Centralized regulation and fisheries co-management institutions, when imposed on Malawian fisherfolk, have failed due to poor local legitimacy, resulting in de facto open access regimes for most Malawian fisheries. Case studies from a range of locations indicate that a key to success or failure in co-management lies in understanding the roles played by both traditional (chiefs) and representational (stakeholder) institutions. Formally instituted, conservation-driven co-management institutions are unlikely to be effective if they try to replace informally derived, consensus-based social and cultural institutions. Additionally, institutional legitimacy requires downward accountability, and both traditional and decentralized institutions must remain vigilant against the threat of elite capture. An analysis of fisherfolk and chieftain roles in Malawi's fisheries management regimes illustrates how these stakeholders may either promote or undermine socially and ecologically sustainable outcomes. We conclude that both decentralized representational institutions and “institutional bricolage” facilitated by chiefs are critical to maintaining resilient co-management institutions and fishing livelihoods.

The climate analogue approach is often considered a valuable tool for climate change impact projection and adaptation planning, especially for complex systems that cannot be modelled reliably. Important examples are smallholder farming systems using agroforestry or other mixed-cropping approaches. For the projected climate at a particular site of interest, the analogue approach identifies locations where the current climate is similar to these projected conditions. By comparing baseline-analogue site pairs, information on climate impacts and opportunities for adaptation can be obtained. However, the climate analogue approach is only meaningful, if climate is a dominant driver of differences between baseline and analogue site pairs. For a smallholder farming setting on Mt. Elgon in Kenya, we tested this requirement by comparing yield potentials of maize and coffee (obtained from the IIASA Global Agro-ecological Zones dataset) among 50 close analogue sites for different future climate scenarios and models, and by comparing local ecological knowledge and farm characteristics for one baseline-analogue pair. Yield potentials among the 50 closest analogue locations varied strongly within all climate scenarios, hinting at factors other than climate as major drivers of what the analogue approach might interpret as climate effects. However, on average future climatic conditions seemed more favourable to maize and coffee cultivation than current conditions. The detailed site comparison revealed substantial differences between farms in important characteristics, such as farm size and presence of cash crops, casting doubt on the usefulness of the comparison for climate change analysis. Climatic constraints were similar between sites, so that no apparent lessons for adaptation could be derived. Pests and diseases were also similar, indicating that climate change may not lead to strong changes in biotic constraints at the baseline site in the near future. From both analyses, it appeared that differences between baseline and analogue sites were mostly explained by non-climatic factors. This does not bode well for using the analogue approach