Piles of garbage left by humans thousands of years ago may have helped form tree-covered biodiversity hot spots in the Florida Everglades (USA), according to a new study.
     The so-called tree islands of the Everglades are patches of relatively high and dry ground that rise from the wetlands. They stand between 1 and 2 meters higher than the surrounding landscape, can cover 100 acres or more, and host two to three times the number of species living in the surrounding marsh. Previously, scientists had presumed that many of the larger tree islands, which have a veneer of peaty soil, formed atop topographical high spots in carbonate bedrock underlying the Everglades. But research reported at a meeting of the American Geophysical Union in Santa Fe suggests that the real foundation for some of these islands may be prehistoric trash heaps, known as middens.
     These middens - a mix of discarded food, charcoal, shell tools, and broken pottery - would have been slightly higher and drier than the surrounding marsh, offering a foothold for trees, shrubs, and other vegetation, says paleoecologist Gail Chmura of McGill University in Montreal, Canada. Bones in the trash would have been a good source of phosphorus, a nutrient otherwise in short supply in the Everglades, she and her McGill colleague Maria-Theresia Graf contend. Their fieldwork at a couple of tree islands, as well as previous research, indicates that what previously had been considered to be elevated bits of bedrock are actually concretelike layers of carbonate left within the peaty soil when water drawn upward by the trees evaporates, leaving behind its dissolved minerals. Beneath these so-called perched carbonate layers the researchers found more peat and a midden.
     But just how and when these perched carbonates form is still up for debate, says ecohydrologist Pamela Sullivan of Florida International University in Miami. And the fact that the layers form within soils, not as part of a chronological sequence of sediment layers, makes it particularly difficult to determine their age, she says. What's more, although trash middens may have been the foundation for some tree islands, they didn't trigger the growth of all of them, says palynologist Christopher Bernhardt of the U.S. Geological Survey in Reston, Virginia.
     Bernhardt surveys at two tree islands in the eastern Everglades indicate that those islands, like the ones Chmura and Graf studied, have perched carbonate layers and middens that overlie layers of peat. Yet analyses of the pollen and spores in the peat underlying the trash middens suggest that the island hosted trees, shrubs, and ferns well before people arrived. So residents' trash may have helped the island grow, but it didn't get the ball rolling, he says, adding that although bones in a midden are a good source of nutrients, it's just as likely that the phosphorous that nourishes a tree island's vegetation comes from the prodigious amounts of guano left by birds attracted to the ecosystem there.

Edited from Science Magazine (22 March 2011)