Local | PANORAMA

South America

Manejo integrado de vicuñas que promueve la salud, bienestar y medios de vida de comunidades en el área protegida de Apolobamba

Fabian

Beltran

Coordinación multisectorial para el monitoreo, vigilancia y manejo adecuado de vicuñas

Monitoreo y vigilancia del estado de las poblaciones y de la salud de vicuñas

Desarrollo de capacidades en bienestar animal, bioseguridad y obtención de fibra de mayor calidad

Implementación de estrategias para fortalecer la conservación y la salud de vicuñas y de su hábitat

South America

Manejo integrado de vicuñas que promueve la salud, bienestar y medios de vida de comunidades en el área protegida de Apolobamba

Fabian

Beltran

Coordinación multisectorial para el monitoreo, vigilancia y manejo adecuado de vicuñas

Monitoreo y vigilancia del estado de las poblaciones y de la salud de vicuñas

Desarrollo de capacidades en bienestar animal, bioseguridad y obtención de fibra de mayor calidad

Implementación de estrategias para fortalecer la conservación y la salud de vicuñas y de su hábitat

South America

Manejo integrado de vicuñas que promueve la salud, bienestar y medios de vida de comunidades en el área protegida de Apolobamba

Fabian

Beltran

Coordinación multisectorial para el monitoreo, vigilancia y manejo adecuado de vicuñas

Monitoreo y vigilancia del estado de las poblaciones y de la salud de vicuñas

Desarrollo de capacidades en bienestar animal, bioseguridad y obtención de fibra de mayor calidad

Implementación de estrategias para fortalecer la conservación y la salud de vicuñas y de su hábitat

South America

Manejo integrado de vicuñas que promueve la salud, bienestar y medios de vida de comunidades en el área protegida de Apolobamba

Fabian

Beltran

Coordinación multisectorial para el monitoreo, vigilancia y manejo adecuado de vicuñas

Monitoreo y vigilancia del estado de las poblaciones y de la salud de vicuñas

Desarrollo de capacidades en bienestar animal, bioseguridad y obtención de fibra de mayor calidad

Implementación de estrategias para fortalecer la conservación y la salud de vicuñas y de su hábitat

Image recognition algorithms for jaguar detection and identification

Jaguar identification model

Tech4Nature Mexico

The image recognition component, based on Convolutional Neural Networks (CNN) in the framework of the Tech4Nature Mexico pilot, plays a fundamental role by: i) Automatically detecting the presence of jaguars in camera trap captures, thus speeding up data processing; and ii) Automatically identifying individual jaguars in the region, which enhances the understanding of local populations. This approach is of vital importance in conservation by taking advantage of advanced methods that allow a faster and more detailed analysis.

Data collected from the devices and strategic alliances with Huawei, UPY, and other expert conservation organizations for data sharing. During a full school year, a group of young data engineering students from UPY were dedicated to the development of the image recognition models. Given the possibility that the models could be biased in the recognition of jaguars due to the students' lack of experience in monitoring this species, the group received training and feedback from a biologist specialized in feline conservation in Yucatan.

We faced a considerable challenge in developing automatic models for the detection and identification of jaguars in images. This task is complicated not only by the scarcity of available data, but also by the limited amount of images captured by camera traps containing the species of interest, due to its critical conservation status. These obstacles have been notable in the initial stages of the project, prompting us to collect animal images from a variety of sources to expand our dataset. However, complexity increases at this stage due to additional factors.

Implementation

Local

Bioscanner platform

Camera traps and eco-acoustic monitoring devices deployment

Benjamin Campos deploying an audiomoth

Tech4Nature Mexico

The local team strategically placed 15 camera traps and 30 eco-acoustic monitoring devices (audiomoths) within the mangrove and lowland rainforest habitats where jaguars have been previously sighted. This deployment effectively captured the region's biodiversity and generated valuable data for subsequent analysis.

Field research, in conjunction with active participation from the local community and insights gained from co-design efforts, pinpointed the optimal locations for deploying cameras and audiomoths. These devices were strategically positioned in less disturbed areas of the mangroves, jungles, and savannahs, ensuring the success of our scientific survey.

Collaborative site characterization and mapping with the local community served as a crucial foundation for the successful deployment of these devices. However, we also encountered challenges, including wildfires and extreme events, which temporarily impeded both device placement and data collection efforts.

Implementation

Local

Community co-design and engagement

Local community

Tech4Nature Mexico

The engagement of local leaders was integral from the project's inception, entrusting them with the characterization and selection of sampling and monitoring sites. Their insights and requirements were actively incorporated into the project's analysis. Timely presentation of results, widespread dissemination of their work and expertise, and inclusion in working meetings were paramount.

The Ministry of Sustainable Development from Yucatan has been engaging and working with the local communities living in and around the Reserve for several years, ensuring cross-pollination of knowledge, good governance and justice.

Moreover, the C Minds' AI for Climate initiative established a robust four-year collaboration with the Yucatan government, essential local stakeholders representing academia, innovation, and civil society sectors.

The comprehensive involvement of the local community across all project stages, encompassing design, deployment, data collection, and analysis, emerged as a pivotal and indispensable factor contributing to the project's successful implementation and the acquisition of valuable biodiversity information within the reserve.

Entirety

Local

Generation of inputs for the strengthening of AI tools and resources for biodiversity protection

Multi-stakeholder alliance

Tech4Nature Mexico

Among the strengths of the pilot is the ability to translate learnings into opportunities and recommendations, especially on issues of innovation, digital transformation and technological ethics for biodiversity protection. For this reason, we closely monitored the implementation of the pilot to develop a public report with a recommendations section, fed by the experiences, inputs, achievements and learnings of the implementing team.

What was learned at each step and with each partner contributed to strengthening AI tools and methodologies for biodiversity protection.

Beyond the boundaries of the Reserve, the Tech4Nature Mexico project has sparked a transformative wave in regional conservation efforts. The fusion of advanced technology with multi-stakeholder collaboration is redefining biodiversity protection. Innovative tracking algorithms have revealed crucial data confirming the presence of threatened species in an unprecedented way. These revelations enrich our understanding of regional ecology and empower local communities, driving lasting commitment to conservation.

Planning phase

Local

Multi-national