Volume 4. No 3
Information Technology News of the University of California, Davis November 1995
Information Technology News of the University of California, Davis November 1995
From her computer station on the second floor of Wickson Hall, Karen Beardsley is helping count elephants half a world away.
Using the Geographic Information System (GIS) software ARC/INFO, Beardsley is refining estimates of how many elephants remain in the Central African rainforest - a region where the closed tree canopy and dense undergrowth prevents elephant counting by either air or ground.
GIS - which is akin to a map on a computer - makes it possible to analyze data and display it visually on a computer screen. The advantage of GIS is that various data layers can be combined to yield an overall analysis. In this case, Beardsley is using a digital representation of the roads that criss-cross the rainforest of Gabon and incorporating other data layers to come up with elephant estimates. Because human activity drives elephants deeper and deeper into the forest, knowing where the roads and navigable rivers lie and factoring in other data can help predict elephant populations in any give forest region.
The information is urgently needed. Beginning in the mid-1970s, an upsurge in ivory poaching decimated African elephant populations, killing an estimated 10,000 elephants a year, and cutting the numbers of Central African forest elephants about 44% overall by 1994 - an attrition rate that if it were to continue would eliminate elephants from the equatorial forests of Africa by around the turn of the century.
Beardsley and her colleagues hope the GIS analysis, by providing more accurate estimates of how many elephants remain, will spur Central African countries to establish protected areas and help save the elephants from extinction.
Researchers have been using GIS to estimate elephant populations since 1986, when the United Nations Environment Program developed its African Elephant Database Project. That project began by surveying people throughout the continent who estimated elephant populations from a wide range of information, including aerial counts.
But because actual counts were available for only one-third of the known elephant range, scientists have since been engaged in a long process of refining the database as new information is developed. Now incorporated into the earlier GIS database is information on vegetation, rainfall, protected areas, human density, and tsetse fly range. The latter is significant because tsetse flies discourage cattle production and human presence, but do not affect elephants. Tsetse fly range can therefore help predict where elephants might be found.
"Having this database has helped push forward a lot of conservation efforts," says Beardsley. "GIS enables scientists to store information about elephant populations on a continental scale, which allows people to update, retrieve, and store estimates and to do modeling. That, in turn, allows us to see better what population trends are going on."
Now the GIS Coordinator in the UC Davis Information Center for the Environment, Beardsley became involved with the project in 1989 after a two-year stint with the Peace Corps in Kenya. Living in Nairobi, she worked with Dr. Richard Barnes of the Wildlife Conservation Society to apply a new model Barnes devised for estimating elephant populations.
By building on and combining existing data, GIS makes it possible to calculate elephant numbers in remote areas faster and at a fraction of the cost of field work. Although some field work is still necessary, GIS significantly speeds the process and reduces the overall expense.
Barnes, Beardsley, and their colleagues now estimate the elephant population in Central Africa to stand at 171,334, down from 306,077 estimated before poaching became widespread.
Scientists are now anxious to apply the GIS model to estimate populations of other species, since field studies in Gabon indicate that, like the elephants, animals such as the spot-nosed monkey tend to concentrate more heavily in remote areas away from roads and rivers.
Says Beardsley, "Future use of GIS for estimating species populations will allow researchers to spend more time analyzing and interpreting population estimates and less time with expensive field expeditions."
| ietpubs@ucdavis.edu |