Erosion of a salt marsh at Fort Pulaski National Monument threatens a pier of historical significance along the North Channel of the Savannah River, the main shipping channel for the Port of Savannah, Georgia. The erosion was initiated by the alongshore and onshore migration of a supratidal oyster shell ridge that exposed a marsh scarp that is susceptible to lateral retreat in response to both wind-generated and vessel-generated waves undercutting the upper-marsh surface. An instrumented field study was completed between October 2007 and February 2008 to examine the relative importance of wind-generated and vessel-generated waves to the retreat of the scarp by 0.21 m along the instrumented transect. On average, there were 14 sailings by container ships per day and up to twice as many sailings by pilot boats ferrying the harbor pilots to container ships offshore, but only those vessels that sailed within i m of the high-tide water level had waves capable of reaching the scarp. The vessel-generated waves accounted for ~5% of the cumulative wave energy over the study, but because of their larger height and longer period, they accounted for almost 25% of the cumulative wave force. It is the locally generated wind waves that account for most of the wave force acting on the exposed scarp and which are largely responsible for the observed retreat. Waves generated during both frontal and tropical storms tend to be associated with storm tides that maintain water levels at or near the elevation of the scarp for several days. Now that the shell ridge has largely migrated past the site and the areas of greater retreat historically are protected by a wedge of sand, it is argued that an increase in vessel traffic and/or the use of larger, post-Panamax ships will not significantly accelerate the retreat of the marsh.
ADDITIONAL INDEX WORDS: Vessel-generated waves, wind-generated waves, salt marsh erosion, Georgia.
Identifying causes of shoreline change is a difficult challenge, particularly in modified environments, where both response and recovery may be triggered by both ecological and hydrodynamic forces at a range of spatial and temporal scales. In salt marshes, a cycle of retreat and growth is common in response to changes in wave climate, tidal prism, and currents (Cox, Wadsworth, and Thomson, 2003); all of which may be associated with not only a change in incident forcing but also a feedback to the morphological development and seaward extension of the marsh. Development of the marsh exposes it to greater hydrodynamic forces that may eventually initiate a period of erosion that will continue until the marsh retreat is sufficient to reduce the hydrodynamic forces. The rate of scarp retreat depends on the amount of wave and current activity relative to the stability afforded by the root layer, which in turn depends on the density and type of marsh grass (van Eerdt, 1985). Where the height of the cliff is greater than the depth of root development, retreat tends to proceed through erosion of the underlying mud and undercutting of the root layer by waves (Schwimmer, 2001).
Continued erosion leads to a large overhang and tensional cracking that causes the undercut section to topple. This "beam failure" mechanism (Pizzuto, 1984) is exacerbated by water being forced against the root layer, causing holes to develop in the marsh surface and a reduction in the tensile strength of the surface (Schwimmer, 2001).
The rate of marsh retreat varies widely in response to the local wave climate, and Schwimmer (2001) showed it to be directly related to the wave power and correlated with storm events at elevated water levels. In many estuarine and fetch-limited environments, vessel-generated waves are often identified as another cause of shoreline change and retreat, particularly where commercial and recreational boat traffic is frequent (Anderson 1974; Bauer, Lorang, and Sherman, 2002; Castillo et al., 2000; Johnson 1957, 1969; Kos and Bielby, 1995; Nanson et al., 1994; Osborne and Boak, 1999; Pardy, 1995; Parnell, McDonald, and Burke, 2007). Despite these studies, the importance of vessel-generated waves to shoreline change is not clear and ranges from being a significant control (e.g, Castillo et al., 2000; Nanson et al., 1994; Parnell, McDonald, and Burke, 2007) to no impact (e.g., Cox, Wadsworth, and Thomson, 2003). In general, vessel-generated waves only contribute significantly to shoreline change where boat activity is regular, concentrated, and close to the shore. When these conditions are met, vessel-generated waves are capable of resuspending significant quantities of bottom and bank sediment because of their grouped nature and in spite of their short duration relative to wind-generated waves (Osborne and Boak, 2002). Once entrained, fine sediment can remain in suspension after the wave group has passed (Schoellhamer, 1996) and can be advected downstream with the mean current (Bauer, Lorang, and Sherman, 2002). Although this leads to a net loss of fine sediment and to bank erosion, coarse sediment can be transported landward in response to the asymmetric and skewed characteristics of the waves in the nearshore (Osborne and Boak, 2002). Depending on the sediment size and type of shoreline, the changes that occur in response to vessel-generated waves may persist even when vessel wakes are reduced or even eliminated (Parnell, McDonald, and Burke, 2007).
[FIGURE 1 OMITTED]
With the notable exceptions of Cox, Wadsworth, and Thomson (2003) and Allison (2005), there is a paucity of field data describing the role of vessel-generated waves in salt marsh erosion and retreat. As noted, Cox, Wadsworth, and Thomson (2003) found no evidence of erosion by vessel-generated waves, whereas Allison (2005) documented net sediment transport into tidal creeks (leading to erosion) in response to barge-induced drawdown and surge. Recently, concerns about vessel-generated waves along the Savannah River have been raised in light of plans to deepen the channel and to allow larger, post-Panamax ships to use the Port of Savannah. It is believed that an increase in vessel-generated waves will increase the rate of marsh retreat at Fort Pulaski National Monument (Figure 1), which now threatens a historically significant pier. In a recent study, Maynord (2007) concluded that a deeper channel would increase bow and stern waves along the Savannah River, but it was not clear whether an increase in vessel-generated waves would increase the rate of marsh retreat at Fort Pulaski. Cox, Wadsworth and Thomson (2003) suggested that the retreat of a salt marsh scarp within the modified Westerschelde Estuary (the Netherlands) was not related to ship activity. However, the limited role of vessel-generated waves at that site may reflect the large distance between the retreating marsh scarp and the shipping lanes, the presence of extensive sandflats to attenuate wave heights, and the greater importance of changes in the tidal prism following recent channel dredging. …