Berkeley -- A clever experiment conducted along the Eel River in northern California sheds new light on a seeming contradiction in the field of ecology about whether a diverse ecosystem is inherently more resistant to invasion by alien species.

	A tussock in the study supporting seven native species, including three species of monkey flower (yellow flowers).

The question has become more important in recent years as scientists campaign to preserve the planet's biodiversity, focusing on areas with the largest concentration of species, or attempt to restore native habitats and native species in the face of dramatic increases in the number of non-native invaders.

Paradoxically, while many studies have shown that more diverse communities are better at resisting invaders, surveys of diverse ecosystems show they contain the highest number of exotic, non-native species.

In a series of experiments reported in this week's issue of Science, Jonathan Levine, a graduate student in integrative biology at the University of California, Berkeley, shows clearly that both are true. Small communities with high diversity are better at resisting invading species, but at larger scales, invaders are more likely to get established in diverse ecosystems.

"Places are diverse for a reason. They are rich in resources, they lack predators, they have a moderate amount of natural disturbance - in general, they are hospitable to a broad range of species. What's good for native species is good for invading species," Levine said. "So, you can't rely on biodiversity alone to prevent invasion. You have to keep exotic species from getting a foothold in the first place."

Levine and one of his faculty advisors, Carla D'Antonio, associate professor of integrative biology at UC Berkeley, wrote a review article for the ecological journal Oikos last year that questioned the theoretical foundation and scientific support for the claim that diversity reduces susceptibility to invasion.

"This is an interesting and potentially scary finding," D'Antonio said of Levine's study. "People want to believe that, if we have more diversity in the ecosystem, it will be more resistant to invasion. Restoration ecologists want to create invasion-resistant communities. But Jonathan's research suggests that even if you get diversity-induced resistance on a small scale, it won't necessarily translate to the scale of the whole ecosystem.

"Hotspots of native diversity are often going to be hotspots of invader diversity if invader propagules are abundant in a region."

Propagules are seeds or pieces of plants capable of developing into mature adults.

Levine's experiment was designed to distinguish between the effects of diversity and the effects of other factors in the ecosystem that are associated with or encourage diversity.

He focused on a seven-kilometer section of the South Fork of the Eel River in northern California where a type of sedge (Carex nudata) grows in small tussocks that create tiny island communities he could easily manipulate. These tussocks, most only a half-foot across, are submerged and dormant most of the year, but when the waters subside between May and September they blossom with some 60 separate native plant species.

Study site at a riffle along the South Fork of the Eel River. The yellow flags indicate individual tussocks, each a mini-ecosystem unto itself.

Selecting 65 tussocks at a single 20-meter riffle of the river, he removed all species except for the sedge itself, then planted the tussocks with one to nine different native species, creating artificial communities of varying diversity. The following season, after the natives had become established, he seeded each tussock with three invasive non-native species commonly found in the area - Canadian thistle, creeping bent grass and common plantain, all invaders from Europe.

As other experiments have shown, the more diverse tussocks resisted the invasions better than the less diverse. The commonly accepted reason, he said, is that more species exploit more fully the resources of the habitat, leaving fewer resources and, thus, less opportunity for invaders.

Conversely, when Levine looked at tussocks along the entire stretch of river, he found invading species to be more abundant on the most diverse tussocks. The fact that the intrinsic protective effects of diversity did not agree with the observed patterns of diversity along the river suggested that something other than diversity was involved in driving the ecosystem-wide patterns.

Levine noticed that downstream tussocks were both more diverse and more likely to be invaded than upstream tussocks. Perhaps, he thought, this is because more seeds are dumped on downstream tussocks as a result of the natural stream flow.

To rule out other alternatives, he conducted a second experiment in which he planted the same number of seeds on all tussocks under study. Choosing 190 natural tussocks of a range of diversities all along the seven-kilometer study site, he added large numbers of seeds on each of them - enough seeds to swamp the numbers delivered naturally by the stream. The invaders colonized all tussocks about equally, suggesting that the observed pattern was due to the supply of seeds and not resource availability or some other characteristic of the tussock.

"Overall, this suggests that the number of seeds dumped on the downstream tussocks overpowers the intrinsic protective effect of diversity," he said.

This is an important conclusion for wildland managers to consider, D'Antonio said.

Although in this particular ecosystem, seed supply was the important factor affecting the number of invasions, other factors probably operate in other systems. These could range from the availability of nutrients or light to the abundance of herbivores or frequency of disturbance.

"For a long time biologists have been repeating the idea that diversity offers protection against invasion without really understanding the theoretical underpinnings or support for it," D'Antonio said. "Many of the confirming experiments have been one-on-one competition-type studies, some of them even in laboratory flasks. But on the ecosystem scale, there are many other factors that come into the equation besides competition."

Levine is supported by a National Science Foundation Graduate Research Fellowship and Dissertation Improvement Grant.