Sea-Level Rise And Coastal Forest Retreat On The West Coast Of Florida, Usa - Statistical Data Included

Ecology, Sept, 1999 by Kimberlyn Williams, Katherine C. Ewel, Richard P. Stumpf, Francis E. Putz, Thomas W. Workman

INTRODUCTION

Although sea level has risen and fallen many times in the past (e.g., Cooke 1939), extensive human settlement of coastal areas and conservation concerns have spurred increased interest in the effects of sea-level rise on the retreat inland of terrestrial ecosystems (e.g., Mehta and Cushman 1989, Wanless 1989, Pezeshki et al. 1990, Harris and Cropper 1992). Due in part to anthropogenic contributions to global warming, the rate of sea-level rise is expected to increase (Warrick and Oerlemans 1990, Edgerton 1991, Wigley and Raper 1992, Titus and Narayanan 1995, Warrick et al. 1996). In addition, other predicted environmental changes, including changing rainfall patterns, changing frequencies of violent storms, and decreasing groundwater supplies of freshwater, may interact with sea-level rise to affect the rate of change in terrestrial vegetation along coastlines (summarized in Schneider [1993]).

Much effort has been focused on responses of marsh vegetation to sea-level rise (e.g., Titus 1988, McKee and Mendelssohn 1989, Day et al. 1995), of mangrove forests to sea-level rise (e.g., Ellison and Stoddart 1991, Snedaker 1995), and of individual tree species to flooding and salinity, the major factors mediating forest response to sea-level rise (see Pezeshki et al. [1989, 1990], and references therein). A few studies have examined recent historical changes in coastal upland forest distribution in relation to sea-level rise (Clark 1986, Ross et al. 1994), linking patterns in stand decline to elevational gradients. Other studies have linked stand decline to elevated soil and groundwater salinity (Brinson et al. 1985, Ross et al. 1994), which are affected by drought, rainfall, and fresh groundwater flow as well as sea level. It has been postulated that soil flooding, resulting in low oxygen availability, reducing conditions, and perhaps [H.sub.2]S formation, may contribute to the forest decline caused by sea-level rise (Salinas et al. 1986, Brinson et al. 1995). Competition from encroaching marsh vegetation may likewise contribute to the death of forests.

The mechanisms by which rising seas cause forest retreat may vary with geomorphological and hydrological characteristics of the coast. Although low-lying limestone coastlines and islands are considered susceptible to submergence by sea-level rise (Hendry 1993, Biljsma 1996), most research on the impacts of sea-level rise on coastal upland forest has been carried out on deltaic coastlines or sandy coastlines (see Williams et al. [1999] for review). On deltaic coasts with high freshwater outflows, such as the Mississippi Delta, rising seas may boost freshwater tables, eliminating tree regeneration through increased freshwater flooding (e.g., Baumann 1987, DeLaune et al. 1987, Conner and Day 1988). Zonation patterns in several coastal floodplain forests are consistent with gradients in flooding stress, with coastward stands being dominated by the most flood-tolerant tree species and abutting freshwater or brackish marsh, rather than salt marsh (White 1983, Doumlele et al. 1985, Pratt et al. 1989, Hackney and Yelverton 1990). Saltwater intrusion has been linked to stand decline in these forests (e.g., Penfound and Hathaway 1938, Platt and Brantley 1990, Allen 1992). However, this saltwater intrusion is often attributable to canal building, dredging, and other manipulations of water flows, making the role of sea-level rise difficult to discern (Penfound and Hathaway 1938, Hackney and Yelverton 1990). On sandy coasts, erosion and shifting sands, often caused by storms, are generally linked to forest retreat (e.g., Penfound and O'Neill 1934, Kurz 1942, Alexander and Crook 1974, Hayden et al. 1991).

The mechanisms by which rising seas affect forest retreat on stable coastlines with low freshwater outflows have not been well studied. On carbonate coastlines of the southeastern United States, forest zonation correlates with both elevation and salinity of soil- and groundwater (Kurz and Wagner 1957, Ross et al. 1994). On such coasts, sea-level rise may affect forests primarily through increased salt exposure rather than through increased flooding stress or erosion, but conditions associated with initial stages of forest decline been not been described.

Regardless of the mechanism by which rising seas eliminate coastal forest, tree regeneration may be much more sensitive to rising seas than mature-tree survival. Increases in hydroperiod in the Mississippi Delta have eliminated tree regeneration in forest stands (DeLaune et al. 1987, Conner and Day 1988, Conner and Brody 1989). Failure of tree regeneration has also been linked to saltwater intrusion (e.g., Penfound and Hathaway 1938). On sandy coasts, catastrophic erosion may wash away entire forest stands. However, in the absence of catastrophic erosion, researchers have speculated that tree regeneration on sandy coasts is eliminated by seedling burial by shifting sands (Penfound and O'Neill 1934, Brown 1973). Because tolerance of flooding and salinity often increases with plant age (e.g., Briscoe 1957, Peterson and Bazzaz 1984, Conner and Askew 1992), and because shifting sands are known to impose restrictions on seedling recruitment (hence, the dominance of dunes by vegetatively reproducing species; e.g., Rainwell 1972, Van der Valk 1974), one would expect sea-level rise to eliminate forests primarily through impacts on seedling regeneration. Although failure of regeneration in coastal forest has frequently been noted (e.g., Brown 1973, Clark 1986, Conner and Day 1988), the generality and implications of this pattern have not been much explored. If coastal stands are relict (i.e., non-regenerating), then causes of canopy death at the coastal margin are relatively unimportant: these stands are effectively dead already. If canopy death lags regeneration failure, then aerial monitoring of coastal forest may fail to detect all but the very last stages of forest demise.