2020 | Global Shark Populations

FACULTY SEED GRANT | Global Change Center

Taking the pulse of global shark populations

INVESTIGATORS:

• Dr. Francesco Ferretti, Fish and Wildlife Conservation
• Dr. Edward Fox, Computer Science
• Dr. Trevor Hastie, Statistics, Stanford University

Sharks are among the most endangered vertebrates in the ocean. Many shark populations are declining worldwide because of fishing and a quarter of the extant species are threatened with extinction. Meanwhile, for about half of the species we do not have enough data to inform their conservation status, a condition hampering management and conservation.

Surveying sharks over large ocean regions is expensive and impractical. However, today’s connected world gives us a tremendous opportunity to rely on crowdsourcing for gathering biological information. People engage in outdoor activities, capturing their experiences with media devices able to record GPS position, date and time of videos or photos, and other auxiliary data (Fig. 1). Smartphones, in particular, are the all-in-one devices people use to communicate, capture media, and share them in social networks (SN). With 3.5 billion smartphone users worldwide, 60% of the global human population living close to the coast, and 8 to 9.6 billion people a year traveling to coastal destinations, it is clear that tapping this observation effort would fill the data gaps in sharks very quickly.

Here we propose a multidisciplinary data science initiative called sharkPulse to crowdsource information on sharks’ abundance and distribution. We will create the largest global shark sighting database by mining and cataloging shark images disseminated throughout the web, SN, and personal archives. Through an operational workflow, we are transforming unorganized shark photographs dispersed online into biologically relevant data. Images obtained through web and smartphone apps, sourced from online repositories, and acquired from private archives, will be processed to extract biologically information, cataloged taxonomically, and transformed into occurrence records. These data will inform statistical models to address specific life-history, ecology, and conservation questions, and finally inform applied management, conservation, and policy.

Reference citations for project proposal description available upon request.