This post is written by Kate Longley, Ed Walters, and Omeno Suji, from the Feed the Future Global Supporting Seed Systems for Development activity (S34D). 

Short-term interventions to provide seeds to farmers affected by disaster are a common response in emergency contexts. But how impactful are they? Do they improve food security and enhance livelihoods? And what impacts do they have on seed systems? These were some of the questions posed by a recent study conducted by the USAID Feed the Future Global Supporting Seed Systems for Development (S34D) activity. The findings suggest that emergency seed provisioning does not necessarily lead to increased food security and thus represents a very high-risk investment for donors and aid agencies. Greater impacts may be achieved through longer-term, more strategic interventions that allow for greater attention to varieties, seed systems, and the range of factors that influence crop production such as access to land, soil quality, drought, climate change and access to markets. 

Background

There is currently a lack of strong evidence to show whether emergency seed interventions are achieving their objectives. The aim of the S34D activity was to help generate evidence on the impacts of selected emergency seed interventions on livelihoods, food security and seed systems. Participatory impact assessment (PIA) was identified as the most appropriate methodology because it can measure food security and livelihood impacts where no baseline data are available. 

Overview of the Interventions Assessed

Participatory impact assessments of three different seed interventions were undertaken in Uganda and the Democratic Republic of the Concgo (DRC). In Uganda, S34D partnered with World Vision and Lutheran World Federation to assess their respective interventions among refugee and host communities in different parts of Adjumani District, Northern Uganda. In the DRC, the assessment focused on an intervention implemented by Catholic Relief Services (CRS) to support farmers who had returned to their homes after displacement by conflict in Central Kasai Province. In determining the criteria for the selection of the interventions to be assessed, it was agreed that at least one agricultural year, but not more than two years, must have passed. If more than two years had passed since the intervention, farmers and other stakeholders may have found it difficult to recall the details. This was necessary to be able to determine the complete utilization of the harvest following the seed intervention, including whether any of the harvested output from the seed provided by the intervention had been saved and planted in subsequent seasons. 

Both interventions in Uganda involved the direct distribution of seed of improved varieties for a range of different crops to existing and new farmer groups, along with training in good agricultural practices and Village Savings and Loan Associations, as well as support for refugee farmers to access additional farmland, among other activities. Although the interventions were considered emergency projects, their durations were two and three years respectively and had been preceded by earlier similar project phases. Broadly speaking, the projects aimed to enhance food security and support sustainable livelihoods. 

In DRC, the one-year emergency intervention provided a combination of food assistance and seeds and tools support for agricultural recovery in an effort to address severe food insecurity. The project was in its fourth year, having been preceded by three earlier one-year phases, each implemented in different geographical areas. All households in the targeted communities received quality seeds of improved varieties for a range of crops through voucher-based seed fairs. Those involved in the assessment had, in fact, returned to their homes two years prior to the seed intervention and had been cultivating their farms for over three seasons since their return.

Assessment Findings

World Vision Intervention in Southwest Adjumani District, Uganda 

There was improved food security for both male and female refugees and host smallholder farmers, but the role of seed in contributing to this increase was minimal. Increased access to fertile land was the main reason for the increase in food security among refugee farmers. Livelihood impacts were assessed according to changes in expenditure from the income derived from the sale of crops. Based on this indicator, there was some evidence of increased investments in education, land and livestock by refugees, and an increase in financial investments in Village Savings and Loans Associations, especially by women, and especially from vegetables. However, the level of increases in these investments was small and was not accompanied by reduced expenditure on food, as might have been expected. 

Many seed distributions by many different partners over many years in Adjumani District have introduced new varieties into local cropping systems, giving farmers an appreciation of their particular characteristics (e.g. drought resistance, marketability, duration, etc). Consequently, the varieties provided by World Vision were already available through local grain markets and from other farmers. Anticipated increases in yield due to the use of improved varieties could not be attributed to the World Vision intervention because many of the varieties provided were already being cultivated by the beneficiary farmers. The yield from groundnut seeds provided by the project declined due to their poor genetic quality, which stemmed from the poor quality of the foundation seed used in the seed production process. This was undetectable through the standard seed quality checks normally undertaken. With the exception of groundnuts, feedback from farmers suggested the introduction of new, appropriate varieties over many years by many partners had strengthened the resilience of local cropping systems by broadening the range of varieties cultivated, thus reducing the risk of crop failure. 

Lutheran World Federation Intervention in Northeast Adjumani District, Uganda

It was difficult to draw any conclusions on the impacts of the seed interventions on food security, mainly due to low production caused by low rainfall levels experienced in the project area[1] following the seed distribution. The apparent increase in food security (as indicated by a decrease in the number of hunger months reported in a year) was not matched by increases in crop production for all farmers. Refugee households reported decreases in production for all key crops except for tomatoes, and experienced challenges in accessing additional land. Host farmer households reported increases in production for all key crops after the intervention as compared to before the intervention, largely due to increased access to fertile land. In many cases, this additional land had been cleared through the labour of refugees. As above, there was some evidence of small increases in education and financial investments by host and refugee households, notably savings in Village Saving and Loans Associations, especially by women.

The intervention supported local seed production efforts by Farmer Seed Producers and Local Seed Businesses. While both have important roles to play in providing affordable seed to farmers, there were concerns over the sustainability of the current seed production and marketing models. More broadly (beyond the specific intervention assessed), the many emergency and developmental seed projects implemented by many different partners had led to an increase in the number of agro-input dealers in Adjumani Town. However, there appeared to be a disconnect between many of these agro-input dealers and most smallholder farmers, who rarely — if ever — purchased seed from agro-input dealers. While the capacity and seed sales of many agro-input dealers had increased due to demand from nongovernmental organizations (NGOs) and development projects, there were concerns about market distortions due to over-dependence on NGO/project demand, displaced seed sales and weak seed marketing efforts.

CRS Intervention in Central Kasai Province, DRC

The impacts on food security, incomes and livelihoods varied by season, ecology and gender. Farmers reported increased production and income as a result of the seed intervention in Season A, but irregular rains during the 2021 Season B led to minimal harvests or total crop failure for many households. In Season B, crops planted in the forest area (generally by men) and wetlands performed better than those planted in the drier savanna area (generally by women). Those who received seed in Season A reported increased incomes from the sale of groundnuts, cowpeas and maize harvested from the seed provided. The selection of crops benefited women farmers by enabling them to increase sales of cowpeas and groundnuts. The additional income provided multiplier effects as farmers invested in other productive activities such as livestock, bicycles (for transport) and education, thus enhancing livelihoods.  

The assessment results showed that the seeds received and cultivated in Season A were, in general, saved by farmers for subsequent planting and incorporated into the local seed system. Most of the varieties distributed by CRS had been developed and disseminated over 30 years ago and were therefore already known to farmers, but displacement and other factors had affected seed quality and local availability. Farmers were positive in their evaluations of the varieties received through the CRS intervention, noting that their productivity and drought resistance were higher than traditional varieties. Looking forward, maintaining the quality of the seed within local seed systems will necessarily rely on the ability of farmers and grain traders to manage their seed stocks in ways that ensure varietal and physical purity and seed health.   

Conclusions

Impacts on food security

Out of the three interventions assessed, only one (DRC) showed a clear increase in food security that could be attributed to the seeds provided, and this increase was only realized in one out of the two seasons due to the effects of drought. In the case of Uganda, increases in food security in both projects were attributed to access to additional land, not to the seed provided. Unreliable rains affected production in the relatively drier north-eastern part of the district, resulting in production decreases among refugee households. 

Impacts on livelihoods

Despite adjustments made to the indicators and data collection tools, the impacts on livelihoods proved to be difficult to measure, attribute and triangulate. The overall conclusion was that any impacts on livelihoods that could be attributed to seed provisioning were relatively small. The exception to this was the case of Season A in DRC where increased crop sales were used to support investments in productive assets (livestock), transport and education. In Uganda, some vegetables (tomato, okra) cultivated with irrigation (from watering cans) appeared to have greater impacts than field crops due to their short duration and marketability, provided that the varieties met the market requirements. 

Impacts on seed systems 

In the case of DRC, the intervention usefully re-introduced quality seeds of appropriate improved varieties that were already known to farmers but had deteriorated over time. The projects in Uganda also provided seeds of appropriate improved varieties, though most of these varieties were already locally available. The varieties were incorporated into the informal seed system[2], which successfully allowed for the seed to continue to be available at a reasonable cost (i.e., through farmers and grain traders). In Uganda, the projects provided support to intermediary seed systems,[3] which effectively enhanced the capacity for local seed production but also led to unrealistic expectations in terms of seed sales and income generation. Within the study areas of both Uganda and DRC, seed providers within the formal seed system[4] (i.e., agro-input dealers in Uganda, agri-multipliers in DRC) were geared more towards seed sales to aid agencies rather than seed sales to farmers. 

Timing of the Interventions in Relation to the Crisis

All three of the projects assessed were implemented some years after the initial crisis, when farmers had already re-established their farming activities. This leads to two questions: (i) could the interventions really be considered as emergency interventions?, and (ii) would the impacts have been greater if the study had been able to assess interventions that were implemented in the season immediately following a crisis?[5] The nature of emergencies has shifted over time to become more and more complex and long-lasting; long-term crisis is increasingly normal. Although the implementing partners regarded the projects as emergency interventions, they also recognized the need to replace repeated emergency seed distributions with a more strategic approach to build resilience.

In relation to the second question, the assessment teams compiled data on what the beneficiary farmers planted in the first season following their arrival in the refugee camps (Uganda) or their return home (DRC) and how they acquired the seed. In both countries, farmers had purchased or exchanged seeds of all crops from local markets and from other farmers. Some farmers also reported having selected grains of improved maize varieties from their food rations for planting.  None of the farmers met by the assessment teams reported that they were unable to plant due to lack of seed. While this does not directly answer whether the impacts may have been greater if the study had been able to assess seed interventions immediately following a crisis, it does question the rationale on which many emergency seed interventions are based. Both this study and others suggest that seed is often locally available to farmers following a crisis.   

Resilient Food and Seed Systems Require Longer-Term Interventions

The study shows that different factors apart from seed contribute to crop production and food security, and these factors (e.g., access to fertile land, rainfall) directly affect the overall impacts of a seed intervention. The provision of seed alone cannot be expected to increase food security within a single season or a single year. Given the increasingly protracted nature of emergencies, longer-term interventions that can more accurately assess and address the challenges faced by farmers and enhance the resilience of local food and seed systems would appear to be more impactful. 

The resilience of food and seed systems can be enhanced by the introduction (or re-introduction) of appropriate varieties (e.g., those that have resistance to drought; those that meet market demand) and the incorporation of these varieties into local seed systems so that farmers can effectively access them through their normal supply channels (i.e., from own-saved seed, other farmers and local markets). While the interventions assessed were generally successful in providing a range of appropriate varieties, they were less successful in sustainably enhancing farmers’ access to quality seeds of these varieties through informal, intermediary or formal seed systems. Given the importance of informal seed systems, much greater emphasis should be given to seed-saving by farmers, and seed management by local (grain) traders who also provide seed. Appropriate support for different seed systems requires longer-term interventions that are based on a good understanding of the nature of demand and supply in informal, intermediary and formal seed markets.