Many technologies have been developed during the past 20 years by NARS and CGIARs. Those technologies ranges from new varieties to sol protection, fertilizer usage or water conservation technics. They are built in the intent to increase production and protect environment.

However each farm has its own policy to reach food security and to reduce poverty. To acheive these goals farmers adopt technologies based on several conditions. Sikasso is the most agricultural productive area in Mali. Farmers benefit from many extension services and are the most advanced ones in the country.

This study aims to determine factors affecting adoption of different technologies in the area.

To explain and forecast farmers reaction toward new technologies, mostly micro-economic data are used on individuals. Many factors affect farmers behavior regarding technologies.

Many theoretical and empirical studies have been conducted on adoption or new agricultural technologies choices. On empirical wise many analytical methods have been applied.

To measure factors affecting adoption of new technologies, it is necessary to compute adoption rates. These rates are computed by dividing areas under new technologies by total area concerned. Data collection and analysis are presented as follow.

The study has been conducted in the Mali-Sud zone and concerns the geographical area of Sikasso. Within it seven extension sectors and 12 villages are concerned. In each village farmers have been selected randomly for adoption survey.

Data collection involved 122 farmers. The head of each family or household or farm is questioned. This was done by 5 enquetors and two researchers.

Data were put in Excel and SPSS is used to analyze them. A descriptive statistic is done to test variable significance and Logit model is used to determine factors affecting adoption of new technologies.

To assess farmers decision to adopt new technologies the Logit model is used. A simple description of the model is as follow: y is the decision to adopt new technologies and x the vector of independent variables related to adoption. Farmers decision to adopt is specified as y = f(x,e), where e is an error term with a logistic distribution. Logit model has been several times used in adoption studies.

The conceptual model is given as :

y_ik = F(I_ik ) = for Z_ik = xikb ik and -¥ < z_ik< +¥

Where yik is the dependant variable which could be 1 if the farmer is an adopter and 0 if he does not adopt. Xik is a matrix of independent variables affecting adoption of new technologies and b ik are vector parameters to estimate. Iik is an implicit adoption variable index. The Logit model is estimated using SPSS.

A descriptive statistic of all variables in the empirical model is in the following table. The qualitative variable adopt is 1 if the farmer adopt a new technology (water conservation, land management, new gene, etc.) and 0 if does not adopt. Independent variables and their justification is discussed below. SEX is a variable indexing the adopter’s gender, it takes the value 1 if it’s a male and 0 if it’s a female. Several studies argued the weak proportion of women household heads in Mali. Persons surveyed are generally men therefore adopters are in majority males. In hypothesis sex is positively correlated to adoption. AGE measures farmer’s age. Studies proved that young farmers are more innovative because their openness. Yapi et al found that young farmers adopt more new genes of millet and sorghum in Mali. By hypothesis AGE is negatively related to adoption. Fare’s marital status determines whether the farmer is married or else. The farmer could be married, widow or single. The status of the farmer is important for assessing some information on technologies and their adoption. In hypothesis the marital status is positively related to technologies adoption. Education measures the education level of farmers. Education increases farmers capacity to create or innove. Farmers having a good education level are more open to new technologies. By hypothesis education level is positively related to adoption. The household size determines the number of people and workforce in the household. New technologies are sometimes demanding in labor. The main source of labor remain the household, big households with more farm workers are supposed to adopt new technologies. But large households use less area per labor force and heavy dependency to food. They would rather increase areas to meet food requirement than to intensify production. In hypothesis the household size is negatively related to adoption of new technologies. Experience in agriculture is a variable that measures farmers experience in production. Farmers with good experience in agriculture are more risk averse than newer ones. Those farmers with experience are more induced to adopt new technologies. But this experience migth push them to stick to old technologies. By hypothesis farmers experience in agriculture is positively related to adoption.

Existence of non agricultural income source is a dichotomous variable. The existence of non agricultural income source could allow farmers to handle implementation costs of some technologies. This could include fertilizer costs, labor, equipment, etc. Non agricultural incomes can reduce risk associated with trial of new technology. Studies have shown that income affect positively adoption.

Contact with extension services (NGO, public services, research, etc.) is a variable which takes value 1 if farmer has a contact with any extension agent and 0 if not. Contact allows farmer to access information on new technologies. Often extension agents do demonstration fields where farmers learn and experiment new technics. Hypothetically contact is positively related to adoption of new technologies.

Number of farmed fields belonging to the household is a variable indexing the share of labor force of the household. Total number of fields could show gender issues in land attribution within the household. Generally households with more fields tell existence of individual farms. Individual farms belongs to men and/or women. These fields are generally of small size; in hypothesis farm number is positively related to adoption.

Total area of farmed fields belonging to the household is a variable determining the labor force capacity. This capacity is linked to human, material resources available to the household. This variable measures the intensification or extensification degre of production in the household. Many households seek to satisfy their food security needs by increasing farmed areas and not par increasing land productivity. The first option is more frequent because more accessible to farmers in areas where land is abundant. In hypothesis total farmed area is negatively related to adoption of new technology.

Total fallow area belonging to household is a variable showing land availability or pressure on land in the area. It indicates the production system in the household. Total fallow area show land usage by the household and let to know if the household is using all it’s capital in land or not. Land ownership could push farmers to experiment a new technology. The hypothesis is that fallow land area is positively related to adoption of new technologies.

Table: Statistic description of variables used in the econometric model

RESULTS AND DISCUSSIONS

Field Survey Results

The table below shows the status of new technologies adoption by farmers. Among 122 farmers surveyed 93% adopted at least one of water conservation technologies; 82% adopted at least one of the new technologies used to improve soil fertility; 82% adopted one of the new technologies used for soil management and 63% adopted at least one new gene. Farmers learn new technologies from research services (69%), from NGOs (20%), from extension services (96%).

Results show that households in the study area are generally large (21 persons in average by household). This large size could be explained by the fact that each household has in it many small households. If the number of person by household is large, the number of active worker remains relatively small. The average active number by household is from 9 to 10; this number represents only 50% of household total population, therefore one active worker should produce for himself and for another person in order to meet total household needs. This should make us to think that agricultural intensification is necesary to increase production by active worker.

The study area is covered by mainly CMDT as extension service and others (NGOs). Near 94% of farmers get advice from extension. And all farmers belong to an extension organization (village association, contact group, etc.).

Farmers education level in the area is still average. Only 56 % of surveyed farmers know how to write or read in their native language or other. This level is in defavor of adoption of new technologies conducting to their partial adoption or non adoption.

The size of farms is relatively small based on average population by household and active worker number. On average, each active worker farms near one and half hectares of land in a year (all type of farms considered). This small farmed area is due to not only at a lack of appropriate equipment but also at land availability. Fallow areas and loan areas are relatively small, but their rational use could increase land area by active worker if adequate means are used by producers.

In most all surveyed households (89 %) there is at least one agricultural material. However, few of them possess a complete set of equipment. This lack of equipment can limit the execution of some activities in time as plowing, transport, etc.... As a matter of fact more than 75% of households use manure in their fields, manure is accessed without any cost in the area. However only 59 % of households use chemical fertilizer. This small use of fertilizers is due to its difficult financial access (mainly).

Surveyed farmers experience in agriculture is at least 40 years in average. This experience varies from one area to an other is a good indicator of farmers knowledge on thenologies.

REASONS FOR ADOPTION OF NEW TECHNOLOGIES

Reasons for adopting new genes in the study area by farmers are as follow : earlyness of the variety (66%), height yield (55%) and good taste (38%). These reasons variy by agro-ecological zone. In Sahelian zone its the height yield and earlyness of the gene which are very important. In Sahelian zone rain fall is average (400-600 mm). While in Soudano-Sahelian zone its earlyness and taste which are important in Soudanian zone its the grain size and grain color which are important to farmers.

Concerning water conservation technologies farmers adopt deep plowing, tide ridge and ridging. Reasons for adoption are various from agroecological zones and from farmers. Tide ridge are practiced in sahelian zones, ridging in Sahelosoudanian zones and deep plowing in Soudanian zones. The reason frequently cited by farmers is moisture conservation due to rainfall hazard. Other reasons are to combat weeds, to protect plants against wind, etc. Farmers learnt those technologies from extension services such as CMDT.

For soil management technologies farmers adopt furrow diking, stone diking, headgerows, contours to fight against erosion. Farmers attest these technologies allowed them to use some marginal lands or to crop abandoned lands. Furrow diking is frequent in Soudanian zones, while stone diking and headgerow are frequent in the other zones.

CONSTRAINTS TO ADOPTION OF TECHNOLOGIES

Constraints limiting technology adoption depends on technology, farmers and agroecological zones. For new genes adoption farmers cite essentially the lack of improved seeds or high costs involved in using the seeds such as fertilizer costs, chemicals costs, labor costs and availability of inputs. For water conservation technologies farmers cite the lack of equipment to execute the technology and the timing of execution. Most of time the technology is executed then animals don’t have enough power to pull the plow or the soil is heavy to be plowed. For stone diking the constraints are the lack of man power, lack of equipment and lack of capital.

Many other reasons are cited by farmers. Among them the most important is rainfall hazard, pressure on land by people and animals, etc.

-2 Log Likelihood 86.117

MODEL RESULTS

In the above table are presented the empirical model results. The logit model is significant at 7%. The model predict correctly at 77% for non adopters and at 84% for non adopters. Six variables are significant in explaining new genes adoption: contact with extension services; farmers experience in agriculture ; farmers experience with new genes; existing of other income source of farmers; agro- ecological zone.

The negative sign of age means that young farmers are more willing to adopt new genes than elders. This could be related to their availability to try new genes and their risk averseness is weaker because they are more open to innovations. The positive sign of total area of farmed fields inform on the positive correlation with adoption of new genes.

The negative sign of fallow area indicates that fallow iss negatively correlated with adoption of new genes. The negative sign of experience of farmer in agriculture indicated that the more farmers have experience the less they adopt new genes. This could be explained by the fact that experience is dependant on age. Education is positively correlated to adoption of new genes.

The positive sign and significance of farmer’s contact with extension services indicates that farmers with contact with extension on new genes are more available to adopt. Because new genes are innovations in the area and farmers have more confident to extension agents for information. These kinds of contact are very useful at first years of introduction of new genes which are frequently abandoned by farmers after some years of trial.

The positive sign and significance of other income source indicates that the possibility to a farmer to access to a new source of income than agriculture is a favorable condition to adopt new gene. New technologies are more demanding for inputs, their adoption depends on cash availability to purchase seed, fertilizers, chemicals. Activities generating income allow farmers to adopt new technologies which are input demanding. Many farmers don’t adopt new genes because their costs of adoption is high.

The positive sign and significance of farmers’ experience with new genes indicates the more farmers use new varieties the more they are willing to adopt. This could be explained by the fact that farmers growing new genes get more returns and consequently are willing to accept more effort to improve their production.

The agro-ecological zone plays enough in technology adoption. The positive sign and significance of agroecological zone in the model could be explained by rainfall hazards and traditional varieties can’t mature. New genes have generally a vegetative short cycle and are more suited to rainfall hazards. Consequently farmers are available to accept new genes to improve their production and productivity.

The number of fields of the farm is positively correlated to adoption of new genes. This indicates that households with many fields could face at less risk in adopting new genes than those who don’t have. Because one or many fields are available on which they can experiment new genes; therefore they are more apt to adopt.

The size of the household and or the number of active workers by household plays on adoption of new genes. In the model the number of active men workers is negatively correlated to adoption while the number of women is positively correlated to adoption of new genes. This could be explained by the fact that women adopt more new genes than men. Women in the household own small land area where they produce intensively to get the maximum return. In rural areas women are responsible for many expenses to maintain household members. They do need income to face those expenses.

For soil management technologies, adoption is determined by the type of farm. Since the typology of farm is based on equipment availability in the household, farms with no complete equipment don’t adopt soil management technologies. This is consistent with our hypothesis. Type A has at least two sets of traction equipment, type B at least one set, type C incomplete set and type D manual. As in the table below the significance of variables in the model increases with equipment.

CONCLUSIONS

This report determines factors affecting adoption of new technologies in the southern Mali area. The econometric analysis shows that farmers adopt new technologies based on agro-ecologies. Adoption is higher with farmers having a contact with extension services and for farmers with another income source than agriculture. Adoption is low with old farmers and with those having more land. Adoption is high with farmers having experience with new genes

Farmers adopt new technologies for many reasons (see above sections). In adopting new genes they substitute to other crops. Those are mostly millet. Substitution reasons are economic. The price of millet is low comparatively to cash crops such as cotton, cowpea, ground nut, potatoes.

Farmers emntioned important constraints to adoption of new technologies: They are the lack of seed, the cost involved with adoption and marketing problems. These constraints cited by farmers could be translated as follow: lack of well organized input market, low income of farmers and lack of information from extension services.

However farmers continue to use their old genes and other technologies in major parts of their fields. It is then necessary to find new orientations and establish research and extension priorities in order to have bigger impact on population well being.