afrique sub-saharienne où la moitié de la population consomme cette céréale sous diverses formes. Les variétés

411

normales de maïs communément cultivées dans la région contiennent 10% de protéine grain mais cette protéine est déficiente dans deux amino acides essentiels, la lysine et le tryptophan. La consommation d’aliments à base de maïs normal sans apport supplémentaire adéquate en protéine, conduit une malnutrition répandue, particulièrement parmi les nourrissons, les femmes enceintes et les mères qui allaitent. Afin de résoudre ce problème, une recherche interinstitutionnelle et multidisciplinaire a été initiée en 1989 pour (i) développer du maïs riche en deux amino - acides essentiels, à haut rendement stable (QPM) et (ii) promouvoir la production et l’utilisation de ces variétés au Ghana et dans d’autres pays. Cet effort a conduit au développement et à la vulgarisation de Obatanpa, un composite QPM de maturité intermédiaire en 1992. En 1997, Obatanpa a été semé dans 21% des 650,000 ha réservés à la culture du maïs au Ghana. A part le Ghana, Obatanpa a été officiellement vulgarisé en Guinée, Mali, Mozambique, et en Ouganda. La production à grande échelle des hybrides QPM à 3 voies, Mamaba, Dadaba et CIDA- ba, vulgarisés en 1997 sont actuellement promus au Ghana. Actuellement l’accent est mis le développement de variétés QPM de trois cycles de maturité (extra-précoce, précoce, et intermédiaire/tardive) qui possèdent la tolérance aux stress biotiques et abiotiques majeures qui limitent la production maïsicole dans la sous - région. Deux nouvelles variétés expérimentales QPM dont deux blanches extra-précoces et une jaune intermédiaire ont été développées. Les données préliminaires ont montré qu'une variété extra-précoce désignée GH99- 80DWDQPM-S a donné un rendement de 5,22 t/ha en comparaison avec 4,50 t/ha pour la variété extra-précoce normale améliorée à travers six sites pendant l’année 2000. La variété GH99 - 80DWDQPM-S a été nominée pour être testée dans les essais régionaux uniformes de variétés cordonnés par le WECAMAN en 2001.

Introduction Maize (Zea mays L.) is an important staple crop produced in Ghana and in many other countries in sub-Saharan Africa. It is estimated that about 50% of the population in the subregion consumes maize as a staple food (Okoruwa, 1997). The grain is mainly processed and used in the preparation of several indigenous dishes. In many countries, maize is widely fed to weaning infants without protein supplement of any form. Infants fed on foods from normal maize without balanced protein supplements often suffer from protein-energy malnutrition and develop the disease known as Kwashiorkor. The main reason is that the protein in normal maize is deficient in two essential amino acids, lysine and tryptophan which humans and monogastric animals require in their diets.

412

Earlier attempts were made by the Ghanaian Ministry of Health (MOH) to resolve weaning malnutrition problems in Ghana by promoting the use of milk and maize–legume composite flour (MOH, 1988). This effort was not successful because the recommendations were not practicable. A logical solution, therefore, is to develop maize varieties that have higher levels of lysine and tryptophan to minimize malnutrition in the region. The objectives of this approach were to (i) develop stable, highyielding QPM varieties with characteristics desired by maize farmers and consumers, (ii) demonstrate the nutritional superiority of QPM in feeding trials, and (iii) promote production and utilization of QPM in Ghana and, possibly, in other countries.

Institutional role in QPM development and promotion Development and promotion of QPM were achieved through the joint efforts of several organizations whose roles were complimentary. The various organizations and their roles are indicated below: (i)

(ii) (iii)

(iv) (v)

(vi) (vii) (viii) (ix) (x) (xi) (xii)

Governments of Ghana and Canada through the Ghana Grains Development Project (GGDP)-provision of financial support for research, extension and training; Sasakawa Global 2000 (SG 2000)— provision of financial support for research, QPM laboratory, irrigation facilities and extension; Crops Research Institute (CRI)—lead Institute for variety development, testing, promotion, and breeder seed production; Nyankpala Agricultural Experiment Station (NAES)-collaboration in variety development and testing as well as breeder seed production; Ministry of Food and Agriculture (MOFA)-on-farm testing, sensory evaluation, seed production and promotion through extension; Nutrition Department, Ministry of Health (MOH) --infant feeding studies; Kwame Nkrumah University of Science and Technology (KNUST)-Animal feeding studies using pigs and poultry; International Maize and Wheat Improvement Centre (CIMMYT)-supply of germplasm; International Institute of Tropical Agriculture (IITA)-supply of germplasm and conversion to streak resistance; West and Central Africa Collaborative Maize Research Network (WECAMAN)-provision of supplementary funds for variety development and testing; National Agricultural Research Project (NARP)-provision of supplementary funds for varietal development and testing; Privates sector partnership: Asare Farms and Company Limited-commercial feeding of poultry

413

(xiii)

Privates Seed production.

Growers

Association-certified

seed

Historical background of QPM development in Ghana Interest in improving the quality of the protein in maize began in the mid-60’s when mutants including opaque -2 which exhibited higher levels of lysine and tryptophan were discovered by Mertz et al. (1964) in USA. In Ghana, QPM research started in the early 1970s with little progress because of problems associated with the initial opaque -2 materials available. Some of the problems were low yield potential, poor plant vigor, low seed viability, and high susceptibility to ear rot and storage pests. These problems were largely overcome through intensive research at CIMMYT resulting in the development of modified opaque-2 maize germplasm (Vasal et al. 1993). The modified opaque-2 materials had desirable characteristics that were not present in the opaque2 version. Recent advances in QPM variety development in Ghana Availability of a large pool of germplasm with modified opaque-2 endosperm (QPM) from CIMMYT (Vasal et al. 1993) rekindled the interest of the Ghana Maize Program and consequently, the QPM variety development effort was intensified in 1989. Two CIMMYT QPM populations, Populations 63 and 62, selected for superior yield potential from multilocation trials involving a large collection of QPM germplasm, formed the base materials for further improvement. These populations, as well as the other QPM materials, could not be released directly to farmers because of susceptibility to the maize streak virus disease, loose husk cover and exposed ear tips, segregating grain types, and low yields compared with the popular normal maize varieties. A combination of crossing, selection, and evaluation for these desirable traits on-station and on-farm in population 63 resulted in the development of GH8363-SR, an intermediate maturing (105–110 days), streak resistant, white, semi-dent, composite variety. Improvement for resistance to the maize streak virus was done in collaboration with IITA, Ibadan, Nigeria. Farmers, extension staff of MOFA, and researchers (breeders, agronomists, socioeconomists) collaborated in the evaluations in farmers’ fields throughout Ghana in 1991 and 1992. Apart from its superior quality protein (Table 1), GH 8363-SR was also superior or comparable to the popular normal maize varieties in yield and other agronomic traits. For example, GH8363-SR had a yield potential of 5.5 t/ha similar to improved intermediate and late maturing normal maize varieties while the local variety (landrace) yielded 3.5 t/ha on research station (Sallah et al. 1997a; Twumasi -Afriyie et al. 1992). In farmers’ fields, the new QPM variety yielded 3.2 t/ha compared to 3.0 t/ha for the improved normal maize varieties, and 3.0 t/ha for farmers’ variety (data not shown). In this case, farmers’ variety denotes improved normal

414

maize varieties volunteered by farmers who collaborated in the on-farm evaluations. GH8363-SR was named Obatanpa which literally means Good Nursing Mother and was released to farmers in 1992. Table 1. Essential amino acid contents of quality protein maize (QPM) and normal maize varieties in Ghana and amino acid requirements for children and adults. Amino acis Normal maize requirements Oko1-yr Pre-school Amino acid masa Abeleehi olda or adultb ----------------------g/100g protein------------------------Threonine 3.50 2.47 2.78 4.3 2.5 Cystein + Methionone 4.21 3.70 5.02 4.2 2.5 Valine 4.93 3.39 5.25 5.5 3.5 Isoleucine 3.08 2.36 3.42 4.6 3.5 Leucine 9.05 7.92 11.43 9.3 6.5 Tyrosine + Phenylalanine 7.40 6.59 5.61 7.2 6.5 Histidine 3.60 2.26 3.66 2.6 Lysine 3.70 2.36 3.10 6.6 5.0 Tryptophan 1.03 0.62 0.61 1.7 1.0 Leucine/Isoleucine ratio 2.93 3.35 3.34 Protein (%) 9.73 9.86 9.87 a FAO/WHO (1991) recommendations. bYoung and Pellet (1991) suggested requirements. QP M Obatanpa

During the development of the open-pollinated QPM varieties, a QPM hybrid maize development program was concurrently initiated in 1991, using populations 62 and 63 as sources of inbred lines. Several QPM hybrids developed in this program were evaluated extensively on-station and in farmers’ fields in Ghana from 1995 to 1997 (Twumasi-Afriyie et al. 1997a). In addition, six QPM hybrids, Obatanpa and open-pollinated normal maize checks were evaluated in a CRI -coordinated international variety trial in over 20 countries in Africa, Central America, South America and Asia in 1995 and 1996 (Twumasi -Afriyie et al. 1997b). High and stable yielding three-way hybrids were identified from these evaluations. These were named Dadaba (Daddy’s child), Mamaba (Mother’s child) and CIDA-ba (CIDA’s child) with yield potential of 6.8, 7.3 and 6.3 t/ha, respectively. The hybrids were released by the Ghana National Variety Release Committee in 1997 (Twumasi-Afriyie et al. 1997a). Current programs and achievements

Development of extra-early and early QPM Extra-early (75–80 days) and early (90–95 days) maturing QPM varieties are available in normal maize germplasm for early planting to help break the hunger gap. The early crop is often harvested as green maize, which is boiled or roasted and eaten on the cob. One normal maize extra-early variety named Dodzi has been released in Ghana (Sallah et al. 1997b) and farmers

415

continue to show keen interest in this variety. Extra-early and early QPM will have an added advantage of being more nutritious, because green maize is generally consumed without protein supplement. The objective of the program is to develop and make available to farmers high yielding, disease resistant early and extra-early QPM varieties using the backcross breeding procedure. Two extra-early QPM experimental varieties have been developed. Three new extra-early maturing QPM experimental varieties were evaluated along with nine normal extra-early and early maize varieties at 10 locations in 2000. Results from six sites showed that the yield of two extra-early QPM conversions averaged 5.1 t/ha compared with 4.5 t/ha for Dodzi, the normal maize variety used as the recurrent parent (Table 2). One promising extra-early QPM variety, designated GH 99-90DWQPM-S, has been nominated for inclusion in the regional uniform variety trials coordinated by WECAMAN.

Development of yellow QPM varieties Yellow QPM is important for the poultry industry in Ghana and for countries where yellow maize is the preferred grain type. The development of yellow maize varieties had not received much attention in our program in the past because of lack of demand. Poultry farmers are now demanding high and stable yielding yellow QPM varieties to substitute imported yellow maize. This program aims at developing high yielding and disease resistant yellow QPM composites for the livestock industry. Table 2. Grain yield an d agronomic traits of early and extra-early maturing maize varieties evaluated at 6 locations during the 2000 major season. Variety

EV EJ9190DWDP Dorke SR Safita-2 90 DWD Pop Kawanzie 90 DYF Pop EV9980DWD QPM-S EV9980DWD QPM-T NAES Pool 16 DT Dodzi Local Catete Hi-Fe Mean LSD (0.05) CV (%)

Grain yield Mid-silk kg/ha

days

Plant height cm

5805 5767 5545 5507 5237 5227 5220 4987 4752 4502 4071 3747 5031 349.1 3.3

48 53 49 47 49 48 45 45 47 46 59 47 49 0.8 2.8

190 214 199 201 195 191 187 188 184 191 221 193 196 7.9 7.1

Ear Total aspect lodging score % 2.0 2.2 2.3 2.2 2.2 2.2 2.7 2.5 2.6 2.5 2.8 3.1 2.4 0.3 18.9

27 36 31 25 22 24 31 40 29 39 30 33 31 6.9 39.0

416

The QPM populations under improvement are EV 8766 SR BC6, a yellow, streak resi stant, intermediate QPM variety obtained from IITA; and BR454 and BR 474 from Brazil. These populations are being improved for yield, disease, pest, and lodging tolerance through recurrent selection. In addition, a program was initiated in 1999 to convert Obatanpa to yellow QPM using the backcross procedure. Obatanpa is a highly productive composite that is popular among farmers in Ghana and other countries in West, Central and Eastern Africa. Livestock farmers in Ghana now prefer Obatanpa in feed formulations because of its high nutritional quality. Conversion of Obatanpa to yellow grain color will further enhance its acceptance and adoption, especially by poultry and livestock farmers. One experimental yellow QPM variety, GH9866-SR has been developed from population 66 -SR. Across six sites in 2000, GH9866-SR exhibited a yield potential of 4.85 t/ha compared to 6.24 t/ha for Obatanpa (Table 3). Further improvement for higher productivity needs to be carried out in this variety. Table 3. Grain yield and agronomic traits of medium-late maturing maize varieties evaluated at 6 locations during the 2000 major season. Variety

Grain yield kg/ha

(GH3 x 1368) x 5012 GH132-28 Dobidi Obatanpa 8321-18 GH2328-88 (GH28 x 1368) x 5012 GH110-28 Okomasa GH110-5 Abeleehi Aburohoma GH9866 SR BR 474 YQPM BR 454 YQPM Local Mean LSD (0.05) CV (%)

6907 6435 6313 6244 6155 6123 6076 6036 6000 5874 5305 5081 4853 4831 4771 3949 5685 504 15.6

Mid-silk Plant height days cm 54 55 57 53 55 52 55 53 58 52 54 54 52 54 54 57 55 1.0 3.3

221 216 223 227 213 201 215 205 246 189 195 251 208 205 196 238 215 9.9 8.1

Ear aspect score

Total lodging %

2.5 2.3 2.4 2.3 2.3 2.3 2.6 2.3 2.2 2.5 2.3 2.4 2.5 2.7 2.8 2.7 2.4 0.3 19.5

19 28 19 28 34 20 30 32 26 47 23 30 31 19 26 26 27 7.0 45.1

Hybrid development Line development in GH 9163 SR and in EV 8762 SR populations. The QPM populations 62 and 63 from CIMMYT are two heterotic, white -endosperm materials with potential for high yield that are currently under improvement in the breeding program. Inbred lines from these populations have been released

417

as parents of the three -way hybrids Mamaba, Dadaba and CIDAba. However, the released lines have several defects such as low yields of the lines, susceptibility to the maize stre ak virus and lodging. A program was initiated in 1997 to develop superior inbreds from advanced cycles of the two populations. The lines are at the S5 stage in GH 9163 SR and S6 stage in EV 8762 SR, respectively. Based on general combining ability evaluati ons, 10 best S3 lines were selected from GH 9163 SR to form experimental variety designated GH 9963-SR and 10 from EV 8762 SR to form GH 9962 -SR. In addition, six drought tolerant S3 lines from EV 8762 SR were recombined to form GH 9962 SRDT. Conversion of popular normal maize inbreds to QPM . Several normal maize inbreds with resistance/tolerance to MSV and Striga from IITA and Ghana are widely used as parents of important hybrids. A program was initiated in 1997 to convert these normal maize inbreds that combine well with QPM materials to QPM. The inbred lines are GH20, GH22, GH24, IITA 1368, Pop10-1, 4058, TZMi301 and IITA 9071. Each line was crossed with QPM inbred line Entry 5 as the donor parent and as female. These lines are being converted to QPM using backcrossing and selection to develop superior QPM lines. Improvement of released QPM lines. Several QPM inbred lines have been developed in our breeding program. Eight of these lines namely, ENT 5, ENT 6, ENT 24, ENT 27, ENT 70, ENT 88, P23, and P28 are the parents of the three -way hybrids Dadaba, Mamaba and CIDA -ba that were released in 1997. However, these lines have some undesirable attributes that are transmitted to their hybrid progenies, such as low yield potentials of the lines per se, suscepti bility to lodging, common rust, and drought stress. A program was therefore initiated in 1998 to recycle these important inbred lines. Initially, crosses involving elite QPM inbred lines were made to improve on the performance of the lines per se as well as their hybrids, using a combination of sib-mating and selfing.

Promotion of QPM Following the release of Obatanpa and the QPM hybrids, several strategies were adopted to promote the production and utilization of QPM in Ghana and in other African countries. Examples of the strategies used are as follows: On -farm demonstrations. On-farm demonstrations of Obatanpa and QPM hybrids were established throughout Ghana by SG 2000, CRI-GGDP, and MOFA. Sensory evaluations. Consumer acceptability is an important criterion for the adoption of new varieties. Prior to its release, sensory evaluations of Obatanpa in local dishes were conducted

418

in collaboration with Women in Agricultural Development (WIAD) of MOFA, maize consumers, indigenous maize processors, food scientists, and SG2000. Results showed that the overall acceptability of local dishes made from QPM varieties was high and comparable to those made from the preferred local maize variety (Ahenkora et al. 1995; 1999). Infant feeding studies. Infant feeding studies were conducted using QPM and normal maize in the Ejura-Sekodumase District of Ashanti Region of Ghana, in collaboration with the Nutrition Department of Ministry of Health, Kumasi. In addition, feeding trials were conducted on pigs and chicken in collaboration with the Animal Science Department of Kwame Nkrumah University of Science and Technology, Kumasi which demonstrated the superior nutritional and growth promoting qualities of QPM over normal maize (Okai et al. 1994; Osei et al. 1994a, 1994b, 1994c). These studies showed that (i) QPM improved the nutritional and health status of children when maize constituted the major component of the diet, (ii) children fed on QPM gained more height and had better chances of escaping death from diarrhoea and other diseases than those fed on normal maize, (iii) piglets fed solely on Obatanpa gained 13.9 kg body weight at 16 weeks compared to 5.9 kg for those fed on normal maize for the same period, and (iv) cost per kilogram feed was reduced by 29.4% for broilers and by 18.0, 12.6, and 12.8% at starter, grower, and finishing phases for pigs, respectively, when QPM was substituted for normal maize in these feeds (Okai et al. 1994; Osei et al. 1994a, 1994b, 1994c). Production of breeder’s seed. Breeder’s seed of Obatanpa and the inbred parents of Mamaba hybrid are supplied annually to the seed industry in Ghana. Breeder’s seed of Obatanpa is also supplied to several countries that have released or intend to release Obatanpa. In addition, breeder’s seed and seed increases of single and three-way crosses of Mamaba, Dadaba and CIDA -ba were produced to meet local demands and requests from abroad. Distribution of Fact Sheets. Fact sheets on production and utilization of QPM were produced and distributed in Ghana and other countries. Demonstrations of nutritive value. Demonstrations of (i) pigs fed on QPM diets compared with those fed on normal maize diets, (ii) ears of QPM and normal maize varieties, and (iii) local dishes prepared from QPM were mounted at important functions such as the Annual Farmers’ Day, African Day of Scientific Renaissance, and agricultural shows at national, regional and district levels.

Impact of QPM technology in Ghana Research on QPM in Ghana continues to have profound positive effects on mai ze production and utilization and on agricultural

419

development in general in the country. Specific areas of impact are as follows: Adoption of improved maize varieties. Adoption of improved maize varieties is high, averaging about 90% in the major maize growing zones of Ghana. Since its release, Obatanpa, has been widely adopted by farmers throughout Ghana, accounting for over 90% of improved seed sales. In Ghana, 1000–1400 tons of Obatanpa commercial seed was produced annually from 1996 to 1998. However, m ost of the Obatanpa now planted by farmers is from seed that has been saved by farmers and distributed to other farmers. In 1997, Obatanpa was planted to about 21% of Ghana’s maize production area of 650 000 ha (Morris et al. 1999). Seed industry development. The release of Obatanpa greatly promoted the development of the seed industry in Ghana. Many private, small-scale seed production and distribution enterprises have sprung up to market certified seed throughout the country. Increased job opportunity. There is a very high demand for certified seed of Obatanpa as well as the grain due to its nutritional attributes. The QPM technology thus provides jobs and income for millions of maize farmers, seed growers, seed distributors, and grain sellers. Increased utilization of maize . There has been a tremendous increase in utilization of maize in Ghana due to increased production of maize of high nutritional quality. Since the release of Obatanpa in 1992, maize production increased from 703 600 metric tons to over one million metric tons in 1998 (PPMED, 1999). This increased production of maize definitely contributes to national food security in Ghana. Several agencies now utilize QPM in various forms, such as in infant formulas, maize grits for the brewery industry, dried fermented flour, relief donations, feeding malnourished infants, livestock feed, and in educational institution feeding using QPM grain produced from school farms. Increased production and consumption of maize. Maize is a major staple cereal in Ghana and its consumption per capita is approximately 55 kg/person/year. Increased production and consumption of QPM is, therefore, enhancing the nutritional status of farm households and the general public, particularly children. Infant nutritional studi es with QPM and normal maize have shown that QPM improves the nutritional and health status of children when maize formed a major portion of the diet. Utilization of maize by industries. Ghana Agro-Food Company Limited (GAFCO) plans to purchase 600,000 tons of QPM grain for the production of poultry feed beginning from year 2000/2001 and this demand is expected to further promote QPM production and utilization in Ghana. Analyses of maize grain by General Mills Ghana Ltd. has shown that Mamaba and Obatanpa are more

420

suitable for the production of grits for brewery than the normal maize varieties in Ghana. The company now purchases Mamaba and Obatanpa grain in preference to normal maize for grits production for the breweries.

Impact of Ghana’s QPM in other countries When our research findings on quality protein maize were presented at the Regional Maize Workshop organized by WECAMAN in 1995 at IITA-Cotonou, Benin Republic (Twumasi -Afriyie et al., 1997b), it was viewed as a success story by the participants. At the end of the Workshop, maize scientists from West and Central Africa made a strong recommendation for seed multiplication and immediate release of Obatanpa throughout the sub-region. Since that time, overwhelming demand for Obatanpa seed have been received from several countries within and outside the African continent, including, Benin, Burkina Faso, Nigeria, Togo, The Gambia, Senegal, Sierra Leone, Mali, Ethiopia, Tanzania, Congo, Mozambique, Guinea, Uganda, Cote d’Ivoire, South Africa, and Guatemala. To date, Obatanpa has been released officially in Guinea, Mali, Mozambique and Uganda. Testing of Obatanpa has been going on in several other countries. A seed grower in southern Burkina Faso purchased Obatanpa foundation seed from Ghana and produced 12.0 and 20.0 tons Obatanpa commercial seed in 1998 and 1999, respectively. This grower claims that there is considerable demand for Obatanpa seed in his country and purchased 1350 kg foundation seed for the 2000 growing season. Maize breeding programs in Ea stern and Southern Africa have also expressed keen interest in Obatanpa and our new QPM hybrids. Several countries in the subregion continue to demand seed of our QPM varieties. Tanzania and Ethiopia are undertaking further field evaluations prior to release while the QPM three-way hybrid Dadaba has been released by Quality Seed Company in South Africa. This company intends to release Dadaba throughout southern and eastern Africa and has been granted permission to do so. Similarly, National Tested Seeds in Zimbabwe has shown keen interest in producing seed of Obatanpa for relief agencies in southern Africa. In 1994, Guinness Brewing Worldwide tested four QPM varieties developed in our program for their malting properties in laboratories in the Republic of Ireland. Two QPM hybrids were identified based on the quality of the protein needed to provide the required amino acid base for the growth of yeast for the brewery. Arrangements were almost concluded with the company to use these QPM hybrids for malting in countries where there were limitations on importation of barley malt. Unfortunately, import restrictions were later relaxed in those countries and the planned large-scale production of the hybrids did not take off.

421

Seed production and distribution of Oba tanpa was initiated in the Republic of Congo in 1996 under the direct supervision of the First Lady of that country at that time.

Plans for future QPM research and extension The tremendous success of Obatanpa assures us that if high and stable yielding, nutritionally superior, and consumer preferred varieties are developed and made available, they are likely to be adopted by growers, processors and consumers. The Ghana Maize Program now devotes about 80% of the breeding effort to the development of nutritionally superior QPM varieties. ?? Apart from lysine and tryptophan, several other nutrients are limiting in maize grain, for example iron. Iron is the most common micronutrient deficient in infants, young children and pregnant and lactating mothers in deve loping countries. Iron is a constituent of red blood cells. Iron also plays an essential role as a component or cofactor of several metabolic enzymes involved in processes such as energy production, fatty acid biosynthesis, and immune response. Iron defici ency, therefore, results in conditions such as anaemia and reduced immune response. For example, it has been estimated that seven out of ten pregnant women in Ghana are anaemic. Improved maize varieties that combine protein quality with high iron content need to be developed to further enhance the nutritional quality of maize. The QPM maize varieties now grown by farmers are susceptible to several abiotic and biotic stresses. The common stress factors that contribute to the low maize yields on-farm in sub-Saharan Africa include: ?? Drought stress, particularly in the southern and northern savanna zones as well as forest and transition zones in the minor season; ?? Striga parasitism in the savanna zones; ?? Low soil fertility; ?? Diseases such as Curvularia leaf blight, maize mottle virus, downy mildew; ?? Field and storage pests such as stem borers during the minor season in the coastal savanna, forest and transition zones; termite, larger grain borer and weevils; ?? Yield loss due to root and stem lodging The development of QPM varieties that possess tolerance/resistance to these stresses needs emphasis to ensure yield stability. Availability of stress tolerant QPM varieties will greatly enhance adoption not only by Ghanaian farmers but by farmers throughout the subregion.

422

Production and utilization of QPM varieties in West and Central Africa will continue to be vigorously promoted.

Conclusion This program sought to develop high and stable yielding quality protein maize varieties for maize farmers and consumers in Ghana. This activity was one of several breeding projects undertaken by the Ghana Maize Breeding Program during the implementation of the Ghana-CIDA Grains Development Project. One open-pollinated QPM variety, Obatanpa, developed through this effort has been well adopted in Ghana and other countries. To a large extent, Obatanpa has replaced the improved normal maize varieties that were previously very popular with growers in Ghana. Some of the factors that contributed to the success of Obatanpa are: •

•

•

• •

Well trai ned and dedicated research team convinced about the nutritional qualities of QPM and were all out to prove the point that QPM is beneficial; Continuous support by the GGDP and SG 2000 in training, breeding, agronomic research, demonstration, seed production and extension of the QPM technology; Strong research–extension–farmer linkages developed during the implementation of the Ghana-CIDA Grains Development Project and complemented by efforts from the Ghana–German GTZ Project at Nyankpala in northern Ghana; Convincing demonstrations of the nutritional superiority of Obatanpa through livestock feeding; Intense promotional campaigns for the production and utilization of QPM in Ghana and elsewhere.

Acknowledgements We are grateful for the financial assistance provided by the Governments of Ghana and Canada through the Ghana-CIDA Grains Development Project (GGDP). The financial assistance and the provision of the QPM laboratory and irrigation equipment specifically for QPM development at the Crops Research Institute by SG 2000 is gratefully acknowledged. The personal interest and encouragement of Dr. Norman Borlaug and by Dr W.L. Haag in QPM helped to sustain our enthusiasm in the program. Last, but not the least, we are grateful to our numerous technical officers and MOFA collaborators for a job well done.

423

References Ahenkora, K., S. Twumasi-Afriyie, W. Haag, and B.D. Dzah. 1995. Ghanaian kenkey from normal and quality protein maize: comparative chemical composition and rat growth trials. Cereal Res. Commun. 23: 299–304. Ahenkora, K., S. Twumasi -Afriyie, P.Y.K. Sallah, and K. ObengAntwi. 1999. Protein nutritional quality and consumer acceptability of tropical Ghanaian quality protein maize. Food and Nutr. Bull. 20: 354–360. FAO/WHO (Food and Agriculture Organization/World Health Organization). 1991. Protein quality evaluation. Food and Agricultural Organization of the United Nations. Rome: FAO: 61. Mertz E.T., L.S. Bates, and O.E. Nelson, 1964. Mutant genes that change protein composition and increase lysine content of maize endosperm. Science 145: 279–280. MOH, 1988. Preliminary report on evaluation of weaning food production project. Nutrition Division, Ministry of Health, Accra, Ghana. Morris, M.L., R. Tripp, and A.A. Dankyi. 1999. Adoption and impacts of improved maize production technology: A case study of the Ghana Grains Development Project. Economics Program Paper 99–01. Mexico, D.F.: CIMMYT. Okai, D.B., S.A. Osei, A.K. Tua, S. Twumasi -Afriyie, W. Haag, B.D. Dzah, K. Ahenkora, and L.E.K. Osafo. 1994. The usefulness of Obatanpa, a quality protein maize variety in the feeding of pigs in Ghana. Proc. Ghana Anim. Sci. Symp. 22: 37–43. Okoruwa, E.A. 1997. Enhancing maize processing and utilization in West and Central Africa. Pp 108-119 in B. Badu -A praku, M.O. Akoroda, M. Ouedraogo, and F.M. Quin (eds.) Contributing to food self-sufficiency: maize research and development in West and Central Africa. Proceedings of a Regional Maize Workshop, May 29–June 2, 1995, IITACotonou, Benin Republic. WECAMAN/IITA. Osei, S.A., D.B. Okai, K. Ahenkora, B.D. Dzah, W. Haag, S. Twumasi -Afriyie, A.K. Tua, and L.E.K. Osafo. 1994a. Quality protein maize as the main source of energy and amino acids in the diets of starter pigs. Proc. Ghana Anim. Sci. Symp. 22: 31–36. Osei, S.A., A. Donkoh, C.C. Atuahene, D.B. Okai, A.K. Tua, W. Haag, B.D. Dzah, K. Ahenkora, and S. Twumasi-Afriyie, 1994b. Quality protein maize as a broiler feed ingredient. Proc. Ghana Anim. Sci. Symp . 22: 45–49. Osei, S.A., C.C. Atuahene, A. Donkoh, K. Kwarteng, K. Ahenkora, B.D. Dzah, W. Haag, and S. Twumasi -Afriyie. 1994c. Further

424

studies on the use of quality protein maize as a feed ingredient for broiler chickens. Proc. Ghana Anim. Sci. Symp . 22: 51– 55. PPMED. 1999. Production estimates of some major crops in Ghana: 1970–1997. Policy Planning, Monitoring & Evaluation, Ministry of Food and Agriculture, Accra, Ghana. Sallah, P.Y.K., S. Twumasi -Afriyie and C.N. Kasei. 1997a. Grain productivities of four maturity groups of maize varieties in the Guinea savanna. Pages 173–178 in B. Badu-Apraku, M.O. Akoroda, M. Ouedraogo, and F.M. Quin (eds.) Contributing to food self-sufficiency: maize research and development in West and Central Africa. Proceedings of a Regional Maize Workshop, May 29-June 2, 1995, IITA-Cotonou, Benin Republic. WECAMAN/IITA. Sallah, P.Y.K., S. Twumasi -Afriyie, E.A. Asiedu, K. Obeng-Antwi, K. Boa-Amonsem, K. Ahenkora, A. Agyemang, and E.K. Lampoh. 1997b. Development and release of Dodzi, an extra-early maturing maize variety in Ghana. Crops Research Institute, Kumasi. 23 pp. Twumasi -Afriyie, S., B. Badu-Apraku, P.Y.K. Sallah, W. Haag, E.A. Asiedu, K.A. Marfo, S. Ohemeng-Dapaah, and B.D. Dzah. 1992. Development and release of Obatanpa, an intermediate maturing quality protein maize variety in Ghana. Crops Research Institute, Kumasi. Mimeo. 19 pp. Twumasi -Afriyie, S., P.Y.K. Sallah, K. Ahenkora, E.A. Asiedu, K. Obeng-Antwi, P.P. Frimpong-Manso, S. Osei-Yeboah, A.O. Apau, A. Mensah-Ansah, W. Haag, and B.D. Dzah. 1997a. Development and release of three quality protein maize hybrid varieties, Dadaba, Mamaba and CIDA-ba in Ghana. Crops Research Institute, Kumasi. Mimeo. 31pp. Twumasi -Afriyie, S., P.Y.K. Sallah, M. Owusu-Akyaw, K. Ahenkora, R.F. Soza, W. Haag, B.D. Dzah, D.B. Okai and A. Akuamoah-Boateng. 1997b. Development and promotion of quality protein maize in Ghana. Pages 140-148 in B. BaduApraku, M.O. Akoroda, M. Ouedraogo, and F.M. Quin (eds.) Contributing to food self-sufficiency: maize research and development in West and Central Africa. Proceedings of a Regional Maize Workshop, May 29 -June 2, 1995, IITACotonou, Benin Republic. WECAMAN/IITA. Vasal, S.K., G. Srinivisan, S. Pandey, F. Gonzalez, J. Crossa, and D.L. Beck. 1993. Heterosis and combining ability of CIMMYT’s protein maize germplasm: Lowland tropical. Crop Sci . 33: 46– 51. Young, V.R., and P.L. Pellet. 1991. Protein evaluation, amino acid scoring and Food and Drug Administration’s proposed food labelling regulations. J. Nutri. 121:145–150.

425

Factors affecting maize grain quality and fumonisin content in some villages of the western highlands of Cameroon Z. Ngoko1, K.F. Cardwell*2, F. Schulthess2, W.F.O. Marasas 3, J.P. Rheeder3, G.S. Shephard3 , and M.J. Wingfield4 1

IRAD Box 80, Bamenda, Cameroon I I T A, Cotonou, Republic of Benin 3 Programme on Mycotoxins and Experimental Carcinogenesis, Medical Research Council, PO Box 19070, Tygerberg, South Africa 7505 4 Forestry and Agricultural Biotechnology Institute (FABI), Faculty of Biological and Agricultural Sciences, University of Pretoria, Pretoria, South Africa 0002. 2

Abstract A survey was conducted in three areas of the Western Highlands (WHL) of the Republic of Cameroon to assess the biological and physical constraints to maize grain quality. Thirty-two of the 36 samples analyzed tested positive for fumonisin with concentrations ranging between 0 and 8.2 ppm. Questionnaires were administered to 36 farmers in 1997. A backward regression analysis of agricultural practices associated with the infection of maize grains by Fusarium species and subsequent contamination by fumonisin was done. The results showed that harvesting maize in June (11.1%), sorting harvested ears in the field (16.7%) and drying maize over the fireplace with husk (19.4) or without husk (33.3%) or in cribs significantly reduced the infection and the risk of contamination by fumonisin. Yellow maize was less contaminated with fumonisin compared to white maize. The storage weevil, Sitophilus zeamais, significantly increased the risk of contamination by fumonisin. Other factors such as harvesting in August, storing in bags, continuous cultivation of maize, and the education level of the farmers were nonsignificant factors retained by the regression analysis. Résumé Une enquête a été conduite dans trois régions montagneuses occidentales de la République du Cameroun afin d'évaluer les contraintes physiques et biologiques de la qualité des graines de maïs. Trente - deux des 36 échantillons analysés étaient positifs pour le fumonisin, dont les concentrations variaient entre 0 et 8,2 ppm. Les questionnaires étaient administrés à 36 agriculteurs en 1997. Une analyse de régression par élimination des pratiques agricoles associées avec l'infection des graines de maïs par les espèces de Fusarium et la

426

contamination consécutive avec fumonisin a été effectuée. Les resultats ont révélé que la récolte du maïs en juin (11,1%), le tri de la récolte au champ (16,7%), le séchage du maïs sur la cheminée avec les spathes (19,4) ou sans les spathes (33,3%) ou dans les "cribs" réduit considérablement l'infection et le risque de contamination par le fumonisin. Le maïs jaune était moins contaminé par le fumonisin en comparaison au maïs blanc. Les charançons de stockage, Sitophilus zeamais, augmenteraient considérablement le risque de contamination par les fumonisins. D'autres facteurs tels que la récolte en août, le stockage en sacs, le maïs comme précédent cultural, et le niveau d'instruction des agriculteurs étaient des facteurs non - significatifs retenus par l'analyse de régression.

Introduction Maize (Zea mays L), one the most important commodities produced in Cameroon, is cultivated on about 600,000 ha for a total annual grain production of 800,000 metric tons (AyukTakem 1996). Since the devaluation of the CFA (Communauté Financiere Africaine) currency, maize production has increased markedly, although it is still not keeping pace with the demand. Production could be higher if factors such as soil infertility, lack of know-how by farmers, high cost of inputs, environmental stresses, pests and diseases in the field and in storage were not constraining grain yield (Ayuk-Takem et al. 1982). Leaf, stem and ear diseases occur in all ecologies where maize is produced in Cameroon (Cardwell et al., 1997; Ngoko, 1999). Leaf and stem diseases decrease maize grain yield by lowering photosynthetic activities, interrupting nutrient transport and causing lodging. Incidence of ear rots in the field and/or in storage affect grain quantity and quality especially in areas where maize is harvested under high rainfall conditions or with high grain moisture content. Although 40% of the farmers remove damaged grain before selling or consuming, contaminated grains are frequently found in maize delivered to local breweries, feed mills and food grain marketers (McHugh 1994). The infection of grain by fungi that produce mycotoxin has been reported in many parts of the world. Marasas et al. (1988) reported Fusarium moniliforme Sheldon as the predominant fungus contaminanting maize in South Africa. Booth (1971) noted that F. moniliforme occurred commonly on maize throughout the world. Logrieco et al. (1995) reported that F. proliferatum (Matsushima) Nirenberg (34%) and F. moniliforme (54%), were the predominant fungi infecting maize ears in Italy. F. moniliforme produces fumonisin, a carcinogenic mycotoxin (Gelderblom et al. 1988; Marasas 1996; Miller 1996). The contamination of maize and maize products by mycotoxins has been associated with several human and animal diseases

427

(Marasas 1996). It is well documented in the medical and veterinary literature that mycotoxin related diseases occur in many parts of the world, particularly in Africa (Adhikari et al. 1994; Allen et al. 1992; Cova et al. 1990; Gelderblom et al. 1988; Marasas 1995, 1996, 1997; Marasas et al. 1988; Rheeder et al. 1992; Ross et al. 1993; Shephard et al. 1992, 1995; Sydenham et al. 1990; Thiel et al. 1991; Wild 1993; Yang, 1980; Zarba et al. 1992). Apergillus flavus Link: Fr. is another fungus that produces mycotoxins. Occurrence of the fungus and associ ated aflatoxin production have been reported from studies conducted in Nigeria (Udoh 1997) and Benin (Hell 1998). Considering the farming practices and the weather conditions under which maize is produced, harvested, dried and stored, it is likely that mycotoxin contamination occurs also in Cameroon. The objective of this study was to identify traditional farming, harvesting, drying and storage practices related to fungal invasion and mycotoxin contamination of stored maize in the Western Highlands (WHL) of Cameroon.

Methods Sample collection A questionnaire was administered to 36 farmers selected at random from three villages (Bamunka, Bali, Njinikom) in the WHL in 1997. The questionnaire survey was conducted in collaboration with an extension worker who served as the facilitator at each location. The altitude in the WHL ranges between 800 m and 2500 m above sea level (MINAGRI 1996). Bamunka is located at 1100 m above sea level in the Ndop valley surrounded by mountains with peaks at 2600 m. The rainfall ranges from 1000 mm to 1500 mm per annum with temperatures from 180 C to 35 0 C. Bali, with two cropping seasons, receives 1500-2000 mm rainfall a year and is located in the Guemba plateau, approximately 1000 m above sea level. The temperature at Bali is similar to that of the Ndop valley. Njinikom, one of the most important maize production and consumption areas in the province, is situated at 1500 m above sea level. The rainfall is also bimodal and the temperature varies between 180 C and 30 0 C. The wettest months are August and September, which correspond to the main harvesting period in the region. Fumonisin analysis Fumonisin concentrations were determined by polyclonal antibody (PAb) -Based competitive direct enzyme -linked immunosorbent assay as descri bed by the manufacturer (Agri-

428

Screening Kit Catalogue no. 70/8830, Neogen Corp.; Lansing, MI, USA). Aliquots of 50 g of corn were extracted with CH3 OH/H2 O (70:30). Aliquots of each extract were applied to individual coated microtiter wells which were then incubated for 10 minutes to allow any free toxin and the toxin-peroxydase conjugate to compete for binding to the available antibody sites. Reagents were washed from the wells with distilled water. Bound toxin-conjugate levels were measured colorimetrically, following the addition of an enzyme substrate, incubation for 10 min and a stop reagent. Fumonisin concentrations were measured by recording optical density (OD) readings at 650 nm, using a Bio-Tek EL301 microwell strip reader. Sample toxin levels were compared to the standards received from Veratox, Neogen Corporation, Lansing, MI. Data analysis Data obtained from the field survey were subjected to frequency counts using GENSTAT 5 for Windows, release 3 (GENSTAT, 1993). Data obtained on the grain samples were matched with the responses to the questionnaire in each village. On the basis of the answers, the percentage of farmers’ responses was calculated. A stepwise backward linear regression was used to identify constraints that were significantly associated with fumonisin contamination. Fumonisin B1 concentration was used as the dependent variable, and the crop management practices as the binomial independent variables.

Results Fumonisin contamination Of the 36 samples analyzed, 32 tested positive for fumonisin with the concentration ranging from 0 to 8000ng/g. The concentration of most samples was less than 500ng/g. Farming practices Several agricultural practices associated with harvesting and postharvest storage were recorded. Generally in the WHL, 11% of the farmers harvested their maize crop in June, especially the farmers in Bali where most maize is harvested at that period. The peak period of harvesting was August when 64% of the crop was harvested. Only about 24% was harvested in July and a negligible proportion of the farmers harvested late planted maize in September (Table 1). Farmers did not take into consideration the weather conditions of the day of harvest. About 33% of the farmers harvested and transported their maize from the field to their houses on a wet day. About 3 % of the farmers reported that they stacked the cobs in the field, covered them with banana (Musa accuminata Colla) leaves or tarpaulin for a night pending transportation.

429

Table 1. Farmers’ agricultural practices in three villages of the western highlands of Cameroon, 1997 (n = 12 per location). Practice

Bamunka

Bali

Njinikom

Average

Harvest period June 6.9 17.6 8.8 11.1 July 13.6 56.5 10.7 23.6 August 77.5 25.6 78.6 63.9 September 2.0 0.3 1.9 1.4 Cleaning stores Yes 90.1 88.9 78.9 85.9 No/Uncertain 9.9 11.1 21.1 14.1 Harvesting With husk 80.4 85.6 90.7 85.6 Stacka 3.2 1.4 3.0 2.5 Wet b 16.4 13.0 6.3 11.9 Sorting Yes 21.3 18.2 10.6 16.7 No/Uncertain 78.7 81.8 89.4 83.3 Drying Banda w/fire 21.5 30.6 26.1 26.1 Banda wo/fire 18.9 18.4 13.4 16.9 Crib 35.2 14.1 4.6 17.9 Others 24.4 36.9 55.9 39.1 Storing Bandas 42.1 54.0 51.9 49.3 Crib 33.5 18.0 7.0 19.3 Bags 14.4 4.8 11.1 10.1 Boxes/Drums 7.5 12.9 9.4 9.9 Others 2.5 10.3 20.6 11.4 Period of drying 1 month 67.9 80.0 50.1 66.0 2 month 20.2 15.3 33.5 23.0 >2 month 11.9 4.7 16.4 11.0 Use of bad maize Animal feeds 34.2 30.7 35.0 33.3 Sales 19.0 23.0 18.0 20.0 Consumption 11.4 9.9 8.7 10.0 Others 35.4 35.4 38.3 36.7 a Cobs were stacked in the field for at least one day pending transportation. bCobs were harvested and/or transported on a wet day.

Different harvesting procedures (with husk, stacked in the field and harvested on a wet day), were identified (Table 1). Most farmers (77.8 %) harvested with husk; others dehusked in the field. In general, harvesting was done manually as soon as the crop reached the hard dough stage. Most farmers harvested some green ears for family consumption, especially from fields near the homestead. Nearly 17% of the farmers sorted the harvested ears in the field. For these farmers, sorting consisted of removing rotten cobs and ears with bad tip cover from the lot before drying. The other farmers did not sort at all perhaps out of ignorance or because

430

they thought it would reduce their crop. Bad maize was fed directly to livestock, mostly chickens and pigs, by 33.3% of the farmers. Another 20% sold the bad maize in the local market. Some of the bad maize, probably mixed with some good maize, were used to make meals or snacks by 10% of the farmers. The remaining farmers (36.6%) either threw it away or did not give a clear response. Four drying methods were identified in most locations (Table 1); that is: i. cobs dried with husks in a “banda” (an elevated bamboo structure in the house) or under the ceiling of the house (19.4%); ii. cobs without husk over the fireplace (16.7%); iii. air- or sun-drying by hanging the maize unde r the ceiling or in a “banda”, with or without husk (33.3%); iv. drying in the crib (18.1%). The drying period varied between one (66%) and two (23%) months; only about 11% of the farmers dried their maize for periods longer than two months. Most (88.9%) of the farmers indicated that they usually cleaned the drying and storing structure every year before the new maize was brought in. Almost 50% of the farmers disinfected these structures using either actellic powder (perimiphos-methyl), a common commercial insecticide sold in the region, or fresh leaves of a common local tree (Cypressus sp.). These fresh leaves were burnt inside the store about an hour before cobs were brought in for drying and/or storing. Seven kinds of storage were identified (Table 1). Most (70.6%) of the farmers used storage that facilitated drying of the maize. Maize cobs harvested without husks were kept in a wellconstructed crib with bamboo or wire mesh and covered with a corrugated iron roof by 18.0 % of the farmers. Another 20% of the farmers reported that they used platforms constructed around their houses for storing, and 31.0% stored the maize in bandas in their kitchen. Some farmers preferred different types of bags (9.7%) and boxes or drums (9.7%). Few farmers stored cobs on the bare ground for a short period; 79.2% heaped all the cobs on the bare ground before sorting, while 12.5% sorted the maize on concrete floors before drying and hand shelling. Compared to 16.7% of the farmers that stored maize in houses with thatched roofs, 80.6% had storage constructed with corrugated iron sheets that they claimed enhanced the drying of grains. About 50% had white, 8% yellow, and 41% mixed color maize (data not shown). Four categories of farmers were identified in the WHL (data not shown). Those who did not attend school (27.4 %), those who attended primary school only (44.4%), those who attended secondary and high school (13.9%) and 11.1% that had university

431

degrees. The university graduates were either actually farming or were landlor ds that hired laborers for their farming activities. Analysis of factors affecting fumonisin contamination in WHL The relationships between the potential biological and physical constraints and the fumonisin concentration in the maize samples are summarized in Table 2. Table 2. Parameters from a stepwise multiple regression analysis (backward selection) of crop management factors affecting fumonisin contamination of maize in the Western Highlands of Cameroon in 1997. Variable b-value Drying over fire -170.5 Drying in the crib -264.5 Sitophilus zeamais 81.9 Sorting rotten cobs -672.5 Harvesting in June -130.4 Yellow maize -81.9 Harvesting in August 249.8 Storing in boxes -470.1 Storing in bags 938.7 Gender -329.5 Maize as previous crop 56.2 Farmer’s education level 23.2 Intercept = 145.0 R2 = 0.75 P>F = 0.05

t-value 3.54 2.31 0.66 0.81 0.44 0.45 0.31 -1.33 -1.14 0.10 0.61 0.80

P>t 0.001 0.006 0.012 0.033 0.041 0.052 0.149 0.192 0.259 0.359 0.530 0.701

Drying maize over the fireplace (P = 0.001) or in a crib (P = 0.006) significantly reduced fumonisin contamination. Insect attack, especially Sitophilus zeamais Motschulsky (Coleoptera: Curculionidae), significantly (P = 0.0012) increased the contamination of maize grain. Removing rotten cobs in the field also reduced fumonisin contamination significantly (P = 0.033). Harvesting in June was significantly (P = 0.041) related to lower fumonisin contamination. Yellow maize was less contaminated with fumonisin than white maize, but was only barely significant (P = 0.052). Fumonisin levels were numerically reduced in maize handled by women, but not significantly (P > 0.05) from those handled by men. There was no significant difference between storing in boxes and storing in bags. Maize as the previous crop had no significant effects on the fumonisin contamination of the samples analyzed. Maize handled by non -educated persons was not significantly different from maize collected from educated farmers. Field production factors (soil texture, soil fertility, and weeds) had no significant relationship with grain quality. In summary, several groups of potential factors affecting the contamination of maize by fumonisin were identified in the WHL

432

in 1997 and some recommendations are made about these in Table 3. Table

3. Summary of potential factors affecting fumonisin contamination of maize from rural areas of the Western Highlands of Cameroon, 1997.

Factor Harvest time

Effect on fumonisin contamination Reduces

Recommendation Harvest at physiological preferably in June.

maturity,

Harvesting period

Reduces

Harvest on sunny days. Avoid harvesting and transportation on wet days.

Drying method

Reduces

Dry over the fireplace or any other source of heat immediately after harvesting.

Sorting time

Reduces

Sort immediately after harvesting, before drying and storing. Practise continuous sorting before consuming.

Storing period

Reduces

Shell, treat and store grain in bags/boxes in cool, dry rooms. In the alternative, treat cobs and store in cribs.

Weevil outbreak

Increases

Treat maize cobs or grains with insecticide to reduce fumonisin contamination.

Previous crop

May increase

Avoid continuous maize, maize-legume rotation.

Yellow maize

Reduces

Grow yellow maize.

Gender

May reduce

Encourage sorting by women.

practise

Discussion The results of the present study have clearly shown that it is important for maize farmers in the WHL of Cameroon to harvest maize on sunny days and dry the maize immediately after harvest in order to reduce fumonisin contamination. Drying with or without husk over the fireplace or in cribs was associated with reduced fumonisin contamination. Hell (1998) noted that aflatoxin levels increased throughout storage in poorly dried maize stacks. Harvesting of maize during wet days increased the labor, grain moisture content and grain susceptibility to infection by fungi and subsequent mycotoxin contamination. The use of cribs should be improved in areas at 1600 m and higher because

433

drying of grain is affected by crib size and air circulation between the cobs. This study also demonstrated that the storage weevil, S. zeamais significantly increased fumonisin contamination of maize. The role of insects in the dispersal of fungi is well established. Setamou et al. (1998) found that damage by the cob-borer Mussidia nigrivenella Ragonot (Lepidoptera: Pyralidae) was positively related to aflatoxin contamination. Flett & Van Rensburg (1992) found that Busseola fusca Fuller (Lepidoptera: Noctuidae) and physical damage increased the incidence of F. moniliforme. Consequently, the control of S. zeamais and B. fusca should reduce fumonisin contamination and lead to better grain quality. Sorting bad cobs before drying reduced fumonisin contamination. Borgemeister et al. (1994), Udoh (1997) and Hell (1998) found that sorting bad grain reduced losses due to insect attack and aflatoxin contamination after harvest. The contamination of maize stacks by molds is related to the dispersal capacity of the fungi within the lot. Therefore, removal of the moldy cobs increases the chances of preserving good quality grain in storage. Sorting should therefore be considered as a continuous process in grain management from harvest to final utilization. Since moldy grain is an important source of mycotoxin contamination in maize, the extension services should emphasize the sorting of bad cobs prior to leaving the field and continuous sorting before consumption. In this study, contamination by A. flavus was negligible and aflatoxin B1 was detected in very few samples. Other fluorescing compounds detected may have been the by-products of aflatoxin. Although aflatoxins B1 and G1 were detected in only about 8% of the samples, it should not be considered as a proof that A. flavus and A. parasiticus Speare invasion is not important in the WHL of Cameroon. Weather conditions prevailing in the WHL in 1997 may not have favored the development of Aspergillus spp. and subsequent aflatoxin contamination. The infection of grain by fungi and subsequent mycotoxin contamination depends on weather conditions as well as pre- and post-harvest handling of the crop. Harvesting in June and July resulted in decre ased fumonisin concentration in maize in 1997 compared to harvesting in August. The temperature in the WHL in June and July varied between 20 and 250 C and the relative humidity was approximately 70%. It is important to note, however, that harvesting is possible in June and July only if planting is done early in March. In turn, planting in March depends on rainfall and farmers’ priorities. Harvesting in August was detrimental to maize grain quality in 1997. This corresponded to the peak of the rainy season with rains falling almost every day. There is an urgent need to conduct on-station baseline research to determine the time favorable for the

434

development of Fusarium species. The information obtained from the research will be useful in designing a new croppi ng calendar for farmers in the WHL. Fumonisin contamination was lower in yellow maize compared to white maize as reported earlier by Rheeder et al. (1992). In the WHL of Cameroon, 51% of the population prefer white maize (McHugh 1994). It becomes necessary, therefore, to emphasize breeding white maize for resistance to F. moniliforme ear rot. No significant relationship was found between storage methods and fumonisin contamination. The negative relationship identified between storing in boxes and fumonisi n contamination was not significant. Conversely storing in bags was positively related to fumonisin contamination, but not significantly. Boxes are made of wood and the chances of damage from rodents, insects and household tools are reduced. This may explain why better quality maize grain is found in this type of structures. Hell (1998) and Udoh (1997) investigated maize storage structures and aflatoxin contamination in Benin and Nigeria and found that storing in bags reduced the infestation of maize grain by Aspergillus flavus and subsequent aflatoxin contamination. The conditions conducive to the development of F. moniliforme and A. flavus could be used to explain this apparent contradiction. Although the control of maize grain by synthetic chemicals or local insecticides was not retained by the regression analysis in this study, Awuah (1996) and Cardwell and Dongo (1994) suggested some natural products to control insects and storage fungi. These products could improve the quality of the maize grain store d in bags or boxes. Caution should, however, be exercised in using natural insecticides. Hell (1998) reported that the use of Khaya senegalensis as a stored grain protectant against insects increased the risk of aflatoxin contamination. This plant may have stimulated the breeding of insects, which enhanced the dispersal of aflatoxin-producing fungi in the stores. Therefore, the insecticidal and fungicidal properties of local plants should be well investigated before the products are applied on maize in stor age. A negative relationship between gender and fumonisin contamination of maize grain revealed that maize handled by women had lower fumonisin levels compared to maize handled by men. This may be explained by the fact that most women usually sort bad cobs and grain before drying, storing or cooking. More investigations are needed on the role of women in determining maize grain quality. Although emphasis was placed on the post-harvest practices in this study, it must be understood that inappropriate agronomic practices, such as repeated maize cropping in the same field, may lead to increased incidence of mycotoxin producing fungi in the

435

field. Maize as previous crop was retained by the regression analysis as a factor that contributes to fumonisin contamination. Therefore, continuous maize cropping in the same field should be discouraged and rotation with legumes should be encouraged. The analysis of factors contributing to fumonisin contamination showed that the level of education was not significantly related to fumonisin contamination. Similarly Jaminson & Lau (1982) and Mook (1981) found that levels of education were not necessarily an improvement over zero education in predicting technology adoption by small-scale farmers in the developing world. In contrast, Carlson & Mueller (1987) found a strong, positive effect of formal education on farmers’ perception of the cost-benefit of farming activities. Although our study did not reveal any role of education of farmers in reducing fumonisin contamination, the results may be misleading because some university graduates hired other people to farm for them. Consequently, the skills and knowledge of the graduates may not have been used in actual farming. Postharvest factors such as period of harvest, methods of drying and storing techniques had significant effects on the contamination of maize grain by fumonisin. There is a need for on-station research to identify the combination of factors that will reduce the contamination of maize by mycotoxins. Also, education campaigns should be initiated in rural areas to create awareness about the relative importance of the post-harvest handling of maize grains in order to minimize the risk of fungal infection and subsequent mycotoxin contamination.

Recommendations To the Institute of Agricultural Research for Development (IRAD) ?? Efforts should be concentrated on reinforcing the maize improvement programs to facilitate their capacity to breed for resistance to the major maize diseases and pests in Cameroon in collaboration with other national and international research and teaching institutes. ?? An illustrated manual should be prepared to assist farmers in in identifying maize diseases and insect pests. To the extension services and farmers ?? Use improved maize seeds tolerant of the major diseases of your region. ?? Plant maize in a well-fertilized soil under reduced stress. ?? Plant the maize crop as early as rains permit in March. ?? Harvest on sunny days at physiological maturity in June or July. ?? Avoid stacking maize overnight in the fields after harvest.

436

?? Avoid spreading maize cobs without husks on the bare ground. ?? Sort bad cobs before drying and continue to sort during drying and in storage. ?? Dry maize as soon as possible after harvest. ?? Shell cobs as soon as the grain is dry. ?? Treat maize grain with insecticides. ?? Store maize in a dry environment with good air circulation. ?? Always sort bad grain before consumption.

Acknowledgements This is IITA manuscript 99/131/JA. We thank IITA reviewers C. Nansen and S. Korie. This study was supported in part by IFAD and PROMEC.

References Adhikari, M., G. Ramjee, and P. Berjak.. 1994. Aflatoxin, kwashiorkor, and morbidity. Natural Toxins 2:1–3. Allen, S.J., C.P. Wild, J.A.G. Wheeler, E.M. Riley, R. Montesano, S. Bennet, H.C. Whittle, A.J. Hall, and B.M. Greenwood. 1992. Aflatoxin exposure, malaria and hepatitis-B infection in rural Gambian children. Transaction of the Royal Society of Tropical Medicine and Hygiene 86:426–430. Awuah, R.T., 1996. Possible utilization of plant products in grain storage. P 32 in K.F. Cardwell (ed.) Proceedings of the Workshop on Mycotoxins in Food in Africa, IITA: Cotonou, Benin. Ayuk-Takem, A.J. 1996. Agricultural research and extension resources. Country: Cameroon. Pages 44–45 in K.F. Cardwell (ed.) Proceedings of the Workshop on Mycotoxins in Food in Africa, IITA: Cotonou, Benin. Ayuk-Takem, J.A., J.P. Eckebil, and H.R. Chheda. 1982. Problems and potentials of maize (Zea mays L.) research and production in Cameroon. Revue Science et Technique 2: 5– 16. Booth, C. 1971. The Genus Fusarium. CMI: Kew, Surrey, England. Borgemeister, C., C. Adda, B. Djomamou, P. Degbey, A. Agbaka, F. Djossou, W.G. Meikle, and R.H. Markham. 1994. The effect of maize cob selection and the impact of field infestation on stored maize losses by the larger grain borer (Prostephamus truncatus (Horn) Col., Bostrichidae) and assorted storage pests. Pages 906–909 in H. Highley, E.L. Wright, H.J. Banks H, and B.R. Champ (eds.) Proceedings of the 6 th International Working Conference on Stored Product Protection, Canberra, Australia.

437

Cardwell, K.F., and L. Dongo. 1994. Effects of extracts from nine plant species found in Africa on the mycelia growth of A.flavus Link ex Fries. Pages 978–980 in H. Highley, E.L. Wright, H.J. Banks H, and B.R. Champ (eds.) Proceedings of the 6 th International Working Conference on Stored Product Protection, Canberra, Australia.. Cardwell, K.F., F. Schulthess, R. Ndemah, and Z. Ngoko. 1997. A systems approach to assess crop health and maize yield losses due to pests and diseases in Cameroon. Agriculture, Ecosystem & Environment 65: 33–47. Carlson, G. A., and R.A.E. Mueller, 1987. Farmers’ perceptions, education and adoption of pest management strategies by small scale farmers. Pp 147-154 in E. D. Magallona (ed.) Proceedings o f the 11th Congress of Plant Protection Manila, Philippines. Cova, L., C.P. Wild, R. Mehrotra, V. Turusov, T. Shirai, V. Lambert, C. Jacquet, L. Tomatis, C. Trepo, and R. Montesano. 1990. Contribution of aflatoxin B1 and hepatitis-B virus infection in the induction of liver tumors in ducks. Cancer Research 50: 2156–2163. Flett, B. C., and J.B.J. Van Rensburg, 1992. Effect of Busseola fusca on the incidence of maize ear rot caused by Fusarium moniliforme and Stenocarpella maydis. South African Journal of Plant and Soil 9: 177 -179. Gelderblom, W. C. A., K. Jackiewicz, W.F.O. Marasas, P.G. Thiel M.J. Horak, R. Vleggaar, and N.P.J. Kriek. 1988. Fumonisinsnovel mycotoxins with cancer- promoting activity produced by Fusarium moniliforme. Applied Environmental Microbiology 54:1806–1811. GENSTAT. 1993. GENSTAT 5 Release 3 Reference Manual, Oxford Science Publications, Oxford University Press Inc., New York. 796 pp. Hell, K. 1998. Factors contributing to the distribution and incidence of aflatoxin producing fungi in stored maize in Benin. Ph.D. Dissertation: University of Hannover, Germany. Jaminson, D.T., and L.J. Lau. 1982. Farmer education and farm efficiency. John Hopkins Press: Baltimore, Md, USA. Logrieco, A., A. Moretti, A. Ritieni, A. Bottalico, and P. Corda. 1995. Occurrence and toxigenicity of Fusarium proliferate from pre-harvest maize ear rot and associated mycotoxins in Italy. Plant Disease 79: 727–731. McHugh, D. 1994. Evaluation of stored maize Cameroon. Experimental Agriculture 30: 45–55.

losses

in

Marasas, W.F.O. 1995. Fumonisins: Their implications for human and animal health. Natural Toxins 3:193–198.

438

Marasas, W.F.O. 1996. Fumonisins: History, worldwide occurrence and impact. Pages 1–17 in L. Jackson (ed.) Fumonisins in Food Plenum Press: New York. Marasas, W. F.O. 1997. Risk assessment of fumonisins produced by Fusarium moniliforme in corn. Cereal Research Communications 25: 399–406. Marasas, W. F. O., K. Jaskiewicz, F.S. Venter, anf D.J. Van Schalkwyk. 1988. Fusarium moniliforme contamination of maize in esophageal cancer areas in Transkei. South African Medical Journal 74: 110–114. Ngoko, Z. 1999. Mycotoxin contamination of maize in relation to insect infestation, agricultural practices and agroecology in the Republic of Cameroon. Ph.D Thesis. The University of the Orange Free State, Blomfontein, South Africa. Miller, J.D. 1996. Mycotoxins. Pages 18–20 in K.F. Cardwell (ed.) Proceedings of the Workshop on Mycotoxins in Food in Africa, IITA: Cotonou, Benin. Mook, P.R. 1981. Education and technical efficiency in small farm production. Economic Development and Cultural Change 9: 723–739. Rheeder, J. P., W.F.O. Marasas, P.G. Thiel, E.W. Sydenham, G.S. Sherphard. And D.J. Van Schalkwyk. 1992. Fusarium moniliforme and fumonisins in relation to human esophageal cancer in Transkei. Phytopathology 82: 353–357. Ross, P.F., A.E. Lodet, E.D. Owens, L.G. Rice, H.A. Nelson, G.D. Osweiler, and T.M. Wilson. 1993. Experimental equine leukoencephalomalecia, toxic hepatosis, and encephalopathy caused by corn naturally contaminated with fuminosins. Journal of Veterinary Diagnostic Investigations 5: 69–74. Sétamou, M., K.F. Cardwell, F. Schulthess, and K. Hell. 1998. Effect of insect damage to maize, with special reference to Mussidia nigrivenella (Lepidoptera: Pyralidae), on Aspergillus flavus (Deuteromycetes: Moniliales) infection and aflatoxin production in maize before harvest in the Republic of Benin. Journal Economic Entomology 91: 433–435. Shephard, G. S., P.G. Thiel, and E.W. Sydeham. 1992. Initial studies of the toxicokinetics fumonisin B1 in rats. Food and Chemical Toxicology 30: 277–279. Shephard, G. S., P.G. Thiel, and E.W. Sydeham, and M.E. Savard. 1995. Fate of a single dose of 14 C-labelled fumonisin B1 in vervet monkeys. Natural Toxins 3: 145–150. Sydenham, E. W., P.G. Thiel, W.F.O. Marasas, G.S. Shephard, D.J. Van Schalkwyk, and K.R. Koch. 1990. Natural occurrence of some Fusarium mycotoxins in corn from low and high esophageal cancer prevalence areas of the Transkei . Journal of Agricultural and Food Chemistry 38: 1900–1903.

439

Thiel, P.G., G.S. Shephard, E.W. Sydeham, W.F.O. Marasas, P.E. Nelson, and T.M. Wilson. 1991. Levels of fumonisins B1 and B2 in feeds associated with confirmed cases of equine leukoencephalomalacia. Journal of Agricultural and Food Chemistry 39: 109–111. Udoh, J.M. 1997. Aflatoxin content of maize grain as affected by agricultural practices in five agroecological zones of Nigeria. Ph.D Thesis: University of Ibadan, Nigeria. Wild, C.P. 1993. Aflatoxin as a human hepatocarcinogen and the possible interaction with hepatitis B virus. African Newsletter on Occupational Health and Safety Supplement 2: 24–31. Yang, C.S. 1980. Research on esophageal cancer in China: A review. Cancer Research 40: 2633–2544. Zarba, A., C.P. Wild, A.J. Hall, R. Montesano, G.J. Hudson, and J.D. Groopman. 1992. Aflatoxins M1 in human breast milk from The Gambia, West Africa, quantified by combined monoclonal antibody immunoaffinity chromatography and HPLC. Carcinogenesis 13: 891–894.

440

Qualité comparée des grains de maïs béninois issus des écotypes locaux et des cultivars améliorés : mise au point de tests rapides de sélection C. Mestres 1 , M. Nago2, J. Hounhouigan2, et N. Akissoë2 1

2

CIRAD-CA/CERNA-UNB UNB/FSA, DAVRIEUX Fabrice, CIRAD-CP

Résumé Une grande variabilité est observée dans la qualité des maïs cultivés au Bénin, et en particulier en ce qui concerne leurs caractéristiques physiques : les variétés locales ont des petits grains à tendance farineuse et tendre tandis que les cultivars améliorés ont plutôt des gros grains vitreux et peu friables. Ces caractéristiques vont conditionner leur aptitude à la transformation ; les grains durs sont adaptés aux technologies faisant appel au dégermage (production de gritz de brasserie ou de mawè) tandis que les grains tendres seront facilement transformés en farine complète (lifin) ou en ogui. Un test de friabilité, mis au point au laboratoire, s’avère être un bon descripteur de la qualité technologique et culinaire du maïs. Par ailleurs, une calibration selon la méthode en spectroscopie en proche infrarouge pour la détermination des caractéristiques chimiques et physiques des grains de maïs entier a été mise au point au CIRAD. Elle permet de prédire l’ensemble des caractéristiques du grain en moins d’une minute avec une précision proche de celle des méthodes classiques tout en préservant intact l’échantillon. Elle peut ainsi être utilisée dans les schémas de sélection précoce sur un grand nombre d’échantillons.

Abstract A wide variability has been noted with respect to the quality of maize varieties grown in Benin, and particularly to their physical properties : the local varieties yield small, soft and rather floury grains, while improved cultivars produce large, flint and unflaky grains. These traits determine their performance in terms of processing. For example, hard grains are adapted to degerming technologies (production of brewers grits or mawè) while soft grains can be easily processed into whole flour (lifin) or ogui. A laboratory- developed test for assessing corn flakiness appears as a good descriptor of the technological and cooking quality of maize. Furthermore, a calibration system based on near infrared spectroscopy for the determination of chemical and physical properties of whole maize grain has been developed by CIRAD. It makes it possible to predict all the grain properties almost as precisely as conventional methods and in less than one minute, while

441

preserving the sample. Thus, the method can be used in early selection schemes for a large number of samples.

Introduction Dans la zone guinéenne, l’adoption par les paysans des variétés améliorées de maïs reste limitée ; ainsi, au Bénin, moins de 10 % des surfaces consacrées au maïs sont emblavées avec des cultivars améliorés (Anonyme 1992). Les utilisateurs reprochent en effet aux grains des cultivars améliorés leur mauvaise aptitude à la conservation, leur dureté qui les rend difficiles à réduire en farine fine, ainsi que la mauvaise qualité des pâtes préparées à partir des farines (Agossou et al. 1986; Koudokpon 1991; Tchamo 1993). Ces raisons conduisent ainsi les paysans à refuser les cultivars améliorés malgré leur intérêt agronomique (Kydd 1989 ; Koudokpon 1991). Si l’on veut favoriser l’adoption des cultivars améliorés par les paysans, il est donc important de pouvoir mesurer de façon simple et fiable leurs qualités technologiques et organoleptiques et d’intégrer ces critères de qualité dans les sché mas de sélection. Cette démarche commence à être mise en œuvre en Afrique et notre étude vise à y contribuer en proposant des tests de sélection basés sur la qualité d’utilisation des maïs simples, rapides et fiables utilisables par les sélectionneurs.

Aptitude des maïs aux différents modes de transformation utilisés au Bénin ; intérêt du test de friabilité Au Bénin, trois modes de transformation primaire du maïs coexistent (Nago and Hounhouigan 1990): - la transformation en lifin, une fari ne entière de maïs, concerne 50 % du maïs grain, - la transformation en ogui, une farine fermentée partiellement dégermée, concerne 35 % du maïs, - la transformation en mawè, une pâte fermentée dégermée, pour 15 % du maïs. La production de lifin Le lifin est obtenu en milieu urbain par broyage à sec du maïs entier. Les consommateurs recherchent en premier lieu un lifin de granulométrie fine. Nous avons testé récemment l’aptitude des grains de maïs issus de 21 cultivars récoltés au bénin en 1993 pour la production de lifin (Nago et al. 1997). Il apparaît que les grains des cultivars améliorés (16 cultivars ont été testés) présentent en moyenne après mouture un pourcentage de particules de taille inférieure à 150 µm de l’ordre de 42 %, alors que cette moyenne est de 51 % pour la moyenne des écotypes locaux (5) testés : les écotypes locaux permettent donc bien de produire une farine plus fine comme l’indiquaient les utilisateurs.

442

Nous avons par ailleurs pu mettre en évidence que cette différence de comportement est essentiellement liée aux propriétés mécaniques des grains. Nous avons ainsi mis au point un test de friabilité (broyage/tamisage standardisé) simple, rapide et ne nécessitant que peu de matière première (50 g) dont les résultats sont for tement corrélés avec la granulométrie du lifin (Figure 1) : les écotypes locaux ont une friabilité généralement supérieure à celle des cultivars améliorés. 75

% passant à 150 µm

65 2

R = 0.72 55 Ecotypes locaux

45

Cultivars améliorés 35

25 35

40

45

50

55

60

65

70

Friabilité (% bs)

Figure 1. Relation entre friabilité et finesse des farines de maïs (lifin).

La production d’ogui Pour l’obtention de l’ogui, les grains de maïs sont trempés puis broyés au moulin à disques. L’ogui est constituée par la fraction fine recueillie après un tamisage humide du broyat : le principal critère de qualité recherché par les utilisateurs est, dans ce cas, le rendement de la transformation. Nous avons ainsi testé (Nago et al. 1998) parmi ceux utilisés pour le lifin, 4 échantillons issus de cultivars améliorés (2) et d’écotypes locaux (2). Il apparaît (Fig. 2) que les cultivars améliorés, dont les grains ont une friabilité plus faible, donnent des rendements en ogui inférieurs (65 % environ) à ceux obtenus pour les écotypes locaux (75 à 80 %). La production de mawè et de gritz Pour la production de mawè, les grains sont lavés, concassés et les fractions grossières, grosses semoules (ou griots), sont recueillies par tamisage ; les sons (germes et enveloppes, sous produits du procédé) sont, au contraire de la production d’ogui, obtenus dans la fraction fine. Un procédé similaire dans le principe est utilisé pour obtenir les griots de brasserie. Le critère de qualité principal est ici aussi le rendement de transformation en produit dégermé. Nous avons testé des grains originaires du Mali et de France quant à leur aptitude à produire des griots de

443

brasserie. Il apparaît dans ce cas une corrélation négative entre la friabilité des grains et le rendement en griots (Figure 3) : plus le grain est friable, plus faible est le rendement obtenu. 65

RendementRendement ogui (% bs)

en grits (% bh) Fiabilité (% bs)

60 55 82 50 45 40

78

35 30

74

Gnonli

Gbogboué

DMR-ESR-W

Pirsabak 7930SR

85

70

80

R2 = 0.89

75

667 0 65

626 0 32

Gnonli

34

Gbogboué

DMR-ESR-W

36 38 Friabilité (% bs)

Pirsabak 793042 SR

40

Figure 2. Friabilité et rendement en ogui de quel ques cultivars béninois. Figure 3. Relation entre friabilité et rendement en grits.

La friabilité comme test de sélection La friabilité des grains paraît être un test prometteur pour prévoir l’utilisation potentielle d’un maïs grain : - les grains friables, comme par exemple les écotypes locaux rencontrés au Bénin, sont adaptés à la transformation en lifin et en ogui, - les grains durs, ou peu friables, comme la plupart des cultivars améliorés, sont adaptés à la production de mawè ou de griots de brasserie. Toutefois, ce test nécessite l’utilisation de 50 g de grains qui sont détruits lors de la mesure. Un outil très rapide : la Spectrométrie dans le Proche InfraRouge (SPIR) Principe et mise au point et développement d’une calibration. Le principe de la Spectroscopie dans le Proche InfraRouge (SPIR) repose sur la mesure d’absorption du rayonnement proche infrarouge par la matière organique. Dans le domaine observé (entre 800 et 2500 nm), les bandes d’absorption sont le fait des

444

liaisons C-H, O-H et N-H ; la quantité d’énergie absorbée est proportionnelle au nombre de groupements et donc à la concentration en un constituant donné. 1.6

Log (1/R)

1.2

0.8

0.4

0 400

900

1400

1900

2400

longueurs d'onde (nm)

Fig. 4. Exemple de SPIR de grains de maïs

Pour cela, on enregistre dans un premier temps le spectre d’absorption d’une série d’échantillons (Figure 4). On détermine ensuite les caractéristiques (chimiques ou physiques) de chacun des échantillons, puis on établit des régressions multiples entre les différentes bandes d’absorption et la teneur en chaque composant. Cette étape de calibration nécessite typiquement 200 à 300 échantillons de référence. On valide ensuite les équations de prédiction sur un plus petit nombre d’échantillons, en particulier pour chaque nouvelle série d’échantillons analysée. Application au maïs Dans le cas du maïs, cette première étape de calibration a été réalisée sur grains entiers (50 g sont nécessaires) pour les caractéristiques suivantes : teneurs en protéines, amidon, amylose, lipides ainsi que sur la friabilité (Tableau 1). Tableau 1. Calibration SPIR sur maïs. Caractéristique Amidon Amylose Protéines Lipides Friabilité

Né chantillons 200 209 340 355 213

ETr 0,9 0,7 0,3 0,1 2,0

ETp 1,3 0,8 0,5 0,3 2,8

La calibration a été obtenue à partir d’une population de 200 à 355 échantillons. Une validation des résultats a été réalisée sur

445

Figure 5. Corrélations entre friabilité prédite et mesurée pour les 30 échantillons de validation.

une trentaine d’échantillons ; on remarque alors que l’erreur de prédiction (ETp) est à peine plus élevée que l’erreur de mesure (ETr) pour la plupart des caractéristiques mesurées. En particulier pour la friabilité, les prédictions sont très proches des valeurs réelles mesurées expérimentalement (Figure 5). Les potentialités de la SPIR comme outil de sélection La SPIR apparaît donc comme un outil intéressant pour intégrer des critères de qualité des grains dans des schémas de sélection variétale. Ces avantages sont nombreux : - c’est une méthode très rapide (moins d’une minute pour effectuer l’acquisition du SPIR), - qui demande peu de matière première (50 g), - et qui est non destructive (les grains peuvent être semés après mesure), - elle permet des prédictions de multiples caractères simultanément, - tant chimiques que physiques (friabilité), voire organoleptique, Elle présente toutefois quelques contraintes : - l’équipement nécessaire est coûteux (environ 60 millions de F CFA), - la phase de calibration est longue (200 échantillons au minimum), - une validation est nécessaire (20 échantillons environ) pour chaque nouvelle série d’échantillons.

Bibliographie Agossou,Y., D. Afoda, L. Lare, K. Sewonou, et B. Tagba. 1986. Tests de technologie traditionnelle de transformation des vivriers. Synthèse des travaux 1982-1985. Société Togolaise du Coton: Lomé (Togo). 37 p

446

Anonyme. 1992. Le Benin en chiffre, 88. ; Université Nationale du Bénin. Section Economie Rurale: Cotonou (Bénin). Koudokpon, V. 1991. Pourquoi les variétés améliorées de maïs ne sont-elles pas largement adoptées par les paysans ? Bulletin de la Recherche Agronomique du Bénin 2:6-9. Kydd, J. 1989. Maize research in Malawi: lessons from failure, 112–144.Vol. 1. Nago, M., N. Akissoë, F. Matencio, and C. Mestres. 1997. End use quality of some African corn kernels. 1. Physico-chemical characteristics of kernels and their relationship with the quality of "lifin", a traditional whole dry-milled maize flour from Benin. Journal of Agricultural and Food Chemistry 45:555–564. Nago, M., E. Tétégan, F. Matencio, and C. Mestres. 1998. Effects of maize type and fermentation conditions on the quality of Beninese traditional ogi, a fermented maize slurry. Journal of Cereal Science 28:215–22. Nago, M.C., et J. Hounhouigan. 1990. La technologie traditionnelle de transformation du maïs en pâte fermentée au Bénin. UNB/FSA: Cotonou (Bénin), 1990: 30 p. Tchamo, P. 1993. Stratégie d’amélioration du maïs destiné à la consommation humaine dans l’Est du Cameroun. Pp 347–357 in Le progrès génétique passe-t-il par le repérage et l’inventaire des gènes ? AUPELF-UREF: Paris (France)

447

Utilisation du maïs en patisserie au Mali B.A. Bengaly Laboratoire de Technologie Alimentaire IER, BP 2058, Bamako, Mali Résumé La consommation des pâtisseries gagne une place de plus en plus importante dans notre alimentation. La principale céréale utilisée dans les confiseries est le blé qui n’est pas suffisamment produit dans la sous région. Cela engendre une dépendance vis à vis de l’étranger conduisant à des importations massives de blé estimées à 27100 tonnes en 1994-1995 et à 38600 tonnes pour l’année 1995-1996 (Direction Nationale des Indus tries). Une étude menée par la filière maïs de l’IER en 1992 a estimé qu’une substitution de l’ordre de 5% de la farine de blé par celle du maïs, générerait un besoin de 1500 tonnes de farine de maïs par an. Par ailleurs, WECAMAN de concert avec L’IER a mené une étude sur les farines composées au Mali dans la préparation des pâtisseries . Ces produits ont été transférés aux niveaux des pâtissiers. De très bons produits ont été produits avec des substitutions de 20%, 30%, 50%. Plus de 70% de la population ont accepté ces snacks et pâtisseries à base de mais. Une étude économique a montré que les produits à base de maïs est profitable pour tous les acteurs de la filière maïs. Certaines pâtisseries sont en train d’utiliser la technologie; mais une adoption totale dépendrait de l’implication du gouvernement.

Abstract Pastry products are gaining popularity in our food system in Mali. Wheat is the main cereal crop used in pastry-making but wheat production is not a common sight in the subregion, hence our dependence upon massive wheat importations from abroad, estimated at 27,100 tons in 1994-1995 and 38,600 tons in 1995-1996 (National Directorate of Industries DNI). A study conducted by the Maize program of IER in 1992 estimated that for a 5% substitution of wheat flour by maize flour, 1,500 tons of maize flour will be required every year. Morever, a WECAMAN/IER collaborative study was conducted in Mali on composite flours in pastry-making. The products are now commonly made by confectioners. Very good products have been obtained with 20%, 30%, 50% substitutions. More than 70% of the population have accepted maize-based snacks and pastry. An economic study has shown that maizebased products are profitable to all the stakeholders of the maize sector. Some pastry makers are now adopting the technology but total adoption would hinge on the government involvement.

448

Introduction La consommation des pâtisseries gagne une place de plus en plus importante dans notre alimentation. La principale céréale uti lisée dans les confiserie est le blé qui n’est pas suffisamment produite dans la sous région. Cela engendre une dépendance vis à vis de l’étranger conduisant à des importations massives de blé estimées à 27100 tonnes en 1994–1995 et à 38600 tonnes pour l’année 1995–1996 (Direction Nationale des Industries (DNI). Une étude mené par la filière maïs de l’IER en 1992 a estimé qu’une substitution de l’ordre de 5% de la farine de blé par celle du maïs, générait un besoin de 1500 tonnes de farine de maïs par an. Par ailleurs, le prix du blé ne cesse d’augmenter depuis un certain temps sur le marché international ce qui pèse lourdement sur la balance de paiement des pays consommateurs non producteurs de blé. Les objectifs de cette étude étaient : 1) Changer les habi tudes des boulangers par l’utilisation du mélange de farine de maïs et de blé; 2) Evaluer la rentabilité économique de l’action et son acceptabilité par les consommateurs. Materiels et Méthodes Le test a commencé le 8 Septembre, 1996 par la réunion des boulangers et patissiers. L’aspect préparation de pain à base de farine composée étant couvert par le programme national de l’IER nous nous sommes attelés sur le second volet qui est la préparation de pâtisserie à base de farine composée . Sept pâtisseries ont été retenues: la pâtisserie de l’Étoile, Ali Baba, Relax, Sabbach et Phenicia, Arawane et la pâtisserie de Koutiala. Les différents pourcentage de substitution étaient 15, 30, et 50%. Pour cette année nous avons travaillé avec trois pâtisseries Étoile, Arawane et Sabbague. Le moka et le croissant ont été préparés dans ces trois pâtisseries, les produits ont été évalués. Le matériel se compose de deux variétés du programme maïs Sotubaka (maïs jaune) et tuxpeno (maïs blanc) Les échantillons ont été décortiqués et finement broyés( taille des particules 180 à 200 microns). Le test a été conduit sur le moka, les croissants, les pâtés, et les gâteaux chocolats.

Resultats Le test n = a pas été soutenu par une de campagne médiatique ce qui fait que nous sommes limités aux consommateurs touchés

449

par les enquêteurs. Le suivi a été effectué sur une population cible qui fréquente les pâtisseries et les services publics. La majorité des dégustateurs ont apprécié le moka à 50% de maïs par rapport au moka de blé pur du point de vue texture et goût.

Tableau 1. Paramètres statistiques de l’évaluation sensorielle du moka à 100% blé. Paramètres sensoriels Paramètres statistiques

AcceptCouleur

Odeur

Goût

Arôme

Texture

abilité

Minimum

3

2

3

4

3

4

Maximum Moyenne SD

9 6.07 1.01

9 6.91 0.97

9 6.18 1.67

9 7.15 1.54

9 5.73 1.32

9 5.16 1.13

CV

0.78

0.56

0.83

0.67

0.61

0.78

Tableau 2. Paramètres statistiques de l’évaluation sensorielle du moka à 15% de maïs. Paramètres sensoriels Paramètres statistiques

Couleur

Odeur

Goût

Arôme

Texture

Acceptabilité

Minimum Maximum

2 9

2 9

3 9

5 9

6 9

6 9

Moyenne SD CV

6.48 1.71 1.12

6.69 1.43 0.87

6.6 1.45 0.23

7.21 1.34 1

6.73 1.20 1.1

6.56 1.76 1.3

9= J’aime énormément, 8= j’aime beaucoup, 7 =modérément 6= j’aime un peu, 5= Je suis indifférent, 4= je n’aime pas beaucoup, 3=Je n’aime pas, 2=Je n’aime du tout, 1=Je déteste

Tableau 3. Paramètres statistiques de l’évaluation du moka à 30% de maïs. Paramètres sensoriels Paramètres statistiques

Couleur

Odeur

Goût

Arôme

Texture

Acceptabilité

Minimum

2

4

4

4

5

5

Maximum Moyenne

9 7.13

9 6.45

7 7.46

9 7.18

8 6.92

8 6.82

SD CV

1.67 1

0.98 2

0.76 5.18

0.78 1.45

0.34 1.98

1.18 0.78

450

Tableau 4. Paramètres statistiques de l’évaluation du moka à 50% de maïs. Paramètres sensoriels Paramètres statistiques

AcceptCouleur

Odeur

Goût

Arôme

Texture

abilité

Minimum

4

5

7

5

7

6

Maximum Moyenne SD

9 8.02 1.67

9 7.58 0.98

9 7.87 0.76

9 7.56 0.78

9 7.92 0.34

9 8.32 1.18

CV

0.89

0.77

1.78

2.10

1.03

0.55

Il ressort de l ‘analyse de différents paramètres organoleptiques que le moka de farine à 50% était mieux apprécié par rapport au moka de blé pur. Du point de vue de la conservation aucune différence n’ai été observée entre les différents moka ( 15%, 30%, et 50%). Analyse de marché Une analyse de marché concernant ces produits à base de farine composée a été faite par ECOFIL les résultats sont consignes au Tableau 9. Tableau 5. Paramètres statistiques de l’évaluation du croissant de blé pur. Paramètres sensoriels Paramètres statistiques

AcceptCouleur

Odeur

Goût

Arôme

Texture

abilité

Minimum

4

4

4

5

6

6

Maximum Moyenne

9 6.85

8 5.89

8 6.76

8 5.56

9 7.13

9 7.67

SD

1.23

1.65

1.85

1.98

1.47

1.24

CV

1.12

1.45

4.61

1.54

1.71

2.32

Il ressort des différents paramètres statistiques du croissant que les substitutions supérieures à 30% affectent la qualité du croissant qui decroit considérablement; il a même été observé une nette dimunition du volume du croissant à 50% farine de maïs et une plus grande consistance.

451

Tableau 6. Paramètres statistiques de l’évaluation du croissant à 15% maïs. Paramètres sensoriels Paramètres statistiques

AcceptCouleur

Odeur

Goût

Arôme

Texture

abilité

Minimum

5

4

4

5

5

6

Maximum Moyenne

9 6.13

9 5.17

8 6.12

9 5.32

9 7.06

9 6.77

SD

1.04

2.42

1.66

1.76

1.33

1.17

CV

2.15

2.13

1.45

1.19

1.71

0.97

Tableau 7. Paramètres statistiques de l’évaluation du croissant à 15% maïs. Paramètres sensoriels Paramètres statistiques

AcceptCouleur

Odeur

Goût

Arôme

Texture

abilité

4 9

4 9

4 8

6 9

6 9

6 9

SD

6.23 1.8

5.34 1.95

6.55 1.78

5.12 1.73

6.65 1.86

6.45 1.26

CV

2.34

1.65

0.97

0.89

0.89

1.12

Minimum Maximum Moyenne

Tableau 8. Paramètres statistiques de l’évaluation du croissant à 30% maïs. Paramètres sensoriels Paramètres statistiques

AcceptCouleur

Odeur

Goût

Arôme

Texture

abilité

Minimum

3

4

3

4

5

5

Maximum Moyenne

9 5.98

9 5.12

8 6.15

7 5

7 6.62

8 6.39

SD

1.23

1.62

1.55

1.44

1.34

1.11

CV

1.56

1.17

0.75

1.13

0.55

1.12

452

Tableau 9. Paramètres statistiques de l’évaluation du croissant à 50% maïs. Paramètres sensoriels Paramètres statistiques

Couleur

Odeur

Goût

Arôme

Texture

Acceptabilité

Minimum

3

4

3

4

3

4

Maximum

8 5.77

8 5.07

8 6.03

7 5.01

7 6.22

8 6.11

1.09 1.66

1.31 2.31

1.42 0.99

1.14 1.05

1.12 1.89

1.02 2.57

Moyenne SD CV

Évaluation du coût des différents produits Table 10. Coût- bénéfice de la préparation des pâtisseries de farine composée et de blé pure. Patisseries

Gâteaux

100% BLE Croissants

50

50

50

40

40

40

Moka

Gâteaux

Farine composee Croissants Moka

Quantité de blé(g) Quantité de la farine de mais (g) Coût de la farine de blé(FCFA) Coût de la farine de sorgho(FCFA) Coût des ingrédients (CFA)

0

0

0

10

10

10

15 000

15 000

15 000

12000

12000

12000

0

0

0

2000

2000

2000

2200

2275

8813

2000

2275

8813

Total coût (x)(CFA) Nom bre de produits

17 200 3000

17 275 2500

23 813 1002

14 000 3000

14 275 2500

22813 1002

Revenue provenant des produits (y)(CFA) Bénéfice net (CFA)

60 000

50 000

40080 16 267

60 000

50 000

40080

46 000

35 725

17 267

42 800

32725

Conclusion En conclusion nous pouvons affirmer que ce premier test a été concluant, la substitution de la farine de blé par le maïs est possible jusqu’à 50 % sans affecter la qualité du moka. Les croissants comme le pain subissent une substitution limitée.

Bibliographie Touré, A., A. Bengaly, J.F. Sheuring, D.T. Rosenov, and L.W. Rooney. 1998. The potential of local cultivars in sorghum improvement in Mali. African Crop Science Journal 6(1):1–7.

453

Mechanization of maize degerming for mawè production D.J. Hounhouigan1 , H.G. Sekpe1 A.P. Kayodé1, C. Mestres 2 , and C.M. Nago1, 1

Faculté des Sciences Agronomiques, Université Nationale du Bénin, 01 BP 526 Cotonou, Bénin. 2 CIRAD-CA,BP 5035, 34032 Montpellier cedex 1 Abstract Maize degerming was conducted on three maize varieties (Gnonli, Gbogboué, and DMR-ESR-W) to produce mawè (a traditional West African fermented dough) using a Maquina d’Andrea type 2 degermer. The performance of the equipment was assessed and the physico-chemical and sensory qualities of mawè and a derived product were analysed and compared to the product obtained by the traditional local method. The best operating parameters were 7 min residence time and a batch load of 13 and 15 kg for DMR-ESR-W and Gbogboué respectively. Under these conditions, the degerming rate ranged between 66 and 74% and the residual fat contents of the grits were 0.79 and 0.94, respectively. The optimum yields were 66% for Gbogboué and 68.6% for DMR-ESR-W. These values were similar or slightly higher than that obtained by the traditional method. A sensory triangular test carried out on aklui prepared with maquina-mawè showed that the properties (acidity, whiteness, and viscosity) were higher than that of traditionally prepared aklui. The result of a hedonic test showed that aklui prepared with maquina was acceptable. Résumé La dégermination du maïs a été réalisée sur trois variétés de maïs (Gnonli, Gbogboue, et DMR –ESR-W) pour produire le mawe (une pate traditionnelle fermentée de l`Afrique de l’Ouest) en utilisant la machine à dégermer Maquina d’Andrea type 2. La performance de cet équipement a été évaluée et les qualités physico-chimiques et sensorielles du mawè et d’un produit dérivé ont été analysées et comparées au produit obtenu par la méthode locale traditionnelle. Le meilleur paramètre opérationnel fut 7 minutes de temps de résidence et des chargements de 13 et 15kg pour DMR-ESR-W et Gbogboue respectivement. Sous ces conditions, le taux de degermination se situe entre 66 et 74 % et le taux résiduel des graisses des graines écrasées ont été 0,79 et 0,94, respectivement. Les rendements optimum ont été 66 % pour Gbogboue et 68,6 % pour DMR- ESR-W. Ces valeurs sont identiques ou légèrement supérieures à celles obtenues par la méthode traditionnelle. Le test du triangle sensoriel utilisé sur aklui préparé avec la

454

maquina-mawè a montré que les propriétés (acidité, blancheur et viscosités) sont plus élevées que celles de aklui préparé traditionnellement. Le résultat du test hedonic a montré que aklui prépare avec la Maquina a été acceptable.

Introduction One of the most important problems confronting the development of agro-based micro-enterprises in developing countries is the non -availability of suitable equipment for local food processing. The available types of equipment are often inadequate to produce the quality of local food products that would meet the taste of consumers. In many cases, the cost of the available equipment is prohibitive. For these reasons, the mechanisation of traditional food processing needs basic studies that establish their technical and economical feasibility. Mawè is a traditional West African fermented dough produced by dehulling and degerming maize. The traditional process includes crushing of maize kernel, sieving and washing of the grit (Hounhouigan et al. 1993a). Grit sieving and washing, which result in dehulling and degerming are manually accomplished and are very tedious and time-consuming operations. For example, it takes about four hours for a woman to process 100 kg of maize kernel (Kayodé 1999). Nago et al. (1989) reported that these operations can be successfully mechanized if their economic and technical feasibility justify the cost of mechanization. There is little or no information available in the literature on the mechanization of maize dehulling and degerming for mawè production. Previous studies showed that the Mini -PRL dehuller was wasteful of raw material. Moreover, this equipment provides only partial mechanization since manual washing of the grit is necessary to achieve adequate removal of germs (Hounhouigan et al. 1999). Studies involving the use of “Maquina d’Andrea” degermer indicated yield and defatting rates of the grit close to those obtained traditionally in mawè production (Capelle 1995; Hounhouigan et al. 1993a). The present study was carried out to determine the technical feasibility of dehulling and degerming of locally grown varieties of maize for mawè production, using “Maquina d’Andrea” Type 2 degermer.

Materials and Methods Three varieties of maize (Gnonli, Gbogboué, and DMR -ESR-W), with different physico-chemical characteristics and sensory properties (Nago et al. 1997b) were used in the study. Gnonli and Gbogboué were obtained from local markets and the International Institute of Tropical Agriculture (IITA) Benin Station provided DMR-ESR-W. Conditioning. 13 or 15 kg of maize was manually cleaned and conditioned to 13% moisture content. The latter was achieved by

455

adding a quantity Q of water to the dry grain with less than 13% moisture as follows: Q = m (Wf - Wi ) / 100 - Wf , with Q = weight of water to be added; m = weight of sample, Wi = initial moisture conte nt of grain, Wf = final moisture content of grain. Kernels with more than 13% moisture content were primarily dried and conditioned to attain the required moisture content. Dehulling and Degerming: The equipment used was the Type 2 degermer designed by Industria Maquina d’Andrea (Sao Paulo, Brasil). The equipment consists of a degerming chamber to which a sieving unit is attached. Both units are driven by a single motor of 1450 rpm speed. The degerming chamber is fitted with a rotor equipped with 36 shape d knives. The bottom of the chamber is equipped with an alternative sifter composed of two sieves: the upper one with circular apertures of 6 mm in diameter and the lower one with square sieve openings of 4 x 4 mm. The rotor operated at 850 rpm. Each sample of maize was introduced into the degerming chamber for 6 or 7 min through a feeding hopper. Three fractions: the over (whole and broken grains), the grits and the bran meal were discharged through three outlet hoppers. The over and the grits were blende d and the output was considered as grits yield of the equipment. Bran meal (hulls and germs) was discarded. Mawè production. Traditional commercial mawè was produced as described by Hounhouigan et al. (1993a) using 13 kg or 15 kg of cleaned grains. The grits resulting from grinding of grains, sieving and washing were soaked in water for 4 hr before milling and kneading to a dough, which was then covered with a polyethylene sheet and allowed to undergo natural fermentation for 72 hr. Grits from the mechanical degerming were processed following the same procedure to obtain a “maquina-mawè”. A preliminary experiment was done to determine the soaking duration of the “maquina grits” that would have similar fineness as the mawè flour. All the experiments were replicated three times.

Performance measurements The energy consumption per kg of grits was calculated as E = P x T/ 3600, where E = energy consumed by the motor (kWh), P = power drawn by motor to process a sample (kW), T = time needed to degerm a sample (S). The degerming rate (Dg) was calculated as the ratio of the amount of fat removed from the grain to the amount of fat in the grain before degerming as follows:

456

Dg = (fo – f) / fo x 100, where fo = fat content of the whole grain (% db), f = fat content of the grit (% db). The dehulling rate (Dh) was calculated as the ratio of the amount of hulls removed from the grain to the amount of hulls of the grain before degerming as follows: Dh=Wo-W /Wo x 100, where Wo = percent of hulls (whole grain), W = percent of hulls (grits). The percent of hulls on whole grains or grits was determined following the method described by Nago et al. (1997): 50 g of grains or 15 g of grits was soaked in 50 ml of water for 1 hr and manually dehulled. The hulls were dried at 105°C until constant weight and expressed in percent.

Chemical analysis Titratable acidity and pH were measured as described by Hounhouigan et al. (1993a). Protein, fat, ash and moisture content were determined using AACC methods (AACC 1984). Crude fibre content was determined as described by Osborne and Voogt (1978). Carbohydrate was calculated by difference.

Physical analysis Flour particle size measurement was performed as described by Hounhouigan et al. (1993a), using sieves of increasing apertures; 90, 250 and 355 µm.

Results and Discussion Tables 1 and 2 show the effect of variety, batch load and residence time of kernel on fine grits and total grit yields, respectively. While only the residence time affected the fine grit yield of DMR, the total grit yield was significantly affected by the maize variety whatever the operating conditions. The lowest yield was produced by Gnonli (39.6%) while DMR gave the highest yield (74%) followed by Gbogboué (70%). Table 1. Effect of residence time and sample size on fine grit yields of three maize varieties produced from the Maquina degermer compared with the traditional processing method.

Maize variety

DMR-ESR Gbogboue Gnonli

Residence time (mn) 13 kg

6 7 6 7 6 7

56.2 (a) 60.3 (b) 59.9 59.1 39.6 -

Yields (% dwb) 15 kg Traditional

53.6 (a) 58.8 (b) 61.5 61.5 42.8

64.9 66.7 -

457

Table 2. Effect of residence time and sample size on total grit yields of three maize varieties produced from the Maquina degermer compared with the traditional processing method.

Maize variety DMR-ESR Gbogboue Gnonli

Residence time (mn) 6 7 6 7 6 7

13 kg 73.1 (d) 67.7 (e) 65.6 (a) 63.0 (c) 39.6 (f) -

Yields (% dwb) 15 kg Traditional 74.1 (d) 68.7 (e) 64.9 70.2 (b) 66.2 (a) 66.7 42.8 (f) -

The batch load also affected the grits yield; the higher the batch load, the higher the grit yield. Furthermore, the residence time affected the grit yield: the grit yield was lower at 7 min residence time than at 6 min. Gnonli was less affected by the residence time than the other two varieties. The differences among the varieties might be related to their friability. Gnonli, a floury variety, was the most friable of the three varieties, followed by Gbogboue and DMR (Nago et al., 1997b; Hounhouigan et al. 1999). In corn dry milling, the amounts of grits and meal produced depend upon the proportion of the vitreous endosperm in the kernel (Manoharkumar et al. 1978; Chassaray, 1991). The low grit yield recorded on Gnonli maize variety in the present study could be an indication that the equipment used is not suitable for floury corn. The grit yields recorded with Gbogboué at 15 kg/6 min and with DMR (15 kg/6 min) ranged between 70 and 74%. These values were higher than that achieved by traditional method (66%) and that obtained with Mini -PRL dehuller in mawè production, using DMR maize variety (Hounhouigan et al. 1999). The best operating conditions were chosen on the basis of grit yield similar or higher than that obtained in traditional mawè processing with fat content of grits less than 1%, which is the recommendation for a long shelf-life of cereal flour. These operating conditions were 15 kg/7 min for Gbogboué and 13 kg/7 min for DMR (Table 3). Table 3. Efficiency of the Maguina d’Andrea degermer under the best operating conditions. Energy consumption (kWh/kg)

Grit yield (%w/w)

Fat (%)

Degerming rate (%)

Dehulling rate (%)

Unbroken grit (%w/w)

Gbogboué 15 kg/7 min

0.05

66.2

0.9

78.7

89.3

4.7

DMR 13 kg/7 min

0.05

67.7

0.8

82.2

86.8

7.4

Type of mawè

458

The optimum grits yields were 66.2% for Gbogboué and 67.7% for DMR with a corresponding energy consumption of 0.5 kWh. The fat contents of grits obtained under these conditions were 0.9% for Gbogboué and 0.8% for DMR. Fat content of traditional market mawè is about 1% (Hounhouigan et al. 1993a). The yield of mawè obtained from Gbogboué grit at 15 kg/7 min was similar to that of traditional mawè but that from DMR was higher. Comparative study of particle size distribution of the different mawè produced (Table 4) showed that longer soaking time (15 hrs. vs 4 hrs.) was necessary for Maquina grits to provide a flour of particle size similar to that of traditional mawè. This is due to the bigger particle size of Maquina grits compared to the grits obtained by the traditional method. Table 4. Particle size distribution of traditional mawè and maquina mawè at different soaking duration. Particle size (µm) Type of mawè Mawè 4hrs (Traditional) Mawè 6 hrs Mawè 8 hrs Mawè 10 hrs Mawè 12 hrs Mawè 15 hrs Mawè 24 hrs Mawè 48 hrs

x > 355

250 ? x < 355 90 ? x < 250 x < 90

5,5 15,3 14,4 14,0 12,2 5,6 3,5 3,9

4,4 3,8 8,6 8,6 4,6 5,4 4,12 4,6

19,7 20,5 20,4 18,6 20,4 18,9 21,6 19,1

70,4 60,4 56.6 58,8 62,8 70,1 70,7 72,4

Colour parameters recorded for the different types of mawè are given in Table 5. The processing method significantly affected the colour parameter L* (luminosity), b* and ? E (total colour difference with standard tile). Maquina mawè were brighter than traditional mawè probably because the traditional mawè retains more hulls and tip cap than the other processing method. Maquina mawè will be more appreciated by Beninese consumers who normally consider colour, fineness and acidity as the major quality criteria in mawè consumption (Hounhouigan et al. 1993a). Table 5. Comparison of colour parameters of traditional and maquina-mawè.

Colour parameters L* a* b* ?E

Mawè of Gbogboué Traditional Maquina 80.6? 0.0a 89.8? 0.4b –2.7? 0.1a –2.6? 0.0b 12.6? 0.5a 10.2? 0.2b 26.5? 0.0a 18.6? 0.1b

Mawè of DMR-ESR-W Traditional Maquina 83.9? 0.1c 90.6? 0.1d –2.1?0.05c –1.7?0.0d 8.3?0.0c 5.1? 0.1d 21.3? 0.1c 14.1? 0.1d

The processing method affected the pasting behaviours of mawè (Table 6). The stability of starch (Vmax-Vf95) and the index of

459

gelatinization (Vfin-Vf95) were higher in the maquina than the traditional mawè. Table 6. Comparative study of pasting characteristics of traditional and maquina-mawè. Pasting parameters

Mawè of Gbogboué Traditional Maquina

Mawè of DMR-ESR-W Traditional Maquina

Vmax 76.1? 0.3a 87.9?2.0b 54.5?0.7c 62.5? 2.0d Vf95 57.1? 0.0a 68.6?1.3b 45.4?0.5c 53.6? 1.6d Vfin 91.6? 0.0a 109.4? 2.2b 78.0?1.2c 93.5? 2.2d Vmax-Vf95 19 19.3 9.1 8.9 Vfin-Vf95 39.9 40.8 32.6 39.9 Vmax = maximum viscosity during heating, Vf95 = viscosity after 15 min at 95°C, Vfin = viscosity after cooling at 50°C, Vmax-Vf95 = stability of starch, Vfin-V95f = index of gelatinization.

The high fat and fibre content of the traditional mawè may have reduced its viscosity as previously reported for ogi obtained from dry milled maize compared with the viscosity of ogi from steeped whole maize (Akingbala et al. 1987). High fat content in maize flour decreases the viscosity of the derived products (Nago et al. 1997b). Furthermore, all the pasting values recorded were higher in mawè of Gbogboué than those of DMR. Greater particle size in DMR-mawè due to the hardness of its endosperm may be responsible for this difference as reported by Nago et al. (1998) for ogi and Akissoé (1992) for mawè. pH and titratable acidi ty of all samples were similar (Table 7). The average pH recorded was 3.4. This value is lower than that of commercial mawè which ranged between 3.6 and 4.2 (Hounhouigan et al. 1993a). The pH is close to that of Nigerian Ogi (3.5 and 4.5) (Banigo and Muller 1972; Akinrele 1970). There were no significant changes in overall protein, carbohydrate and ash content of the various types of mawè (Table 7). There was a reduction in fat and fibre content of maquina mawè compared to the traditional product. Gbogboué kernel contains 4.7% crude fat, which was reduced to about 1% in traditional product and to 0.9 in maquina mawè. Triangular test assessment showed that maquina and traditional mawè were different in colour, acidity and viscosity. The hedonic test indicated that both products were acceptable.

Conclusion This investigation has shown that maquina d’Andrea type of degermer is suitable for the mechanization of maize dehulling and degerming for traditional mawe production. Medium and hard endosperm varieties of maize provide best yield and should be selected for maquina processing. The grain moisture content should be about 13%, which corresponds to the average of the moisture content of dry maize for storage in the tropics. The dry-

460

460

Table 7. Proximate compo sition of corn, traditional and maquina-mawè. Chemical composition pH Titratable acidity Crude protein (% dw) Crude fat (% dw) Crude fibre (% dw) Ash (% dw) Carbohydrate (%dw)

Mawè Gbogboué Traditional Maquina 3.4? 0.0 3.3? 0.0 2.1? 0.01 2.3? 0.02 9.4? 0.04 9.1? 0.14 1.2? 0.01 0.9? 0.0 0.7? 0.02 0.5? 0.01 0.32?0.00 0.33? 0.02 88.4 89.2

Mawè DMR-ESR-W Traditional Maquina 3.4? 0.0 3.3? 0.0 2.1? 0.02 2.3? 0.02 9.8? 0.1 9.6? 0.11 1.1? 0.03 0.8? 0.0 0.8? 0.02 0.6? 0.02 0.36? 0.02 0.34? 0.01 87.9 88.6

Corn Gbogboué

DMR

-

-

10.8? 0.0 4.7?0.0 1.1?0.0 1.5?0.0 81.9

10.6? 0.0 4.5?0.0 2.0?0.0 0.5?0.0 82.4

461

dehulling/degerming of maize opens new opportunities for utilisation of maize grits for brewery and dry-milling for biscuit factories.

Aknowledgement The authors thank the International Foundation for Science for financial support.

References AACC. 1984. Approved methods of the American Association of Cereal Chemists, 8th Edition, St. Paul, MN, USA. Akingbala, J.O. E.U. Onochie, I.A. Adeyemi , and G.B. Oguntimein. 1987. Steeping of whole and dry milled maize kernel in ogi preparation. J Food Process. Preserv. 11:1-11. Akinrele, I.A. 1970. Fermentation studies on maize during the preparation of a traditionnel African starch-cake food. J. Sc. Food Agric., 21: 619–625. Banigo, E.O.I., and H.G. Muller, 1972. Manufacture of ogi (a Nigerian fermented cereal porridge): comparative evaluation of corn, sorghum and millet. Canadian Inst. Food Sc. Technol. J., 5 (4):217–221. Capelle, J., 1995. Essais d’une dégermeuse à maïs. Rapport de stage. Département des systèmes agro-alimentaires et ruraux CIRAD-SAR. Chasseray , D., 1991. Caractéristiques des grains et leurs dérivés. Pages 397–421 in Godon et C. William (eds.) Les industries de premières transformations des céréales. Collection technique et Documentation. Ed. Lavoisier. Hounhouigan, D.J., M.J.R. Nout, C.M. Nago, J.H. Houben, and F.M. Rombouts. 1993a. Composition and microbiological and physical attributes of mawè, a fermented maize dough from Bénin. Inter. J. Food Sc. Technol., 28 : 513–517. Hounhouigan, D.J., M.J.R. Nout, C.M. Nago, J.H. Houben, and F.M. Rombouts. 1993b. Changes in the physico-chemical properties of maize during natural fermentation of mawè. J. Cereal Sci. 17(3): 291–300. Hounhouigan, D.J., P. Kayode, C.M. Nago, et C. Mestres. 1999. Etude de la mécanisation du décorticage du maïs pour la production du mawè. Annales des Sciences Agronomiques du Bénin 2:99–113. Kayode, A.P. 1999. Etude de la mécanisation du décorticagedégermage du maïs pour la production du « mawè». Thèse d’Ingénieur Agronome, FSA/UNB. Abomey-Calavi, Bénin, 128p.

462

Manoharkumar, B., P. Gerstnkorn, H. Zwingelberg, and H. Bolling. 1978. On some correlation between grain composition and physical characteristics to the dry milling performance in maize. J. Food Sci. Technol. 15 : 1–6. Nago, C.M. 1989. Technologies traditionnelles et alimentaires au Bénin: aspects techniques, biochimiques et nutritionnels. Identification et caractérisation des principales filières et technologies du secteur traditionnel de transformation alimentaire. Document FSA/UNB, Abomey-Calavi, Bénin. Nago, C.M., C. Thuillier, et D.J. Hounhouigan. 1992. Etude des systèmes techniques de transformation artisanale du maïs au Bénin. Communication aux journées scientifiques du CIRAD, Montpellier (France), Novembre 1992. In J. Muchnik (éd.) Alimentation, techniques et innovations dans les régions tropicales. Editions l’Harmattan. Nago, C.M. 1997. La transformation alimentaire traditionnelle du maïs au Bénin : Détermination des caractéristiques physicochimiques des variétés en usage ; relations avec l’obtention et la qualité des principaux produits dérivés. Thèse de Doctorat D’Etat. Académie de Paris, Université Paris 7-Denis Diderot UFR de Biochimie. 201 p. Nago, C.M., E. Tetegan, F. Matencio, and C. Mestres. 1998. Effects of maize type and fermentation conditions on the quality of beninese traditional ogi, fermented maize slurry. Journal of Cereal Science 28, 215–222. Nago, C.M., N. Akissoe, and F. Matencio. 1997. End use quality of some African corn kernels. 1. Physico-chemical characteristics of kernels and their relationship with the quality «lifin», a traditional whole dry-milled maize flour from Benin. Journal of Agricultural and Food Chemistry 45: 555–564. Osborne, D.R., and P. Voogt. 1978. The analysis of nutrients in food . Academic Press Inc. London, England. Pages 151–153. Sahay, K.M. 1990. Evaluation of a general purpose abrasive mill for pearling of coarse and dehusking of pulses. Intern. J. Food Sci. Technol. 25: 220–225. Watts, B.M., G.L. Ylimaki, L.E. Jeffery, et L.G. Elias. 1991. Méthode de base pour l’évaluation sensorielle des aliments. Ottawa, Ont., CRDI. Pp 57–68.

463

Effect of synthetic and botanical products on seed viability and seedling vigor of maize from two agro-ecological zones of Cameroon C.G. Zonkeng, A. Tagne, C. Thé, and P. Feudjio Institute of Agricultural Research for Development (IRAD) P.O. Box 2067 Yaounde, Cameroon Fax: (237) 222 59 24 E-mail: [email protected]/[email protected] Abstract Two laboratory experiments were conducted in 2000 to identify the best chemicals and crude natural powder for seed dressing that would allow high and durable preservation of maize seed viability by the small-scale seed producer. Seed for two openpollinting maize varieties were obtained from the humid forest zone of Cameroon and used in Experiment 1. Treatments consisted of four synthetic chemicals, each applied at four doses with two untreated controls. Seed for the two varieties were obtained from the savanna zone and used in Experiment 2. Each variety was treated with five synthetic chemicals, one mixture of chemicals, crude powder from four aromatic plants (Ocimum gratissimum, Thymus vulgaris, Azadirachta indica, and Cymbopogom citratus), and four combinations of chemicals and aromatic plants. The experiments started two months after harvest and continued for eight months. Samples were drawn from each treatment on a monthly basis for germination test and data were collected on germination percentage, percentage of abnormal seedlings, height of seedlings, and length of seedling radicle. After 240 days of storage, the average germination percentage of maize seed from the savanna zone was 72 while that from the humid forest zone was 44.3. Therefore, the savanna zone is a better ecology for maize seed production in Cameroon. The optimum duration for maize seed storage without significant loss of viability was 180 days after harvest. The seed dressings investigated in this study were equally effective during this period. Beyond this period, the seed dressings that produced maximum germination rates were Marshall 35 ST, Benlate + Marshall 35 ST, Marshall 35 ST + T. vulgaris, Marshall + O. gratissimum. The use of 30 g of Marshall 35 ST per 100 kg of seed, or a mixture of Marshall 35 ST and a crude natural powder would redu ce the cost of seed dressing considerably without reducing the maximum germination percentage. Résumé Deux expériences de laboratoire ont été conduits en 2000 pour identifier le meilleur produit chimique et la poudre crue naturelle pour l`enrobage des semences qui peut permettre une haute et durable préservation de la viabilité des semences

464

de maïs par le petit producteur de semences. La semence de deux variétés de maïs à pollinisation libre a été obtenue de la zone de la forêt humide du Cameroun et utilisée dans l’expérimentation. Les traitements étaient constitués de quatre produits chimiques synthétiques, chacun appliqué à quatre doses avec deux témoins sans application. La semence pour les deux variétés a été obtenue à partir de la zone de la savane et utilisée dans l`expérimentation 2. Chaque variété a été traitée avec cinq produits chimiques synthétiques, un mélange de produit chimique, une poudre crue provenant de quatre plantes aromatiques (Ocimum graticimum, Thymus vulgaris, Azadirachta indica et Cymbopogom citratus) et quatre combinaisons de produits chimiques et des plantes aromatiques. Les expériences ont commencé deux mois après la récolte et ont continué pendant huit mois. Des échantillons ont été prélevés à partir de chaque traitement mensuellement pour le test de germination et les données ont été collectées sur le pourcentage de germination, le pourcentage des semences anormales, hauteurs des plants et longueur des racines. Apres 240 jours de conservation, le pourcentage moyen de germination des semences de maïs provenant de la zone de savane a été de 72 tandis que celui en provenance de la zone de forêt humide a été de 44,3. Par conséquent, la zone de savane est une meilleure écologie pour la production des semences de maïs au cameroun. La durée optimum pour la conservation des semences de maïs sans une perte significative de la viabilité est de 180 jours après la récolte. L`enrobage des semences expérimentée dans cette étude a été effective durant cette période. Après cette période, l`enrobage des semences qui a donné le taux maximum de germination a été Marshall 35 ST, Benlate + Marshell 35ST, Marshall 35 ST + T. vulgaris, Marshall + O. gratissimum. L`utilisation de 30g de Marshall 35ST pour 100kg de semence, ou un mélange de Marshall 35 ST et une poudre crue naturelle pourrait réduire considérablement le coût d`enrobage des semences sans réduire le pourcentage maximum de germination.

Introduction In West and Central Africa (WCA) maize (Zea mays L) is a major crop and is extensively used for human consumption, as animal feed, and raw material in food industry. According to Okoruwa (1995) the demand for maize in WCA will increase tremendously in the new millennium. In Cameroon, the average grain yield of maize is 2.0 t/ha (FAO 1998). The annual national production, estimated at 750 000 t, is less than the demand; therefore 30 000 t are imported each year (Toussi et al. 1999). Some of the factors that contribute to the low grain yield are poor seed quality and post-harvest losses. Seed quality is highly influenced by seed viability, which in turn, is highly influenced by seed moisture

465

content and storage conditions. High grain moisture and poor seed storage conditions usually favor insects and fungi proliferation and lead to a reduction in the viability of seed stocks (Samgam et al. 1987; Moreno Martinez et al. 1981; Adegoke et al. 1995). Maude (1993) considered temperature below 5oC and relative humidity less than 50% to be the best conditions for storage of orthodox seed. In humid tropical areas, these conditions are hardly achieved at the level of resource -poor seed producers. For these seed producers, insects and fungi constitute a serious threat to seed viability. Under such situations, seed treatment with synthetic chemicals or extracts from natural plants appears to be an alternative to alleviate these problems. Moreno and Vidal-Gaona (1981) reported that after 150 days Captafol and Carbendazim + Maneb were efficient in maintaining the germinabiliy of maize seed inoculated with spores of storage fungi. In 1994 The National Cereals Research and Extension Project (NCRE) in Cameroon identified Marshall 35 ST and Actellic 2% dust as potential insecticides for seed preservation. Essential oils and powders derived from aromatic plants are known to possess insecticidal and antimicrobial activities and their importance in seed and food conservation has been reported. Adegoke and Odesola (1995) found that within a storage period of 10 days, maize and cowpea treated with lemon grass (Cymbopogom citratus) showed no mould and bacteria growth and had no physical damage. Bhaskara et al. (1998) found that strawberry fruit decay could be controlled up to 75.8% by essential oils from Thymus vulgaris. From a study consisting of extracts from five West African plants, Awuah (1989) concluded that extracts from leaves of Cymbopogom citratus were effective as a fungitoxicant and completely inhibited the growth of all four fungi studied. Similarly, extracts from fresh leaves of Ocimum gratissimum reduced radial growth of the fungi by 10–60%. Yehouenou (1995) investigated the response of maize to seed treatments and found a yield increase in plots sown with seed treated with Azadirachta indica leaves over the control. Gwinner (1996) also found that extracts from A. indica leaves had repulsive and inhibitory effects on infestation by and development of storage insect pests. Although synthetic chemicals have protective effects on seed, they have the disadvantage of having toxic residual effects that pollute the environment. In Cameroon, for example, some farmers have reported toxic effects of Marshall after seed treatment and long term (unspecified period) storage. Natural extracts from plants have the advantage of being less polluant to the environment, but the appropriate doses to be used by farmers and small-scale seed producers in Cameroon are not known.

466

The objectives of this study were to: (a) identify the best seed treatment (insecticides, fungicides, natural compounds, or a mixture of treatments and doses) that allows high and durable preservation of maize seed viability and seedling vigor at the farmer and small-scale seed producer level; (b) determine the storage period beyond which toxic effects can be detected.

Materials and Methods Two laboratory experiments were conducted in 2000 with maize seed obtained from the humid forest and savanna zones of Cameroon. Experiment 1. Treatments in the first experiment consisted of two widely grown open pollinated maize varieties (CMS 8704 and CMS 8501), four synthetic chemicals: Actellic 50 EC (AEC), AD = Actellic 2% dust (AD), Marshall 35 ST (MA), and AD + BE = Actellic 2% dust + Benlate (AD + BE). Each chemical was applied at four doses: D1= recommended commercial dose (10 ml of AEC/1 l of water for 200 kg of seed, 50 g of AD for 100 kg of seed, 800 g of MA for100 kg of seeds, 800 g of MA +60 g of BE for 100 kg of seed); D2= 0.5 D1; D3 = 2 D1; D4 = 0.03 % of grain weight (6 ml of AEC/1 l of water for 200 kg of seed, 30 g of AD for 100 kg of seed, 30 g of MA for 100 kg of seed, 30 g of MA + 30 g of BE for 100 kg of seed); Experiment 2. Seed of the two open-pollinated maize varieties used in Experiment 1, CMS 8704 and CMS 850, were also used Experiment 2. Seed of each variety was treated with five synthetic chemicals (Ridomil plus 72, Hydrox, Marshall 35 ST, Benlate, Oncol), a mixture of synthetic chemicals (Benlate + Marshall 35 ST), powders prepared from four natural aromatic plants and four mixtures of chemicals + natural powders. The plants were O. gratissimum, T . vulgaris, C. citratus, and A. indica. The dose used for all the treatments was 30 g of product to 100 kg of seeds. Fresh mature leaves of the aromatic plants were harvested, washed with sterile distilled water and dried. Dried leaves were then crushed and sieved to obtain the powder. Apart from Actellic 50 EC which was applied as suspension in water, all other treatments were applied as dust. Data collection. For each of the two experiments, two untreated controls were added. Each sample was made of 3500 seeds that were stored at room temperature. The initial seed moisture

467

content was 12% for both zones. Each experiment started two months after harvest with a 95% initial seed germination rate. Germination test was done monthly for eight months. Data were collected on germination percentage, percentage of abnormal seedlings (i.e., seedlings showing abnormalities that may not allow them to grow to a normal plant under favorable soil, moisture, temperature and light conditions), and seedling vigor measured as seedling height and the length of seedling radicle. Data analysis. A 4-factor (storage period, variety, chemical, and doses) randomized complete block design (RCBD) with two replications was used in Experiment 1 while a 3-factor (storage period, variety, treatments) RCBD with 2 replications was used in Experiment 2. For each experiment a combined analysis of variance was performed across all storage periods and separate factorial analyses were then done for the five consecutive storage periods (120, 150, 180, 210, and 240 days). Orthogonal contrasts were also performed for each storage period to test the effects of various treatments.

Results and Discussion Experiment 1 In Experiment 1, no significant interaction was found between varieties and see d treatment factors. Therefore, varietal means were used for the 3-factor (storage period, chemical, and doses) analysis. Highly significant storage period x chemical interaction mean squares were detected for percentage germination and seedling height (data not shown). This indicated that the ranking of the effectiveness of the treatments was not consistent among the seed storage periods. During the first 120 days of storage, the germination percentage remained relatively high (=90%) for all treatments except the control that was 88% at 30 days. Thereafter, significant differences were observed among the treatments for each of the remaining storage periods (Table 1). Table 1. Mean squares from orthogonal contrasts for germination percentage between groups of treatments applied to maize seed harvested from the humid forest zone of Cameroon and stored for different periods in 2000. Storage period (days) Orthogonal contrast 120 150 180 210 240 [AEC+AD+MA+ 0.188** 0.651** 0.743** 0.942** 0.988** (AD+ BE) ] vs control (AD+BE) vs (AEC+AD 0.445 0.432 0.255 -1.409 0.734 + MA) MA vs [AEC+AD+ 0.560 4.502** 9.618** 11.468** 11.368** (AD + BE)] [AD+(AD+BE)] vs -0.498 -2.234 -4.279** -3.412** -4.468** [AEC + MA] *, ** Significant at P = 0.05 and P = 0.01 respectively AEC = Actellic 50 EC, AD = Actellic 2% dust, MA = Marshall 35 ST, BE = Benlate

468

As expected, the diffrences between the untreated control and the chemical treatments were highly significant for all storage periods (Table 1). The mean squares associated with this contrast increased from 0.188 at 120 days to 0.998 at 240 days. This indicated that the effectiveness of the seed treatments in sustaining high seed germination increased as the storage period i ncreased. The orthogonal contrasts further showed that comparisons involving Actellic 2% dust and Benlate (AD + BE) did not differ significantly from the other chemicals. Conversely, the effectiveness of Marshall 35 ST (MA) differed significantly from that of all other chemicals for 150- to 240-day storage periods (Table 1). Mixture of Marshall 35 ST and Actellic 50 EC (ABC + MA) also differed from the other chemicals during the last three storage periods. On the average, germination percentage decreased from 90% at 120 days to about 44% at 240 days of storage (Table 2). For all storage periods between 120 and 240 days, the control had the lowest germination rate with 87% at 120 days and 28.5% at 240 days (Fig. 1), a consequence of the high colonization of germinating seed by fungi. A high infestation rate by storage insects was also observed in the control (50% at 120 days and 80% at 240 days). Treatment x dose interaction was not significant in any storage period for germination percentage. Actellic 50 EC/D4

100

Actellic 2 % Dust/D1

90

Marshall 35 ST/D4

Germination %

80

Actellic Dust + Benlate/D1 Control

70 60 50 40 30 20 10 0 120

150

180

210

240

Storage period (days)

Fig. 1. Germination % of humid forest maize seeds as influenced by seed dressing and length of storage period.

469

Table 2. Effects of chemical seed treatment, dosage, and length of the storage period on the germination percentage of maize seed harvested from the humid forest of Cameroon in 2000. Storage period (days) 120 150 180 210 90.8 76.3 66.8 52.5 88.3 80.5 67.5 52.3 88.0 80.8 65.8 53.0 91.8 73.0 64.0 47.8 Actellic 2% dust 90.0 77.3 68.5 54.3 (AD) 90.8 79.0 68.3 52.3 90.3 77.3 66.3 56.5 87.8 78.8 66.5 55.3 Marshall 35 ST 90.8 85.5 80.3 70.3 (MA) 89.8 84.3 79.8 70.3 90.0 85.5 83.0 68.8 92.5 84.0 82.0 68.8 AD + Benlate 90.5 82.3 73.5 55.0 (AD + BE) 89.0 78.3 71.0 55.3 89.3 80.3 71.8 61.3 93.5 82.3 71.3 55.0 Control 87.0 69.3 59.0 42.0 Mean 90.0 79.6 70.9 57.0 C.V. 1.8 5.3 4.7 5.7 LSD (0.05) ns ns 7.0 6.9 D1= Recommended commercial dose (10 ml/1l H 20 for AEC, 50 g/100 kg of seed for AD, 40 g/5 kg seed of seed for MA, 60 g/100kg of seed for BE), D2=0.5 D1, D3= 2D1, D4= 30 g/100 kg of not significant at the 0.05 probability level. Treatment Actellic 50 EC (AEC)

Dosage D1 D2 D3 D4 D1 D2 D3 D4 D1 D2 D3 D4 D1 D2 D3 D4

240 34.3 38.8 36.0 41.5 43.3 40.5 39.5 40.0 57.0 53.0 64.3 52.5 49.0 45.3 45.0 45.0 28.5 44.3 10.6 9.9 seed ns,

Therefore, the germination percentage of Treatment D3 (Marshall at 1600 g/100 kg of seed) was not significantly different from that of Treatment D4 (Marshall at 30 g/100 kg of seed) For each storage period, the chemical treatments did not differ significantly for percentage of abnormal seedlings (Table 3). Although percentage of abnormal seedlings increased with increased storage period, the treatment x dose interaction was not significant. For all storage periods, Treatment D3 (Marshall at 1600 g/100 kg of seed) had an adverse effect on seedling growth and radicle elongation (Table 4 and 5). It could be inferred, therefore, that the higher dose of Marshall had an harmful effect on the maize seedlings. Marshall at D4 enhanced roots and seedling growth. This dose of Marshall, which gave 82% germination (28% increase over the control) 180 days after the beginning of the experiment, could be recommended for commercial seed treatment. Highly significant treatment x dose interaction was observed for seedling height and radicle length. For each treatment the best dose for optimum germination percentages after 240 days of storage at room temperature were D4 for Actellic 50 EC (41.5% germination), D1 for Actellic 2% dust (43.3 %), D3 for Marshall 35 ST (64.3 %), and D1 for Actellic dust 2% + Benlate (49%).

470

Table 3. Effects of chemical seed treatment, dosage, and length of the storage period on the percentage of abnormal seedlings of maize seed harvested from the humid forest of Cameroon in 2000. Storage period (days) 120 150 180 210 240 2.7 2.5 4.7 5.8 7.3 3.0 2.3 4.2 3.8 4.0 3.0 2.0 4.3 4.3 5.3 2.3 2.3 3.0 4.0 11.0 Actellic 2% dust 3.8 3.5 2.8 5.3 7.0 (AD) 3.3 3.7 3.5 3.5 6.3 2.5 3.0 3.3 4.8 9.8 2.3 3.0 4.5 4.0 9.0 Marshall 35 ST 3.3 3.3 5.0 6.3 8.5 (MA) 2.3 2.5 8.5 6.8 8.0 1.8 3.0 4.8 9.5 8.5 1.8 2.3 4.0 2.8 11.0 AD + Benlate 2.0 2.5 3.3 2.8 7.8 (AD + BE) 2.0 3.0 5.0 3.3 7.5 2.0 4.5 4.3 3.8 6.0 1.5 3.0 3.5 4.5 7.5 Control 3.0 4.5 3.8 2.3 4.3 Mean 2.5 2.8 4.3 4.5 7.5 C.V. 46.6 30.6 34.7 38.3 30.7 LSD (0.05) ns ns ns ns ns D1= Recommended commercial dose (10 ml AEC/1l H20 for 200 kg of seed, 50 g of AD for 100 kg of seed, 800 g of MA for 100 kg of seed, 60g of BE for 100 kg of seed). D2=0.5 D1, D3= 2D1, D4=0.03 % of grain weight. ns, not significant at the 0.05 probability level. Treatment Actellic 50 EC (AEC)

Dosage D1 D2 D3 D4 D1 D2 D3 D4 D1 D2 D3 D4 D1 D2 D3 D4

Table 4. Effects of chemical seed treatment, dosage, and length of the storage period on the height (cm) of seedlings of maize seed harvested from the humid forest of Cameroon in 2000. Storage period (days) 120 150 180 210 240 5.4 3.8 6.2 6.6 8.6 6.8 4.5 5.6 7.6 8.0 5.6 4.8 5.7 7.5 7.4 6.6 2.8 5.9 7.5 7.4 Actellic 2% dust 5.2 2.7 6.1 6.2 7.4 (AD) 6.1 3.2 5.4 7.0 7.7 5.3 2.3 6.1 6.0 8.2 4.8 2.0 5.3 6.9 8.8 Marshall 35 ST 4.3 3.6 2.0 3.3 5.7 (MA) 3.3 4.1 2.7 5.3 7.4 2.6 2.0 1.9 2.5 3.6 5.8 2.6 5.1 5.7 9.7 AD + Benlate 5.5 3.1 5.4 6.0 8.0 (AD + BE) 5.5 2.6 6.3 6.6 7.8 7.0 2.6 6.1 7.3 7.0 6.0 2.9 5.5 7.0 6.4 Control 4.1 3.0 5.4 5.3 3.9 Mean 5.3 3.1 5.1 6.1 7.2 C.V. 18.8 34.1 8.2 19.1 16.2 LSD (0.05) 2.1 ns 0.8 2.4 2.4 D1= Recommended commercial dose (10 ml AEC/1l H20 for 200 kg of seed, 50 g of AD for 100 kg of seed, 800 g of MA for 100 kg of seed, 60g of BE for 100 kg of seed). D2=0.5 D1, D3= 2D1. D4=0.03% of grain weight. ns, not significant at the 0.05 probability level. Treatments Actellic 50 EC (AEC)

Dosage D1 D2 D3 D4 D1 D2 D3 D4 D1 D2 D3 D4 D1 D2 D3 D4

471

Table 5. Effects of chemical seed treatment, dosage, and length of the storage period on the radicle length (cm) of seedlings of maize seed harvested from the humid forest of Cameroon in 2000. Storage period (days) 120 150 180 210 240 9.8 9.4 10.2 10.8 10.8 10.4 9.3 9.9 10.5 10.8 9.4 10.0 10.0 8.6 10.4 9.6 9.9 9.5 9.1 9.6 Actellic 2% dust 9.2 10.0 10.2 8.3 9.2 (AD) 9.9 9.5 9.6 10.4 10.9 8.2 8.5 10.2 9.8 10.5 8.5 9.3 8.5 10.7 13.2 Marshall 35 ST 2.3 2.4 2.0 2.6 6.3 (MA) 2.7 5.5 3.4 4.0 8.7 1.7 2.2 1.4 1.7 2.2 8.7 9.6 9.1 8.5 10.4 AD + Benlate 7.4 8.7 9.5 8.3 12.6 (AD + BE) 8.4 9.3 9.7 10.7 11.3 9.7 8.7 8.4 9.4 9.8 8.4 8.9 10.0 10.2 9.2 Control 6.7 8.2 10.2 9.1 5.1 Mean 7.7 8.2 8.3 8.4 9.4 C.V. 14.2 17.1 14.9 15.9 16.2 LSD (0.05) 2.3 2.9 2.6 2.8 3.2 D1= Recommended commercial dose (10 ml AEC/1l H20 for 200 kg of seed, 50 g of AD for 100 kg of seed, 800 g of MA for 100 kg of seed, 60g of BE for 100 kg of seed). D2=0.5 D1, D3= 2D1, D4=0.03% of grain weight. ns, not significant at the 0.05 probability level. Treatments Actellic 50 EC (AEC)

Dosage D1 D2 D3 D4 D1 D2 D3 D4 D1 D2 D3 D4 D1 D2 D3 D4

Experiment 2 Highly significant treatment x storage period was detected for all variables except percentage abnormal seedlings (data not shown). From the beginnig of the experiment to the 90-day storage period, there were no significant differences among treatments for germination percentage, seedling height and radicle length. All contrasts among treatments were highly significant for germination percentage (Table 6). Table 6. Orthogonal contrasts for germination percentage between groups of treatments applied to maize seed harvested from the savanna zone of Cameroon and stored for different periods in 2000. Orthogonal contrast

120

Crude plant powder (alone and in combination) vs other treatments

-1.390**

Storage period (days) 150 180 210 -3.877**

-1.485*

240

-5.632**

-3.257**

Crude plant powder -1.764** -4.806** -4.028** -7.528** alone vs Crude plant powder + synthetic chemicals Treated vs Control 0.274** 0.833** 1.265** 1.011** *, ** Significant at P = 0.05 and P = 0.01, respectively

-7.118**

0.986**

472

The orthogonal contrasts also indicated that seed treatment effect was positive on germinability while the effect of natural crude powders was lower than that of synthetic chemicals. In addition, the performance of crude plant powder used alone was smaller than that of the crude powder combined with synthetic chemicals (Table 6). All treatments, including the control, exhibited good germination percentage (>90%) at the 120-day storage period (Table 7). However, significant differences were detected among tratments (Table 7). Seed dressing with Marshall 35 ST, Benlate + Marshall 35 ST, Marshall 35 ST + T. vulgaris, and Marshall 35 ST + O. gratissimum consistently produced higher germination percentage than other treatments for all storage periods. At the 240-day storage period, the germination percentages for these treatments were 87, 92.5, 86.3, and 84.8, respectively (Table 7). The germination percentage for the control at the 240-day storage period was 58.3. Although germination percentage decreased as storage period increased for all treatments (Fig. 2), the rate of decrease was not as drastic as that for seed harvested from the humid forest zone (see Fig. 1). Apart from the 240-day storage period, the treatments had no effect on the percentage of abnormal seedlings (Table 8). On the average, however, there was a slight increase in the percentage of abnormal seedlings as the length of the storage period increased. Table 7. Effects of seed treatment and length of the storage period on the germination percentage of maize seed harvested from the savana zone of Cameroon in 2000.

Treatments Ridomil Hydrox Marshall 25 ST Benlate Oncol Benlate + Marshall O. gratissimum T. vulgaris A. indica C. citratus Marshall + T. vulgaris Marshall + O. gratissimum Benlate + T. vulgaris Benlate + O. gratissimum Control Mean C.V. (%) LSD (0.05 %)

120 98.0 93.5 99.0 98.0 95.5 97.5 92.0 91.0 95.0 91.5 95.5 98.5 90.5 95.8 91.0 94.8 1.8 3.7

Storage period 150 180 81.5 73.3 90.0 75.5 91.0 87.0 87.7 69.5 88.0 77.8 92.3 92.0 81.3 77.5 81.5 69.8 78.3 74.5 63.0 64.3 91.3 88.3 88.8 86.8 61.8 53.3 89.0 79.8 70.8 56.8 82.4 74.4 4.4 4.7 7.7 7.5

Ns: not significant at 0.05 level of probability. * dosage for all treatments is 30g for 100 kg of seed.

(days) 210 72.8 74.5 86.5 75.8 72.0 91.0 67.0 54.5 69.8 51.5 85.8 81.8 45.5 68.3 56.0 70.1 5.4 8.1

240 73.5 77.3 87.0 65.8 72.8 92.5 67.5 64.0 73.3 56.0 86.3 84.8 53.5 73.0 58.3 72.0 6.5 10.0

473

120 100

Marshall 35 ST

Germination %

Benlate + Marshall

80 Marshall + T.vulgaris Marshall + O. gratissimum Control

60 40

A. indica

20 0 120

150

180

210

240

Storage period (days)

Fig. 2. Effect of seed dressing and storage period on the germination percentage of maize seeds harvested from the savanna zone of Cameroon. Table 8. Effects of seed treatment and length of the storage period on the percentage of abnormal seedlings of mai ze seed harvested from the savana zone of Cameroon in 2000. Storage period Treatments 120 150 180 Ridomil 1.0 5.3 7.0 Hydrox 4.0 3.0 7.8 Marshall 25 ST 0.5 2.8 3.0 Benlate 1.5 3.5 3.5 Oncol 1.0 5.3 4.8 Benlate + Marshall 1.0 3.5 5.3 O. gratissimum 2.5 4.5 4.3 T. vulgaris 2.0 4.3 3.3 A. indica 1.5 9.3 6.0 C. citratus 2.0 3.0 4.7 Marshall + T. vulgaris 3.5 3.5 3.3 Marshall + O. gratissimum 1.5 3.3 1.5 Benlate + T. vulgaris 4.0 2.8 4.8 Benlate + O. gratissimum 1.5 2.8 3.3 Control 2.5 5.0 6.8 Mean 2.0 4.1 4.8 C.V. (%) 84.3 48.2 37.5 LSD (0.05 %) ns ns ns ns, not significant at the 0.05 probability level. *, dosage for all treatments is 30 for 100 kg of seed

(days) 210 7.8 6.5 4.5 6.8 7.0 2.5 2.5 3.8 4.5 7.5 5.8 3.5 4.8 7.5 6.0 5.2 35.8 ns

240 5.5 6.0 3.5 5.8 6.5 1.8 7.5 7.3 6.0 6.0 2.5 3.0 5.5 3.3 6.5 4.8 36.0 3.7

Seed dressing with Marshall 35 ST, Benlate + Marshall 35 ST, Marshall 35 ST + T. vulgaris, and Marshall 35 ST + O. gratissimum did not have any negative effect on seedling height and radicle length (Tables 9 and 10). In general, each natural

474

crude powder applied alone was not very effective for seed dressing. A. indica powder with 73.3 germination percentage after 240-day storage period, was the most effective natural crude powder. Table 9. Effects of seed treatment and length of the storage period on the seedling height (cm) of maize seed harvested from the savana zone of Cameroon in 2000. Storage period (days) Treatments 120 150 180 210 240 Ridomil 3.5 4.2 5.1 5.7 6.6 Hydrox 4.4 4.3 5.7 5.9 7.1 Marshall 25 ST 4.0 5.9 5.6 6.1 8.1 Benlate 3.6 7.0 7.0 9.5 8.0 Oncol 3.3 6.9 4.1 8.7 8.1 Benlate + Marshall 5.3 6.8 6.3 6.2 9.2 O. gratissimum 4.1 4.9 6.7 7.6 8.4 T. vulgaris 5.2 5.6 5.3 5.2 9.3 A. indica 5.4 7.0 5.6 7.4 8.3 C. citratus 3.2 4.7 5.8 6.6 9.5 Marshall + T. vulgaris 3.1 6.4 4.3 6.9 9.2 Marshall + O. gratissimum 4.2 5.4 6.1 6.1 8.2 Ben late + T. vulgaris 5.1 4.0 5.2 6.5 8.7 Benlate + O. gratissimum 3.8 6.7 4.6 7.6 8.7 Control 4.6 8.0 6.9 4.9 9.4 Mean 4.2 5.8 5.6 6.7 8.4 C.V. (%) 15.7 12.1 23.0 14.6 9.6 LSD (0.05 %) 1.4 1.5 ns 2.1 ns ns, not significant at the 0.05 probability level. *, dosage for all treatments is 30 g for 100 kg of seed.

Table 10. Effects of seed treatment and length of the storage period on the radicle length of the seedlings of maize seed harvested from the savana zone of Cameroon in 2000. Treatments Ridomil Hydrox Marshall 25 ST Benlate Oncol Benlate + Marshall O. gratissimum T. vulgaris A. indica C. citratus Marshall + T. vulgaris Marshall + O. gratissimum Benlate + T. vulgaris Benlate + O. gratissimum Control Mean C.V. (%) LSD (0.05 %)

120 4.6 6.2 10.6 7.8 9.7 10.2 8.9 9.7 9.0 8.3 9.0 10.2 8.6 5.8 10.5 8.9 11.3 2.2

Storage period (days) 150 180 210 3.0 5.6 4.6 2.6 4.5 5.8 10.2 9.0 9.7 11.1 9.6 10.4 11.5 9.6 11.5 8.6 10.9 9.6 9.3 10.8 7.6 8.7 9.0 8.0 10.0 9.8 9.4 6.7 8.0 7.6 9.2 9.4 10.9 9.1 10.7 9.8 5.2 8.5 8.5 10.3 9.9 9.4 10.2 11.4 5.9 8.8 9.1 8.7 36.3 16.1 15.8 6.9 3.1 2.9

ns, not significant at the 0.05 probability level. *, dosage for all treatments is 30 for 100 kg of seed

240 5.0 6.4 12.4 11.2 12.3 11.8 10.0 12.3 10.9 9.3 10.9 10.1 10.8 11.8 10.7 10.4 13.8 3.0

475

Conclusion After 240 days of storage, the average germination percentage of maize seed from the savanna zone (72 %) was significantly higher than that of the seed from the humid forest zone (44.3 %). Therefore, the savanna zone is a better ecology for maize seed production in Cameroon. The results of this study indicated seed dressing as one way to improve the germination rate of maize seed in both the humid forest and savanna zones of Cameroon. Differential germination rates were associated with the type of seed dressing, the dosage of the chemical or natural product used, and the seed storage period. The optimum duration for maize seed storage without significant loss of viability is 180 days after harvest. If the seed is to be planted within 180 days from harvest, any of the seed dressings investigated in this study would be effective. If the targeted sowing period starts 240 days or more after harvest, the seed dressings that would produce maximum germination rates are Marshall 35 ST, Benlate + Marshall 35 ST, Marshall 35 ST + T. vulgaris, Marshall + O. gratissimum. The use of 30 g of Marshall 35 ST per 100 kg of seed, or the substitution of a fraction of Marshall 35 ST by a crude natural powder will considerably reduce the cost of seed dressing without reducing the maximum germination percentage. Economic analyses of the different types of seed dressing studied here are necessary before definite recommendations could be made to the farmers.

References Adegoke, G.O., and B.A. Odesola. 1995. Storage of maize and cowpea and inhibition of microbial agents of biodeterioration using the powder and essential oil of lemon grass (Cymbopogom citratus). International Biodeterioration and Biodegradation: 81–84. Awuah, R.T. 1989. Fungitoxic effects of extracts from some West African plants. Ann. Appl. Biol. 115: 451–453. Bhaskara Reddy, M.V., P. Angers. A. Gosselin, and J. Arul, 1998. Phytochemistry. 47:1515–1520. FAO. 1998. Bulletin trimestriel de statistiques Vol.11 No 3/4. Gwinner, J., R. Harnisch, et O. Muck. 1996. Manuel sur la manutention et conservation des grains après récolte.GTZ (Deutsche Gesellschaft fur Technische Zusammenarbeit).368 p. Maude, R. B. 1993. Seed borne diseases and their control: Principles and practice. Horticultural Research International. Wellesbourne. 280 p.

476

Moreno-Martinez, E., and G. Vidal-Gaona. 1981. Preserving the viability of stored maize seed with fungicides. Plant Disease 65(3) : 260–261 NCRE. 1994. Annual Report. Okoruwa, A.E., 1995. Enhancing maize processing and utilization in West and Central Africa. Pages 108–119 in B. Badu -Apraku, M.O. Akoroda, M. Ouedraogo, and F.M. Quin (eds.) Contributing to food self-sufficiency: Maize research and development in West and Central Africa. Proceedings of a Regional Maize Workshop, 29 May–2 June 1995, IITA, Cotonou, Benin Republic. Samgan, L., and H.K. Saksena. 1987. Pathogenic potential of fungi associated with seeds and seedlings of maize. Seed Research 5(1) : 52–60. Toussi, A., B. Ngniado, et C. Zonkeng. 1999. Etude diagnostic de la filière maïs au Cameroon. Yehouenou, A. 1995. Réponse du maïs au traitement de semences dans le nord Borgou. In B. Badu-Apraku, M.O. Akoroda, M. Ouedraogo, and F.M. Quin (eds.) Contributing to food self-sufficiency: Maize research and development in West and Central Africa. Proceedings of a Regional Maize Workshop, 29 May–2 June 1995, IITA, Cotonou, Benin Republic.

477

Diversification de l’utilisation du maïs dans l’alimentation humaine au Tchad P. Kagne, F. Namba, D. Nadjiam, et K. Mbayhoudel ITRAD, N’Djaména Résumé Le Tchad, comme les autres pays du Sahel, dépend beaucoup des céréales pour son alimentation. Il s’agit essentiellement du mil et du sorgho, le maïs étant marginal. Cependant, on constate une progression dans la production du maïs ces dernières années, mais sa consommation est peu diversifiée faute de technologie de transformation appropriée. C’est pour accroître davantage la production qu’on a utilisé trois variétés de maïs pour élaborer quelques produits qui d’habitude se font à base de blé ou d’autres céréales : il s’agit de spaghetti locale ou « douédé », de gâteau moundang «soumbis », de galette « kissar », de beignet, des biscuits et des farines infantiles. Ces produits sont analysés et testés pour mieux appréhender le comportement technologique du maïs et son acceptation par les consommateurs. Les résultats ont montré qu’il n’y a pas de problèmes majeures pour utiliser le maïs. Abstract In order to feed her population, Chad, like most Sahelian countries heavily relies on cereal crops, mainly millet and sorghum, maize being a marginal crop. However, maize cultivation has been gaining ground in recent years but consumption is less diversified because of lack of appropriate processing technology. For higher increase in production, three maize varieties were used to develop a few products that are traditionally derived from wheat or other cereal crops : local spaghetti or « douédé », moundang cake « soumbis », pancake « kissar », doughnut, biscuits and baby food. These products were analysed and tested in order to have a better grasp of the technological performance of maize as well as of its acceptance by the consumers. Results have shown that there are no major constraints to maize utilization.

Introduction Au Tchad, le maïs qui était au départ une culture de case est en train de devenir une culture de champ. De ce fait, il occupe une place non négligeable parmi les céréales cultivées. Sa production de la campagne agricole de 1999–2000 est estimée à 94. 161 tonnes pour 134 .678 ha emblavés en zone soudanienne et en zone sahélienne. Ceci le place au 3e rang après le sorgho et le pénicillaire.

478

En raison des conditions de transformation primaire difficiles (décorticage et mouture), sa consommation était surtout limitée aux épis frais, au stade laiteux, sous forme cuite au grillée. Cependant, depuis l’introduction des décortiqueuses et des moulins, la production augmente et la consommation aussi. Le maïs est alors utilisé comme les autres céréales après transformation en farine. Contrairement aux autres céréales, la consommation du maïs est limitée aux produits traditionnels tels que la bouillie, la pâte ou quelquefois les boissons alcoolisées ou non. Les raisons généralement évoquées concernent l’inaptitude de cette céréale à d’autres utilisations. Le travail mené actuellement dans le cadre du projet WECAMAN vise à accroître la demande en maïs au Tchad, en montrant la diversification de son utilisation, et par voie de conséquence favoriser l’augmentation de la production. Apportée de la valeur ajoutée au maïs par la recherche de nouvelles formes d’utilisation et l’amélioration des techniques traditionnelles de transformation est l’objectif visé pour le présent projet, ce qui pourrait à terme: -

Accroître le revenu des artisans vendeurs ; Améliorer l’alimentation des populations ; Garantir un meilleur prix aux producteurs

Pour ce faire, plusieurs variétés de maïs ont été testées pour réaliser les produits qui d’ordinaire se font à base d’autres céréales comme le blé par exemple. Le présent rapport décrit les différents tests réalisés tant au laboratoire qu’en milieu réel avec les groupements notamment féminins.

Matériels et Méthodes Matériels Trois variétés de maïs ont été utilisées dont une variété locale : « Kouri » et deux variétés améliorées : « CMS 8602 et IB84A 202 ». Ces variétés ont été analysées pour connaître leur caractéristiques physico-chimiques et leur rendement en farine. Les analyses concernant les caractéristiques physico-chimiques ont été faites au laboratoire avec les matériels habituels comme l’étuve ou les balances. La vitrosité, la couleur et la forme ont été caractérisées par observation à l’œil nu. Quant au rendement en farine, il a été calculé à partir de la farine obtenue par les moulins du quartier : quatre sortes de farines ont été obtenues : farine crue, farine fermentée, farine de maïs germé et farine cuite fermentée.

479

La mise au point des différents produits a été réalisée avec les ustensiles généralement utilisés en cuisine.

Méthodes Teneur en eau : environ 3g d’échantillon de grains de maïs ont été mis à l’étuve à 110°C pendant au moins six heures puis retiré pour peser avant de remettre à l’étuve jusqu’à l’obtention d’un poids constant. La différence entre le poids de départ et le poids final donne la teneur en eau. La vitrosité est déterminée en réalisant une coupe transversale des grains pour estimer de visu le pourcentage d’albumen vitreux par rapport à l’albumen farineux. La couleur et la forme ont été également jugées visuellement. Pour l’obtention des farines, il a été fait appel aux détenteurs des moulins à marteaux et des décortiqueuses comme prestataires de service. Le rendement en farine a été calculé à partir de poids de départ des grains crus germés ou fermentés. Les différentes farines ont été obtenues de la manière suivante : - Pour les farines crues, le maïs après décorticage est vanné, lavé et trempé pendant trois heures. Il a ensuite été égoutté et séché au soleil pendant environ deux heures (cela dépend de l’intensité du soleil) avant la mouture. - Pour la farine fermentée, les grains après décorticage, vannage et lavage ont été trempés dans l’eau pendant vingt quatre heures; ils ont ensuite été lavés, égouttés et séché s pendant trois heures avant mouture. - Pour la farine germée, les grains lavés ont été trempés dans l’eau pendant vingt quatre heures puis égouttés avant d’être étalés en couche de 5cm environ entre deux plastiques pour germination pendant trois jours ; pendant ce temps, les grains sont humidifiés une fois par jour. Les grains germés ont été séchés au soleil avant la mouture. - Pour la farine cuite et fermentée, les grains lavés sont cuits jusqu’à explosion du péricarpe. Ils ont ensuite été refroidis avant d’être saupoudrés de 10% en poids de farine de granulométrie grossière de maïs germé. Après quoi on laisse au repos pendant une nuit pour fermentation avant de procéder au séchage et à la mouture. La farine crue sert à faire des produits comme le « kissar », une sorte de crêpe tchadienne, le « douédé » sorte de spaguetti , le beignet et le biscuit, tandis que la farine germée entre dans la fabrication de « Soumbian », mets spécial fait généralement à base de sorgho, de farine infantile et du kissar. La farine cuite et fermentée rentre dans la composition du « soubiam » également.

480

Mise au point des produits finis: Le « douédé ». Produit fabriqué essentiellement à base de farine de blé à l’aide d’une machine à façonner les spaghetti, le douédé se vend un peu partout au Tchad et il se vend bien depuis la dévaluation car un paquet de spaghetti de 250g coûte 500 Fcfa alors que le paquet de douédé pesant environ 100g coûte 50 Fcfa. Compte tenu de ce qui précède, l’objectif est de substituer partiellement la farine de maïs au blé pour la production de douédé. Ainsi la farine de maïs cru est incorporée à la farine de blé à La figure 1 schématise les étapes d’obtention de ces farines

Maïs grains Nettoyage - vannage Grains sains Lavage - égouttage

Etalage

Aspersion

Germinati

Lavage

Décorticage

Vannage - lavage

Trempage - Egouttage

Etalage

Mouture - tamisage

Farine de maïs germé

Refroidissemen t Pilage

Séchage

Mouture - tamisage

Cuisson

Fermentat

Mouture -Tamisage Farine maïs cru

de

farine de maïs cru et fermenté

481

raison de 30%. La farine composée est hydratée à l’eau bouillante à raison de 55% (V/P) puis pétrie en une pâte consistante qui elle, est découpée en morceaux de 20 g environ. Chaque morceau de pâte est aplati à la main et laminé successiveme nt à la machine à nouille dont l’écart entre les 2 cylindres est graduellement réduit à chaque fois. A la fin du laminage, la pâte est immédiatement tréfilée puis séchée à l’ombre pour éviter le craquage Des tests de cuisson ont ete realises dans le but d’apprecier la capacite d’absorption d’eau, la tenue a la cuisson, l’aspect et la couleur du « douede » cuit. ces tests ont ete faits sur les trois varietes de maïs en comparaison au « douede » pur ble. l’essai consiste a faire cuire pendant 7 mn, 5 g de chaque echantillon dans 1000 ml d’eau bouillante. les produits sont ensuite peses apres egouttage et refroidissement. Le « Soumbiam » Le « Soumbiam » ou « soumbis » est un mets très prisé dans l’ethnie moundang (ethnie peuplant une partie de la région Sud-Ouest du Tchad et du Nord-Est du Cameroun). Il est obtenu par incorporation de la pâte d’arachide à la farine de sorgho cuite et fermentée, avec ajout du sucre ou miel et quelque fois du piment. De grande valeur nutritive et énergétique, le produit est un véritable coupe–faim consommé entre les principaux repas et peut se conserver pendant plus de 2 mois. Plus souvent, il est d’abord délayé dans l’eau avant d’être consommé. C’est dans ce contexte que nous avons essayé les potentialités technologiques de maïs dans l’élaboration du soumbi et de contourner la phase de délayage de celui –ci en mettant au point le soumbiam. Pour ce faire, la farine de maïs cuite et fermentée, la pâte d’arachide et le sucre ont été mélangés dans les proportions respectives de 30 : 50 : 20 . Cet ensemble a été pilé pour obtenir une pâte onctueuse qui est le ‘‘soumbi’’. Ce dernier est moulé et cuit au four à 250°C pendant 25 à 30 mn. La figure 3 décrit le procédé. Le test consiste à comparer le soumbiam à base des 3 variétés de maïs par rapport au sombi original à base de sorgho. (goût, couleur, aspect). Le « kissar » C’est une crêpe locale préparée généralement à base de maïs, riz ou mil fermenté. Le « kissar » est surtout consommé dans le milieu musulman Tchadien. Sa durée de conservation est au plus 48 heures, et les invendus sont séchés pour être consommés dans l’eau ou du lait. Les variétés de maïs ont été testées pour faire du kissar et le comparer au kissar traditionnel au point de vue acceptabilité par les consommateurs.

482

Une première pâte consistante est obtenue en pétrissant dans 28% du total d’eau à utiliser, une farine composée de maïs cru et du blé dans les rapports respectifs de 40 et 15% (par rapport au total de la farine à utiliser). Une pâte onctueuse obtenue en diluant 9% de farine crue dans 12% d’eau est ajoutée sous agitation durant 5 mn à l’eau bouillante, estimée elle à 40% pour former une bouillie visqueuse . 27% de farine de maïs cru sont ajoutés à cette bouillie qui est laissée cuite pendant 5 mn avant d’être malaxée en une deuxième pâte consistante. Cette pâte est aussitôt transférée sur 9% de farine de maïs germée, malaxée jusqu’à totale liquéfaction puis, laissée refroidir à 50°C environ. La première pâte apprêtée est alors incorporée à la pâte liquéfiée par malaxage et, cet ensemble formé est laissé au repos pendant une nuit pour fermentation. Au lendemain on procède à la dilution de la pâte fermentée avec de l’eau tiède ajoutée à concurrence de 20% pour obtenir la pâte à kissar. Une louche de cette dernière est transférée à la surface d’une plaque surchauffée et graissée au préalable par la moelle épinière de bovin. La pâte est alors uniformément étalée du centre vers l’extérieur de la plaque et laissée cuite pendant 3 mn environ. Le kissar ainsi préparé est enfin retiré et conservé chaud sous un film de plastique. Les paramètres, couleur, goût, souplesse, élasticité et poids sont testés. La figure 4 montre le procédé. Le beignet. Au Tchad, en milieu urbain comme rural, le beignet fait à base de farine de blé est rentré dans les habitudes alimentaires. La farine de blé n’étant pas toujours disponible en milieu rural surtout, pourquoi ne pas utiliser les céréales locales, en l’occurrence le maïs ? L’essai consiste à mettre au point des beignets enrichis. Ainsi on ajoute dans 1 verre de farine de maïs un verre de farine de blé, ½ verre de sucre de 1,5 cuillères à soupe de margarine, 1 œuf, 1 cuillérée de levure chimique, une pincée de sel et un sachet de sucre vanillé. L’ensemble des composants est mélangé avant d’ajouter graduellement de l’eau, tout en pétrissant pour obtenir une pâte consistante. La pâte à beignet obtenue est frite dans l’huile. Le biscuit. Il existe à N’Djaména une petite unité de production de biscuits dénommée « Alfirdos ». qui produit des biscuits à base de farine de blé importé. L’objectif visé par le choix de ce thème, est de mettre au point un procédé simple de production de biscuits de qualité acceptable et ce, à partir de farine composée de maïs, et du blé et des ingrédients localement disponibles. A peine une mesure de farine de maïs cru (15% ) est mélangée à 2 mesures de farine de blé, à peine 1 mesure de sucre, 1 mesure de margarine et à des quantités marginales des additifs suivants : sel, sucre vanillé, levure chimique et eau. Les composants sont en suite pétris suffisamment et la pâte obtenue, découpée et cuite au four porté à 250°C pendant 25 à 30 min.

483

Farine infantile. L’objectif visé ici est de trouvé un procédé moins onéreux de production de farine infantile permettant d’accroître à la fois sa densité énergétique et son contenu nutritionnel. Ainsi, la farine de maïs fermenté, la farine de soja cuit, le sucre et la farine de maïs germé sont mélangés dans les proportions respectives de 60 ; 20 ; 12 et 8%.

Résultats et Discussion Caractéristiques physico–chimiques et technologiques Tableau

1. Caractéristiques physico–chimiques

utilisées.

Variétés IB 84 A 202 KOURI CMS 8602

Humidité % 11,18 10,26 9,69

Rende -ment en farine % 49,50 48,70 44,80

Vitrosité % 75 80 85

des

Couleur Blanche Jaune Jaune foncé

variétés

Pureté 97 98 98

de

maïs

Poids de 1000 grains (g) 206,24 164,08 208,07

Il ressort de ce tableau que les trois variétés sont vitreuses, ce qui explique le faible rendement e n farine obtenue à chaque fois. Un trempage de 3 heures des grains décortiqués n’a pas pu améliorer de façon sensible le taux d’extraction de farine. Cependant, un trempage de 24 heures des grains décortiqués dans l’eau a permis d’extraire 51,02% de farine . Notons que ces rendements sont exprimés par rapport au poids de l’échantillon prélevé directement du sac.

Le « douédé » Le comportement des pates hydratees, obtenues a

partir de 30% de la farine de chaque variete de maïs et de 70% de la farine de ble, s’avere bon au cours des operations de petrissage, laminage et trefilage, les pertes etant du meme ordre de grandeur pour les 3 varietes utilisees (ib 84 a202 : 13,13%; kouri : 12,96%; cms 8602 : 11,09%) , compares a celle du temoin 10,01%). Le comportement de douede apres sechage, s’est revele egalement acceptable (Tableau 2). Tableau 2. Paramètres Technologiques des « douédés » mixtes par rapport au « douédé » à base de blé pur. Types de ‘‘douédé” IB 84 A 202 CMS 8602 KOURI BLE PUR *PF = produit

Prise d’essai 300 300 300 300 final

Volume d’eau ( ml ) 165 165 165 165

Poids moyenne de P F* 260,61 265,94 261,10 273,92

Perte moyenne ( %) 13,13 12,09 12,93 10,01

Couleur des PF* Blanche Jaune Beigne Brunâtre

484

Les tests de cuisson, réalisées avec les ‘‘douédés’’ mixtes en comparaison au ‘‘douédé’’ pur, ont permis de déceler quelques différences d’une variété de maïs à une autre en ce qui concerne la couleur et le poids des produits cuits. Tandis que les couleurs des ‘‘douédés’’ de Kouri et de CMS 8602 se rapprochent plus ou moins de celle du douédé pur blé (beige), la couleur des autres à base de IB 84 A 202, est blanche et par conséquent peu acceptée par le consommateur. Par ailleurs, en faisant les rapports de taux d’absorption d’eau de cuisson, on observe que la quantité d’eau absorbée par Kouri et CMS 8602 se rapproche de celle absorbée par le témoin douédé pur (2,9 et 3,1 respectivement contre 3,0). Par contre le douédé de IB 84 A202 absorbe beaucoup d’eau (3,5). Sa faible vitrosité pourrait expliquer ce constat (Tableau 3). Le comportement du douédé mixte à la cuisson, comparé à celle du témoin, est bon pour les 3 variétés étudiées, il y a moins de perte de matière sèche dans l’eau de cuisson. Ces constats faits par quelques consommateurs et productrices de douede et des collegues chercheurs, permettent de dire que : le Kouri et CMS 9602, sont plus aptes pour la production des ‘‘douedes mixtes”. Tableau 3. Résultats des tests de cuisson des ‘‘douédés mixtes” comparés au témoin. Variétés

PI* douédé ( g)

Kouri IB 84 A202 CMS 8602 Blé

100 100 100 100

Vol eau de cuisson (ml ) 100 100 100 100

PF(**) douédé (g)

PF/PI

Couleur après cuisson

Aspect eau de cuisson

Comporte -ment

288,04 345,01 307,99 298,99

2,88 3,45 3,08 3,0

Beige Blanche Beige brunâtre

Trouble Trouble Trouble Trouble

bon bon bon bon

(*)PI : Poids Initial (* ) PF : Poids final ( après cuisson ). Le « Soumbiam » Les tests portant sur la mise au point de soubi à partir duquel s’obtient le soumbi n’ont pas montré de variantes significations entre les 3 variétés de maïs utilisées en ce qui concerne le goût et l’aspect du dit produit délayé dans l’eau pour les proportions retenues : 30% de farine de maïs cuit et fermenté, 50% de pâte d’arachide et 20% de sucre. Le « soumbi » est délayé dans l’eau pour être consommé sous forme de boisson désaltérante La couleur de cette boisson est fonction de la couleur de la variété utilisée ; c’est-à-dire blanche pour IB 84 A 202, jaunâtre pour le Kouri et jaune pour le CMS 8602.

485

Néanmoi ns, il a été observé que les couleurs de « soumbiam » de CMS 8602 et de Kouri se rapprochent parfaitement et sont d’un brun – doré tandis que celle de ‘‘ soumbiam’’ de IB 84 A202 est brunâtre (Tableau 4). Le problème majeur rencontré reste la fiabilité des ‘‘ soumbiam” dans l’ensemble car, souvent au démoulage, il arrive que le ¼ des produits finis se cassent ou s’effritent. Ce constat reste aussi vrai pour les soumbiam de sorgho. Nous retenons que les trois variétés de maïs sont très bien indiquées dans la production de ‘‘soumbi’’ et peuvent faire l’objet de diffusion en milieu réel, à l’image des ‘‘soumbi” assez bien appréciés lors de nos formations. Par contre, les ‘‘soumbiam” dans leur ensemble, nécessitent quelques travaux de recherche dans l’optique d’améliorer leur stabilité notamment. Tableau 4. Paramètres observés sur les ‘‘ soumbiam” de maïs par rapport à ceux de sorgho. Variétés

Composition des ‘‘ soumbiam’’

Couleur

Goût

IB 84 A 202

30: 50:20

Brunâtre

Bon +

Perte en moyenne en moyenne (%) 23

Kouri 30: 50:20 Brun - doré Bon + CMS 8602 30: 50:20 Brun - doré Bon + Berberé Rouge 30: 50:20 Brun – doré + Bon ( *) - Farine de maïs cuit et fermenté : 30% - Pâte d’arachide : 50% -Sucre : 20%

27 25 22

aspect

attrayant attrayant attrayant attrayant

Le "kissar” Les essais de production de kissar avec les 3 variétés de maïs ont pour but de déceler les différences technologiques éventuelles entre les produits finis. Les paramètres couleur, nombre, poids et tenue des kissar frais sont ainsi comparés entre eux ( Tableau 5) Tableau 5. Paramètres observés sur les kissars de CMS 8602, Kouri et IB 84 A202. Variétés

P.T. Nb. des Des PF composants ( ** ) IB 84 A202 39 27 g 14 Kouri 39 27 g 14 CMS 8602 40 77 g 18 ( * ) Poids total des composants (** ) Produits finis .

P.T des PF

PU des PF

1.600 g 114 1.686 g 116 2.141 g 119 utilisés.

Couleur

Tenue

goût

Blanche Brunâtre Jaune

Bonne+ Bonne+ Bonne+

Acceptable ++ Acceptable ++ Acceptable ++

Il ressort du tableau V que le goût et la tenue des ‘‘ kissar ‘‘ frais produits ne présentent aucune enjeu majeur quant à l’emploi de toutes les 3 variétés de maïs dans le procédé retenu. Leur texture,

486

assez souple et élastique est bien celle recherchée par les producteurs et consommateurs de kissar et, également, leur saveur légèrement sucrée et acidulée. Cependant, la couleur jaune des kissars CMS, 8602 s’est plus affirmée dans le choix des producteurs et dégustateurs de kissars par rapport au reste. La couleur blanche des kissars de IB 84 A 202 la rapproche de celle du riz et lui accorde ainsi le second rang dans le choix. S’agissant enfin des poids des produits finis par rapport à ceux des différents composants de base, le rapport et la faveur des kissars de CMS 8602 est de 0,53 suivi de ceux de Kouri avec 0,42 et de IB 84 A202 avec 0,41. Les différences observées dans la vitrosité des grains utilisés pourraient expliquer les écarts constatés entre les rendements en ‘‘ kissars ‘‘ Pour la suite de nos travaux sur le kissar, nous proposons les variétés IB 84 A 202, pour sa couleur blanche uniquement et CMS 8602, pour à la fois sa couleur jaune et son rendement élevé en produits finis. Les beignets. La conduite des opérations unitaires avec les 3 variétés de maïs dans la production des beignets s’est révélée similaire. En effet, la tenue des pâtes hydratées, comparées entre elles au cours du malaxage et du laminage, ne montre aucune différence perceptible. La couleur des beignets, dans l’ensemble, est d’un brun– doré et ne s’apparente guère à celle des matières premières utilisées, ce qui pourrait s’expliquer pour la réaction de Maillard. Cependant, le poids des produits finis, exprimé par rapport au poids de l’ensemble des composants utilisés, représente en moyenne 106,34%. L’absorption de l’huile au cours de la friture des pâtes pourrait bien expliquer cette augmentation de poids. Un test sensoriel sommaire, réalisé avec le personnel du laboratoire d’Analyse des Sols, Eaux et Plants (LASEP) permet d’attester que le goût, la couleur, l’odeur et la texture des trois types d’échantillons de beignets présentés, sont assez bons et par conséquent, l’ensemble des produits peut bien être accepté en milieu réel. Cette affirmation a été également dite par les consommateurs. Les 3 variétés de maïs testées, s’accommodent bien de la mise au point des beignets enrichis et peuvent valablement être vulgarisées. Le biscuit. En se conformant au procédé technologique de biscuit mis au point pendant la campagne précédente avec la variété Kouri, il nous a été possible de produire sans difficultés les

487

biscuits de CMS 8602 et de IB 84 A 202. Ces 2 dernières variétés n’ont pas également montré des particularités négatives au cours des opérations unitaires de production en ce qui concerne la tenue de leur pâte au laminage, au pétrissage et à la cuisson. Les produits finis élaborés à partir des 3 variétés de maïs n’étaient point distinctifs dans leur couleur, goût et texture. ils ont reçu dans l’ensemble une appréciation favorable. Il en résulte que ces 3 variétés sont équitablement aptes à la production de biscuit selon la formulation ci -après : - Farine de maïs cru : - Farine de blé : - Sucre : - Margarine : - autres additifs mineurs : e au.

15% 35% 25% 25% sucre vanillé, sel, levure chimique et

Farine infantile. Les tests comparatifs portant sur les bouillies infantiles préparées à base des 3 variétés de maïs n’ont pas, eux non plus montré des spécificités en dehors de leur couleur qui s’apparente à celle de la matière première utilisée. La vitrosité apparente des 3 types de bouillies exténuées visuellement, s’avère bonne en ce référant aux bouillies infantiles habituellement préparées par les mères allaitantes. La formulation retenue est donc : - Farine de maïs fermenté : - Farine de sorgho : - Farine de maïs germé - Sucre

60% 20% 8% 12%

Partout la surprise et la satisfaction de participants de constater les potentialités technologiques du maïs dans l’élaboration des recettes variétés et hautement compétitives, ont été grandes. Ils ont émis , par contre, le souhait de posséder les petits matériels de transformation comme la presse à souder, la machine à douédé, etc.

Conclusion Les travaux de recherche menés pendant la campagne 2000– 2001 permettent d’affirmer que les variétés Kouri, CMS 8602 et IB84 A 202 sont vitreuses ( 80 % en moyenne ) et donnent un faible rendement en farine ( 47,66 en moyenne). Cependant leur farine présente une grande aptitude technologique pour l’élaboration de la plupart des produits.

488

Par comparaison entre elles, la variété CMS 8602 principalement et la variété Kouri, dans une faible mesure nous semblent indiquées pour la préparation de ‘‘douédé”. Pour la production des ‘‘kissar’’ nous proposons IB 84 A202 et CMS 8602. Le volet sur le transfert de nos acquis sur la transformation de Kouri en milieu réel a été riche d’enseignements à plus d’un titre eu égard au nombre des participants formés (101) et à la bonne collaboration avec les ONG qui ont pour mission d’assurer le suivi de l’après formation.

Bibliographie Enquête Démographique et de Santé (EDST), Tchad. 1997. Nutrition des jeunes enfants et de leurs mères au Tchad. MC Dop, D. Benbouzid, S. Trêche, B. de Benoist, A. Verster, et F. Delpeuch. 1999. Complementary feeding of young children in Africa and the Middle East. IRD, WHO. Minguéyambaye, Naïban, 2000. Mémoire de Maîtrise Université de Bangui: Contribution à la valorisation des céréales au Tchad : cas de farine de sevrage. Mbayhoudel, Koumaro. 1999. Recensement sur les unités de transformation artisanales à Moundou, Pala et Sarh.

489

Effect of dehumidification and storage conditions on the longevity of maize seed E.A. Asiedu1, A.G.J. van Gastel1, P.Y.K. Sallah2, and P. Adusei-Akowuah1 1

West African Seed Development Unit (WASDU) Crops Research Institute. PO Box 3785, Kumasi, Ghana. 2 Crops Research Institute, PO Box 3785, Kumasi, Ghana. Abstract High cost of production, drying, processing and storage are the causes of high prices of certified seed purchased by farmers. In Ghana, seed growers dry their seeds to 11.0-12.0% moisture content and store either in the coldroom or in a ventilated warehouse. In an attempt to investigate means of reducing storage costs without compromising physiological quality of seed, a quality protein maize (QPM) variety, Obatanpa was harvested, shelled, cleaned and dried to two moisture contents, 8.0 and 11.0% representing dehumidified and conventional moisture levels, respectively. The two sets of seed were hermetically sealed in polythene bags and stored under two conditions, the cold room (10?C/80% r.h.) and the warehouse (21 31?C/68-86% r.h.) for 18 months. Results showed that high germination percentages (above 95%) were maintained throughout the 18-month period of storage in seeds dried to 8.0% moisture content and kept in either the cold room or the warehouse. The percentage germination of the seed lots was comparable to those dried to 11.0% and stored in the cold room. Seeds dried to 11.0% moisture content and stored under the warehouse conditions, however, showed low germination percentage below the acceptable level of 85% after 9 months in storage. Economic and financial analyses showed high profitability in opting from cold storage to ambient storage of dehumidified seed. Thus, it was technically feasible and financially cost effective to dehumidify maize seed to 8% moisture content and store in a warehouse for periods up to 18 months without losing viability and vigor. Résumé Le coût de production élevé, le séchage, la transformation et le stockage sont la cause des prix élevés des semences certifiées achetées par les producteurs. Au Ghana, les producteurs semenciers sèchent leur semence à 11,0-12% de taux d’humidité et les stockent soit dans une chambre froide ou à la température ambiante sous ventilation. Dans une tentative d’investigation des moyens de réduction des coûts de stockage, sans compromettre la qualité physiologique des semences, la variété de maïs hyperprotéïque, Obatanpa a été

490

récoltée, égrainée, nettoyée et séchée à deux taux d’humidité, 8 et 11% representant respectivement les taux d’humidité avec déshumidification et conventionnel. Deux lots de semences ont été hermétiquement enfermés dans du polythène et conservés sous deux conditions, la chambre froide (10°C/ 80% r.h.) et la pièce ambiante (21-31°C/68-86%) pendant 18 mois. Les résultats ont montré que les pourcentages de germination audessus 95% étaient maintenus pendant une période de 18 mois de stockage dans des séchoirs de graines à 8,0% d'humidité et conservés dans l’une des chambres froide ou ambiante. Les graines séchées à 11.0% d'humidité et stockées dans une pièce ambiante ont cependant montré des pourcentages faibles de germination en dessous du niveau acceptable de 85% après 9 mois de stockage. Le pourcentage de germination des lots de semences était comparable à celles séchées à 11% et s tockées dans une chambre froide. Les analyses économiques et financières ont montré une rentabilité élevée en changeant de l’option stockage à la chamber froide au stockage à la temperature ambiante des semences deshumidifiées. Ainsi, le coût de stockage de graines pourrait être radicalement réduit si les graines sont déshumidifiées à 8,0% et stockées dans une pièce ambiante pendant des périodes longues allant jusqu'à 18 mois sans perdre leur pouvoir germinatif.

Introduction The purpose of seed storage is to ensure that viable seeds that produce vigorous plants are available at the planting time. A good seed storage system, therefore should maintain the physiological quality, particularly viability and vigor, of seeds. Longevity of seed in storage is controlled primarily by seed moisture content and storage temperature (Asiedu et al. 1998). A major cause of poor seed quality is seed ageing or deterioration during storage, particularly, at high moisture content, high temperature, and high relative humidity (Roberts 1972). Seed deterioration, which entails physiological ageing and insect and fungi activities, increase with increases in seed moisture content. These processes are the major causes of loss of seed vigor and of viability in the warm and humid tropi cal environments, such as the humid regions of West and Central Africa (WCA). In order to prolong the storage life of seeds, seedsmen often opt for sun or mechanical drying to reduce seed moisture content, followed by storage under low temperature and low relative humidity. The target reduction in temperature and humidity depends on the length of the storage period, the initial viability of the seed, desired seed quality at the end of the storage period as well as the type and value of the seed (van Gastel et al. 1999a). In many tropical climates, average temperature is about 30?C and relative humidity is often above 75% (Asiedu and Powell 1998),

491

which means that, using natural air (sun) drying, the equilibrium seed moisture content of maize may not be lower than 11.5% (Asiedu et al. 2001). Also using mechanically heated air blown by fan to dry seed in bins, seed moisture content may only be reduced to 11 12% (van Gastel et al. 1999b). The rule of thumb for seed storage suggests that 1.0% decrease in seed moisture content or 10?F decrease in storage temperature doubles the storage life of seeds (Harrington 1972). Thus, seed moisture content, which is a function of the relative humidity of the storage environment, is more important than temperature in determining storage life. It has been well established that seeds that are dried to very low moisture contents by dehumidification and packaged in sealed moisture -proof bags have very little respiratory activity. Such seeds, therefore, lose little energy with no toxic conditions built up to cause loss of vigor and viability (van Gastel et al. 1999b). Also, insect and fungi activities are significantly reduced in dehumidified seeds and therefore the production of moisture and heat by these organisms that enhance seed deterioration are also reduced to very minimum levels. It is also a well established fact that when dried seeds are kept in humid environments, they tend to absorb moisture, resulting in increases in seed moisture content until equilibrium is reached between the seed and the environment. Thus, the advantages of dehumidification will be lost if seeds are not kept in moisture-proof packages. On the contrary, if moist seeds are kept in a dry environment, they lose moisture to the environment and thus store longer. Reports of studies on the effects of seed moisture and ambient storage conditions on the viability and vigor of maize seed in WCA are scanty. The objectives of this study were to determine the: (1) effect of dehumidifying on maize seed storage life under ambient warehouse conditions, (2) effects of dehumidifying and storage conditions on the storage period beyond which maize seed may be considered unacceptable for commercial purpose, and (3) economic feasibility of dehumidified drying of maize seed.

Materials and Methods Seed of the most popular maize variety in Ghana, ‘Obatanpa’ a quality protein maize (QPM), was harvested in August 1999, shelled, dried to 11.0% moisture content and cleaned to remove unwanted materials. Part of the seed was then dehumidified to further reduce the moisture content from 11.0 to 8.0% using a Munters type mechanical dehumidifier Model MX 1500E, Fabrication No. 98/24/17046 manufactured by Munters Europe AB, Tobo Sweden. The entire period for dehumidifying 300 kg (one full bin) of seed was 5 hours.

492

Description of the dehumidified dryer A dehumidifying dryer consists of several parts inside a sealed, airtight closed-circuit system. A brief description of the dehumidified dryer is as follows: ?? A seed chamber, which holds a relatively thin layer of seed (a thin layer requires less static pressure to force the air to flow through the seed). ?? A desiccant chamber, through which the drying air flows to have its moisture removed. ?? A ducting system that carries dry air from the dehumidifier to the seed chamber and return the "moist" air to the dehumidifier for redrying. This includes a fan, which can develop sufficient static pressure to force the air to flow through the seed mass. ?? A system for removing moisture from the desiccant and discharging it as moist air from the inside of the system to the outside. This process is called “regenerating” the desiccant. The operation of the dehumidified dryer is simple, consisting of the following procedure: ?? The air (inside the system) is passed through a desiccant (silica gel) and dried to very low RH. ?? The dehumidified air is then forced through the seed mass inside the sealed closed dryer system. ?? As the air passes through the seed mass, it absorbs moisture from the seed. ?? After it has absorbed moisture from the seed, the air flows back to the dehumidifier. ?? In the dehumidifier, the air again passes through the desiccant and is further dried to a very low RH. ?? This redried air again passes through the seed, and continues drying the seed.

Seed storage experiment Using the dehumified seed of Obatanpa, 200 g seedlots were weighed into 72 moisture -proof polythene bags (0.2 mm thickness, measuring 6 x 10 cm) and heat-sealed. For purposes of comparison with the conventional practice of seed growers, a second set of 100 kg seed from the same seed lot as the former, was mechanically dried to 11.0% moisture content and 200 g weighed into each of 72 moisture -proof polythene bags (0.2 mm thickness). The two sets of seed, totalling 144 packages, were stored under two conditions, the cold room (10?C/80% r.h.) and the warehouse (21–31?C/68–86% r.h.) for 18 months. The treatments were as follows: 1. 8.0% moisture content in the warehouse (21–31?C/68-86% r.h.).

493

2. 8.0% moisture content in the cold room (10?C/80% r.h.). 3. 11.0% moisture content in the warehouse (21–31?C/68–86% r.h.). 4. 11.0% moisture content in the cold room (10?C/80% r.h.). The experimental design was a randomized complete block design arranged in a 2 x 2 (two moisture levels and 2 storage conditions) factorial, replicated four times. The seedlots were assayed monthly for germination test. The final germination test was done after 18 months of storage. A second experiment was conducted to determine the storage potentials of popular inbred lines in the Maize Breeding Program and their two-way and three -way hybrids under dehumidified and non -dehumidified conditions. The genotypes consisted of six inbred lines (GH3, 1368, 9071, Entry 5, Entry 6, and Entry 70), two single crosses (GH3x1368 and 6x70), tw o 3-way crosses [(GH3 x 1368) x 9071 and (Entry 6 x 70) x Entry 5], and one OPV (Obatanpa). Seed of the varieties was dried to 11.0% and 8.0% moisture contents and stored in sealed polythene bags under ambient conditions for six months. Germination tests were conducted at the beginning of the experiment and also at the end of 3 and 6 months. A randomized complete block design with four replications was used in the experiment.

Germination test At the end of each month of storage, a sample of 100 seeds was obtained in the four replications of each treatment for germination tests. The seeds were geminated in moist, heatsterilized sand (heated at 105?C for 24 hours) in a circular tray, (30 cm in diameter). This was kept in polythene bags to reduce moisture loss at the ambient temperature of 27–32?C. Germination counts were made on the 4th and 5th day after planting. Seeds were considered germinated only if their seedlings were normal as described by AOSA (1981). In view of the prevailing high cost of cold storage ($38 per ton for six months, which represents a fraction of the cost of electricity), and the high germination percentage of dehumidified seeds, it seems attractive to dehumidify commercial seeds and store them in the warehouse. A cost-benefit analysis was therefore conducted to determine the economic advantage of dehumidified hermetic storage in the ambient.

Results and Discussion After 18 months of storage under ambient conditions, the seed dehumidified to 8.0% moisture content with an initial 96.7% germination showed only a 4.1% decrease in germination (Table

494

1). There was no change in the germination percentage of seeds dried to the same moisture content but stored in the cold room. Table 1. Effects of percent seed moisture content (% MC) and storage conditions on germination of the maize variety, Obatanpa over a period of 18 months. % MC 8 8 11 11 Mean ± SE

Storage condition Warehouse (21-310C/6886% r.h.) Cold Store (100 C/80% r.h.) Warehouse (21-310C/6886% r.h.) Cold Store (100 C/80% r.h.)

3 6 9 12 15 18 Initial mo. mo. mo. mo. mo. mo. -------------------Germination, % -------------------96.7 94.4 95.4 94.7 94.6 95.4 93.6 96.7

96.3

95.4

94.0

96.2

96.1

96.5

96.7

95.1

93.4

91.4

75.8

61.7

40.3

96.7

96.4

95.1

95.5

96.3

96.6

95.8

96.7 0.00

95.6 0.48

94.8 0.48

93.9 0.89

90.7 4.99

87.5 8.59

81.6 13.76

Within the same period of storage, the germination of seed dried to 11.0% moisture content and stored under warehouse conditions dropped from 96.7% to 40.3%, representing 56.4 decrease in germination. Seeds dried to 11.0% and stored in coldroom, however, maintained high germination percentage above 95% throughout the 18-month period. In an attempt to determine the vigor levels of the seeds being studied, 4th day germination count was recorded from 13 months onwards (Table 2). Table 2. Effects of percent seed moisture content (% MC) and storage condition on fourth-day germination of maize seed 13-18 months in storage. % MC 8 8 11 11 Mean ± SE

Storage condition Warehouse (21–310C/68– 86% r.h.) Cold Store (100 C/80% r.h.) Warehouse (21–310C/68– 86% r.h.) Cold Store (100 C/80% r.h.)

13 14 15 mo. mo. mo. --------------Germination, 90.9 90.4 89.9

17 18 mo. mo. % ------------85.7 87.8

95.6

82.9

88.4

92.0

94.5

68.4

46.2

20.6

12.3

23.5

92.8

85.9

85.1

93.7

93.6

86.9 6.25

76.4 10.17

68.5 16.06

70.9 19.62

74.9 17.18

495

As at the 13 th month of storage, there was already an indication that the seed dried to 11.0% moisture content and stored in the warehouse had lost substantial level of vigor, whereas the others still showed high vigor ratings. From 15 to 18 months, the vigor ranking was highest for seed dried to 8.0% moisture content and stored in the cold room. This was followed in descending order by seed of 11.0% moisture content stored in the cold room, seed of 8.0% moisture content stored in the warehouse and seed of 11.0% moisture content stored under ambient conditions. These results clearly indicate that seed deterioration could be minimised by dehumidification. The results also corroborate the established fact that low moisture content and low temperature are important determinants of seed longevity, with moisture content being more important than temperature (Harrington 1972). The combination of the two factors prolonged seed storage life longer than either factor alone. Results obtained from the second experiment indicated that all the varieties maintained high and acceptable germination percentages up to three months (Table 3). At six months, all the seeds dried to 8.0% moisture content maintained germination above 95.0%, whereas those dried to 11.0% moisture content generally showed reduced germination. Table 3. Effects of variety, percent seed moisture content (% MC) and storage period on the seed germination of eleven maize varieties.

Variety Inbred lines GH3 1368 9071 Entry 6 Entry 7 Entry 70 Single-crosses GH 3 x 1368 Entry 6 x Entry 70 Three-way crosses (GH 3 x 1368) x 9071

(Ent 6 x 70) x Ent 5 OPV Obatanpa Mean ± SE

8.0% MC 11.0% MC 3 6 3 6 Initial mth mth Initial mth mth ------------------ Germination, % -----------------97.5 95.5 99.0 96.5 93.0 98.0

99.5 99.0 98.0 97.5 96.5 96.5

95.0 99.5 100 95.0 97.5 95.5

96.5 99.5 98.0 96.5 96.5 98.0

95.5 96.0 98.0 97.5 92.5 97.0

87.0 66.0 92.0 80.5 73.0 83.5

95.0 95.5

100 99.0

98.0 97.0

98.5 98.5

97.5 98.0

80.0 84.0

95.5 100.0

99.0 98.5

99.0 97.5

98.5 99.5

98.5 100

95.0 88.0

98.0 96.7 0.61

98.0 98.3 0.35

98.0 97.5 0.52

99.5 98.1 0.36

97.5 97.4 0.84

92.5 80.5 4.12

Four of the six inbred lines dried to 11.0% moisture content showed reduced germination below 85% whereas two lines (9071 and Entry 70) maintained high germination. All the hybrids and

496

the open-pollinated variety showed high and acceptable germination at 8.0% moisture content. At 11.0% moisture level, only two of the hybrids, ((GH 3x1368) x 9071) and Entry 6 x Entry 70) showed acceptable germination. Basing the cost-benefit analysis of dehumidified drying on the capacity of four bins of the dehumidified dryer (260 tons) and storing for 24 months, cost of storage was determine to be $13 362 for cold storage and $4608 for ambient storage (Fialo 2000). Financial rate of return of the dehumidifier dryer was given as 131% over 10 years for the Grains and Legumes Development Board (GLDB) who could purchase the equipment and offer custom services to private seed producers with the initial four years as payback period. For the private producers, the financial saving on the use of the equipment was estimated as $34 962 per year, which represented 68% saving for opting for the dehumidifier dryer. For GLDB, additional benefit is obtained as a saving on electricity charges up to $5726 every year. The cost of the equipment and installation is $22 942. Thus the dehumidified seed drying system seems to be a feasible option to reduce cost of storage without loss of seed quality. In Ghana, acceptable germination percentage is 90.0% for foundation seed and 85.0% for certified seed (Ocran et al. 1998) below which seeds are not considered acceptable for marketing at commercial level. According to the results of this study, seeds dried to 11.0% moisture content and stored in the warehouse for periods longer than nine months may be considered expired and cannot be marketed as either foundation or certified seed. A significant proportion of seed growers dry their seed to about 11.0% moisture content. However, due to high cost of cold storage or long distances of seed production fields from coldrooms, most seed growers keep their see ds in warehouses. From the results of the present study, it is evident that such seeds may lose viability quickly which could lead to huge financial losses to seed companies. Most seed growers who are located near seed storage centres, however, keep their seeds in the cold room at 11.0–12.5% moisture content. These seed growers may not encounter problems of loss of viability within the period of 18 months, although the storage cost will be as high as $114 per ton of seed. One limitation in setting germination standards for seeds in developing countries is that, most often, initial germination is determined without an indication of expiry dates beyond which seed quality may not be considered acceptable for the purpose of planting. The results of this preliminary investigation give tentative indications as to the dates of expiry of maize seed under different storage conditions.

497

Summary and Conclusions A major challenge in seed storage in the tropics is to maintain physiological quality of seed by drying to 11.0% moisture content or lower. Sun-drying or conventional seed dryers used by most seed producers cannot dry seed to such low moisture contents. A mechanical dehumidified dryer was used to reduce seed moisture content to 8.0% and was compared with see d dried to 11.0%, when both were stored in the cold room or under the ambient conditions. With the exception of the seed dried to 11.0% and stored under warehouse conditions, the seeds showed high vigor and germination for 18 months. Thus it was possible to retain the physiological quality of maize seed under the ambient conditions when seed was dried to 8.0% moisture content. Furthermore, screening of eleven maize varieties comprising six inbred lines, two single cross hybrids, two three-way hybrids and one open-pollinated variety dried to 8.0 and 11.0% moisture and stored for 6 months showed high germination percentages above 90% throughout the storage period for all the varieties, when dehumidified to 8.0% moisture content. Seed of these varieties stored at 11.0% moisture content for 6 months showed high germination (above 85%) in one inbred line, the two three-way hybrids and the one open-pollinated variety. Thus, the two single crosses and 5 inbred lines showed reduced germination below the acceptable level for certified seed. Cost of storage for 24 month was estimated as $13 362 for cold storage and $4608 for dehumidified warehouse storage. Financial rate of returns on the dehumidifier dryer was calculated to be 131% over a period of 10 years. Additional benefit is a saving on electricity charges up to $5726 every year. For private producers, the financial saving on the use of the equipment was estimated as $34,962 per year, which represented 68% savings. Thus, the dehumidifier dryer offers a cost-effective option of storing seed in the tropical environment such as the humid zone of WCA. The economic and financial analyses showed high profit margins of seed enterprises if they opt for the dehumidified dryer.

References Asiedu, E. A. and A.A. Powell. 1998. Comparison of the storage potential of cultivars of cowpea (Vigna unguiculata) differing in seed coat pigmentation. Seed Science and Technology 26, 211–221. Asiedu, E. A., S. Twumasi-Afriyie, P.Y.K. Sallah, J.N. Asafu-Agyei, and J.G.A. van Gastel. 1998. Maize seed production: principles and practices. In C. Osei Kwabena, H. Dapaah, I.S. Banning, and I.O.O. Ansah (eds.) Published by Training Communication & Publication Unit, (TCPU), Crops Research Institute. 58p.

498

Asiedu, E. A., A.J.G. van Gastel, B. Greg, and A. Ebert. 2001. Dehumidifying drying: A viable option for long term seed storage in the humid tropics. Pages 59–67 in B. Badu -Apraku, M.A.B. Fakorede, M. Ouedraogo, and R.J. Carsky (eds.) Impact, challenges and prospects of maize research and development in West and Central Africa. Proceedings of a Regional Maize Workshop, IITA-Cotonou, Benin Republic, 4–7 May, 1999. WECAMAN/IITA. Association of Official Seed Analyst (AOSA). 1981. Rules for testing seeds. Journal of Seed Technology 62 (2):1–26. Fialo, S.C. 2000. Financial and economic analysis of the West Africa Seed Development Unit’s (WASDU’s) seedsaver. A Consultancy Report Submitted to the West Africa Seed Development Unit (WASDU). October, 2000. 8p. Harrington, J.F., 1972. Seed storage and lon gevity. Pages 145– 240 in T. T. Kozlowski (ed.) Seed Biology. Academy Press Inc., London. Ocran, V.K., L.L. Delimini, R.A. Asuboah, and E.A. Asiedu. 1998. Seed management manual for Ghana. A manual on seed production and management recommendations for the major food crops of Ghana. Published by the Department For International Development (DFID), UK. 60p. Roberts, E.H. 1972. Cytological, genetic and metabolic changes associated with loss of viability. Pages 253–306 in E.H. Roberts (ed.) Viability of seeds. Chapman and Hall Limited, London. van Gastel A. J.G., Gregg, B.R. And Asiedu, E.A. 1999. Safe seed storage in West Africa. WASDU Workable Approach No. 1. West African Seed Development Unit (WASDU). Accra, Ghana, 1999. van Gastel A.J.G., B.R. Gregg, and E.A. Asiedu. 1999. Dehumidified seed drying, a cost effective option for seed storage and marketing in hot and humid climates. WASDU Workable Approach No. 4. West African Seed Development Unit (WASDU). Accra, Ghana.

499

An approach to rapid deployment of agricultural technologies—Transfer of downy mildew resistant maize to farmers in Ogbomoso, southwest Nigeria S.O. Ajala, V.M. Manyong, V. Adenle, K.O. Makinde, A. Akintunde, J. Olufowote1, M. Bolaji2 , and B. Bolaji2 International Institute of Tropical Agriculture, PMB 5320, Ibadan, Nigeria 1 World Vision International, PO Box 1490, Kaneshie, Accra, Ghana 2 Oyo State Agricultural Development Project, Zonal Office, Ogbomoso, Oyo State, Nigeria. Abstract A new approach to the deployment of downy mildew resistant (DMR) maize (Zea mays L.) varieties evaluated in selected downy mildew endemic areas of Ogbomoso in southwestern Nigeria. Nine villages were selected for the project during the first season of 1997. Seed and inorganic fertilizers were supplied to three farmers from each village and the farmers were trained to produce high-quality seed on their own. In the following season, each farmer on the project brought in three new farmers from a new village and one from his own village and bachstopped them with seeds and technical know-how imbibed from his earlier training. In 1998, 25 villages and 111 farmers participated in the project. By the third year, 625 farmers in 159 villages had been trained and were producing seed of DM resistant varieties. In addition, the farmers were trained to produce a healthy maize crop. In the 1998/1999 seasons, average yield from the farms of non-participating farmers was 1846kg/ha compared with 2763 kg/ha for participating farmers, a 50% yield increase for the participating farmers. Results from the survey conducted at the termination of the project indicated that within two years, the farmers observed a drastic reduction in the incidence of downy mildew and obtained increases in maize yield and income. Résumé Une nouvelle approche pour la diffusion des variétés DMR de maïs (Zea mays L.) résistantes au mildiou ont été évaluées dans certaines zones de Ogbomoso endémiques au mildiou dans le Sud Ouest du Nigéria. Neuf villages ont été sélectionnés pour le projet durant la première saison de 1997. La semence et les engrais inorganiques ont été fournis à trois paysans dans chaque village et les paysans ont été formés pour produire eux mêmes des semences de haute qualité. A la saison suivante, chaque paysan du projet envoie trois nouveaux paysans provenant de nouveaux villages et un

500

paysan de son propre village et les a doter de semences et des connaissances qu`il a déjà reçu. En 1998, 25 villages et 111 paysans ont participé au projet. A la troisième année, 625 paysans dans 159 villages ont été formés et ont produit des semences des variétés résistantes au mildiou. En plus, les paysans ont été formés pour produire des champs de maïs sains. Pendant la saison 1998/1999 le rendement moyen des champs des paysans qui n`ont pas participé au projet a été de 1846kg/ha comparé avec 2763 kg/ha pour les paysans qui ont participé au projet, une augmentation de 50% des rendements pour les paysans qui ont participé. Les résultats de l`enquête menée en fin de projet ont montré que dans les deux ans, les paysans ont observé une réduction importante de l`incidence du mildiou et ont obtenu une augmentation du rendement et des revenus.

Introduction Technological innovation in low-resource agricultural systems will continue to be a major factor contributing to Africa’s ability to meet the challenges of food security and agricultural development in the foreseeable future. However, technical approaches alone will not solve Africa’s food security problems (US Congress Office of Technology Assessment 1988). The development of appropriate technology is a necessity, but not a sufficient condition for ensuring technology adoption. For agricultural development to occur, a system of technology transfer that provides farmers with the inputs and information needed to enhance productivity, is a prime requirement. Technology dissemination is the foundation for progress if advancement in scientific knowledge is to be used for development (Terbicke and Berhan 1983). In sub-Saharan Africa, considerable emphasis is placed on technology transfer. Over the past three decades, several technology transfer approaches have been tested. The first of these approaches is the traditional extension system of the ministries of agriculture in the early 1960s. Since inception, the system has been criticized for none or poor performance of its services. There were many complaints, including wasteful spending, prescriptive recommendations, non-availability of required inputs to enhance changes and lack of training (Adekunle et al. 2001). The Training and Visit (T&V) system was introduced to solve some of these problems. The central aim of the T&V approach, which the World Bank has supported in 37 African countries, was to ensure that well-trained agents, bearing suitable messages, visit farmers regularly. Accordingly, much emphasis was placed on improved management and training of extension staff and on technology testing on farmers’ fields with full participation of the farmers. Although this approach has had a positive impact in various countries of sub-Saharan Africa, notable among which

501

are Kenya, Burkina Faso and Cote d’Ivoire, the World Bank evaluations of its own efforts indicated that the extension systems were poorly managed and that the technology being promoted were often irrelevant. Farmer participation, meeting the information needs of women, private sector involvement and improving the quality of agricultural education are some of the suggestions made for improving the T&V approach (Cleaver 1993). While governmental efforts led to the establishment of agricultural development projects, a number of voluntary organisations also initiated activities in several communities. A notable example of this is the SG 2000 Program. The SG 2000 projects focus primarily on technology testing and demonstration by establishing field plots managed by individual farmers under the supervision of the extension staff. There also have been several training courses based on themes along the cropping calendar, conducted in what is known as the FAO Field Schools, using a participatory approach. This approach also suffered from over-emphasis on integrated pest management technologies to the detriment of other areas and little or no attention has been paid to input supplies. These efforts have not been generally sustainable due largely to the lack of an effective mechanism for farmer empowerment. There is thus the need for initiatives on completely sustainable technology deployment strategies. The goal of the project described in this paper was to improve the food security of the smallholder farmers in the Ogo Oluwa and Orire Local Government Areas (LGA) in Ogbomoso area of Oyo State of Nigeria. The strategy was to rapidly saturate the area with downy mildew resistant (DMR) maize varieties. The specific objectives were to: a) rapidly disseminate DMR varieties of maize in the area, b) develop a model that can be used directly or modified for similar deployment exercises elsewhere in sub-Saharan Africa, and c) develop the capacity of farmers themselves to implement and sustain the intervention.

Elements of the new strategy The traditional top-down approach of developing technologies at research stations and transferring them to farmers has been largely unsuccessful in Africa (US Congress Office of Technology Assessment 1988). Practical experiences have shown that ‘success’ has often been the result of respecting natural processes with interwoven principles. Major elements of our approach in deploying DMR maize varieties to farmers in Ogbomoso include:

502

?? multi-instituti onal collaboration to put complementary efforts in place, ?? identification of a niche for the technology, ?? active farmer participation, ?? farmer empowerment, ?? interaction among farmers. These elements are briefly discussed in the following sections.

Multi-institutional collaboration to put complementary efforts in place Although improved technology is an important driving force in accelerating the growth in food production, the way in which such interventions will alter the productivity of farmers depends on the extent to which different institutions share information and coordinate their efforts. Every technology is a multidimensional process with each aspect requiring the attention of different experts. This suggests that a holistic or systems approach that recognises and plans for interactions among various components of a technology have the best chance of being successful. Therefore, it was clear from the onset that experts in the different institutions need to come together to provide the institutional collaboration needed for the success of the proposed intervention.

Identification of a niche for the technology A study commissioned earlier by World Vision International (WVI) in 1988 clearly demonstrated an urgent need for a highly costeffective technology transfer system in Nigeria. Based on the results of the study, a 5-year strategic plan was developed to popularise existing improved, adapted, seed-based technologies developed in national and international agricultural research centers for the major food crops: maize, sorghum, cowpea, rice, and root crops. Implementation of seed production activities—multiplication, dissemination, farmers’ field days and varietal testings—became the priority for the ‘Year of Seed’ initiatives of the WVI Food Security Plan. WVI had executed a Child Survival Project (CSP) for about 7 years in two Local Government Areas (LGA), Orire and Ogo-Oluwa, of Ogbomoso. Deployment of DMR varieties of maize for the control of DM disease was identified during initial discussions between the IITA scientists (maize breeder, agricultural economist and pathologist) and WVI representatives. Thereafter, the support of the Oyo State Agricultural Development Programme (OYSADEP), which has the responsibility for agricultural extension services in the target area, and the supervisory councillors for agriculture in the local councils, were solicited for the proposed project. The two LGAs (Orire and Ogo-Oluwa) were to be used as case studies. This

503

choice was quite appropriate because the downy mildew disease of maize is highly endemic to the Ogbomoso area of Oyo State. Failure of local skills and traditional methods was a motivation for the farmers to try out new strategies.

Active farmer participation To effectively assist resource -poor farmers, it is highly essential to enlist their participation in the design and execution of any intervention of technology transfer. Farmers need to be involved more actively in selecting and testing messages and in identifying the farm -level problems that should be addressed by research and extension (Venkatesan and Schwartz 1992). Based on this premise, concerted efforts were made to involve the farmers from the inception of the project. Such interaction between farmers and researchers has benefits for both groups. While researchers gain an understanding of farmers’ needs and the production system into which the intervention must fit, farmers are helped in adopting new techniques that could lead to increased productivity. In this respect, the Ogbomoso project started at an opportuned time. At the beginning of the project, farmers were reluctant to participate because of several problems, such as unkept promises they had received from earlier motives by other organizations to assist them. When they were informed that the project was funded by the WVI as an extension of the CSP, quite a number of the farmers became interested. Farmer participation is particularly important in understanding the complexity and diversity of farming systems of the south-western Nigeria, which is typical of the low-resource agriculture in sub-Saharan Africa, the integrated nature of the agriculture and the relationships between the different components of interventions.

Farmer empowerment There were two aspects to this strategic element–embodied and disembodied empowerment. Embodied empowerment deals with imparting technical skills to participating farmers. Farmers’ field schools (including field days) were organised to impart technical skills. The field days were designed to train a minimum of 20 farmers. Disembodied empowerment consisted essentially of input procurement and distribution. One metric ton of DM resistant maize variety was distributed to 50 or more farmers to enable each farmer plant a multiplication area of 1000 m2. This strategy also recognised the key role played by credit, based on the understanding that poor farmers in Africa typically suffer from lack of funds to procure essential inputs. The DMR seeds were provided free of charge to farmers by the project while fertilisers and insecticides were provided as revolving loans in kind to farmers to be repaid after the crop had been harvested and the grains marketed.

504

Interaction among farmers A final element of the strategy was planning for interactions among farme rs in a manner that guarantees exponential rate of growth in dissemination. This constitutes the main rationale for farmers teaching other farmers. Several field schools on basic concepts of seed production for use in their respective communities were organised and farmers also participated in collecting and inoculating downy mildew spores from infected leaves to test genotypes in a demonstration area. Even if associated farmers do not learn by doing, the experience of observing someone else’s plot may compel them to seek more direct participation in the program and to begin applying some or all of the elements of improved technology on their own. Indeed, group action can help farmers ease the transition from testing technology with the assistance of extensi on staff to applying it routinely with no outside support (Dowswell 1993, p.67). The benefit of working with farmers in this way is that it creates capacity for farmers themselves to implement and maintain such interventions particularly when assisted with funds to run the project.

Case Study The DM disease of maize, incited by the fungus Perenosclerospora sorghi has been reported in Nigeria, Democratic Republic of Congo, Mozambique and Uganda. Symptoms of the disease include the characteristic half le af chlorosis; narrow, stiff, erect leaves and malformation of the inflorescence referred to as ‘crazy top’. Infected plants generally do not produce cobs. A near total yield loss is possible in highly susceptible varieties. Farmers in Nigeria are aware of the effects of DM disease and had contacts with extension agents on the use of resistant varieties and “Apron plus”, a fungicide to control the disease. The extension agents in the project area are often not able to provide resistant varieties but usually are aware of where to purchase the Apron plus due perhaps to the more aggressive sales drive of pesticide sales representatives interacting with the extension agents. The cost of a sachet of Apron plus is high and one sachet is usually recommended for the treatment of 1 kg of maize seeds. Frequently, fake products are sold in the market. This situation has led to a decline in maize production in the area. IITA has produced new DMR maize varieties, which were not reaching the farmers. To solve this problem, the WVI in association with IITA decided on a deployment program for DMR maize varieties in the area as part of the Year of Seed Initiative in 1997. The decision on this exercise resulted from a food security assessment conducted earlier by WVI in late 1996. WV Australia funded this initiative with some contribution from WV New Zealand.

505

Choice of area and deployment strategy The smallholder farmers in the Ogo-Oluwa and Orire LGA in Oyo State, Nigeria (Fig. 1) formed the target population for this project. In the first year of intervention (1997), a survey of the villages and farmers was undertaken and five villages were selected per LGA. From each village, three farmers were selected randomly and backstopped to grow the DMR maize varieties. For the second year, each of the participating farmers was encouraged to backstop a new village in addition to at least one new person from his/her own village. The procedure was repeated in the third year. In effect, for the three years, number of participating farmers increased from 25 in 1997, to 111 in 1998, and 404 in 1999. Field sizes ranged from 0.4 to 1.0 ha. Each farmer was given 15 kg of DMR maize seed to be planted at about 50 000 plants/ha.

N ig e r ia

O r ire

O g o- O l uw a

Figure 1. Map of Oyo State of Niger ia showing the study area.

Fertilizers (NPK 20:10:10) were distributed to farmers according to field sizes and soil fertility status. Application rate was 60 kg N per hectare. Fertilizer application started four weeks after planting. Side dressing and spot application were used. Where

506

necessary, a second application of fertilizer was done when the plants were flowering. The staff of IITA and the Oyo State Agricultural Development Project (OYSADEP) supervised field activities. Demonstration plots were set up for the farmers OYSADEP office in Ogbomoso where sampled farmers were invited to observe how downy mildew infests maize plants. Inoculations were done in the nights and farmers were made to participate fully in the demonstration. The project intervention consisted principally of procurement and distribution of DMR maize varieties and fertilisers charged to the individual farmer’s account. Other interventions included recommendation and supervision of agronomic practices such as land preparation, planting, weeding, and harvesting. The farmers supplied the labor for these operations. In the event of farmers not being able to provide the labor needs, the project provided these and charged the account of the farmers. An account was maintained for individual farmers with a mutual understanding that all ‘loans’ would be recovered by the project immediately after harvest. At about physiological maturity, an area about 20 m x 20 m was sampled from each of the farmers’ fields to obtain data on grain yield, stand count, ear number and moisture content at harvest. Data on labor use for the different farm activities, input use, maize output and prices were also obtained from the farmers. Grain yield and labor use data were also obtained from interested neighboring non-participating farmers for purposes of evaluation of the effect of the intervention. The data were subjected to descriptive statistics, frequency distribution, and partial budget analysis.

Results and Discussion Trends in technology dissemination The primary objective of this project was to quickly saturate a downy mildew endemic area with DMR maize varieties. Therefore, the first evidence of the project’s impact is the extent of spread of the DMR varieties (Fig. 2). The project started off in 9 villages in 1997. The project’s influence rapidly spread to other villages and areas where the downy mildew disease was a problem. The number of villages covered by the project increased by 178% in 1998 and by 296% in 1999. Our results corroborate the conclusion of Byerlee and Heisey (1993) that small-scale farmers in Africa do accept well-adapted technologies, once these are made available, along with appropriate institutional support. The pattern of spread of the DMR varieties shows what happens when there is a perceived

507

need by farmers for solution to a problem—in this case, the downy mildew disease of maize.

120

450 400

100 80

300 250

60 200 40

150

No. of Farmers

No. of Villages

350

100 20 50 0

0 1997

1998 Year Villages

1999 Farmers

Figure 2. Number of villages and farmers that participated in the dissemination of DMR maize varieties in Ogbomoso, 1997–99.

Figure 2 also shows the increase in the number of farmers that participated in the project from 1997 to 1999. In the first year of operation, only 25 farmers participated in the project. The number increased to 111 and 404 in 1998 and 1999, respective ly. On average, each initial farmer backstopped 3.8 new farmers, a 95% achievement of the planned target of four new farmers per initial farmer. Yield gains Maize grain yield increased from 2207 kg ha-1 in 1997 to 2929 kg ha-1 in 1999 in fields with intervention; that is, the participating farmers’ fields (Tables 1 and 2). Flooding and and inadequate use of fertilizers due to very high prices resulted in low yields in 1998. Maize cannot exhibit its maximum yield potential under conditions of excessive rainfall and inadequate use of fertilizers, especially nitrogen. The fact that yield figures were similar on fields with intervention and those without suggests that the potential of the DMR maize was not realised under the conditions of inadequate fertilizers and flooding which prevailed in 1998. Considerable variation in grain yield was observed for different years. In 1999, maize yields were higher on intervention fields than on those without intervention. The combined yield from

508

maize plots in 1999 was about 31% higher than in 1997 and about 56% more than the yields observed in 1998. Table 1. Grain yield and economic profitability of DMR maize varieties in Ogbomoso, Nigeria, 1997-99. (*All fields at full costs). 1997 n=25

1998 n=109

1999 n=487

0.71 2207 18996 40458 21462 2.13

0.66 1847 33647 23258 -10389 0.69

0.57 2887 19858 32929 13071 1.65

BC ratio (% farmers) BC < 1.00 1.00 < BC < 2.00

8.0 40.0

80.7 19.3

20.3 46.8

BC > 2.00

52.0

0.0

32.9

Average area (ha) Yield (kg/ha) Cost (N/ha) Revenue (N /ha) Net revenue (N /ha) BC ratio

*Cost of family laboiur included. Table 2. Grain yield and economic profitability of DMR maize varieties in Ogbomoso, Nigeria, 1997–99 (All fields at cash costs). 1997 N = 25

1998 N = 109

1999 N = 487

0.71 2207 13555 40458 26903 2.98

0.66 1847 21447 23258 1811 1.08

0.57 2887 15040 32929 17889 2.19

BC Ratio (% Farmers) BC < 1.00 1.00 < BC < 2.00

10.5 50.8

45.9 40.4

11.5 35.0

BC > 2.00

38.9

Average area (ha) Yield (kg/ha) Cost (N /ha) Revenue (N/ha) Net revenue (N /ha) BC ratio

13.8

53.5

Economic benefits A direct measure of the impact of a new technology is the economic benefits to the adopters relative to non-adopters. The partial budget analysis was used to investigate the economic performance of the DMR varieties. The analysis was undertaken in four parts, (i) economic benefits over all maize fields (ii) economic benefits due to intervention (iii) effect of project experience on economic benefits and (iv) the spill-over effects of intervention.

509

Economic benefits over all maize fields Results of the partial budget analysis for all fields where full costs are charged (cost imputed for family labor) are presented in Tables 1. The results of the partial budget analysi s when only cash costs are considered (family labor is removed) are presented in Table 2. As depicted in Table 1, production cost per hectare was lowest (N18 996) in 1997 when the project started followed by 1999 (N19 589). In 1998, production cost rose by 77% due to about a 170% increase in fertilizer prices from N400 in 1997 to N1080 in 1998. Net revenue per hectare was highest in 1997 (N21 462), and lowest in 1998 (-N10 389). Consequently, the benefit-cost (BC) ratio was less than unity in 1998 (0.69), indicating that the average farmer could not fully recover his costs in contrast to the BC ratio of 2.13 and 1.65 obtained for 1997 and 1999, respectively. Although this indication of profitability is convenient, it does not show the distribution of the gains among participating farmers. The situation is more clearly demonstrated by the frequency of the BC ratios among farmers also shown in Table 1. The distribution shows that while over 90% of farmers in 1997 and about 80% in 1999 attained break-even point, 80% of the farmers could not recover their costs in 1998. When only cash costs were considered, there was a marked improvement in all of the economic indices over the three years. The proportion of marginal farmers (BC < 1) decreased while that of be tter off farmers increased (BC > 2). We used this analysis to show that if only cash outlays are considered and family labour is not included in the analysis (as farmers often do), then the farmers make more profit.

Economic benefits due to intervention Grain yield was generally higher on fields with interventions than those without interventions, where farmers grew maize according to their traditional practices. Results of partial budget analysis conducted separately for fields with and fields without intervention in 1999 are presented in Table 3. Average revenue per hectare was 33% higher on fields where DMR maize was grown than on traditional maize farms. Net revenue was about 140% higher than non -intervention farms. As shown by the benefit–cost ratios among participating farmers, the proportion of marginal fields was higher on fields without DMR maize varieties. The implication of this result is that the DMR maize is economically superior to the alternative maize varieties grown by project farmers.

510

Table 3. Economic profitability of DMR maize varieties in fields with or without intervention at Ogbomoso, Nigeria, 1999.

Variable

Field without intervention with intervention n=393 n=94

Average area (ha) Yield (kg/ha) Cost (N /ha) Revenue (N/ha) Net Revenue (N /ha) BC Ratio BC < 1.00 1.00 < BC < 2.00 BC > 2.00

0.54 2929 19649 34590 14941 1.76 17.0 47.8

0.71 2713 20736 25982 5246 1.25 34.0 42.6

35.1

23.4

Effect of project experience on economic benefits An important but often overlooked aspect of the profitability of farmers adopting a new technology is the experience gained by farmers in project participation. In the specific case of our study, experience gained from the use of DMR maize varieties may be expected to influence profitability in two ways. First, it is possible that experienced farmers may reduce costs due to increased technical skills in the use of DMR maize. Second, previous participation could enhance managerial competence of farmers in the use of inputs. The relative performances of experienced and new farmers are shown in Table 4. Profitability of DMR maize increased as the farmers accumulated experience in participation. Table 4. Effects of farmers’ experience on the profitability of growing DMR maize varieties in Ogbomoso, Nigeria, 1998– 99. 1998 farmers’ experience New 1 Year

n=87 Average area (ha) Yield (kg/ha) Cost (N/ha) Revenue (N /ha) Net revenue (N /ha) BC ratio BC ratio (% farmers) BC < 1.00 1.00 < BC < 2.00 BC > 2.00

N=22

0.61 0.86 1707 2398 32748 37199 21321 30917 -11427 -6282 0.65 0.83 48.3 40.2 11.5

36.4 40.9 22.7

1999 farmers’ experience New 1 Year 2 Years

n=420

n=47

n=20

0.57 2932 20117 32850 12733 1.63

0.56 2466 17656 32470 14814 1.84

0.65 2936 19615 35670 16055 1.82

21.7 46.0 32.4

10.6 55.3 34.1

15.0 45.0 40.0

511

Analysis of the 1998 data showed that farmers who participated in the project in 1997 had larger fields and higher yields than their counterparts who were new on the project. Average revenue per hectare for experienced farmers was N30 917 compared with N21 321 for new farmers. Although net benefit was negative for the average farmer in 1998 due to factors discussed earlier, previous experience helped some farmers to reduce the monetary losses by about 45%. The effect of experience was more clearly highlighted by the 1999 results. Net revenue per hectare was highest (N16 055) for farmers with 2-year experience, followed by those with 1-year experience (N14 814) and least for new farmers (N12 733).

Spill-over effects of intervention Every technological intervention affects and is affected by numerous outside factors. Unfortunately, however, the complimentarity of traditional farm enterprises and the skills gained by farmers on experimental fields can hardly be disentangled from field to field. That notwithstanding, we examined the spill-over effects of the experience gained by farmers from growing DMR maize on other maize fields. Table 5 shows the effect of experience separately for fields with intervention and without intervention. The pattern observed on fields with intervention (experimental fields) is congruent with the results in Table 4. A similar pattern was observed in nonexperimental fields, indicating that profitability was higher on fields of experienced farmers than on new farmers. Table 5. Spillover effects of growing DMR maize on fields with and without intervention in Ogbomoso, Nigeria, 1997–99.

Variable Average area (ha) Yield (kg/ha) Cost (N/ha) Revenue (N /ha) Net revenue (N /ha) BC ratio

With intervention New 1 Year 2 Years n=338 n=40 n=15 0.53 2988 19948 34617 14669 1.74

BC ratio (% farmers) BC < 1.00 18.6 1.00 < BC < 2.00 BC > 2.00

46.4 34.9

0.53 2381 17828 33386 15558 1.87

0.72 3046 17756 37205 19449 2.10

Without intervention New 1 Year 2 Years n=82 n=7 n=5 0.73 0.74 2699 2955 20811 16674 25566 27229 4755 10555 1.23 1.63

0.44 2607 25191 31065 5874 1.23

7.5

6.8

34.1

28.6

40.0

60.0 32.5

46.6 46.6

43.9 22.0

28.6 42.8

40.0 20.0

Farmers perception of benefits An end-of-project survey was conducted between November and December 2000 to elicit farmers’ assessment of the impact of the DMR maize. The objectives were to assess the effect of DMR

512

varieties on the grain yield of maize and the income received by farmers, and to determine the extent of farmer-to-farmer diffusion. The survey was conducted on 396 of the 404 farmers that participated in the project and 68 other farmers who had received seeds from some participating farmers but did not participate directly in the project. From the farmers’ perception of the impact of downy mildew resistant maize, it can be concluded that the incidence of downy mildew reduced to a minimal level in the project area. Results summarized in Figure 3 show that the incidence had reduced drastically on the fields of 92% of the participating farmers and 96% of non-participating farmers that grew DMR varieties. About 66% of the farmers indicated that they had dramatic increase in grain yield as a result of growing DMR maize varieties (Fig. 4). In addition, 29% of participating farmers and 32% of nonparticipating farmers observed a small increase in yield of DMR maize over their own variety. Taken together, we can infer that the use of DMR maize varieties resulted in higher yields for 95% of the farmers. Increased grain yield is normally expected to translate to proportionate increase in income. Over 90% of the farmers in our study indicated that they had increased income from producing DMR maize varieties (Fig. 5). Responses of the farmers showed that they expect that this performance would be sustained due to the excellent performance of the DMR maize varieties in the field.

100% 90% 80%

Farmers

70% 60%

Others Less severe Disappeared

50% 40% 30% 20% 10% 0% Participating

Non-participating Farmer type

Figure 3. Perception of the current level of downy mildew incidence by participating and non-participating farmers in a DMR maize variety intervention in Ogbomoso, Nigeria, 2000.

513

70

Big increase

60

Small increase Others

% Farmers

50 40 30 20 10 0 Participating

Non-participating Farmer types

Figure 4. Effect of DMR varieties on maize grain yield as perceived by participating and non-participating farmers in a DMR maize variety intervention in Ogbomoso, Nigeria, 2000.

70 Big increase Small increase

60

Others

% Farmers

50 40 30 20 10 0 Participating

Non-participating Farmer types

Figure 5. Effect of DMR varieties on income from maize production as perceived by participating and non-participating farmers in a DMR maize variety intervention in Ogbomoso, Nigeria, 2000.

514

The weaver bird attack on maize just before the harvest and the fall in market prices experienced by farmers reduced the expected income. That notwithstanding, the farmers attributed an estimated 74% of total maize income to the DMR maize. The fall in market prices du e probably to maize surplus could be considered a bonus to consumers. Therefore, while farmers gained by selling more maize, consumers gained by buying at cheaper prices. The uses made of the additional income by the farmers from growing DMR maize are summarized in Fig. 6. Although DMR maize varieties were planted to a relatively small portion of their farms, income generated therefrom was an important source of financing development activities and improving livelihood in the community. A large proportion of the farmers (39% of participating and 54.8% of non-participating farmers) reported that the additional income was used for human capital development such as school fees and health care. The second major use for the additional proceeds was on investment in farming, either to pay short term investment in operational costs or long term investments on new capital items or improvement in material lifestyle such as building or repairing a house and wedding ceremony.

% Farmers

100%

80%

Others

60%

Domestic expenses Farm investment, long term Farm investment, short term Human capital

40%

20%

0% Participating

Non-participating Farmer types

Figure 6. Uses made of additional income by participating and nonpaticipating farmers in a DMR maize variety intervention in Ogbomoso, Nigeria, 2000.

515

The level of diffusion of the DMR maize from farmer to farmer appeared to be low. About 34% of the farmer received a request for seeds while 25% gave out only about 0.52 “kongo” (0.75 kg) of maize seeds on the average. Discussion with the farmers suggests that the low level of farmer-to-farmer diffusion may have resulted from poor extension messages to farmers from the project team. The project team failed to encourage participating farmers to give out seeds to other farmers.

Credit Extension In most developing countries, the supply, distribution and credit functions are separated from extension activities. In Nigeria, the development bureaucracies of the past have recommended that a commercial officer undertake input supply functions. In order to give participating farmers ready access to key components of the recommended package, the project took on the responsibility of supplying farmers with necessary inputs to grow the DMR maize. The farmers were allowed to pay for the supply after they had harvested and marketted their grains. A project account was maintained for each participant and the technical support staff members were responsible for loan recovery in addition to other administrative duties. All loans to the participating farmers were fully recovered except from a farmer that died in 1998 and another that disappeared in 1999. This strategy made it easier for farmers to have timely access to inputs.

Participation in field days Field days were held to showcase DMR varieties. Both participating and interested non-participating farmers attended the field days as well as inoculation of the downy mildew at night where DM transmission was demonstrated. The field days further provided farmers the opportunity to ask questions on several issues, ranging from seed preservation to shortages of fertilizers. They also provided opportunities for policy makers, traditional rulers and opinion leaders, including Local Government Chairmen, to interact with the farmers.

Indirect benefits Several other indirect benefits, difficult to quantify, were achieved in the course of the project, including the following. i. Impact on the non-participating farmers who were secretly observing what was going on in the fields of participating farmers. The relatively large number of farmers that participated in the second and third years of the project is an indication that the interest of many of the non -participating farmers in the project was stimulated.

516

ii. The field schools and the farmer–to-farmer training facilitated by the project are also indirect benefits. Field schools were an opportunity for farmers to exchange their experiences, pr oblems and plans for the future. iii. Farmer-extensionist linkages were also strengthened. The involvement of local extension services in the implementation of the project substantially changed the farmers’ attitude toward the extension services. As the project progressed, farmers became more receptive to the extension agents and even commenced visits to the extension offices nearest to them. At a higher level, the State Government also modified her attitude and became more interested in allocating necessary inputs to the participating farmers. Before the project started, none of the project farmers had access to chemical inputs, such as fertilizers, distributed by the State Government. Execution of the project has now put in place a means for the State Ministry of Agriculture to make fertilizers available to farmers in the project area. The project also facilitated interaction between the farmers and the Local Government authorities. iv. The project staff gained considerable experience in the art of provi ding catalytic stimulation and assistance to the rural farmers for the generation and development of participatory approaches to development. Individual staff members were actively involved in the diagnosis of, and proffering solutions to, farmers’ problems. This way, they became an effective source of help and information to the farmers as well as an essential intermediary in dealing with the farmers.

Lessons learnt The implementation of this project in southwest Nigeria has shown that transformation of rural population is possible provided there is a conscious effort to do so in line with certain minimum conditions. i.

Since technology dissemination is a multiple dimension process, there must be institutional collaboration and cooperation among scientists. In the Ogbomoso case, the commitment and cooperation among intervention agencies progressively provided the impetus for the implementation of the strategy. This commitment was not a monopoly on the part of the scientists, farmers also showed considerable commitment and cooperation in the implementation of the project. Despite the fact that rather disappointing outcomes surfaced in the course of implementation of the project, farmers were consistent in their support for the project. This mutual reinforcement is absolutely essential to the success of this type of project.

517

ii.

For grassroot mobilisation and participation to be effective, the processes of discussion and decision making must accord maximum recognition to the farmers. Meetings have to be held frequently to make the leadership accountable to the led (the farmers) and raise their level of understanding (technical and administrative).

iii.

The feeling of despondency emanating from more conscious farmers at the end of the project suggest that institutional support of some sort is highly essential.

Based on our experience and the success of the deployment efforts at Ogbomoso, we suggest that technology deployment should proceed with cautious optimism. We recommend that technology dissemination efforts should begin with a few farmers and increased progressively to cover more farmers. Dowswell (1993 p.69) indicated this was the most notable mistake in the management of SG 2000 projects in Ghana where the field program was expanded from 17 000 demonstration plots in 1988 to nearly 80 000 in 1989. The expanded program became too large for extension managers and frontline staff to handle, leading to a decline in the quality of demonstrations (manifested in lower yields) and a precipitous drop in loan recovery. The third lesson from the project is that a minimum period of time is required to improve skills of farmers and build confidence among partners. In this case, a minimum of three years was needed to achieve good results. An abrupt ending of the project after a year or two could hardly have produced the kind of result obtained. Donors and private agencies must learn to support development strategies over the period of time required for the efforts to yield fruits. Finally our study provided further evidence to strengthen the current discussion among researchers that farmers want to be empowered instead of remaining on the dole. Farmers recognize farming as a business and will support any technology that contributes to their well being. They also go through a learning process and make attitudinal adjustments in line with what they have been exposed to and accepted.

Conclusion An attempt has been made in broad terms to describe the main elements in the deployment of DMR maize varieties to selected farm communities in the Ogbomoso area of southwest Nigeria. The project focussed on institutional cooperation, farmer participation, and farmer empowerment as necessary conditions for participatory and self-reliant pattern of technology dissemination. The approach was facilitated by a review of the traditional extension system, World Bank T & V system, FAO field schools and minikit demonstration. The overall thrust is to

518

stimulate the involvement of farmers themselves in participatory technology deployment programs built on solid foundation of research which, in turn, leads to spontaneous process of group information and improvement in their social and economic conditions. Although the village structure already provides the framework for functioning of this strategy, farmers were apprehensive of the prospect for increased loan default once scientists pull out of project administration leading to a feeling of despondency. Therefore, despite the fact that the results suggest that this approach is quite promising, definitive conclusions on the success of the approach depend on the ability of farmers themselves to sustain the system. The entire process represents a new approach to technology dissemination although we make no claim of absolute truth. We suggest a repetition of the approach to rapid deployment of technology used in this paper elsewhere. The approach could also be attractive to both donors and recipients since it enables high profile projects to be initiated rapidly and backed up by costbenefit analysis for quantified inputs and outputs.

References Adekunle, A.A., M.O. Adekunle, and D. Fielding. 2001. The Training and Visit system of extension and the challenge of commercialization of extension services. Pages 345–357 in B. Badu -Apraku, M.A.B. Fakorede , M ouedrago, and R.J. Carsky (eds.) Impact, challenges and prospect of maize research and development in West and Central Africa. Proceeding of a Regional Maize Workshop, IITA-Cotonou, Benin Republic, 4– 7 May 1999, WECAMAN/IITA. Byerlee, D., and P. Heise y. 1993. Strategies for technical change in small-farm agriculture, with particular reference to subSaharan Africa. In N.C. Russel, and C.R. Dowswell (eds.) Policy options for agricultural development in subSaharan Africa. Mexico, D.F.: CASIN/SAA/Global 2000. Cleaver, K. 1993. Making agricultural extension work in Africa. In N.C. Russel, and C.R. Dowswell (eds.) Policy options for agricultural development in sub-Saharan Africa. Mexico, D.F.: CASIN/SAA/Global 2000. Dowswell, C.R. 1993. Achieving more effective transfer of crop technology in sub-Saharan Africa. In N.C. Russel, and C.R. Dowswell (eds.) Policy options for agricultural development in sub-Saharan Africa. Mexico, D.F.: CASIN/SAA/Global 2000. Tebicke, H.L., and T. Berhan. 1983. People’s science, appropriate technology and R & D systems in rural settings. In Innovative

519

approaches to rural development. Gris Policy Dialogues, Rome 24–26 June, 1983. 275pp. US Congress, Office of Technology Assessment. 1988. Enhancing agriculture in Africa: A role for US Development Assistance. OTA-F-356 Washington, DC, U.S. Government Printing Office, September, 1988.

520

Enhancing the capacity of National Agricultural Systems for maize research and development in West and Central Africa: Accomplishments of WECAMAN and planned activities for the immediate future B. Badu-Apraku1, M.A.B. Fakorede2, S.O. Ajala3, R. Obubo3 and C. Okafor3 1

IITA, Bouaké, Côte d’Ivoire Obafemi Awolowo University, Ile-Ife, Nigeria 3 IITA, Ibadan, Nigeria

2

Abstract One of the main objectives of the West and Central Africa Collaborative Maize Research Network (WECAMAN) is to enhance the research capacity and capability of the participating national research systems (NARS). The strategy adopted to achieve this objective includes training courses and workshops, research project development and implementation, monitoring tours, consultation visits and a visiting scientist scheme. These activities were the primary skill acquisition methods for the NARS scientists and technicians. The Network has used human resources and facilities available in the subregion to conduct all training courses and workshops. A total of 46 technicians and about 150 scientists have participated in various training courses and workshops organized by the Network. Analysis of participants’ evaluation of the courses and workshops showed that they were effective means of skill acquisition and the technicians and scientists have benefited from the training courses. The Network will henceforth explore the possibility of collaborating with universities in the subregion to train candidates at the M.Sc and Ph.D levels in biotechnology, information technology, geographic information systems (GIS) and computer simulation modeling. Résumé Un des principaux objectifs du réseau de recherche collaborative sur le maïs en Afrique de l`ouest et du centre (WECAMAN) est de renforcer les capacités de recherche et les capacités des systèmes nationaux de recherche (SNRA) des pays membres. La stratégie utilisée pour atteindre ces objectifs inclus les cours de formation, les ateliers, l`initiation et la mise en oeuvre des projets de recherche, les évaluations, les consultations, et les visites d’échanges entre chercheurs. Ces activités ont été les premières acquisitions pour les chercheurs et techniciens des SNRA. Le réseau a utilisé des ressources humaines et des infrastructures disponibles dans la sous région pour réaliser toutes les formations et les ateliers. Un total de

521

46 techniciens et près de 150 chercheurs ont participé dans les différentes formations et ateliers organisés par le réseau. L’analyse de l’évaluation des cours et des ateliers par les participants a montré que les moyens et méthodes utilisés ont été efficaces et les techniciens et chercheurs ont bénéficié des formations. Le réseau explorera la possibilité de collaborer avec des universités dans la sous région pour former des candidats au niveau Msc et PhD en biotechnologie, en technologie de l’information, en système d’information géographique (SIG) et en modèle de simulation informatique.

Introduction One of the major constraints to maize research and development in West and Central Africa (WCA) is inadequate number of welltrained and skilled research scientists and research support staff. This situation often leads to erroneous field designs, poor data collection, spurious results and wasteful spending of research funds. To conduct sustained and well-focused research that would revolutionize maize production in WCA, it is expedient that researchers and technicians are trained from time to time in order to update and sharpen their skills and knowledge. The scientists need to acquire the necessary skills to apply the current internationally accepted research methodologies, such as information technology and biotechnology in their work. One of the objectives of the West and Central Africa Collaborative Maize Research Network (WECAMAN) is to enhance the research capacity and capability of the participating national research systems (NARS). One of the strategies adopted to achieve this objective is the execution of training courses and workshops. Others are: ?? Provision of financial assistance for the establishment of new, or improvement of existing, research infrastructure; ?? Research project development and implementation; ?? Monitoring tours and consultation visits; ?? Visiting Scientist Scheme. Training courses, workshops, research project development and implementation, monitoring tours, and consultation visits were executed in such a manner that they were the primary methods used to acquire skills by the NARS scientists and technicians of the Network. With the introduction of the African Maize Stress (AMS) Project in the Network in 1998, visiting scientist positions have been established as an additional means of acquiring skills. An overview of these activities have been reported in greater detail elsewhere (Badu-Apraku et al. 2002). We present herein a brief overview of the training activities with emphasis on the results of the preliminary analysis of the evaluation of the training courses

522

and workshops by the participants. Planned or proposed training activities for the early part of the 21st Century are also discussed.

Methodology Identification of training needs The Network has used several methods to identify training needs, including: (i) structured and unstructured questionnaires as well as observations made at the initial stages of SAFGRAD/WECAMAN by the more experienced international scientists; (ii) the quality of research proposals and reports submitted by scientists to the Network for funding, (iii) the quality of papers presented at conferences and workshops, (iv) monitoring tours and (v) consultation visits. The choice of subjects for training courses and workshops has been based primarily on NARS requests. The Network also established a Working Group in 1995 whose terms of reference included identification of training needs for the NARS scientists. This Working Group carried out the assignment in 1995 and at every biennial workshop of the Network thereafter; that is 1997, 1999, and 2001 (Ba du-Apraku et al. 1997; 1999; 2001). Additionally, consultants to the Network have made recommendations on training needs of the scientists.

Training facilities and resource persons In most cases, the training activities were held at IITA, Ibadan; IITA-Cotonou, Benin Republic; or the Crops Research Institute (CRI), Kumasi, Ghana. Training workshops were also held at the Agricultural Research Management Training Institute (ARMTI), Ilorin, Nigeria and the Institute of Renewable Resources of Kwame Nkrumah University of Science and Technology, Kumasi, Ghana. The factors considered by the Network in choosing the locations for training activities include (i) halls large enough to accommodate all participants during plenary sessions, (ii) availability of parallel bilingual (English and French) interpretation, (iii) rooms for working group sessions, (iv) secretarial and reprographic services, and (v) audio-visual aids. Resource persons for all training workshops were drawn from IITA, WARDA and the NARS in WCA.

Technicians Training Course The Technician Training Course was initiated in 1988 at Ouagadougou, Burkina Faso, but was later moved to

523

Ferkessedougou, Côte d’Ivoire. The course covers a 5-month period, usually from July to November, that corresponds to the cropping season, to expose the participants to the latest developments in maize production research throughout a full planting season. The course is practical-oriented and the topics covered are identification, layout and implementation of the various expe rimental designs, data collection and analysis, seed production, and report writing. Apart from the lectures and practical classes, each trainee is assigned a project to manage from planting to harvesting. The trainee collects and analyses the data, summarizes results, writes a project report and submits bound copies of the reports to the Network Coordinator, who is the Principal Resource Person. So far, 46 technicians have participated in the course (Table 1).

Scientists Training Courses and Workshops A total of 148 national scientists and extensionists have participated in the various courses and workshops conducted by the Network (Table 2). Table 1. Year, number of participants and countries represented at the WECAMAN’s Technician Training Course, 1988-99. Year 1988

No of participants 6

1989 1990

3 6

1994

8

1996

9

1998

7

1999

7

Total

46

Countries represented Central Africa Republic, Guinea-Conakry, Tchad, Benin, Burkina Faso, Mali Ghana, Guinea-Bissau, Tchad Benin, Burkina Faso, Cameroon, Ghana, Mali, Nigeria Ghana, Nigeria, Cameroon, Benin, Togo, Mali, Côte d’Ivoire, Burkina Faso Burkina Faso, Nigeria, Mali, Togo, Ghana, Côte d’Ivoire, Cameroon, Benin, Kenya Ghana, Togo, Cameroon, Mali, Tchad, Senegal, Guinea Benin, Nigeria, Senegal, Tchad, Burkina Fas o, Guinea, Côte d’Ivoire

Seed Production Course. Most of the WCA countries lack effective, efficient and well-organized seed production, certification and distribution system. This has greatly hampered the adoption of improved maize by farmers in the subregion. Apart from Nigeria and, to a much-limited extent, Ghana, none of the countries in the region has organized seed industries of any kind. The Network, in collaboration with the Group Training Unit of IITA and the Crops Research Institute of Ghana organized a seed production course in Ghana, 14–25 August 1995. The objective of the course was to strengthen the capacity of participants to organize and manage seed production. There were 27 participants

524

from Mali, Nigeria, Togo, Sierra Leone, Cote d’Ivoire, Guinea, Ghana, Burkina Faso, Benin, Cameroon, Kenya and Uganda. Participants reviewed and analyzed the constraints and opportunities of the traditional seed production for their communities, and discussed how to alleviate the constraints of the traditional seed production systems. They discussed how to design and manage seed production systems and design and execute activities for improved communication among seed users and seed producers. Table 2. Title, year, duration and number of participants in WECAMAN training courses for scientists, 1991–2000

Course title Seminar for research agronomists Seed Production Striga control and technology transfer Preparation of extension materials Advanced statistical computing Farmer participatory methods of on-farm testing and variety evaluation Workshop on maize quality, processing and utilization Breeding for stress tolerance in maize Workshop on breeding Striga resistance in cereals and seminar on marker assisted breeding Advanced statistical computing course for breeders and agronomists Impact assessment of maize stress management technologies Total

Year 1991

Duration 2 weeks

No of Participants 11

1995 1995

2 weeks 2 weeks

27 8

1996

1 week

15

1996

2 weeks

17

1998

2 weeks

11

1998

2 weeks

13

1999

1 week

13

1999

1 week

10

2000

2 weeks

14

2000

2 weeks

9 148

The Training, Communication, and Publications Unit of CRI in collaboration with the resource persons produced eight training manuals used for the course. These were supplemented with handouts and training materials from IITA and CIMMYT. The course presentations were in the form of lectures, demonstrations, practical sessions and group assignments. The participants were given the opportunity to make presentations on the organization of seed production in their respective countries. About 40% of the total course time was devoted to hands-on practicals in the field, including informal surveys in farmers’

525

fields. The participants had the opportunity to work in groups and to transform the acquired skills into farmer messages which were later used in a field day organized at Asuyeoboah and Kwadaso near Kumasi. Workshop on production of extension materials. Another constraint to the rapid adoption of new technologies in the subregion is the non -availability of user-friendly information. In an effort to reduce this problem, a workshop was held 9–13 October 1995 in collaboration with the Training Materials Unit of IITA and CRI of Ghana. Resource persons from IITA Training Materials Unit and the Training, Communication and Publications Unit of CRI delivered the lectures and conducted the practical classes. By the end of the course, the participants had produced and tested prototypes of four extension materials on maize production. The participants were tasked to adapt the extension materials produced during the course to their own environments and submit them to the Coordinator for editing and publication. Following this, extension manuals that characterize the released maize varieties in Benin, Cameroon, Mali and Togo were developed and distributed to extensionists and farmers of the respective countries. Workshop on Striga control and technology transfer. Striga, a parasitic weed, has been a major constraint to maize production in the savannas of WCA. Progress had been made at IITA in breeding for host-plant resistance and integrated control. Also, new screening methodologies had been developed. It was necessary to familiarize WECAMAN scientists with the new methodologies. WECAMAN, therefore, organized a special workshop on Striga control and technology transfer for member countries. The objective of the workshop was to provide NAR scientists practical experience in the analysis of Striga problems, research on control techniques, and the dissemination of information and transfer of available technologies. All member countries of WECAMAN were represented. The workshop was organized in two phases: ?? 3-8 October, 1995 – Striga control – IITA, Ibadan ?? 9-14 October, 1995 – Technology transfer – IITA, Cotonou. The workshop had to be held this way because Striga research and control facilities are available only at IITA, Ibadan. Resource persons were drawn principally from IITA. The accelerated participatory research method was used to study the problem of technology transfer at various levels; i.e., the village, the farm and research plot levels. A field trip was organized to Canakphi village in the Bohicon District of the

526

Republic of Benin to identify farmers’ problem and to design a test protocol. One of the topics discussed during the workshop was the different communication technics used in the field for training and extension in order to ensure accurate and effective transmission of information. It was emphasized that it is necessary to have good communication skills for ease of dialogue with farmers and also to forge collaborative links between research and extension. Advanced statistical computing courses for breeders and agronomists. This course was organized in response to the recommendation of the Working Group set up by WECAMAN in 1995. The working group had recommended that the Network, in collaboration with other bodies must, as a matter of urgency, organize a short-term training course to update the knowledge of scientists in the design, analysis and interpretation of experiments. The Steering Committee of WECAMAN endorsed this recommendation and the course was organized for breeders and agronomists, 11-21 June, 1996 in collaboration with the Group Training and Biometrics Units of IITA from where the resource persons were drawn. The Agricultural and Rural Management Training Institute (ARMTI), Ilorin, Nigeria hosted the training course. The course was designed to upgrade the skills of national scientists in the analysis and interpretation of complex statistical data, selection of appropriate statistical methods for data analysis and the use of statistical packages especially GENSTAT. A total of 14 participants from Network member countries, two from the Sorghum Network and one from the IITA liaison office in Ghana attended the course. A similar course was organized and run at IITA-Ibadan, 10 -21 April, 2000. Fourteen scientists from Senegal, Ghana, Benin, Togo, Cameroon, Nigeria, Tchad, Mali, Guinea and Côte d’Ivoire attended the course. Workshop on farmer participatory methods of on -farm testing and evaluation of varieties. The collaboration of WECAMAN with NARS in WCA has resulted in the development of new streak resistant, drought tolerant and high yielding maize varieties with different maturity, grain color, grain texture and response to agronomic practices (rate and time of fertilizer application, population density, etc). However, the adoption of these varieties has not been as high as desirable. In an attempt to alleviate the constraints of technology transfer, WECAMAN is presently organizing a number of activities to provide the fora for reviewing research findings, grower recommendations, and agricultural policies. The measures include:

527

?? strengthening of research –extension–farmer linkages in member countries; ?? organizing workshops on technology transfer; ?? encouraging and assisting member countries to organize annual maize workshops and research planning sessions for researchers, extensionists, policy makers and farmers. Inadequate on-farm testing of WECAMAN technologies has also been identified as one of the major constraints to the adoption of available improved maize technologies. There are a wide variety of methods used for on-farm evaluation of te chnologies. These range from researcher-managed trials, to demonstration of technologies. The level of involvement of farmers varies from country to country. Researchers need to have a continuous dialogue with farmers. We also identified a need for WECAMAN to modify or change the on -farm research approach being used. Farmer participatory approach is an attractive option. For example, farmers may be invited to research stations to evaluate appropriate germplasm and agronomic practices and to select those for testing on their farms. Breeders can learn from farmers’ variety selection and evaluation criteria. WECAMAN Steering Committee appointed a sub-committee in 1996 to review and develop strategies for the promotion and adoption of improved maize technologies. The sub-committee strongly recommended that a workshop on farmer participatory methods of on-farm testing and evaluation of varieties be organized for researchers within WECAMAN. To this end, a workshop on farmer participatory methods of onfarm testing and varietal evaluation was organized by WECAMAN in collaboration with IITA and the Institute of Renewable Resources of Kwame Nkrumah University of Science and Technology, Kumasi, Ghana, 17–28 August, 1998. Eleven agronomists and breeders from Togo, Burkina Faso, Guinea, Ghana, Tchad, Benin, Senegal and Nigeria participated in the workshop. The areas covered in the course included (i) the use of selected participatory rapid appraisal (PRA) methods to communicate with farmers to identify their constraints and opportunities, (ii) use of selected PRA methods to identify and prioritize research and extension themes, and (iii) the application of appropriate PRA tools to analyze problems at different stages of the research process. The resource persons for the course were drawn from IITA and the Republic of Benin. Workshop on maize quality, processing and utilization. An important area in which remarkable progress has been made by WECAMAN is the development of maize varieties with good

528

adaptation to bio-climatic conditions of the semi -arid ecologies and resistance/tolerance to major diseases. Also, improved agronomic practices have been developed for the varieties. The effort has significantly increased maize yields on a sustainable basis. Despite the significant progress that has been made through plant breeding in improving yield, disease and insect resistance of maize varieties, and other improved agronomic practices for farmers, the adoption of the improved varieties by farmers is often not as high as desirable. This is partly due to the fact that some of the improved varieties lack the desired quality for processing. Information regarding the characteristics of maize required for specific end-users is not readily available. The lack of information on the processing and utilization of maize is a constraint to the effective marketing and increased production of maize in the sub-region. Lack of appropriate post-harvest technologies is considered as a major constraint to the production of maize and is noted as an area that needs urgent attention by WECAMAN. So far, the Network is funding a few projects on post-harvest technologies. There is a need for more work on storage, utilization, and nutrition. There are on -going research and development activities already in some member countries that the Network can link up with. In Ghana, Global 2000 is assisting farmers with credit from the Agricultural Development Bank, for the construction of maize cribs. The World Bank and USAID are funding projects on maize transfor mation and ways of improving maize storage in Mali. It would be appropriate for the Network to link up with these initiatives in order to stimulate dissemination of appropriate postharvest technologies. In some Network member countries, there is lack of information on available appropriate post-harvest technologies. There is, therefore, a need for the exchange of information among the member countries on available methods of processing and utilization. In addition, there is a need for training of national program scientists on the development of appropriate maize storage, processing and utilization technologies. Technology transfer has been a weak link in WECAMAN and this is now being addressed through the formation of partnerships among research, technology transfer entities and the private sector. Progress has been made but there is still a lot that can be done in the technology transfer arena. One of the things that has really been lacking, until recently, is an effort to develop partnerships with food technology institutes and food processors to increase the nutritional value of maize products. A high quality protein maize (containing high levels of lysine and tryptophan) has been developed in Ghana and has been used to feed school children and swine. Nutritional results from feeding this QPM have been very encouraging. WECAMAN is distributing seed of this QPM to other Network countries in WCA. The high quality protein maize could be taken a step further and fortified with other mineral, especially micronutrients.

529

WECAMAN would like to involve different partners that would have a major role to play in improving the nutritional status of the region’s children and lactating mothers. WECAMAN would like to bring together agricultural researchers, extensionists, food technology institutes, producers, etc. to work together to produce more food and provide maize products that will improve the nutritional status of the people of the sub-region. In an effort to achieve these goals, a workshop on maize quality, processing and utilization was organized by WECAMAN in collaboration with IITA and the university of Benin at IITAIbadan, 24 August to 4 September 1998. The workshop provided a forum for the exchange of information among the WECAMAN member countries on avai lable improved methods of processing and utilization and offered the national program scientists an opportunity for training in the development of appropriate maize storage, processing and utilization technologies. Furthermore, it emphasized on the promotion of improved post-harvest quality with emphasis on mycotoxins, processing and utilization in order to increase the nutritional standards. Thirteen maize utilization specialists from Senegal, Benin, Guinea, Ghana, Togo, Burkina Faso, Nigeria, Mali, Cameroon, Cote d’Ivoire and Tchad attended the workshop. Resource persons were drawn from IITA and the University of Benin. During the Workshop, Dr Kaml Hyder, a consultant from Winrock International who was visiting selected countries of West Africa in connection with the proposed Micro-nutrient Fortification Project in West Africa presented a paper entitled “Micro-nutrient malnutrition and how to combat it”. Workshop on impact assessment of maize stress management technologies. A workshop on impact assessment of maize stress management technologies for scientists from WECAMAN member countries was held in IITA-Ibadan, Nigeria, 5–16 June 2000. Funded by the AMS Project, the workshop was aimed at equipping the NARS scientists with knowledge to measure the returns to investments in research, to set priorities, to improve program design and to influence agricultural policy in the respective countries. Nine participants from Benin, Cameroon, Togo, Burkina Faso, Ghana, Guinea, Mali, Senegal, and Tchad attended the workshop.

Participants’ Workshops

evaluation

of

training

courses

and

Some important unique features of all WECAMAN training courses and workshops are: ?? Pre- and post-course evaluation of the participants;

530

?? Daily and overall course evaluation by the partici pants Pre- and post-course evaluation of the participants. During the Pre-Course Workshop, the resource persons prepare a set of questions based on the materials to be covered in the course. The questions are administered as pre-course evaluation test. The same set of questions is again given to the participants at the end of the course as post-course evaluation test. The difference in the scores of participants is an indication of the amount of course material that was disseminated to the participants. Usually, participants score very low grades in the pre-course evaluation tests and perform very well in the post-course tests. Daily and overall course evaluation by the participants. A daily activity evaluation form was given to the participants and collected at the end of the day. The participants were requested to take a few minutes after each activity to assess the activity. Assessment was based on the following six criteria: ?? ?? ?? ?? ?? ??

Knowledge acquired – much, some, none? Usefulness/relevance – very, partial, none? Depth of coverage – too deep, appropriate, superficial? Presentation – good, satisfactory, poor? Training material – good, satisfactory, poor? Time allotment – too long, good, too short?

The daily activity evaluation form also provided space for comments on each activity for the day and on the day as a whole. Additionally, at the end of each day, the course coordinator held a brief meeting with all resource persons to review the activities of the day. The purpose was to strengthen whatever was done satisfactorily and improve areas of poor performance. On the last day of activities, the participants were given a terminal course evaluation questionnaire. Participants were to complete the questionnaire at the end of the very last activity of the training course. The questionnaire contained seven sections: ?? ?? ?? ?? ?? ?? ??

Overall assessment of the course; Course objectives; Course design/delivery; Supporting facilities and arrangements; Performance of Resource Persons; Course evaluation procedure; General comments.

Data summarization. Completed questionnaires for the overall course evaluation by participants were available for the 1994 Technician course, 1995 Seed Production course, 1995 Workshop on Preparation of Extension Training Materials and the 1996 Advanced Statistical Computing Course. The completed forms for

531

all of these courses were summarized as frequency counts and the results are discussed herein.

Results Attitudinal changes resulting from some of the training courses and workshops are summarized in Table 3. Table 3. Attitudinal changes of NARS resulting from WECAMAN training courses and workshops in WCA. Course title Technicians’ Training Course

Seed Production

Attitudinal changes ?? Improvement of

data collaborating NARS,

recovery

from

?? Improvement in the data collected from the Regional Trials, ?? Improvement in the management of research plots reported by monitoring and consultation visit teams Following the course, the participants have played leading roles in the establishment of community seed production schemes in the respective WECAMAN member countries. Participants have trained over 1,000 farmers in the techniques of seed production and maintenance of large quantities of breeder, foundation and commercial seed of released varieties and promoted the diffusion and utilization of improved maize varieties (BaduApraku et al., 1999b; Badu-Apraku and Fakorede, 2002)

Preparation of extension materials

Extension bulletins which characterize the available maize varieties in Benin, Togo, Cameroon, Burkina Faso, Mali and Ghana have been published with financial support of WECAMAN and are in use by farmers and extensionists in these countries.

Advanced statistical computing

?? Improvement of capabilities in data analysis as shown by the quality and number of technical reports and papers presented during the Regional Maize Workshops, ?? Good presentations of scientific papers at conferences, ?? Preparation of well structured, focused and convincing research proposals,

Farmer participatory methods of on-farm testing and variety evaluation

Cameroon, Nigeria and Togo have adopted farmer participatory methods of on-farm testing and variety evaluation

Course evaluation by participants. Participants of the various training courses and workshops indicated that the courses were

532

satisfactory (23%) or very good/excellent (77%). The quality of instruction and the planning and management of the courses were also rated highly by the participants (Table 4). A rather high proportion (41%) of the participants felt that the objectives of the courses were only partially met, but most (81%) agreed that the courses were of professional benefit to them (Table 5). Most of the questions asked on course design and delivery received high, positive rating by the participants (Table 6). Table 4. Overall course assessment by participants in WECAMAN courses. Poor

General assessment of the course as a whole (n = 61) Quality of instruction (n = 64) Planning and management (n = 62)

Satisfactory/ Good/ Fair Excellent ………..…% respondent ………… 0 23.0 77.0

3.1

28.1

68.8

6.4

35.5

58.1

Table 5. Extent to which WECAMAN course objectives were met and the benefits of the courses as rated by participants. No Course objectives met? (n=246) Professional benefit to participant (n=62)

Partially

Fully

………..…% respondent ………… 3.7 41.1 55.3 1.6

17.7

80.6

Table 6. WECAMAN Course design and delivery as rated by course participants.

Criterion 1. Knowledge acquired (n=63) 2. Relevance of topics (n=62) 3. Depth of coverage of subject matter (n=65) 4. Integration of lectures with practicals (n=61) 5. Class discussion periods (n=37) 6. Presentation of topics (n=62) 7. Usefulness of field trips (n=45) 8. Size of group (n=23) 9. Overall quality of training Materials (n=51) 10. Course duration (n=62)

Poor

Satisfactory

0 1.6

12.7 19.4

Good/ Excellent 87.3 79.0

3.1

24.6

72.3

23.0 21.6 5.4 6.7 4.3

50.8 75.7 27.0 35.6 91.4

26.2 2.7 67.6 57.7 4.3

23.5 51.6 (two short)

39.2 38.7 (about right)

56.9 9.7 (too long)

However, integration of lectures with practicals, class discussion periods and quality of training materials need to be improved.

533

Similarly, more time may have to be allocated to practical/field trips and perhaps the materials covered in each course should be reduced (Table 7). About 52% of the participants felt that the duration of the courses was too short (Table 6). This may also mean that the participants enjoyed the courses so much that it appeared time went by too fast. About 60% of the resource persons performed excellently in course delivery (Table 8). Table 7. Views of WECAMAN course participants on time allotment for course activities. Course activity 1. Lectures (n=38) 2. Demonstrations (n=35) 3. Field trips/practicals (n=36) 4. Exercises/assignments

Short 21.0 28.6 40.0 22.2

Adequate 73.7 71.4 56.7 77.8

Too much 5.3 0 3.3 0

Table 8. Support facilities and arrangements of WECAMAN courses as rated by participants.

Course execution item 1. Performance of resource persons (n=63) 2. Notice about participation (n=22) 3. International travel arrangements (n=19) 4. Physical facilities - Classroom (n=38) - Laboratory (n=33) - Fields (n=8) 5. Supporting equipment (n=62) 6. Accommodation (n=59) 7. Coffee breaks (n=23) 8. Local transportation (n=61) 9. Meals (n=54) 10. Interpreters (n=14) 11. Social/recreational activities (n=38)

Poor 1.5

Satis factory 41.2

Good/ Excellent 57.3

18.2

40.9

40.9

10.5

31.6

57.9

10.5 9.1 62.5 8.1 18.6 4.3 26.2 24.1 0 52.6

21.0 18.2 25.0 25.8 40.7 60.9 21.3 42.6 21.4 23.7

68.5 72.7 12.5 66.1 40.7 34.8 52.5 33.3 78.6 23.7

Generally, course participants felt that the physical facilities were good or excellent (Table 7). However, field facilities, local transportation, meals and recreational facilities may need to be improved upon in future training courses. Not all resource persons were good or excellent either. One way to adequately prepare resource persons is to hold/adopt the training course workshop approach used by the IITA Training Program. In this workshop, resource persons and the course organizers interact to preview the objectives of the course and discuss how to meet the objectives. The deficiencies of the resources person in relation to the proposed course are corrected in the training workshop.

534

Finally, all participants in WECAMAN training courses would gladly recommend colleagues at their level (77%) or senior colleagues (21%) for the courses. Thus, one may conclude that the courses generally met the needs of the level of participants for whom they were intended.

Discussion WECAMAN/SAFGRAD started research training, but it was primarily limited to Technicians. The present WECAMAN has improved on this and has been training technicians, researchers/scientists, seed production staff and extensionists. A total of 148 scientists, nearly 20 scientists per member country, have been trained. If facilities were available for trained scientists to apply the skills they have acquired and/or train others, the research and extension capabilities of the national programs would be strengthened within a few years. Only a preliminary analysis and evaluation is presented here and the findings presented show that a full impact assessment of WECAMAN training courses is justified. However, observations such as, good presentations of scientific papers at conferences, well written research proposals, properly conducted research projects, and well written progress reports, seem to indicate that WCA Scientists have benefited from the training courses. The effort that goes into the Technician Training Course is quite substantial. Similarly, the Seed Production Course held in 1995 had quite a lot of input. Materials covered during each of these courses were much more than what would be covered in one or two semesters in a university degree program. Similarly, there is no university in the region that has the human resources and infrastructure put together to run each of these courses. Perhaps it would be more cost effective if the training courses are tied with some university degree program so those course credit units are awarded to the participants. The training provided through the courses and workshops have made a valuable contribution towards improved technical capacity of national scientists and te chnicians. During the period 1994-97, a total of 75 scientists from the 8 member countries actively participated in WECAMAN’s collaborative research projects while 105 scientists from eleven member countries participated in the projects during 1998-2000 (Fig. 1).

535

120 100 80 60 40 20 0

1987-1993 1994-1997 1998-2000

Scientist

Ph.D degrees

Figure1. Number of scientists involved in the collaborative research projects of WECAMAN and number with Ph.D. degrees in the 1987–1993, 1994–1997 and the 1998–2000 eras.

Thus the networking approach has made an impact by putting in place a team of well-trained researchers. The impact of the training courses and seminars on the research capability and capacity of NARS include: (i) (ii)

(iii) (iv) (v) (vi) (vii) (viii)

(ix)

Increased research output as evident from the number of collaborative research projects executed at the national level, Improvement of capabilities in data analysis as shown by the quality and number of technical reports and papers presented during the Regional Maize Workshops, Good presentations of scientific papers at conferences, Preparation of well structured, focused and convincing research proposals, Improvement of data recovery from collaborating NARS, Release of improved varieties and agronomic practices, Improved understanding of cropping systems and the need for adopting soil/water conservation, By pooling together research talents through WECAMAN, the NARS working on maize in WCA have been able to attain critical research mass at the regional level that has influenced agricultural development at the national levels, and Group training has helped to build interpersonal relationships and enhanced trust and confidence in exchange of information and maize germplasm.

Planned training activities of WECAMAN for the immediate future. WECAMAN considers training as one of the main activities for strengthening the rese arch capabilities of NARS. Towards this end, about 194 national scientists and technicians have been provided short-term training lasting from a few days to

536

6 months since 1988. However, the problem of shortage of trained manpower remains acute and widespread in many countries. Out of the 105 national scientists from eleven member countries who participated in the collaborative research projects of the 19982000 era, only 37 (about 34%) have Ph.D (Fig. 1). However, 31 out of the 75 scientists (41%) who participated in the projects during the 1984–97 era had Ph.D level training. Thus about 65% of the scientists currently involved in the WECAMAN activities are junior scientists who could benefit from post-graduate level training. Therefore, while the short-term training programs have been effective and useful in meeting national needs, there is a great desire among the NARS for long-term (degree) training. While admittedly, this is beyond the mandate and capacity of WECAMAN, it is believed that WECAMAN should facilitate such training through IITA and provide guidance with thesis research. WECAMAN member countries need to take full advantage of the IITA research facilities in Nigeria and Cote d’Ivoire for thesis research. While it is not considered appropriate that the Network should be used to fund long-term degree training, the Network could facilitate training by providing facilities and access to other scholarships from IITA and donors such as the European Union for long-term training. The technician training course has been organized by WECAMAN since 1988, and several technicians have been trained resulting in the improvement of capabilities and quality of data coming from the regional trials. There is now a need to shift emphasis to computer training of technicians so that they can effectively assist the national scientists in data analysis and interpretation. Table 9 presents an indication of planned in-service and training workshops for the immediate future. Five short-term training workshops lasting 1-2 weeks have been identified in problem areas for improvement. One important area that requires immediate attention is the training of maize breeders for the national maize programs. At present, apart from a few countries such as Nigeria, Ghana and Cameroon all the national maize breeding programs have only one breeder and only a few of them have Ph.D training. Therefore, several maize breeders in WECAMAN member countries would benefit from post-graduate level training. The future challenge of WECAMAN there fore would be to support long-term degree training for the national scientists by providing oversight to thesis research and fellowships and facilitating access to scholarships from donors such as the European Union.

537

Table 9. Number of workshops and training courses planned for WECAMAN, 2003–2005.

Type of training activity Seed Production course

2003

Technician training course Computer course for technicians Technology Transfer Scientific Monitoring Tour Regional Maize Workshop Visiting scientific in IITAIbadan Visiting scientific in IITACote d’Ivoire

x

x x

2004 x

2005

Approximate number of participants 22

x

16 22

x

x x

22 13 120 3

x

x

2

x x

There is a need for the training of national scientists in modeling, GIS and biotechnology, not necessarily for immediate application but to help them stay abreast of technological changes and relate their potential use for future research. The training program of the Network should also draw a wider range of partnerships such as private sector processors and universities.

References Badu-Apraku, B., and M.A.B. Fakorede. 2003. Promoting maize technology transfer in West and Central Africa: A networking approach. (In this Volume). Badu-Apraku B., M.O. Akoroda, M. Ouedraogo, and F.M. Quin. 1997. Contributing to food self sufficiency: maize research and development in West and Central Africa. Proceedings of a Regional Maize Workshop, 29 May–2 June 1995, Cotonou, Benin Republique. WECAMAN/IITA. Badu-Apraku, B., M. A. B. Fakorede, and S.O. Ajala. 2002. Enhancing human resources for maize research and development in West and Central Africa: A networking approach. Submitted to Journal of Natural Resources and Life Sciences Education (Madison, Wisconsin, USA). Badu-Apraku B., M.A.B. Fakorede, M. Ouedraogo, and F.M. Quin. 1999a. Strategy for sustainable maize production in West and Central Africa. Proceedings of a regional Maize Workshop, IITA-Cotonou, Benin Republic, 21–25 April, 1997. WECAMAN/IITA. Badu-Apraku B., B.I. Hema, C. Thé, N. Coulibaly and G. Mellon. 1999b. Making improved maize seed available to farmers in West and Central Africa–The contribution of WECAMAN. Pp 138–149 in B. Badu -Apraku, M.A.B. Fakorede, M. Ouedraogo, and F.M. Quin (eds.) Strategy for sustainable maize

538

production in West and Central Africa. Proceedings of a regional Maize Workshop, IITA-Cotonou, Benin Republic, 21– 25 April, 1997. WECAMAN/IITA. Badu-Apraku B., M.A.B. Fakorede, M. Ouedraogo, and R.J. Carsky, 2001. Impact, challenges and prospects of maize research and development in West and Central Africa. Proceedings of a Regional Maize Workshop, 4-7 May 1999, IITA-Cotonou, Benin Republic. WECAMAN/IITA.