One of the major issues in conservation and natural resource
management is the distribution of species across landscapes. How
widespread are species on average, and how much does the composition
of a community of species change from place to place (known as "beta
diversity")? Obviously, the pattern differs in different kinds of
organisms and habitats, and beta diversity can be extreme across some
environmental gradients.

An international group of entomologists and botanists, including Scott
Miller and Karolyn Darrow from the National Museum of Natural History
and Yves Basset from the Smithsonian Tropical Research Institute in
Panama, with NMNH research associates Vojtech Novotny (Czech Academy
of Sciences) and George Weiblen (University of Minnesota) and many
others, has assembled a large data set representing 500 species of
caterpillars, ambrosia beetles, and fruit flies across 75,000 square
kilometers of contiguous lowland rain forest in Papua New Guinea. The
study published in the 9 August 2007 issue of Nature shows that most
lowland New Guinea insect and plant species are widely distributed.
The results imply that where relatively uniform altitude, climate and
soil support a low beta diversity of vegetation, the plant-feeding
insect community also has low beta diversity.

The results have implications for the design of nature reserves as
well as the control of forest pests. "There are some philosophical
questions that our data should be useful to address," Miller said. "If
you can preserve 10,000 hectares of forest, is it better to preserve
it as 10 small 1,000 hectare plots or one large 10,000 hectare plot?"

Strategies for preserving high and low beta diversity forests might
not be the same. "Our data from the New Guinea lowland forest suggests
that bigger is better," he said. "In this kind of relatively uniform
habitat, you're not losing a lot of beta diversity that would have to
be represented by several smaller sites. But the opposite might apply
on elevation or climate gradient."

There is still much to learn because collecting high intensity data is
a very time-consuming process. "We haven't solved the conservation
management question by any means," Miller said. He noted this kind of
study needs to be replicated across a large spatial scale in the
Amazon Basin and the Congo Basin to see if the low beta diversity
pattern holds.