#main-navigation ul.navbar-nav li a:hover { background-color: #444444 !important; }
Skip to main content

Statistics professor helps Costa Rican researchers improve their study methods

  • By Amanda Bradford
  • 575-646-3223
  • ambradfo@nmsu.edu
  • Sep 09, 2013
Members of a Costa Rican research team prepare to collect animal population data.

Researchers visit Costa Rica all the time to study its rich biodiversity - after all, it has more than 500,000 species of plants, animals and fungi. Most of the time, the researchers collect their data and head back home to summarize and publish the information they've collected.

But an applied statistics expert from New Mexico State University recently traveled to one of the Central American country's most beautiful and diverse areas for a different reason: to help local Costa Rican researchers improve their own studies by evaluating their survey methods and finding more effective ways to track and monitor animal populations.

William Gould is a professor in NMSU's Economics, Applied Statistics and International Business Department in the College of Business. He regularly consults with researchers on improving their data collection and analysis methods and this collaboration sometime requires him to travel- most recently visiting Corcovado National Park in Costa Rica in April. The park is internationally renowned for its biodiversity, but the dense rainforest vegetation makes it difficult to monitor animal populations using traditional track survey methods. Thick groundcover and frequent rain during much of the summer and fall seasons leave little evidence of animal movements for scientists to discover and record.

"Some animals of particular concern, like the white-lipped peccary, don't use the park's limited trails much, so it's extremely difficult to find their tracks off the trail," Gould said. "But just because you visit an area and don't see something, doesn't mean it's not there."

Gould's primary research interest has been in abundance estimation methods - determining how many animals of a species are present in a given area. He was able to bring the Costa Rican scientists up to speed on techniques like camera traps that use motion sensors to photograph passing animals and capture-recapture methods, in which a sample of animals is briefly caught, marked and released. Later, another sample is caught or re-sighted, and researchers use the proportion of tagged animals in this sample to estimate the total population of the area.

Each of these methods can provide a better measurement of the animal population than simple track surveys in which researchers locate and document tracks on the ground. With a more rigorous sample design, Gould suggested that the researchers employ occupancy modeling for monitoring changes in the population distribution- using multiple data collection visits to identify and adjust for imperfect detection of animals in the field.

Gould also introduced occupancy modeling to scientists in Peru in 2009 and in Portugal in 2012. In all of these settings, he said communicating with researchers about the constantly developing methodology of population monitoring helps improve the knowledge base of researchers throughout the global scientific community.

"The only way we make progress in science is by accumulating knowledge," he said. "It makes me feel like I'm doing something worthwhile, being able to introduce researchers to new methodologies that can improve their conservation efforts."

While trudging through the rainforest in 90 percent humidity and scouring the dense, green foliage in search of jaguars and margays might not seem like a typical activity for a statistics professor, Gould said it is important to get in the field and see the landscape and conditions that ultimately affect what can and cannot be done.

"There's always been a natural connection between life and social sciences and statistical analysis," he said. "Almost every field collects data and summarizes information using statistical methods, but most statisticians have a keen appreciation for the limitations of those methods and the importance of correct interpretations."

That's why the cutting edge occupancy models that Gould shared with the researchers will be an important component of future monitoring efforts - resulting in improved collection techniques that the scientists will use to draw more-accurate conclusions about the number or distribution of animals and the patterns and dynamics of those animals over time.

Gould is happy to report that the director of the National System of Conservation Areas has agreed to implement some of his recommendations.

Interim College of Business Dean Kathy Brook said Gould's success in developing international relationships that benefit the college makes supporting his research an easy choice.

"Bill's work with researchers in Costa Rica on animal population dynamics is a great illustration of the interaction of the faculty with the international community in ways that benefit all of us," she said.

Gould said he's grateful for the college's financial support of his outreach efforts, which he hopes will continue to have an impact on future research.

"Being able to visit countries that don't have information on recent methodologies, you're really bringing them along and improving what they can do in these incredibly rich, biodiverse environments that are extremely important, not just to them but also to the world," he said. "The more we know about these highly diverse ecosystems using rigorous survey methodologies, the better off the world is going to be."