What problem were they solving? 

Although Guatemala has not historically suffered from water stress, climate change has disrupted rainfall patterns, and 45 percent of the country is now vulnerable to drought. This poses a particular threat to the country’s smallholder farmers who rely on rain-fed agriculture for their livelihoods. In the volcanic region of Santa María de Jesús in the Southern part of the country, many of them specialize in cultivating snow peas – a key export crop for over 35,000 Indigenous farmers in the Guatemalan highlands. To access export markets, they need a certification that depends on limiting the use of chemical fertilizers. This means that farmers have to rely on smart water management to obtain good yields, something that cannot be determined on a regional basis. Micro-variations in the location of their plots (e.g. on slopes or valleys) determine how much rainfall and water are available. Farmers would like to estimate the effect of rainfall on crop yield for each plot, so as to be able to intervene effectively during droughts. This estimation requires the collection of plot-by-plot data on soil conditions, rainfall and crop yield. 

Some farmers in Santa María de Jesús are adapting to water shortages with DIY storage solutions like plastic water tanks to capture rainfall. While experimentation is yielding small-scale solutions, farmer associations rarely interact, making it difficult to pool this expertise and learn what works. Smallholders are also largely disconnected from the valuable expertise of agronomists, who could guide the discovery of more effective adaptation strategies if they had better insight into the environmental variations experienced by the farmers. This dynamic creates a distance gap between the lived experience of agricultural communities and scientific knowledge on climate adaptation.  

What did they do? 

The UNDP Accelerator Lab in Guatemala, in partnership with civil society organizations Tikonel and UNDP’s Volcanoes Project, developed a citizen science initiative involving 20 farmers from the Santa María de Jesús region recruited through both local associations of smallholders. Aged 18 to 75, farmers also included Indigenous women. With the Lab’s facilitation, they designed a process for data collection simple enough to be sustainable, but standardized enough to ensure data can be aggregated over time and compared between different farm plots. They used low-cost sensors (DIY hygrometers) and KoboToolbox, an open-source platform to capture data, and WhatsApp to communicate with each other.  

After experimenting with different types of sensors, farmers agreed that the most strategically important variables to monitor were rainfall and soil moisture. Farmers aggregate this data and share their readouts through a digital platform developed through this prototype. The platform can be accessed by both farmers, agronomists and decision makers from the Ministry of Agriculture and the Ministry of Natural Resources and Environment. This will also help experts provide the farmers with more personalized guidance, increasing scientific knowledge about farming snow peas in the highlands.  

What was the benefit of using collective intelligence for this issue?  

Collective intelligence provided a way for farmers’ organizations to co-create the information platform and steer design towards data that is needed most to help them make decisions about irrigation and disease management solutions. Using a citizen science approach enabled the initiative to open a conversation with agronomists and experts from outside the region, helping to shrink the distance gap between credentialed experts and farmer scientists.  

Bringing together different smallholders from the region surfaced useful insights into local agricultural practices, including using chlorine for contaminated water sources. The collaborative citizen science method also facilitated peer exchange whereby farmers teach one another in their local languages. The co-creation sessions, leading to the choice of which data to collect and the protocols to collect them, also created conditions for women who do not know how to read and write to share their input. This sharing of expertise fast tracks the adoption of climate adaptation strategies as smallholders learn what works from others whose plots face similar environmental conditions. 

What does this experience tell us about collective intelligence for climate action? 

The emphasis on co-design and working with accessible, locally appropriate technologies like rain gauges and DIY sensors helped this initiative engage a wider range of people and perspectives into the initiative. This included women, older and younger farmers, and Indigenous Peoples, typically underrepresented in agricultural digital innovation initiatives. 

This collective intelligence climate prototype exemplifies and articulates a trade-off inherent to collaborations across different forms of expertise. Farmers, who are in charge of data collection, are constrained to keeping it simple and manageable: they need to use sturdy sensors, which must be affordable and cheap enough not to be subject to theft. On the other side of the collaboration, agronomists need higher-quality and more granular data collected longitudinally which are more burdensome to collect. It is hoped that if the farmer scientists adhere to the standardized protocol and maintain data collection over time, the data quality will keep improving and drive higher acceptance by agro-experts in the long term.