La Selva from Space

Rapidly improving satellite "eyes" are giving tropical biologists more detailed views than ever before of La Selva and other tropical rain forests, says David Clark, the University of Missouri ecologist who is a former scientific director at La Selva.

"It's the trees that are a key to understanding tropical ecology," he says. "It’s certainly my bias, but I look on forests as being built around trees, and everything else flows from the trees. So, really, what you want to know first off is what's happening to tropical trees, and from that you have a good idea what’s happening to everything else."

For example, says Clark, although no scientists could hope to chart from satellite images the loss of animal species from tropical forests, they can see the changes in tree populations that result from loss of plant-eaters, pollinators and seed dispersers.

Thus, Clark and his colleagues are excited by the potential for high-resolution satellite sensing to ultimately give them a tree-by-tree look at tropical forests.
Satellite sensing is especially crucial because of the severe biases and limitations of studying tropical forests using the traditional ground-based measurements of research plots.

"We tend to go to places with big trees, and we tend to put our plots there," says Clark. "And we tend to avoid nasty places on slopes with lots of gaps." Also, plots are typically near air strips or rivers, and in politically stable areas with friendly governments. It's no wonder, then, that the research plots that do exist are far too sparse and small -- numbering only about two dozen, averaging about a few acres -- to accurately reflect the vast reaches of the world's tropical forests.

While aerial remote sensing provides the sharpest resolution, says Clark, airplanes cannot reach many tropical forests because of limited range and restrictions on access to air space.

Clearly, satellites represent the future as ecological eyes in the sky, he says. For example, for the last three decades the venerable Landsat <http://landsat7.usgs.gov/index.php> series of satellites has been providing masses of data on tropical forests. At a resolution of 15 meters square, however, it can't resolve individual trees -- the holy grail of satellite ecology. With its 4-meter resolution in multiple colors, the newer IKONOS < http://www.spaceimaging.com/products/ikonos/> satellite comes closer. Seeing multiple colors of the spectrum is extremely important for distinguishing types and health of trees.

Such power has enabled Clark and his colleagues to use IKONOS to make invaluable comparisons of satellite and ground data on their La Selva plots.
Even better is the Quickbird <http://www.ball.com/aerospace/quickbird.html> satellite, launched in 2001. With its 70-centimeter resolution, Quickbird can easily distinguish individual trees.

The future is even brighter . . . or rather tighter, says Clark. The next few decades will see 10-centimeter resolutions, as well the equivalent of superhuman vision -- "hyperspectral" imaging with hundreds of different wavelength bands.

"The future for remote sensing of the world, and particularly of the tropics, is incredibly bright," says Clark. "This is wonderful news because we’re facing global problems that need to be studied at a global scale, and the only way we can really get global comparative data sets . . . is to do remote sensing.
"The tropical forests of the world are in sad shape and they’re under a variety of assaults from humans in a variety of ways, but our tools for studying these problems, and maybe doing something about them are only going to get better and better."