The Extreme Citizen Science in Agriculture project (eCSAgri) was a six months pilot project helping to facilitate collaborative and participatory Climate-smart Agriculture (CSA) with three smallholder farming communities in three agro-ecological zones in South West Nigeria. The project ran from January to June in 2021 and around 15 farmers in each community were engaged together with 5 agricultural extension officers, who are professionals that support farmers and act as intermediaries between farmers and government and research institutions. Around 4 farmers in each community owned a smartphone – smartphones were not provided and only data bundles and a solar power bank per community was provided where needed. The project was led by the Environmental and Economic Resource Centre (EERC), Nigeria in collaboration with Mapping for Change, UK and ExCiteS and was funded by the UK Research and Innovation (UKRI). eCSAgri can be situated at the intersection of Citizen Science and CSA, which can be briefly described as an integrated approach for promoting productivity-enhancing, ecosystem-friendly, climate change-resilient and inclusive farming systems.
The work focused on: 1) exploring, through the implementation of co-designed ‘working’ prototypes, how digital tools can support – directly or indirectly – any smallholder farmer in reporting and mapping farming issues and receiving timely advice from extension officers and other farmers; 2) discussing and prototyping how the land use data and knowledge shared in these farmer to farmer and farmer to extension officers interactions can be appropriately converted into key information for making evidence-based decisions. To address some technological gaps that were identified at the early stages of the project we explored the implementation of Sapelli and Community Maps in two different ways aiming to contribute to enhance such collaborations. The process and the technology used is represented below and briefly explained in the next sections.
As part of the engagement and Free Prior and Informed Consent process we co-designed, with few iterations due to logistics and other constraints, the Sapelli project (pictorial interfaces with captions in Yoruba) based on the feedback received from farmers and extension officers and the EERC staff’s extensive knowledge on farming related issues such as soil, pests, diseases, weather, storage or market. This first phase focused on engaging farmers and extension officers, and identifying what do people want to report/map. The next phases focused on exploring how this can be mapped and shared. In terms of mapping, we discussed and agreed that information about both location and farm size was important and we tried to adapt the technology accordingly to allow farmers to not only locate their farms using GPS, but also to demarcate them using satellite imagery. The images below represent these two mapping techniques and the text and pictorial interfaces for describing a geometry, i.e., the semantics of a point, line or a polygon in a map – the map legend and meaning.
The image below ‘summarises’ the socio-technical approach of the sharing knowledge phase by showing a screenshot of the Sapelli Nigeria WhatsApp group where farmers can message maps to report farming issues and receive timely advice from extension officers and other farmers who are facing similar issues. Maps can be created and added to WhatsApp (or other messaging apps) while offline, but they will only be sent automatically when online. The maps can also be shared via SMS.
It’s important to highlight here that some extension officers are responsible of supporting hundreds of farmers – digital tools can help.
The dynamic community map resulting from the mapping and sharing phases is generated from the information shared with Sapelli, either privately through GeoKey or WhatsApp, or as open data. The images below show the two different technological approaches for visualising and analysing the data. On the left, the Mapping for Change’s Community Maps with the data being stored in GeoKey, and on the right, a Sapelli Progressive Web App (PWA) mapping prototype showing the contributions shared in the WhatsApp group, which are automatically added to the community map in each device when the URL is clicked in the WhatsApp conversation. Alternatively, in the temporarily password protected PWA prototype, users can share their contributions as open data (Carto database) for anyone to view, comment or delete. This open-for-all approach and functionality was not mentioned nor used – open land user-generated land use data required further discussions and time was limited.
Towards the end of the project we quickly prototyped a Spatial Decision Support System (SDSS) using the ArcGIS Operations Dashboard to explore how the farmer/land user-generated data could be aggregated in order to provide key, validated information for decision makers – the data processing and validation is represented with the ‘transparent box’ shown in the second image of this post. Below is included a screenshot of the SDSS showing climatic data and not-real, aggregated data about total hectares affected by crop type and by farming issue in each of the three local governments. As mentioned, little progress was made in this front and only very few farmers have reported farming issues since the mapping phase started in mid April.
In the last month of the project, EERC together with one of the engaged extension officers organised a meeting with around 40 extension officers from the Odeda local government to present this pilot project and receive feedback. Few days later, for the stakeholders’ meeting, EERC brought together 14 farmers and community heads (4-5 per community), 5 extension officers, members from the Federal University of Agriculture Abeokuta (FUNAAB), the Nigerian Meteorological Agency (NIMET), the State Ministry of Environment, the Ogun State Agricultural Development Program Office (OGADEP) and the Federal Ministry of Agriculture to discuss about the project and the way forward.
Six months go fast. Though the feedback regarding the appropriateness of the methods and the technology used were positive overall, the technical and/or design issues experienced by farmers with the apps, the unequal participation in the WhatsApp group, the sometimes rather inaccurate (spatially, not semantically) contributions when using satellite imagery, and the many other socio-technical challenges… highlights the need of, among other, long-term action research funding and extensive technological support for developing and scaling digital innovations in rural Sub-Saharan Africa (SSA) following human-centred design principles.
In terms of mapping, this pilot project suggests that in this and likely in similar contexts, messaging (text, images, audio, video, etc) for asking and receiving immediate farming advice is more important than mapping in the short term. That said, land use mapping by those who use the land is also important. If land users don’t map their land, sensors and machines will increasingly do it anyway, and these maps made by machines will feed national and planetary-scale environmental management systems that will not capture local knowledge (thus will be incomplete) and might ignore the needs of land users. For these reasons, we tried to adapt mapping to messaging, and not the other way round.
With many thanks to the people involved in the project in one way or another: Ayoade Adetoye, James Michael, Motole Okonkwo, Osilewe Mukaila, Loveth Opeyemi, Favour Nwankwo, Helen Onileimo, Adedoyin Adeleke, Bukola Osuntade, Luke Okojie, Maria Alonso, Hannah Stockwell, Louise Francis, Francisco Sanz, Judy Barrett, Muki Haklay, Claire Ellul, Jerome Lewis; all the farmers, extension officers and stakeholders from Osiele, J4, Saki, Abeokuta and other places in Ogun and Oyo states; the PROCOL Kenya people and Megan Laws, and the crew from the FUNAAB Guest House and immediate surroundings. Also many thanks to the ExCiteS team; ExCiteS, Natural Aptitude and other open source software developers; UKRI and ERC; Carto, Planet, ESA and ESRI.