Farmer perceptions, choice and adoption of soil management technologies in
maize-based farming systems of Malawi

Kabuli, A.M.¹ and M. Phiri¹

Keywords: Farmer adoption, new technologies, maximum livelihood estimation, participatory research methods

Introduction

Soil Fertility Initiatives in Malawi

Declining soil fertility ranks high among the factors limiting smallholder arable crop production in Malawi (Snapp et al., 1998). This problem has been caused by diminishing land holding sizes as a result of the ever-increasing population growth rate currently at 3.2%. The problem has of late been exacerbated by continuous cultivation on the same piece of land as well as cultivation of fragile hill slopes with little or no application of inorganic fertilizers due to escalating prices as a result of removal of input subsidies.

The Risk Management Project and the Soil Fertility Network for Southern Africa have been working in Central Malawi at Chisepo Extension Planning Area between 1999-2004 growing seasons in collaboration with Bunda College of Agriculture, a constituent college of the University of Malawi. The two projects aimed at developing sustainable resource-conserving farming methods in collaboration with smallholder farmers in Malawi by integrating modeling, economic analysis as well as farmer participatory research in identifying viable soil fertility technologies for small scale farmers.

Study Location

The study was conducted in Chisepo, Central Malawi, in 2002-2004 seasons. Chisepo is located in Kasungu Agricultural Development Division (KADD) North of Lilongwe, in the mid-altitude area of the Lilongwe-Kasungu plains. The most common soils are sandy loamy soils (60%) and sandy soils (30%). These soils are generally low in soil organic matter (0.25%), total nitrogen (0.1%) and range in pH from 5.6 to 5.8. Without soil fertility management technology interventions the soils barely produce adequate crops. The main crops grown in the area are maize and tobacco. Maize yields range from as low as 0.1 t ha^-1 to 2.5 t ha^-1. Annual rainfall is 600-800 mm (Kamanga 2002). Farmers produce a wide range of crops such as Maize (dominant food crop), tobacco (dominant cash crop), groundnuts, soyabeans, beans, cassava, sweet potato and vegetables.

Farmer experimentation, trial design and data collection

The Mother-Baby trial approach was used in the experimentation process. This allowed quantitative data from researcher managed on-farm ‘mother trials’ to be systematically cross-checked with farmer-managed ‘baby trials’ with similar themes. The mother trials tested different soil fertility management technologies mainly the incorporation of grain and green legumes and fertilizer management trials.

The mother fields were placed on sandy soils and red (katondo) soils. In each mother trial, farmer chosen legumes were put in intercrop with maize and as sole legumes planted at the same time. The mother trial was replicated three times. Farmers also chose legumes based on what they prefer. The legumes chosen were planted in the baby trials. In this case, farmers wanted all the ten legumes to be subjected to intercropping or rotation with maize. The legumes were mucuna, pigeon pea, bambara, soybean, groundnut, cowpea (Determinate and indeterminate), tephrosia, grahamiana and sunhemp. Each mother trial was set in a simple way intercropping maize with legumes and sole legumes. In the baby trials, each farmer chose four legumes from the mother trials to plant in their fields. Follow up was made to see how each farmer had planted the legumes where they planted them and why they did what they had done.

Two sets of data was collected, namely, agronomic data sets which included soil sampling for nitrogen status, texture, soil pH, phosphorous status; and crop performance as measured by grain yield and harvest indices. Socio-economic data which is the major focus of this report was also collected using farmer participatory methodologies (FPR) which included focus group discussions and pairwise ranking. Farmer groups averaging 20-25 members per group who participated in these on-farm farmer research trials were consulted. Additionally, a short structured questionnaire was administered to all the participating farmers (70 farmers in total). Information collected included farmer derived taxonomies for soil and fertility status, farmer socio-economic characteristics, perceptions of the various soil fertility technologies as well as the factors that constrained increased adoption of these technologies.

Results and discussions

Socio-economic characteristics of farmers in the studied area

The majority of the smallholder farmers working on the project were men (76%). Very few women farmers took part in the experimentation process and the main reason for this was lack of interest from the women farmers. Most of the women farmers were suspicious of the project at first, as they did not understand the benefits of hosting the trials. Most farmers were illiterate and did not use manure in their farming system because of a reduction in stock levels of late due to theft. Very few farmers also applied inorganic fertilizer in their farms mainly due to increased prices of the commodity as a result of the market liberalization and removal of subsidies. More socio-economic characteristics as reported by farmers are presented in Table 1 below.

Table 1. Selected Socio-economic Indicators of farmers in the study area

Soil characterization and fertility management in Malawi

Farmers classified soils based on what they saw and felt about that particular soil. There were certain inherent factors which farmers used to classify and characterize soil types. The most important ones were color, fertility, land type and depth of the soil. Characteristics such as slope, water holding capacity, ease of tilling, physical properties such as stickiness or firmness were also used. The actual fertility status of the soil was also determined by the location and previous soil management techniques that the farmer had been using previously on that piece of land.

The most common soils in the area were sandy clay loam soils (Katondo) and sandy soils (Mchenga). The low fertility of sandy soils was the major factor causing reduced biomass and crop production in the smallholder farming sector. Sandy soils inherently have low amounts of nutrients, low soil organic matter and are weakly structured (Grant, 1981). Farmers reported that most of the fields in the area are infertile due to continuous cropping, which had put pressure on land. However, the farmers were quick to suggest ways of managing and replenishing soil fertility in their fields such as application of compost manure, planting tree species and leguminous plants (e.g. Tephrosia vogelli), incorporating crop residues during land preparations, leaving fields to fallow as well as applying inorganic fertilizer.

Analysis of soil parameters showed no significant differences between sites, but indicated that in both sites there were low levels of SOM and soil N content (Table 2). The levels show that growing maize without fertilizer would be risky. Without alternative means of producing maize in such areas, more and more farmers’ especially female-headed households would be food insecure. At the same time, the fertility status may indicate upon further calculations the level of external inputs needed for smallholders to break even in food production.

Table 2. Initial soil (0-20 cm) fertility status of the Chisepo

Maize response to residual nitrogen

Figure 1 shows maize yields from the one of the mother trials conducted to investigate the effect of different legume incorporation on subsequent maize yields. Farmers planted MH 18 maize varieties in all the plots. The groundnut/pigeon pea intercrop followed by maize recorded the highest maize yields (7,615 kg ha^-1) followed by maize plus fertilizer plots (5,695 kg ha^-1), maize/pigeon pea (5,175 kg ha^-1), maize/tephrosia (5,035 kg ha^-1) and mucuna/maize rotation (4,750 kg ha^-1). Lowest yields came from the continuous maize plots (1,870 kg ha¹). The yield increments were expected since the plots had cumulative effect of residual nitrogen from legume systems planted in the 2001/2002 growing season.

Figure 1. Maize yield response to residual N in the mother trial plots in 2002/2003

Similarly, maize yield response to residual nitrogen in the baby trial plots managed by farmers showed that pigeon pea-based plots compared well with the maize plus fertilizer plots. For example, highest maize yields were obtained from maize/pigeon pea plots followed by maize/tephrosia plots. The groundnut/pigeon pea plots in the farmers’ trials performed the same way as the fertilized maize plots. The maize yields however, varied from one farmer to the other. These variations were expected since some farmers supplemented the residual nitrogen with fertilizer while other farmers did not. Four farmers planted tobacco in the baby trial plots. These farmers said that they wanted to take advantage of the legume residual nitrogen to improve their tobacco yields, which was their major cash crop. This showed that farmer had developing confidence in using legumes to improve their soil fertility even for cash crop production in the area.

Results from the intercrop systems showed that legumes such as pigeon pea, tephrosia, and groundnuts were very suitable for intercropping with maize as was indicated by the maize yields shown above. Since land holding sizes were smaller for most farmers (average 1.90 ha), having more legume best bets suitable for intercropping with maize would increase farmers’ options for improving soil fertility and maize yields.

Farmer perception of the soil fertility and food security benefits of the legumes

Farmers were asked to give reasons for growing a particular legume crop in 2003/2004 growing season. The majority of farmers (48.2%) indicated that they chose the legumes as a source of food followed by those who stated that they grew it to improve their soil fertility status. However, when asked how they knew the legume improved soil fertility, they indicated that extension staff employed by the project organized a field day where they saw higher maize yields in plots where previously there were certain legume crops. Most farmers grew cowpeas than any other legume (23.6%), followed by mucuna and pigeon peas both grown by 19.1% of the farmers (Table 3).

Table 3. Farmer knowledge and practices of improving soil fertility

Traditional legumes were mainly grown for their food values. Cowpeas leaves were eaten as vegetable relish while fresh; they were also boiled and dried for preservation for use in dry season when green vegetables were scarce. The fresh green pods were also boiled and eaten as snacks. Groundnuts and pigeon peas were roasted, salted and eaten also as a snack and in some instances pounded into flour and used to season leaf vegetables. Farmers also used legumes such as pigeon peas, soybean and mucuna as animal feed. Male farmers in particular stated that some legumes were fed to animals in order to boost milk production.

In terms of soil fertility potential of legume crops, farmers perceived that mucuna was the best for soil fertility improvement based on the leafy biomass yield they had in 2002/2003 season and 11.2% opted for growing mucuna for that reason. This was also reflected in the high rating mucuna had on contribution of legumes to soil fertility (Table 4).

Table 4. Farmer rating of soil improving legume techno­logy traits¹

Farmer perceptions on the use of Inorganic fertilizer

These technologies were mostly targeted for areas growing hybrid maize and high value crops to pay for the cost of the chemical fertilizers. While farmers perceived numerous benefits of using the inorganic fertilizers compared to the organic sources of fertilizers such as quick yield response (meeting food security needs), easy to apply and easy to access as most of the fertilizer is sold in many rural markets, the high cost of inorganic fertilizers and the introduction of the structural adjustment program have forced government and non-governmental institutions to explore alternative and economically feasible means of soil fertility improvement (Kanyama Phiri et al., 1998). In Malawi, such efforts include the development of area-specific recommendations for soil fertility improvement which include the evaluation of economic rates of fertilizer application. Most of the farmers involved in the survey expressed concern over the high cost of the chemical fertilizers in the country which is currently selling at MK 3,200.00 per 50 kg bag (approximately 32 US Dollar per 50 kg bag of Urea).

Farmer perceptions on the use animal and compost manure

Farmers with livestock (especially in the Northern Region of Malawi) use farmyard manure to increase soil fertility. The main advantages of this practice according to the farmers were that the effects were seen immediately they applied the manure. Other farmers also stated that manure making that used locally available materials, were cheaper than the inorganic fertilizers. Most of the farmers have the knowledge associated with their preparation. However, the major drawback was that quantities of farmyard manure were generally low and variable. This is due to livestock ownership, particularly cattle, being very low amongst the farmers (about 15% of the sampled households had livestock). In some parts of the country farmers composted household refuse near to the homestead which is used to fertilize vegetable gardens. Here again the constraint is the availability of the manure in required quantities as well as labour and transportation materials of the manure from point of production to the farms.

Conclusions and Lessons learnt

Soil fertility is a major determinant of increased crop yields in Malawi. The use of organic matter technologies offers practical solutions to sustainable soil fertility management in Malawi so long as crop residues and green manures are returned to the soil. The results of this study have indicated that farmer involvement in developing new technologies is crucial in increasing demand for more soil fertility technologies given the perceived benefits and requirements. Farmers have knowledge about the various technology options and their contribution to soil fertility and food security. However, adoption and production of these technologies in Malawi is still low. Farmers indicated a lack of seed, lack of markets particularly for farmers who produced surplus legume crops, poor soils, livestock damage and lack of labour as the factors limiting adoption of these practices. There is also a need to study each farmers socio-economic situation and promote only what is feasible considering the labour, household income, land holding sizes and soil type owned by the different farming households.

References

Grant, P.M. (1981). The fertilization of sandy soils in peasant agriculture. Zimbabwe Agricultural Journal 78: 169-175.

Kamanga, B.C.G. 2002. Understanding the farmers Agricultural Environment in Malawi. Risk Management Project Working Paper 02-01.

Kanyama-Phiri, G.Y., Snapp, S.S. and Minae, S. 1998. Outlook on Agriculture Partnership with Malawian Farmers to Develop Organic Matter Technologies. 27: 167-175.

Snapp, S.S., Mafongoya, P.L. and Waddington, S., 1998. Organic Matter Technologies to improve nutrient Cycling in smallholder systems of Southern Africa. Agriculture, Ecosystems and Environment 71: 187-202.

¹ Rural Development Department, Bunda College of Agriculture, University of Malawi, P.O. Box ²19, Lilongwe, Malawi. E-mail: [email protected]

Suitable approach for sustainable soil fertility management in
Sahelian zone of Niger, West Africa

Hayashi, K.¹; R. Matsunaga² and T. Aboudlaye³

Keywords: indigenous knowledge, soil fertility management, weeding, Niger, West Africa

Introduction

Soil fertility problems remain a high priority for agricultural development in Africa and the role of scientific information is important to improve the situation. However, in most cases, a scientific point of view can only partially reflect the farmers’ point of view in terms of agricultural development. The complexity of farmers’ society creates a gap between the scientist and the farmer. This gap should be bridged in order to facilitate mutual understanding on the problems to be tackled.

In the past years, many studies on soil fertility management have been undertaken in Niger (e.g. Bationo et al. 1998; Ly et al. 2000; Yamoah et al. 1998). Several technologies derived from these studies were disseminated to farmers’ field. Unfortunately, the adoption rate of the disseminated technologies was rather low (Abdoulaye, T. et al. 2000). Therefore, developing a suitable technology to improve soil fertility management on farmers’ fields is still one of the most important issues to raise agricultural production (ICRISAT 2002). Suitable technologies can only be widely adopted if the importance of farmers’ issues is taken into account.

Indigenous knowledge (IK) is receiving considerable attention in recent years in terms of social and agricultural development (UNCED 1992; Ishida et al. 1998). However, most of the information about IK is oral patrimony from generation to generation (Roman et al. 1992). Though obtaining the information behind the reasoning of traditional people in natural resource management is of utmost interest to identify appropriate technologies, there exists a prejudice that IK is against development (Morin-Labatut & Akhtar 1992).

The objective of this study was to have better understanding of existing IK in terms of soil fertility management in Sahel zone of West Africa. The IK is empirical oral patrimony from generation to generation and therefore it should be processed in a scientific manner and evaluated quantitatively to identify the most useful information for agricultural development. It is also necessary to look at farmers’ practice in crop production to evaluate local practice and to identify the possible scenarios to make outputs adaptable and sustainable at farmers’ level.

Materials and methods

Site description

The survey was conducted in three villages of the Fakara region, Dantiandou District of Tillaberi prefecture, Western Niger (60-90 km Northeast of Niamey) in West Africa: Banizounbou (145 households), Tchigo Tegui (135 households), and Kodey (100 households). Principal tribes of this area are Zarma, the agriculturalist who engage in rain fed cereal production such as millet (Pennisetum typhoides) and cowpea (Vigna unguiculata), and Fulani, the pastoralist who engage in livestock production. The prevailing soil type in the Fakara is a Psammentic Paleustalfs, which has a high sand fraction and typical characteristics of an infertile soil. The rainfall pattern of this area is mono modal starting from June till September. Total amount of rainfall is about 500 mm peaking in August.

Gathering indigenous knowledge (IK) in terms of fertility management of agricultural land

A questionnaire was prepared in order to identify the way farmers recognize the agricultural land in the study area. Two steps were taken for gathering the information on land management. The first step was to obtain IK from representative informants and the second step was to confirm recognition as well as perception of this information through different generation and different villages. The prepared questionnaire was developed with the help of an agent who has been working in the area for 11 years in different projects and can speak local languages such as Zarma and Foulfoulde (the language of Fulani). To verify the information obtained in step one, 120 farmers (40 per village) from different generations and locations were interviewed. Three villages, Banizoumbou, Tchigo Tegui and Kodey were targeted and 10 farmers of each age category: 21-30 yr, 31-40 yr, 41-50 yr and >50 yr, were interviewed. In addition to the farmers’ view on agricultural land, the Fulani were also interviewed by the same agent to determine their relationship with the Zarma. A total of 92 Fulani households from Fakara were interviewed. All information gathered through the survey was written in French notation.

Quantitative evaluation on farmers’ practice for crop production

In order to obtain IK through farmers’ practice on millet production, a two-year field experiment was carried out from 2003 to 2004 in Gourou Yena, Fakara region. The experiment was a 2 factorial split-plot with three replications. The main factor was fertilization and weeding frequency was the sub factor. In conventional practice, farmers weed twice during the cropping season, but in most cases don’t apply fertilizer. However, the need for fertilization is increasing due to increased food demand. Through our experiment, we attempted to understand the farmers’ perspective of conventional practice for millet production and determine a rational way to improve this production system through fertilization. The treatments used in this experiment were a combination of weeding with fertilization: inorganic fertilizer (IF) + twice of weeding (2W), IF+1W, IF+0W, inorganic fertilizer with cow manure application through local practice (IM) +2W, IM+1W, IM+0W. Treatments with no fertilizer application (NF) were also established, and treatments were NF+2W, NF+1W and NF+0W. In this experiment, NF+2W was the control and NF+0W was the absolute control. DAP (Di-ammonium phosphate) was used as the source of IF with the dosage of 20 kg/ha (Bationo, A. et al 2000) by hill application. The cow manure was obtained through the corralling system where the livestock are restrained in the field during night. The manure was applied at 6 t DM/ha (William, T.O. et al. 1995).

Results

Farmers’ knowledge on soil classification and land management

The results obtained from the survey to gather IK of soil fertility and agricultural land management showed that the farmers classified the soils in the study area by color and texture. The farmers’ recognized Labu tchirey, a soil with a reddish sandy texture, as the dominant soil type in this area. This soil was classified as Psammentic Haplustalfs or Psammentic Paleustalfs, and has a high sand content (92~96%) and poor chemical characteristics.

Two types of agricultural land in study area were classified dependent on the presence of fertility management practice. Farms that had some form of fertility management were either intensively or extensively managed. Intensively managed land was classified as “Birgui-farey” (which means land with fertility) and had three different sub-classifications based on the organic amendment used. “Birgi-nougou” was land managed by transported farmyard manure, “Gah-zeno” is by corralling and “Farey-djibo” is by crop residues. Of these three different sources of organic amendment, Farey-djibo was the least practical due to high demand for crop residue for livestock feeding, substitute for fire wood, construction, etc. Therefore Birgui nougou and Gah zeno were generally the most common practice for intensively managed land. The extensively managed land is called “Farey”, meaning land restored by using a fallow system. This classification also had different sub-classifications depending on how many years cultivation there had been since the fallow period. The sub-classifications were “Farey-zeno”, “Sakara”, “Lali banda”, “Koiri koiri”, “Koiri zeno”, meaning respectively, fallow land, or the 1^st year, 2^nd year, 3^rd year or 4^th year of cultivation since the fallow. The farms without management were considered as degraded land due to the continuous cultivation diminishing their agricultural potential. There were three levels of degraded land, and Labu Farga was considered the most degraded.

Geographical distribution of agricultural land based on classification of land management practice

The results showed that 66% of the fertility management of the surveyed area was achieved by fallowing, 18% using corralled manure and 16% using farmyard manure. These proportions are indicative of the harsh environment and the low availability of organic matter that can be used for soil fertility management. This highlights the importance of trying to mitigate this situation to improve the livelihood in Sahel. The survey also revealed that the fallowed agricultural land was in a critical situation as fallow periods were being shortened. The land in the study area was usually managed using 6 years cultivation and 3 years fallow. This practice is likely to accelerate soil-land degradation, as a previous study (Hayashi 2005) showed that land fallowed for 2 years was less fertile than that receiving 4 years of fallow, and even 4 years of fallow was less fertile than Sakara (land fallowed for more than 4 years). This indicates that soil fertility can not be restored sufficiently when the fallow cycle is shortened to 2 to 4 years and the resulting chronic decline in fertility leaves the soil completely degraded. This needs to be considered when trying to aim for better conservation technology in fallow system.

From a geographical perspective, the manage­ment systems mentioned above were variously distributed in distance from the residential areas. Transporting manure produced from household waste or farmyard manure generally occurred when the farmer’s residence was less than 1 km away. The fallow system was used in areas more than 2 km away from the residential area, while the corralling system was intermediate being distributed in a range of 1 km to 2 km. This distribution was due to socio-economic factors encountered in the study area. Fallow systems which were managed extensively can not be enhanced by the other systems due to constraints such as difficulties for transporting manure by the farmer, or water scarcity for livestock for the pastoralist. Also, the corralling system can not be managed too close to the farmers’ residence, because of conflict between farmer and pastoralist. Normally the pastoralist manages the corralling system at a distance from the farmers’ residence as livestock can cause damage to the farmer’s possessions, or die from eating non edible materials like plastic bags or cloth. These geographical constraints limit the type of organic material that can be applied to agricultural land, and also determines who takes charge of the agricultural land in these different systems.

Evaluation of productivity through conventional agricultural system

Although the extensively managed fallow system was the dominant form of agricultural land management, farmers still considered weeding as an indispensable input in order to secure crop production to supply the household food demand. On average, crops were weeded twice, however, some crop lands were not weeded at all during the season due to insufficient labour supply caused by economic and social constraints.

Other agricultural inputs such as inorganic fertilizer are not commonly used due to the infra­structure problems in the rural areas of Sahel (Abdoulaye, T. et al. 2000). However, using a com­bination of inorganic fertilizer with organic materials like livestock manure to enhance the millet production on these low fertile sandy soils shows promise. The field experiments carried out in this study illustrate the change in millet production over 2 years under different fertilizer management systems (Figure 2-A, B and C).

When no fertilizer was applied, weeding twice (NF2W) resulted in a better yield than with one (NF1W) or no weeding (NF0W) (Figure 2-a). The yield for second year at NF2W was improved as the weeding was carried out just after the rain and harvest index increased from 0.19 to 0.31. In contrast, the first year of IF and IM treatments resulted in better yields at each frequency of weeding than was obtained from NF2W. It is notable that in the first year, the treatments receiving only one weeding in both the IF and IM management systems yielded as high as those receiving two. However, by the second year, the yield of IF and IM decreased and the yields were only as high as that obtained from NW2W in the second year. There were significant differences in yield between the first and second year in the IM2W and IM1W treatments and these treatments yielded considerably higher than the corresponding IF treatments. However, by the second year the yields for the IM treatments were similar to the IF treatments, indicating that the livestock excrements was only effective for the first year. Overall, the mean harvest index of IM was better at each weeding frequency than that for IF and NF suggesting that the application of livestock excrements can enhance seed production of millet.

Figure 1. Millet yields (kg ha^-1) for different treatments. 2W, 1W and 0W represent frequency of weeding. Graph (A) is for non fertility treatment (NF), (B) is for fertility application through DAP (IF), and (C) is for DAP with livestock excrements (IM)

Discussion

The information about IK obtained by this study demonstrated that soil and agricultural land manage­ment was diverse even in extensively managed fields. This allows the farmers to use their finite resources efficiently to make their production sustainable. According to the results, 34% of agricultural land in study area was for organic matter application through corralling or recycling of domestic waste or farmyard manure, and 66% was extensively managed using a fallow system. The high proportion of fallow management is understandable as farmers in this study area usually owned 36 ha of agricultural land per household (non published data), thus making a large portion of this land at a great distance from the allocation of finite resources. However, another input such as labour required for weeding was allocated evenly for all management systems, although it was hard hand labour using a local hoe. Normally farmers aimed to weed twice during the cropping season, but this frequency declined depending on individual constraints for the various households. The results obtained in the field experiment demonstrate the impact of weeding even when there is no fertilizer input, as yields obtained in the both the first and second years declined as the frequency of weeding decreased. Despite the intensive hand labour necessary for weeding, the yield from the fallow system remained quite low, and may still not be sufficient to motivate the farmer to expend the effort required to improve management practices for increased agricultural production. However, it was notable that in the first year, a single weeding resulted in yields comparable to that obtained with double weeding and the yield was still compromised in the second year by single weeding with fertilizer application.

References

Abdoulaye, Tahirou and J. Lowenberg-DeBoer 2000. Intensification of Sahelian farming systems: evidence from Niger. Agricultural systems 64: 67-81.

Bationo, A., Lompo, F., Koala, S., 1998. Research on nutrient flows and balances in West Africa: State-of-the-art. In Smaling E.MA. (Ed.). Nutrient balances as indicators of production and sustainability in sub-Saharan African agriculture. Agriculture, Ecosystems and Environment 71: (1, 2, 3) pp. 19-36.

Bationo, A. and Ntare, B.R., 2000. Rotation and nitrogen fertilizer effects on pearl millet, cowpea and groundnut yield and soil chemical properties in a sandy soil in the semi-arid tropics, West Africa. Journal of Agricultural Science, (Cambridge) 134: 277-284.

Hayashi, K. 2005. Validation of indigenous knowledge for soil fertility management in Sahelian zone of West Africa. JAICAF Expert Bulletin 25(6): 12-26.

ICRISAT 2002. ICRISAT annual report, Grey to Green revolution, Patasncheru, India.

Ishida F., A. Kamidouzono & T. Wakatsuki 1998. Indigenous rice-based lowland farming systems of Nupe, Nigeria, Tropical Agriculture 42(1): 18-28.

Ly S.A., C.L. Bielders, N. van Duivenbooden, A. Tassiou, A.S. Gouro, K. Anand Kumar (eds.) 2000. Techno­logies diffusables et transférables aux producteurs. 1 ere partie: Dossiers techniques. INRAN/ICRISAT. 83 p.

Morin-Labatut G.M. & S. Akhtar 1992. Traditional environmental knowledge: A resource to manage and share. Development Journal of the Society for International Development 4: 24-29.

Roman R.P., A.S. Jonathan & A.T. Joseph 1992. The role of indigenous soil knowledge in agricultural development. Journal of Soil and Water Conservation 47(4): 298-302.

UNCED 1992. The Rio Declaration on Environment and Development. 22^nd principle.

Yamoah C.F. 1998. Rotation effect on sorghum response to nitrogen fertilizer under different rainfall and temperature environments. Agriculture, Ecosystem and Environment 68:233-243.

Williams T.O., J.M. Powell & S. Fernández-Rivera 1995. Manure availability in relation to sustainable food crop production in Semi-Arid West Africa: evidence from Niger. Quaterly J. Int. Agr. 34: 248-258.

¹ JIRCAS, Department of Crop Production and Environ­ ment, Tsukuba, Ibaraki, Japan [email protected]
² JIRCAS, Department of Biological Resources, Tsukuba, Ibaraki, Japan
³ INRAN, Department of Economic Research, Niamey, Niger