Testing Ecological Patterns
The content you've requested is available without charge only to active Sigma Xi members and American Scientist subscribers.
If you are an active member or an individual subscriber, please log in now in order to access this article.
If you are not a member or individual subscriber, you can:
Ecologists collect information about the distribution of species because the observed patterns often reveal what ecological forces are at work. Intriguing distributions can, however, arise merely by chance, and it often proves difficult to know whether a seemingly interesting pattern is truly significant. This problem has particularly troubled students of island biogeography, who for the past quarter century have debated the right way to evaluate the distribution of species on island archipelagos. The solution described here invokes a simple algorithm used to compute the "knight’s tour," an exercise that has long been taught to undergraduates in computer science. The article illustrates the newly developed method for testing the significance of ecological patterns by applying it to Darwin's finches on the Galápagos Islands.