BioEd Online

One assumption behind attempts to restore coastal wetlands as natural buffers against storms and floods may be wrong, a study suggests.

Experts believe that wetland vegetation could have helped to resist the coastal erosion caused by Hurricane Katrina in 2005 and other recent disasters, such as the Indian Ocean tsunami of 2004 and Cyclone Nargis, which hit Burma last year.

Rusty Feagin of Texas A&M University in College Station and his team devised some experiments that he expected would demonstrate one way in which wetland plants could do this — by resisting the erosion caused by waves beating at the land's edge. He was surprised to find no such effect. It turned out that soil type is much more important, and that the presence or absence of vegetation doesn't make much difference.

Feagin suggests that the conventional wisdom took hold because of studies that show that plants can build wetlands by trapping sediment that would otherwise flow out to sea. "I think we have taken this idea that wetlands can build themselves and jumped to the conclusion that vegetation can prevent erosion," he says. He adds that vegetation may well have other protective effects, such as calming a storm surge that rolls right over a land edge and heads inland.

Shifting sands

Feagin did experiments in large tanks called flumes, in which he could simulate wave action. He took roughly 0.6-metre square pieces of wetland from the Texas Gulf Coast with a home-made 'cookie-cutter' and wedged them in his wave flumes. There was, surprisingly, no significant difference between the erosion rates of the wetland chunks with marsh grasses and the 'bald' squares. This lack of effect was also confirmed by experiments in the field. There was a difference, however, between naturally formed wetland chunks and wetland chunks from a restoration project in Galveston Island State Park. The restored marsh eroded much more quickly.

"If you go stick your hand down in it, the stuff in a natural wetland is gooey and fluffy and light, but also real dark and organic and stinky," says Feagin. "Something that people built ten years ago is real sandy and heavier and there isn't so much organic matter."

Sand turned out to be the biggest predictor of erosion rates. "When you have something that is real sandy and non-organic, it just blows away when you hit it with waves," he says. For soils that are less dense, more clayey and organic, more subtle dynamics predict erosion rates, but for anything denser than about 0.9 g/cm₃, it's the presence of sand that counts. The study appears in today's Proceedings of the National Academy of Sciences.¹

Unfortunately, many restoration projects on the US Gulf Coast have siphoned up sand from offshore sand bars and build it up into land, into which plants have been hastily planted on the theory that they will shore the whole thing up.

Blowing in the wind

Jeffress Williams, a coastal marine geologist for the US Geological Survey at Woods Hole Science Center in Massachusetts, says that Feagin's finding will give him and many of his colleagues pause. "What they are reporting is certainly contrary to what has been in the literature for probably the last 50 years or better," he says. But, he adds, "If you actually look at the literature that that is based on, it is extremely sparse."

Williams would like to see more confirmation of Feagin's finding in other places, and with other vegetation types, such as mangroves. If the non-effect is general, it could change thinking in restoration circles, he says. "[Restoration] is based on general understanding and intuitive observations. What we need to do is more laboratory studies and more deliberative field studies to actually look at the processes that are going on."

Experts at the Louisiana Office of Coastal Protection and Restoration are less sure that Feagin's findings will be a game-changer. Louisiana is sinking, losing sediment and being bashed regularly by hurricanes. The gradual erosion caused by slapping waves typical of a windy day that Feagin simulated may be the least of their worries.

"Within coastal Lousisana, we almost have death by a thousand cuts," says James Pahl, a manager in the applied research and development branch of the office. "A lot of the marsh loss that we see coast-wide you can determine is due to very specific storm events," he says. And overall, adds Richard Raynie, chief of the engineering and science division, "we are seeing, as the wetlands degrade, increased impacts from comparable-level storms". Wetland plants may not be resisting wave erosion on a day-to-day basis, says Raynie, but they seem to be doing something.