Brightly colored corals are supplanted by dark, undulating seaweeds. Familiar fish species vanish — to be replaced by unknown strangers, or nothing at all.

Pushed to the brink by warming oceans and human threats, “the places that we used to recognize as coral reefs no longer look the same,” said Gabby Ahmadia, director of ocean science at the World Wildlife Fund.

It’s a metamorphosis unfolding in ecosystems around the globe.

A sweeping survey published this week looked at tens of thousands of species counts from the past few decades and found that the world’s ecosystems are rapidly reorganizing. On average, more than a quarter of all plant and animal species within an ecosystem are being replaced every decade — probably the result of local extinctions, the introduction of invasive species and migrations motivated by climate change.

In the midst of a global environmental crisis, when an estimated 1 million species are said to be at risk of extinction, the study offers an important look at biodiversity on the level of individual ecosystems, the authors said. It suggests that for now, at least, human activities have resulted not so much in outright losses as in large-scale reorganization.

But the function of ecosystems — their capacity to filter water and clean air, to sustain the plants and animals we rely on and admire — depends on the activities and health of their inhabitants. And those qualities are at risk.

“It’s a little bit like we’re playing musical chairs at the moment,” said macroecologist Maria Azeredo de Dornelas, a co-author on the report. “You have a lot of things moving around, and chances are that some things are going to end up without a chair.”

The study, published this week in the journal Science, draws on a massive new database called BioTime. The database, which Dornelas helped build, contains more than 8 million measures of abundance on more than 40,000 species in roughly half a million locations on land and in the oceans.

This richness, said Sarah Supp, another co-author, allowed the researchers to sift through global biodiversity trends and pinpoint changes happening on a local scale.

“This wasn’t really possible before,” said Supp, a Denison University data scientist who specializes in biodiversity studies. “But it’s important because the scale at which our actions take place are often much more pointed toward specific locations, or political boundaries that are not at the scale of the globe.”

The researchers were surprised to find that the heightened global extinction rate wasn’t reflected at the ecosystem level. In a few extreme cases, the number of species in a habitat declined by as much as 20 percent per year. But on average, species richness — the head count of species present in an ecosystem — didn’t appear to change over time.

What is changing — with alarming speed — are the kinds of creatures present.

“This paper really shows that, more important than just looking at the number of species in an area, we need to look at the identity of those species,” said Kimberly Komatsu, a global change biologist at the Smithsonian Environmental Research Center who was not involved in the study. “And even moving beyond that, we need to think about what the traits of these species would be and what that means for the functioning of the ecosystem.”

Marine ecosystems, such as reefs, are hot spots of transformation, the scientists found. The waters of the western Atlantic and the northwest Australian shelf experienced rates of species turnover much higher than the global average. Tropical regions also seemed to change more than temperate ones, perhaps because these already-warm areas have now been heated to temperatures for which most species are not adapted.

There is not enough data from the preindustrial era for scientists to know how much faster turnover is happening because of humans. But evidence from past research suggests the current average rate of 28 percent turnover per decade is at least two to three times higher than normal, Dornelas said.

To truly understand the consequences of this change, scientists will have to take a closer look at which creatures are vanishing from individual ecosystems — and who is arriving to take their place.

Some turnover, such as New England sugar maples showing up in northern Canada, may be a sign of adaptation to a warming world. Other kinds of restructuring, such as the replacement of coral reefs by algae, could have painful consequences.

Ahmadia, who was not involved in the new study, has surveyed reefs in the Pacific Ocean and coastal Africa. She said this kind of turnover leads to less-productive reefs.

“The composition of the species is going to change. They’re not going to provide the same benefits they used to for local communities,” she said.

In an accompanying analysis published by Science, ecologists Britas Klemens Eriksson and Helmut Hillebrand pointed out that some of this rapid turnover may lead to homogenization of ecosystems. Better-adapted species will spread widely, pushing out native creatures. Habitats that once served as home to unique collections of species may all start to look the same, they wrote.

The pace and scale of the transformations Dornelas and Supp describe are potentially grim. But the study also points to an opportunity, the scientists said. By understanding biodiversity change at a local scale, conservation scientists can figure out how to focus their efforts.

The study argues that regions undergoing rapid transformation should be prioritized for “reactive” conservation measures — interventions that help species adapt to their changing circumstances, rather than attempting to preserve a population the environment can no longer sustain.

On the other hand, more stable systems — like temperate forests — are good candidates for “proactive” protections. By insulating these areas from human disturbance, people might be able to preserve the diversity that makes them unique.

“Climate change is happening now and we need to be able to accommodate the reorganization that is taking place, to some extent,” Dornelas said. “There’s no going back in time … but we can make informed decisions about what kind of future we want to have.”

Read more: