University of Botswana Okavango Research Institute (2017)
Scaling fisheries monitoring in the Okavango Delta with low‐cost technology
Recognized as both a Ramsar Convention wetland site of international importance and UNESCO World Heritage Site, the Okavango Delta isa globally important freshwater wetland. Covering an area of 12,000 square kilometers and spanning three countries, the Delta supports a rich and diverse wetland ecosystem in the otherwise arid southern African region. In addition to the unique flora and fauna supported by the Delta, the ecosystem supports services ranging from crop irrigation to a growing tourism economy. Of particular importance is the Delta’s support of subsistence and small scale commercial fisheries, which are the livelihoods of thousands of people.
Until recently, the Okavango Delta has remained largely intact, with relatively low levels of impact from human induced land‐use change. However, with rapidly increasing agricultural infrastructure in the upstream reaches of the delta, the future health of the fishery and ecosystem as a whole will depend on careful monitoring and conservation management. Unfortunately, the data necessary to assess the health of this ecosystem are not available there; nor is there a strategy or the resources to collect these data over time. This project will help address that gap by partnering with a variety of stakeholders to pilot low-cost fisheries data collection protocols using the iNaturalist mobile biodiversity observation and data collection platform.
Key Objectives and Activities
- Record high quality fish monitoring data (e.g. observations including date, location, verifiable photo, length, and species ID) to iNaturalist from 4 sources: ORI existing monitoring efforts, Department of Fisheries, local fishermen, and recreational fishing tour operators.
- Upload iNaturalist data from the project to the OkavangoDataBase and to the Global Biodiversity Information Facility (GBIF).
- Generate initial estimates of the volume of high-quality data per unit cost from each partner.
- Increase the volume of observations per unit cost while maintaining data quality.
- Increase awareness of using iNaturalist in fisheries monitoring by offering training in Botswana and the surrounding region.
- At least 100,000 ecosystem-scale fish monitoring records from all partners uploaded first to iNaturalist, then to the OkavangoDataBase.
- Work with the iNaturalist team to modify the software to include a reputation system that allows accurate non-expert identification, a Computer Vision program that automates portions of identification, and real-time computer vision identification available to end users.
- Decrease the cost per observation by a factor of ten, while maintaining a quality threshold of 99% accuracy of identifications.
- Report outlining recommendations for most cost effective means of continuing fish monitoring in the Okavango Delta region.
- Host 8 meetings and 2 workshops that bring together fishing organizations, ORI scientists, and iNaturalist staff to iteratively work together to understand drivers of data collection trends and alter incentives and technology accordingly.
The team expects that this pilot project will result in strategies to decrease costs of collection of high quality data that can be used to inform policy. Evidence of successful sustained data collection will be measured as volume and quality of data that flow through iNaturalist into the OkavangoDataBase from each of the three external data providers: the Division of Fisheries, local fishers, and tour operators. If the project is successful, capacity building efforts during the course of this project will ‘institutionalize’ data collection among a growing network of tour operators and motivate data collection from a growing fleet of local fishers. The team anticipates that increased availability of data on the Okavango Delta’s biodiverity will allow ORI to manage and disseminate data products relevant to policy makers, such as annual assessments of fish stocks in the region, resulting in more informed and data-driven decision making throughout the region. The ultimate outcome will be the sustained use of the mobile collection platform in the Delta for routine fisheries monitoring and management.
Last Updated: November 15th, 2017
Project Director Biography
Professor Keta Mosepele is an Associate Professor in Fisheries Biology and Management at the Okavango Research Institute and is currently its Acting Director. He has also been an Acting Deputy Director responsible for Research Services and Training at the institute for over 4 years. Before joining the ORI, he worked as a Research Officer in the Fisheries Division and was responsible for designing the long term fish monitoring protocol currently in use. He has over 50 peer reviewed publications in floodplain fisheries management and ecology, and has participated in several national and regional research projects. His is interested in using inexpensive data collection methods to inform floodplain fisheries management, as the high cost of monitoring programs is a major hindrance to efficient and effective management of most freshwater fisheries in the developing world.
Notes from JRS
This project is our second in as many years to the Okavango Research Institute and adds to capacity development efforts in the region. The iNaturalist technology was successfully deployed in our grant to the Kew Royal Botanical Gardens for work on the Red Listing of plants in Madagascar. Should this project prove successful and the technology prove cost competitive and accurate, JRS would seek its transfer and use in other African fisheries. Success will depend on the technical and social networks that maintain species identification accuracy, aligning the technology and project with the local rewards and incentives, and the economics of the system. It will be challenging to the team to do a full-cost accounting of the method and of the system accuracy to promote its official adoption in the Fisheries Division.