
(image: a 25-centimeter black rhino footprint, courtesy of WildTrack)
Describing their efforts at Duke's weekly Visualization Friday Forum, the founders of the WildTrack project said that monitoring black rhinoceroses movements in Zimbabwe during their 1992-2000 field work first convinced them that conventional "invasive" methods of following wild animals with radio collars can be counterproductive.
For one thing, such collars are not designed for rhinos' bullet-like heads -- easily slipping off or sometimes causing deep neck wounds if they don't. Tests showed collar failure rates of almost 90 percent after two years of use. And the routine animal immobilizations required to fit and maintain the radio collars caused female fertility rates to significantly decline.
"We began to think 'what can we do?'" Jewell said. "The answer was right in front of us."

(image: A black rhino footprint with distinctive landmarks indicated. Courtesy of WildTrack)
That has led to their Footprint Identification Technique" (FIT), a way of using statistics and computer algorithms that can identity individual animals, their species and in some cases their genders with accuracies of up to 90 percent. The researchers are in the Triangle now working with SAS Institute, the Cary N.C. software firm, to develop ways to improve their techniques with SAS's JMP visualization software.
FIT has already been successfully adapted for not only black and white rhinos but also on other iconic species such as Bengal tigers, African lions and polar bears, Jewell and Alibhai said.
1 comment:
This has always been one of my favorite SAS and JMP customer stories. I'm glad to see it here.
Post a Comment