Deploying crowdsourcing and seed diversity in disaster recovery efforts in Nepal

Newly launched CGIAR Research Portfolio tackles growing complexity of agricultural development challenges

With an urgent need for seed in remote farming regions, this second report from Nepal explains why Bioversity International and partners are using a crowdsourcing approach to understand farmers’ preferences for varieties in four project sites.

Bioversity International Senior Scientist Bhuwon Sthapit and colleague Devendra Gauchan, send their second report from Pokhara, Nepal, three months after two devastating earthquakes killed and injured thousands of people, destroying homes, food and seed stores.

Focusing on the urgent need for seed in the affected remote farming regions in Nepal, they explain how Bioversity International and partners are using a crowdsourcing approach to understand farmers’ preferences for varieties in four project sites as part of the disaster recovery efforts.

By Bhuwon Sthapit and Devendra Gauchan

In our first report, ‘Matching seeds to needs in the aftermath of the Nepal Earthquakes’, we spoke about the urgent need to replenish lost and damaged seeds stocks which had been lost in the remote regions that depend on agriculture for food and income. We also explained why, even in the urgent rush to get replacement seeds to farmers, it is vital to ensure that replacement seeds are ones that match the farmers’ needs – both in terms of suitability for local growing conditions and to meet cultural preferences.  For this to happen, the farmers must become part of the selection and evaluation process.

Bioversity International has a long history of working with farmers in this participatory way, including in Nepal, where we have worked closely with a network of researchers, farmers and community groups for more than 30 years. More recently, we have been developing this method into an innovative crowdsourcing approach, through our ‘Seeds for Needs’ Initiative, which works with more than 20,000 farmers in 11 countries on many different crops, such as wheat and barley in Ethiopia, beans in Honduras and rice and wheat in India.

Crowdsourcing places the farmer at the centre of the research process. Instead of having a ‘trial plot’, seeds are distributed to farmers to grow in their fields, so they can be evaluated in localized growing conditions. Each farmer receives a set of three varieties to test, that could have come from a genebank, a breeding programme or be a local farmer variety. They are not told which varieties they are testing, but asked to rank the varieties for specific traits that are important to them, such as early-maturing varieties, as well as a general appreciation score. This helps identify which varieties to prioritize for distribution to farmers.

It is this expertise that formed the basis of a workshop held in Nepal in June, to strengthen the capacity of experts to carry out rapid detection and dissemination of farmer-preferred varieties using crowdsourcing methods in areas affected by disaster or at high risk of being affected in the future.

Bioversity International scientists from Nepal, India and Colombia, with technical support from the CGIAR Research Program on Agriculture, Climate Change and Food Security, led a 2-day training course attended by plant breeders, agronomists, social scientists and other researchers and development professionals. The training gave participants practical knowledge on:

  • the concept and application of crowdsourcing for variety selection and dissemination
  • ClimMob, a data analysis software developed by Bioversity International, to analyse a large amount of farmer preference data quickly
  • how to integrate i-Buttons – innovative environmental sensors that record temperature and relative humidity – into crowdsourcing approaches.

The training stimulated a lot of discussion about the best way to adapt crowdsourcing protocols in the context of Nepal, with the participants concluding that it would best be deployed as complementary to existing participatory variety selection, which could also be further refined.

The final outcome was a decision by participants from the GEF/UNEP project* to pilot crowdsourcing methodology in four project sites – Jumla, Humla, Dolakha and Lamjung – for rapid detection of farmer preferred varieties of target crops in 2015 and 2016.

Also agreed was a priority portfolio of ten varieties for a number of target crops (rice, millet, beans, buckwheat and amaranth), with 300 seed sets for each. This means that each variety will be tested in 100 fields in each community participating in the project, to select the most preferred ones for each location.

Read the full workshop report:
Diversity Kits and Crowdsourcing Approach: Workshop Report, June 29-30, Pokhara, Nepal (pdf)

Read the first report: 
Matching seeds to needs in the aftermath of the Nepal Earthquakes 

View related Nepal photos on Flickr

This work is part of: the GEF/UNEP Project: Integrating Traditional Crop Genetic Diversity into Technology: Using a Biodiversity Portfolio Approach to Buffer against Unpredictable Environmental Change in the Nepal Himalayas; the Global Crop Diversity Trust/Bioversity International/Nepal Agricultural Research Council Project: Rebuilding local seed system: Collection, conservation and repatriation of native crop seeds in earthquake affected areas in Nepal and Netherlands; the Genetic Resources Policy Initiative/Bioversity International project: Rebuilding Family Farming in earthquake affected areas in Nepal.

It is also carried out with support from the CGIAR Research Program on Climate Change, Agriculture and Food Security, the Local Initiative for Biodiversity Research and Development (LI-BIRD), the Nepal Agricultural Research Council, the Department of Agriculture, Development Fund, Norway and the Swiss Development Corporation (SDC), Nepal. 

Photo: Farmers participating in trials of proso and foxtail millet varieties, Nepal. Credit: Bioversity International/B. Sthapit