This article is based on a study of the effects of a "maize extension package" on farmers' indigenous knowledge (IK) in the Jima area of Southwestern Ethiopia. Both qualitative and quantitative approaches were employed to collect information from farmers, extension workers, and researchers. Initially, semi-structured interviews and discussions were used extensively to collect information. The structured questionnaires were administered to 80 randomly selected farmers and to 40 purposively selected extension personnel. The study found that there is a rich wealth of indigenous technical knowledge and practices relating to maize production, though these appeared to be rapidly disappearing and are seriously endangered. Findings revealed that the promotion of modern practices in a package format played a major part in this regard with noticeable effect on local maize varieties and informal seed systems, traditional pest and disease control practices, control over produce and maize utilization. On the other hand, farmers, extension personnel and researchers were somewhat aware of the importance of farmers' IK and the fact that it is at great risk, though the later two tended to favour modern technologies over IK. Age, gender, farming experience, and resource endowment were found to influence the possession and use of IK practices. Based on the findings it was recommended that vigorous efforts should be made to identify, strengthen and integrate IK into formal research and development efforts to make them more relevant, effective and sustainable. To this effect, enhancing awareness, knowledge and skills at different levels; designing flexible programmes, strategies and methods combining IK and modern practices, should be given greater attention.

Many development interventions have failed to generate the desired impact among African resource-poor farmers for reasons including the centralised and top-down approach, lack of active community participation, and overreliance on external information and technology delivery strategies and channels. The migrant African armyworm, Spodoptera exempta, is among the major challenges threatening livelihoods of millions of farmers in East and Southern Africa. Outbreaks occur suddenly and can devastate crops and pasture. National and regional forecasting services have been operational since the 1960s to provide warning of potential outbreaks. This system relies on information from armyworm moth traps usually operated at district level. These centralised services have a number of difficulties and limitations which include: delays in communicating trap catch data to the forecaster; forecasts are not village specific; forecasts do not reach many farmers; lack of responsibility and local ownership of traps. An innovative approach called community-based armyworm forecasting (CBAF) was developed as a response to these limitations. The new approach has been piloted and tested in several East African countries and found to be effective. CBAF establishes a system that allows each village to have its own traps and trained forecasters who collect and interpret data, and provide village specific forecasts. A recent project on CBAF piloted the approach in 10, 5 and 39 villages of Malawi, Zimbabwe and Tanzania, respectively, in the 1 st year, and in a further 25 and 38 villages of Malawi and Tanzania in the 2 nd year. Discussion with stakeholders, field observations and assessments conducted by the authors indicate that the initiative has generated a number of benefits. It built local capacity, and because of the location specific early warning it enabled farmers to combat the pest more effectively. It was noticed by the authors that local communication channels and folk media play an important role in CBAF, complementing and enhancing the effectiveness of exogenous channels. This paper discusses how CBAF makes effective use of different communication channels and capabilities, and highlights preliminary results.