T h e structure of mangrove vegetation, soil salinity, and topographic relief of the mangrove ecosystem were studied along the south coast of Puerto Rico, Culebra, and Mona Island. All systems, with the exception of the forest at Mona Island, were characterized by having a coastal fringe of live vegetation (usually dominated by the red mangrove), a zone of dead trees, and a hypersaline lagoan or dry salt flat o n the landward side. Mean soil salinities were 44 O/00 for the zone of live trees, 72 0100 for the zone of dead trees, and 87 0/00 for the salt flats. T h e Mona Island forest grew over a sandy hardpan that prevented mixing of sea water and the fresh water. Soil salinities were low in this forest, and trees reached a height of 1 3 m. Tree height was inversely proportional to soil salinity (r = 0.72) between 1 7 and 72 O/OO. W h e n soil salinities exceeded 65 0/00, dead tree basal area was higher than live tree basal area. It is suggested that mangrove growth is limited by soil salinities higher than 9 0 0100. Mangrove succession in offshore overwash islands and arid coastlines is described. It is proposed that cyclic rainfall patterns and hurricanes act as regulators of speed and direction of succession. Rainy periods are asswiated with lower soil salinities and expansion of the red mangrove zone. Droughty periods result in high soil salinities, mangrove mortality, and expansion of salt flats. Hurricanes set back succession and reverse successional trends that reduce mangrove areas. A model of mangrove ecosystem function and implications for management are also discussed.

The impact of Hurricane David on the forests of Dominica was measured 40 days after the hurricane struck the island on August 29, 1979, with winds averaging 92 km/h which lasted 10.2 h. Sixteen 0.1-ha plots were studied. They were located on a variety of slope and exposure conditions, representing three life zones and 11 plant associations. Measurements included tree diameter at breast height (dbh), total height, wood volume, and percent canopy closure of stand. Species and damage conditions were noted. The hurricane caused severe damage, particularly to forests on the southern half of the island. Forty-two percent of the standing timber volume in the study areas was damaged at an average intensity of 27 m3/ha. About 5.1 × I06 timber trees were damaged in an area of 246 km2. "Complex" vegetation was more resistant to hurricane damage than "simpler" vegetation. The climax Dacryodesexcelsa Vahl associations suffered the least hurricane damage, while the palm brake suffered the most damage. Only one species (Richeriagrandis Vahl) exhibited significantly higher hurricane damage. Trees in larger diameter classes tended to uproot with more frequency than smaller diameter trees, which tended to snap and break. Regrowth of seedlings and saplings was rapid on wet areas of the damaged sites. We found more species on regenerating plots than on the undisturbed forest plots. The damage caused by Hurricane David is the most extensive reported to date anywhere, and damage to vegetation by hurricanes appears to be more extensive than damage by earthquakes and landslides. Owing to their high frequencies in the Caribbean, hurricanes induce cyclic plant successions.

Uncontrolled releases of Tigris and Euphrates River waters after the 2003 war have partially restored some former marsh areas in southern Iraq, but restoration is failing in others because of high soil and water salinities. Nearly 20% of the original 15,000-square-kilometer marsh area was reflooded by March 2004, but the extent of marsh restoration is unknown. High-quality water, nonsaline soils, and the densest native vegetation were found in the only remaining natural marsh, the Al-Hawizeh, located on the Iranian border. Although substantially reduced in area and under current threat of an Iranian dike, it has the potential to be a native repopulation center for the region. Rapid reestablishment, high productivity, and reproduction of native flora and fauna in reflooded former marsh areas indicate a high probability for successful restoration, provided the restored wetlands are hydraulically designed to allow sufficient flow of noncontaminated water and flushing of salts through the ecosystem.