Project abstract for group packerc

Snapshot Serengeti: Crowdsourcing Camera Trap Data for Serengeti Research

Camera traps - remote, automatic cameras - are revolutionizing ecological research by providing low-cost, non-invasive tools capable of comprehensively monitoring multiple animal species at large scales. This new technology promises to fill a conceptual gap in ecology by providing data on multiple species across broad spatial and temporal scales to test predictions about species coexistence and community dynamics in natural ecosystems. These researchers currently run the world’s largest camera trapping survey to understand the diverse community dynamics in Serengeti National Park, Tanzania, specifically evaluating mechanisms underlying predator-predator, predator-prey, and herbivore-herbivore coexistence. By deploying more than 200 camera traps to simultaneously monitor a 1,100 km2 study area, they are able to evaluate these questions in a natural system to a degree that has not been done before. The camera trapping data also provide the basis for an international citizen science project, Snapshot Serengeti, that has engaged more than 150,000 volunteers and has brought the university of Minnesota widespread media attention. The 225 camera traps produce about 1 million images and over 1 TB of data annually; the researchers have approximately 4.5 TB of images collected to date and anticipate continuing the project for five years. The group uses MSI for storage for the images as well as housing an existing relational database that stores metadata for the images to be used in ongoing research.

Return to this PI's main page.