One option for practically applying innovation systems thinking involves the establishment of innovation platforms (IPs). Such platforms are designed to bring together a variety of different stakeholders to exchange knowledge and resources and take action to solve common problems. Yet relatively little is known about how IPs operate in practice, particularly how power dynamics influence platform processes.This paper focuses on a research-for-development project in the Ethiopian highlands which established three IPs for improved natural resource management.

Adapting through innovation is one way for rural communities to sustain and improve their livelihoods and environments. Since the 1980s research and development organizations have developed participatory approaches to foster rural innovation. This paper develops a model, called the Learning-to-Innovate (LTI) model, of four basic processes linked to decision making and learning which regulate rate and quality of innovation. The processes are: creating awareness of new opportunities; deciding to adopt; adapting and changing practice; and learning and selecting.

This paper takes the viewpoint of a social scientist and looks at agricultural scientists' pathways for science impact. Awareness of these pathways is increasingly becoming part and parcel of the professionalism of the agricultural scientist, now that the pressure is on to mobilize smallholders and their productive resources for (global) food security and for reducing persistent rural poverty. Significant new thinking about pathways is emerging and it is useful to present some of this, even if it is not cut-and-dried.

Most cocoa farmers in Ghana do not adopt research recommendations because they cannot afford the cost, therefore, yields are low. Integrated pest management (IPM) technologies that rely on low external inputs were tried with a group of farmers. The technologies included using aqueous neem seed extracts to control capsids; removing diseased pods to reduce blackpod incidence; controlling mistletoes, epiphytes, weeds; and managing shade. Although yields increased significantly, adoption was constrained by technical, social and economic factors.

Little is known about effective ways to operationalize agricultural innovation processes. The authors of this article use the MasAgro program in Mexico (which aims to increase maize and wheat productivity, profitability and sustainability), and the experiences of middle level ‘hub managers’, to understand how innovation processes occur in heterogeneous and changing contexts. Their research shows how a program, that initially had a relatively narrow technology focus, evolved towards an innovation system approach.

Poor farmers seldom benefit from new agricultural technologies. In response, research and extension approaches based on agricultural innovation systems are popular. Often agricultural research organisations are the network brokers, facilitating the emergence of the innovation system. Based on an analysis of the Sustainable Modernization of Traditional Agriculture (MasAgro) initiative in Mexico, this viewpoint suggests that such organisations are more often suitable network brokers when the objective is the development and scaling out of a technology by itself.

Many small-scale irrigation systems are characterized by low yields and deteriorating infrastructure. Interventions often erroneously focus on increasing yields and rehabilitating infrastructure. Small-scale irrigation systems have many of the characteristics of complex socio-ecological systems, with many different actors and numerous interconnected subsystems. However, the limited interaction between the different subsystems and their agents prevents learning and the emergence of more beneficial outcomes.

This paper (Part I) present a case study of work conducted by the International Centre for Tropical Agriculture (CIAT) to adapt network mapping techniques to a rural and developing country context. It reports on work in Colombia to develop a prototype network diagnosis tool for use by service providers who work to strengthen small rural groups. It is complemented by a further paper in this issue by Louise Clark (Part II) which presents work to develop a network diagnosis tool for stakeholders involved in agricultural supply chains in Bolivia.

Innovations generally arise out of a network of actors and relationships and network structure determines how effective networks are at fostering innovation. This paper (Part II) presents work to develop a network diagnosis tool for stakeholders involved in agricultural supply chains in Bolivia. The prototype method used is based on social network analysis methodology. This paper concludes with a final section to identify lessons learnt and makes recommendations for future research.

This paper presents a case study of the work carried out by CIAT to facilitate the creation of a community of practice, using Dgroups and taking advantage of this virtual space to apply a qualitative monitoring technique called Most Significant Change. The experience reported here mixed key ingredients to create and facilitate a community of practice to facilitate knowledge sharing and communication flow among 14 learning and knowledge sharing centres in Latin America and the Caribbean.