When Teddy Roosevelt created the first national wildlife refuge at Florida’s Pelican Island in 1903, plume hunters were the main threat to the rookery island’s birds. Warden Paul Kroegel fought back with simple weapons: a boat and a gun.

Now Pelican Island Refuge faces subtler threats, including a significant sea level rise by 2100. Refuge Manager Charlie Pelizza says salt–tolerant mangrove trees will protect part of Pelican Island; but if sea level rises more than one meter, everything within the refuge’s original 5.5 acres will be underwater.

Pelizza knows some parts of the modern 5,400–acre refuge will be lost to the sea. But he and refuge partners are protecting its essential habitats by drawing information from a new weapon: the Sea Level Affecting Marshes Model, or SLAMM. Since 2005 the Service has used this computer model to predict how sea level rise will affect coastal refuges. Coastal marshes are especially vulnerable to sea level rise, which can erode underwater banks, drown vegetation and cause increased salinity that kills plants.

SLAMM was developed in 1989 to predict these changes. Now in version 6, it is “the workhorse model for sea level rise planning on the Refuge System,” says conservation biologist Brian Czech, the SLAMM analysis coordinator in the Refuge System’s Division of Natural Resources and Conservation Planning.

The model starts with a landscape image and divides it into small cells, like pixels in a digital photograph. Elevation data are used to make the image three–dimensional. Each cell is coded as open water, beach, tidal flat, wetland or high ground. As the computer runs through various scenarios, the cells’ coding changes and a new image emerges, showing the land as it would be at various sea heights.

Factors that go into the analysis include erosion, inundation, saturation of soils, barrier island overwash by storms, and accretion of sediment. Accretion is often the hardest to estimate and a source of uncertainty, Czech says. To get accretion data, scientists need semi–permanent instrument arrays that cost about $10,000 apiece to set up and monitor. Only 30 refuges have them.

The U.S. Fish and Wildlife Service and its contractors have projected sea level rise effects for 132 of the 171 coastal refuges. Two more studies are underway. Other coastal refuges won’t be modeled because they don’t have marshy coasts or lack the necessary elevation data.

Pelican Island Refuge is using SLAMM results to guide restoration of degraded wetlands on the refuge’s mainland portion. The results are also useful in land acquisition and for public outreach on climate change.

Now the Service is using SLAMM data from all 75 Atlantic coast refuges to assess how sea level rise will change migratory bird habitats along the Atlantic Flyway. Similar multi–refuge analyses are planned for the other major flyways in future years.

Heather Dewar is a writer in the Refuge System Branch of Communications.