The study used the same mist-netting method used in short-term surveys, but over 30 years, to analyse bat populations. "Several species seem to come and go and I ask myself, is this migration? No. Is this extinction? No. What is actually happening is that as populations fluctuate over time, they simply become rare enough to become 'temporarily invisible' to our human biases and technology."
Another researcher said different ecological pictures have appeared depending on whether he sampled in the wet or the dry season and whether he sampled for two years instead of one. Even sampling for two years wasn't enough.
Surveys of flora and fauna may be flawed
Bat study raises doubts over our understanding of Earth's ecosystems.
Matt Kaplan, Nature 12 Dec 08;
One of the most common techniques for diagnosing the ecological health of a region may be painting an inaccurate picture of biodiversity, a study of the bats on the tiny volcanic island of Montserrat suggests.
To understand an area's ecology, researchers are often asked by funding agencies to conduct a short survey, known as a rapid biodiversity assessment.
Such surveys are convenient: they fit easily into the typical 3-5-year timescale of a PhD, match the length of time within which grant-giving agencies expect to see results, and are relatively quick to write up and publish.
Yet an ongoing study on Montserrat is yielding data that suggest these short-term surveys may not always paint an accurate picture.
Slipping through the net
Montserrat was devastated by Hurricane Hugo in September 1989, causing near total defoliation of the island. Ecologist Scott Pedersen at South Dakota State University in Brookings and his colleagues captured some of the island's bats in mist nets before and after the hurricane and found a 10-fold decrease in the population. They further noted that the composition of the bat community shifted from small fruit-eating species to more omnivorous and larger fruit-eating species.
Over 30 years of using the same mist-netting method to capture and analyse bat populations, Pedersen has seen as few as four species and as many as ten on the island at any one time. The findings, he says, show that short-term surveys could be misleading ecologists.
"Several species seem to come and go and I ask myself, is this migration? No. Is this extinction? No. What is actually happening is that as populations fluctuate over time, they simply become rare enough to become 'temporarily invisible' to our human biases and technology," he says. "If this is the case in the pocket-sized system we are studying, then I really don't know what to make of all the rapid biodiversity surveys taking place in larger habitats like Amazonia."
"We've found similar problems in our own work," says Robert Ewers, an ecologist at the University of Cambridge, UK. Data on birds in the Atlantic forests of Brazil have shown Ewers and his colleagues very clearly that snapshots do not give a good representation of diversity. Different ecological pictures have appeared for Ewers depending on whether he sampled in the wet or the dry season and whether he sampled for two years instead of one. Even sampling for two years, which was all Ewers could manage, wasn't enough. He believes he probably needed to double the survey time to get the full picture.
"Ecologists are aware of the sampling problem, it's just that it is so difficult to avoid because granting agencies restrict sampling to a maximum of two years," Ewers adds.
Penelope Firth, deputy director of the US National Science Foundation (NSF) Division of Environmental Biology, says the agency does not limit the duration of biodiversity surveys. "Many are five-year projects, quite a few are three years," she says. The NSF even allows investigators to submit proposals for renewing their awards, and thus six-year and longer surveys are not uncommon, according to Firth.
"Ecologists and statisticians have invested considerable effort into understanding the consequences and implications of employing a finite sampling effort," she adds.
But Pedersen is sceptical. "Almost everybody in the bat world that I know seems cynical regarding the efficacy and accuracy of rapid surveys, but we are all quite happy to take the cash and do what we can," he says.
"Unquestionably, long-term research studies are lacking in nearly every ecosystem," says Jeff Foster, an ecologist at Northern Arizona University in Flagstaff. "But the real question here is, what is the best bang for the buck?". For trophic interactions or population dynamics, a rapid assessment is not suitable. But for determining how many species are likely to be present, a rapid assessment will be far more appropriate even if you happen to miss a few species, Foster explains.
The problem, he adds, is that "we do not know how much spending is going into rapid assessment programmes and whether long-term studies are being underfunded because of this allocation".