Theme: 'Agriculture'. In this research we investigate if and how extension services can influence levels of spread of a new agricultural technology through the selection of different lead farmers, i.e. farmers who implement and maintain demonstration plots of a new technology. These farmers learn about the most effective way to implement the new technology and spread this information to family and friends. Different methods to select lead farmers may have profound effects on subsequent learning on and spread of a new technology. This we propose to investigate in a rural African setting, and analyse if one selection method works better than others (in some locations), in terms of subsequent spread and adoption levels. We randomly select sixty different villages in our region of study, around lake Kivu in the Democratic Republic of Congo. Here, our local partner implements a regular extension program. In each selected village they promote the use of nitrogen-fixing legumes. However, for the purpose of this approach, the method by which lead and satellite farmers are selected will differs from one village to the other. We thereby built upon approaches to select lead farmers as are commonly used by extension services and compare two different approaches, each in 30 villages. The methods included selection, or election, based on a participatory group meeting, versus a selection in consultation with village heads. To ensure that our research is well connected to actual practice, a final decision on treatments is taken at the start of the project in close consultation with various local stakeholders. In order to better understand if, and why some selection methods work better than others, for example, in reaching specific farmer groups such as female farmers, or work better in selected areas only, we collect additional data on village and household characteristics before and after the implementation of the extension program. A baseline survey documents village and household characteristics. In addition, we map social networks by documenting which farmers are linked, and the nature of such links, and we measure social characteristics of the villages experimentally that could explain, or could be explained from, characteristics of the households and social networks surveyed. Such measures include trust or the willingness to invest in a public good. At two points in time after the initial extension activity -one and two years after the initial training- we document levels of awareness, experimentation and adoption levels of the new technologies in an end-line survey. For each village we determine the total number of adopters and number of fields on which the technology is practiced. We use statistical analyses to determine if one of the treatments, i.e. the methods for selecting lead farmers, has led to a significantly higher level of awareness on, or adoption of the technologies. Subsequent statistical analyses, combining data from the baseline and end-line surveys, is deployed to better understand how information on technologies diffuses, and why diffusion processes may have been different across the treatments. We thereby analyse if variation in the treatment response can be related to specific village characteristics and we determine if the structure of some social networks, and or decisions by others in social networks, has played an important role in facilitating adoption or information spread. These analyses provide important information on the processes of technology diffusion and serve to identify best approaches of agricultural technology promotion and diffusion. This in turns helps agricultural development programs to increase their reach amongst rural African smallholders.

Common bean (Phaseolus vulgaris L.) is the most important food legume in the human diet, providing protein, micronutrients and complex carbohydrates for >300 million people in the tropics. Climate change scenarios predict that heat/drought and pests and diseases will be major pressures on bean production in the future. As with most major crops, and because of its domestication history, cultivated common bean lacks genetic diversity. Wild relatives can be used to introduce this diversity for key traits of interest, and in some cases have already been used successfully to provide novel sources of resistance to pests and diseases in beans. This is by no means an easy process, as producing the next generation of plants from these "wide crosses" is difficult, making it hard for breeding programs to make use of the opportunities offered by these wild plants. However, wide crosses occur naturally where farmers grow cultivated beans adjacent to wild populations, as happens across the natural range of beans from Mexico to Argentina. A number of these naturally occurring cultivated-wild hybrids populations have already been collected from sites throughout Central and South America and are stored in the CIAT genebank. We will characterise plants from twenty hybrid populations for priority breeding traits linked to climate change e.g. pest and disease resistance, and heat and drought tolerance. We will explore their genomes and provide all of this information to the bean breeding and research communities in an accessible way, to help users select the most suitable plants for their purpose. We will hold workshops and demonstrations to make sure that breeders and researchers are aware of this resource and understand how best to make use of it. By reducing the barriers to inclusion of wild plants into bean breeding programmes, we will help breeders to produce better beans in a shorter time, which will have a positive impact on global food security.

In Eastern and Central Africa beans are a vital crop because they enrich the soil with fixed nitrogen (an indispensable natural fertilizer), which in turn supports the cultivation of other important crops such as corn (maize) and cassava. Beans are an essential part of the regional diet because they are rich in protein and crucially important micronutrients like iron and zinc. Since beans are predominantly grown and traded by women, this crop provides additional direct economic benefits to women and their children. Unfortunately, a number of viruses attack bean plants, causing severe crop losses from disease. Although there are a few bean varieties with resistance to one of these viruses (called bean common mosaic virus), a closely related virus that occurs widely in, and is indigenous to, Africa (bean common necrotic mosaic virus) causes plants of these 'resistant' lines to die. Thus, it is important to develop new strategies to defend this vital crop. Our approach is to attack the insects (aphids) that transmit these viruses from plant to plant. We have found that aphid-plant interactions are controlled in large part by the plant's 'small RNA pathways'. Small RNA pathways are a recently discovered regulatory system used by plants and many other organisms to control the expression of their own genes as well as to fight off disease. The discovery of small RNA pathways has revolutionized biology and medicine. In plants, small RNA pathways control, among other things, the production of natural signal chemicals that attract or repel insect pests, including aphids. Intriguingly, viruses produce factors called 'silencing suppressors' that modify small RNA pathways. We have found that silencing suppressors affect the interactions of virus-infected plants with aphids in a way that is likely to enhance the rate at which these insects acquire viruses and transport them to other plants. We have assembled a team of scientists based in Uganda, Kenya and at two centres in the UK to translate these findings from the laboratory to the field and exploit them for protection of beans. However, aphids and the viruses they transmit are problems for all major crops and the work will yield vital data for the wider field of crop protection. Our multinational team will collaborate to: A. Identify potential factors involved in mediating plant-aphid communication in plants: small RNAs, the genes they control, and the chemical signals whose production they regulate. B. Use the modelling methods provided by the discipline of mathematical epidemiology to help us design experiments (under lab and later field conditions) to predict how altering the attractiveness of plants (whether engendered by changes in the plant or deployment of signal chemicals as traps or decoys) could be used to help minimize or prevent the transmission of viruses. C. Utilize the work from 1 and 2 to design experiments to test the effects on virus transmission of modifying plant responses to aphids or utilization of purified signal chemicals to trap or deter aphids under controlled and simulated field conditions. D. To achieve impact by disseminating information gained from our work to African crop scientists, growers and consumers through the Pan-Africa Bean Research Alliance, a network of the national bean research programmes of 28 African nations.

[Proposal EE112/ K1396905] Predicting the impacts of global change on rural communities is increasingly challenging due to the accelerating pace of climate change and social and economic development. The combined demands of ensuring food, energy and water security have been described as a "Perfect Storm" by Prof Sir John Beddington, HM Government's Chief Scientific adviser. It is clear that food security will continue to remain a critical issue in developing countries due to the unpredictable nature of food chains and the effects of climate change. Food security in poor rural communities often relies significantly on flows of ecosystem services from 'natural' environments. For millennia mankind has engaged in thinking and learning experiences which have shaped the processes underpinning the production of food and the management of land, addressing multiple factors and tradeoffs. However, many food production systems require intensive management and are prone to failure outside of the range of their optimal environmental conditions. Concerns are growing about the ability of current agricultural systems to support rising human populations without further degrading critical ecosystem services (such as water provisioning, pollination). During extreme events, such as drought, or other shocks or crises (environmental, social or economic), the dependence of rural communities on ecosystem services to meet their nutritional and livelihood needs often increases. This highlights the importance of minimising the impacts of agricultural systems on ecosystems and the services they provide. Strategies for coping with food insecurity may, in turn, have an impact on the capacity of ecosystems to deliver ecosystem services as the spatial and temporal nature of feedbacks between socio-economic and ecological systems can be complex. Addressing the sustainability of natural resource management and rural livelihoods requires integrated thinking across disciplines. The complex transformations which can, or have already occurred from natural forest to managed landscapes must be fully understood so that systems can be adopted which promote sustainable transformations and/or can mitigate any negative impacts. This proposal therefore brings together expertise in social sciences, economics, ecology, risk management, spatial planning, climate change and complexity sciences to design and integrate a suite of models and methods to analyse how dynamic stocks and flows of ecosystem services translate to local-level food security and nutritional health. The study will examine the multiple (and multi-directional) links between ecosystem services, food security and maternal and child health outcomes in poor rural communities, addressing three main themes: 1. Drivers, pressures and linkages between food security, nutritional health and ecosystem services; 2. Crises and tipping points: Past, present and future interactions between food insecurity and ecosystem services at the forest-agricultural interface; 3. The science-policy interface: How can we manage ecosystem services to reduce food insecurity and increase nutritional health? Analysis of household and intra-household nutritional status and assessment and mapping of ecosystem services at the relevant spatial scales will be conducted in sites in Colombia and Malawi, which are characterised by mosaics of forests and agricultural lands, to explore the trade-offs and tipping points associated with managing these dynamic landscapes under climate and socio-economic change. Powerful new models will predict how ecosystem services will be changed by drivers and pressures for human wellbeing and food security. This will allow risk management/mitigation models and strategies to be developed which can inform national and regional policy in order to maintain ecosystems and support human wellbeing.

Cassava (Manihot esculenta) is the third most important source of calories in the tropics and a key staple for millions in Africa, Asia and Latin America. Predictions of rising temperatures linked to climate change favour the productivity of cassava, a crop that performs well under stressful growing conditions linked to drought and high temperatures. However, cassava is susceptible to a number of pests and diseases, and projections of movement of cassava pests linked to climate change predict substantial yield losses. As such, cassava breeders are prioritising traits linked to pest and disease resistance within their breeding programmes. Within Latin America, the largest pest and disease threats to cassava production are whitefly, cassava frogskin disease and bacterial blight. Bacterial blight is also prevalent throughout cassava growing regions in Africa and Asia. These can cause yield losses of 76-92%. We can explore cassava's wild relatives to introduce novel disease resistance traits through conventional or molecular breeding techniques. Crop wild relatives have been used in wheat, rice, tomato, potato, sunflower and other crops to provide sources of novel resistance to pests and diseases. The Genetic Resources Program (GRP, genebank) at the International Center for Tropical Agriculture (CIAT) in Colombia currently houses ~4,900 accessions of cassava and its wild relatives. The mission of genebanks includes the conservation of materials, but also aims to make the material available for use. However having limited information for the vast majority of accessions makes it difficult for users, or even genebank managers, to select materials that may contain useful properties for breeding. Through sequencing the genomes of cassava wild relatives we will reveal the genetic diversity of their resistance genes. In addition we will check the performance of these species against whitefly, cassava frogskin disease and bacterial blight. We will also develop novel tools to explore these datasets, making it easier for breeders to choose which wild materials to include in their programs. All the information generated through this project will be made publicly available for breeders and researchers to access via the GRP portal. Breeders tend to be cautious about including wild materials into their breeding programs as wild species bring a mixture of desirable and undesirable characteristics into the offspring. Breeding is a slow process, and removing the undesirable traits can take many generations. Therefore, having more information about the physical and genetic characteristics of these wild species will help breeders to make the most appropriate choices of plants to use and encourage the inclusion of wild relatives to provide novel sources of pest and disease resistance into cassava. CIAT has centres in Colombia and Vietnam, with researchers working in 53 countries worldwide. Having novel sources of pest resistance will benefit cassava breeders and growers in many countries, however we will target breeders in Colombia, Brazil, Vietnam, Cambodia and Lao People's Democratic Republic. Through the reduction of losses due to pests and diseases, this project will contribute to sustainable agriculture and increasing the resilience of food systems to climatic variability.