It is now widely recognised that agricultural innovation usually requires ideas from many sources, not only research. Indigenous knowledge of rural people is one valuable source, but information from development organisations, the private sector, and technical service providers is also important. As a result, linking organisations and individuals together is seen as important, with their interaction facilitating sharing of ideas and learning from each other. Recent thinking about this gives emphasis to learning as it concerns the changes in behaviour that result from putting ideas into use. This often elusive step is where ideas and technologies are adopted and used, and where innovation takes place.

Learning is also the process by which organisations work out which cluster of organisations they need to work with to bring about innovation. Since in ever-changing conditions the configuration of these clusters is always altering, ability and speed at which players learn are critical elements of innovation capacity. Evidence suggests that in situations where different organisations and individuals have learned to develop the patterns of interaction needed to share ideas and learn, innovation can happen quickly in response to the continuously changing set of challenges and opportunities emerging in the agricultural sector. However, ways of organising players for innovation are highly context-specific since the interaction between the nature of challenges, locations and histories often determine the clustering of players, ways of working, policies and politics. This means there is no single model or pattern of linkages that can enable agricultural innovation.

The history of fodder research and technology transfer, and its poor record of inducing widespread fodder innovation, suggest that ways of working and other practices and policies have not enabled different players to learn how to organise for innovation. Furthermore, habits and policies have often skewed exiting groupings and patterns of interaction so that innovation processes are not responsive to the welfare needs of the most marginalised in society — such as the livestock-dependant poor.

An apparent conclusion is that fodder and agricultural innovation — and associated welfare gains — is not dependant on technology development and transfer, per se. Rather, it is dependant on organising players, including research, in ways that facilitate the sharing of ideas and technology and learning.  Organising players in this way requires changes in ways of working (institutional change), and specific attention needs to be given to practices that make these arrangements responsive to the welfare needs of the poor.   

The implication of this is that fodder innovation, and the articulation of fodder challenges and opportunities, needs to be framed as an integrated, technical, organisational, institutional and policy challenge. The capacity to innovate needs to be viewed similarly from this holistic perspective. Building this capacity will largely involve enhancing the ability of players in a particular location to learn the habits and practices that allow them to organise themselves in ways that promote innovation in response to changing context. In other words, agricultural research, including research on fodder, needs to locate itself in this learning-based process of organisation for innovation.