Landscape scale genomic-environment diversity data to model existing and novel agri-systems under climate change to enhance food security in Ethiopia

Context: Ethiopia has historically been the world's largest recipient of targeted food aid, yet little food-insecurity has been reported for the southern Ethiopian highlands even during the devastating famines of the 1980s. Today, the agri-systems of the southern Ethiopian highlands successfully support one of the highest rural population densities in Africa (up to 1000persons/sqkm). Here, we investigate the landscape scale dynamics, interactions and resilience of these agri-systems, using interdisciplinary environmental modelling, crop genomics and natural capital approaches to understand how best to manage their response to climate change whilst continuing to provide food security for a growing Ethiopian population - predicted to reach 172 million by 2050.
Ethiopia is an important center of diversity for food crops, with an agricultural history defined by the domestication of numerous species including coffee, tef and enset. Southern Ethiopian agri-systems include more than 78 cultivated species encompassing roots, tubers, cereals, vegetables, fruits and pulses, including a very high proportion of indigenous crops. Typical farms average 19 different crop and livestock species underpinned by over 120 species of useful trees and shrubs co-occuring across the homegarden landscape. Research on individual crop species also indicates extremely high diversity, for example we have recorded >600 enset varieties, including up to 24 on a single farm, and 37 varieties of yam. This diversity aggregated at multiple scales may be the key to the past resilience of the southern Ethiopian highlands in times of famine, and the source of future resilience to climate change.

Aims: Building on previous research, we hypothesize that the biotic drivers of high agri-system productivity and resilience are: (1) cultivation of a high crop diversity within farms, (2) cultivation of high genetic diversity within crop species and (3) cultivation practices that commonly involve diverse mixes of annual and perennial, indigenous and alien, semi-domesticated and domesticated crops. This rich diversity at multiple scales can in principle support food security and sustainable intensification whilst buffering seasonal food deficits, emerging pests and diseases and facilitating agronomic adaptation; despite an average farm size of only 0.9 hectares and very few off-farm inputs such as irrigation systems and fertilizers.
In contrast to this indigenous diversity, farmers also grow highly domesticated introduced crops such as maize, avocado and banana, providing an ideal opportunity to evaluate these hypotheses. These crops are high yielding but likely to contain less genetic diversity. This may limit their capacity for adaptation to new or altered environments and their resilience to climate change. The prevalence of these introduced crops is increasing, together with a reported loss of indigenous crop diversity and a shift away from agro-forestry. The impacts of these trends as well as the projected impact of climate change on the resilience of Ethiopian agri-systems is unknown.

Applications and benefits: Our research will generate landscape scale environmental suitability, genomic and natural capital data to underpin a decision making tool for sustainable agri-system development and climate adaptation in the region. By enhance future resource provision and resilience, we will generate clear economic and social impact on the livelihoods they support. The novel methods employed here will be of both broader academic interest in the fields of agronomy, crop breeding and conservation and provide immediate knowledge-transfer and resources to enhance Ethiopia's research capability. Most importantly, capitalizing on our strong existing UK-Ethiopian partnerships and links to regional government we will seek development and implementation of science-based regional agri-systems strategy to bring immediate impact within the life of the project.

The performance of agri-systems (i.e. yield and other ecosystem services) depends on multiple drivers across multiple scales, including; the genomic composition of the crops (i.e. genetic variation, local adaptation), the environment (i.e. soils, temperature, rainfall, annual variability) and management practices (i.e. the composition of species grown, interactions).

Field observation together with landscape scale environmental data will be analysed using a rigorously ground-truthed ensemble environmental niche modelling framework, supported by a network of climate stations. Multiple species-models will be combined to understand agri-system distributions.

High-density genome-wide markers will be developed using tuneable Genotyping by Sequencing approach (tGBS) across ten key study species (2500 samples). In addition to phylogenetic analyses among crop cultivars and metrics of neutral and adaptive variation these data also provide the basis for genome-environment association (GEA) analysis. The principal of GEA methods is to identify genetic differentiation resulting from biotic or abiotic selection, on the basis that similar selective pressures should result in similar genomic patterns of allele frequencies across populations or landraces. By correlating allele frequencies with environmental variables, GEA seeks replicated signatures of selection across independent populations to identify adaptive loci. We have recently developed a way to extend this approach by combining GEA with climate projections to measure the current- or future- risk of non-adaptedness in plants.

Agro-ecological and natural capital data will be underpinned by landscape-wide soil sampling (500 samples), quantification of inputs (i.e. fertiliser), total carbon stocks, and yield. Potential net primary productivity will be assessed through measures of LAI (vertically through the vegetation profile) and NDVI (drone based) combined with sentinal2 imagery.

Who might benefit from this research?
Ethiopia has historically been the world's largest recipient of targeted food aid and is ranked 174/193 on GDP per capita by the UN and 104/119 in the most recent Global Hunger Index, with 28.8% of the population undernourished from 2014-16. The annual costs of malnutrition have been estimated at $4.7 billion, equivalent to 16.5% of GDP, resulting in significant long-term socio-economic consequences.
Ethiopia currently has the second highest population in Africa (>108 million) where agriculture accounts for 80% of exports and 75% of Ethiopians are smallholder farmers. Thus food security is a priority for the Ethiopian government, particularly with the population predicted to reach 172 million by 2050 and increasing climatic uncertainty due to climate change.

Our research will fill a critical knowledge gap delivering key insights on the drivers of sustainability and resilience of current and future Ethiopian agri-systems under climate change. The project outcomes outlined here will be critical to develop strategies to safeguard food and nutrient security, maintain sustainable livelihoods and enable climate-smart adaptation for the agricultural economy with impact specifically targeted to benefit the millions of farmers across the region who depend on the resource provisioning of their indigenous agri-systems.

More broadly, sustainable intensification and resilience to climate change are global agricultural challenges faced by millions of small holder farmers and the value chains that depend on them. Our findings are anticipated to have high applicability to other diverse farming systems as well as providing insights to sustainable agricultural intensification. Our series of workshops will not only seek to engage regional stakeholders, but also invite participation by researchers and policymakers from other African LDCs and LMICs with similar high-diversity agri-systems.

How might they benefit from this research?
Our principal goal is to synthesise our research outcomes into a i) a format that can inform and influence agri-system policy in the Ethiopian highlands, ii) a toolkit for improving agri-system resilience and food security, particularly under climate change and iii) a resource that provides the basis for accelerated and enhanced agricultural research by stakeholders in the region. This includes much needed phylogenetic classification of diverse indigenous crop landraces.

To achieve these outcomes, development of a decision making toolkit will be based on our combined modelling (of current and novel agrisystems), ecosystem provisioning (i.e. yield) and genomic data (i.e. diversity, local adaptation) to map the performance of different agri-systems (combination of crops) at the landscape scale under both current and future climate conditions. Through stakeholder workshops and agri-system information pamphlets targeted across districts, farmers will benefit from agro-ecological advice on how to optimise existing agri-systems in their local area and pathways to future climate adaptation (i.e. climate-appropriate crops, climate-optimised varieties). Supporting and as an early-project pilot strategy, will be nurseries geared towards dissemination of climate-optimised enset cultivars based on the outcomes of our existing enset GCRF research programme.

In the longer term our evaluation of diversity across important indigenous crops will enable us to better safeguard that diversity through existing germplasm collections and seedbanks together with in-country partners. Finally, the training and knowledge-exchange within the network developed over the course of the project will build skills and capacity in Ethiopia to sustainably enhance and monitor it's indigenous agri-systems.