loading page

Wildlife Insights: How Camera Trap Data Can Foster Global Biodiversity Conservation
  • +6
  • Fabiola Iannarilli,
  • Ruth Oliver,
  • Tanya Birch,
  • Sara Beery,
  • Eric Fegraus,
  • Nicole Flores,
  • Roland Kays,
  • Jorge Ahumada,
  • Walter Jetz
Fabiola Iannarilli
Yale University

Corresponding Author:fabiola.iannarilli@yale.edu

Author Profile
Ruth Oliver
Yale University
Author Profile
Tanya Birch
Google Inc
Author Profile
Sara Beery
California Institute of Technology
Author Profile
Eric Fegraus
Conservation International
Author Profile
Nicole Flores
Conservation International
Author Profile
Roland Kays
North Carolina State University,North Carolina Museum of Natural Sciences
Author Profile
Jorge Ahumada
Conservation International
Author Profile
Walter Jetz
Yale University
Author Profile

Abstract

Human activities are driving environmental and climatic changes, affecting the distribution and diversity of species worldwide. Limiting the negative impacts of these activities on wildlife requirestimely knowledge of status and trends in populations over large scales. Camera trapping providesopportunities to simultaneously collect information on several species over large spatio-temporalscales. However, the time required to process large collections of images, the statistical andprogrammatic skills needed to analyze large sets of data, and a general lack of homogeneity in metadata standards hinder the use of camera trapping for local and global conservation. WildlifeInsights (http://wildlifeinsights.org/; WI) is a web platform that promotes and supports the use andsharing of camera-trap data for species conservation and promotes the mobilization of records thatotherwise might be permanently siloed in private data-storage units or lost over time. WI speeds upthe processing of images via an AI model trained to classify >700 species, and automates commonstatistical analysis through a standardized, accessible user interface. It also provides tools to addresscommon issues faced by camera trappers, such as the need of hiding locations of sensitive speciesand removing images of humans, and has a transparent infrastructure to request, share and citedatasets. Although only recently open to the public, the platform already hosts tens of millionsrecords, most of which publicly accessible, from more than 50 countries and 1000 species. Using datashared in WI, we assessed whether information collected using camera traps improved the spatial,temporal, taxonomical, and ecological coverage of many species compared to records available inmore traditional open-access repositories such as GBIF. Birds and mammals, and countries with ahigh proportion of remote areas and biodiversity had the largest increases in coverage. Compared toother traditional methods, camera traps also provided fi ner-resolution temporal information, oftenreplicated over time. Our results showed the importance of sharing camera-trap data for conservationand highlights WI’s role as an invaluable resource to timely inform biodiversity conservation in achanging world