Analysis and Simulation of Propagule Dispersal and Salinity Intrusion from Storm Surge on the Movement of a Marsh-Mangrove Ecotone in South Florida
In wave-protected shorelines of southern Florida, mangroves are the dominant wetland ecosystem (Lugo and Snedaker 1974; Bertness 2007; Zhang et al. 2012). These mangrove communities usually form landward transition zones (ecotones) with freshwater marshes. The mangroves of southern coastal Florida have been observed to be expanding inland at the expense of both salt marsh and freshwater marsh over several decades (Ball 1980; Ross et al. 2000; Doyle et al. 2003; Gaiser et al. 2006; Krauss et al. 2011; Raabe et al. 2012). Sea level rise of more than 20 cm since 1930 (Maul and Martin 1993) is thought to be the primary driver of the changes (Krauss et al. 2011; Saha et al. 2011). In his classic study, Egler (1952) documented distinct zonation bands of vegetation, primarily from mangroves to sawgrass marsh, along a transect from the coast inland. He noted the movement of Rhizophora mangle into Cladium jamaicense marsh, facilitated by storm tides that carry both salinity and Rhizophora seedlings far inland on the flat landscape. Ross et al. (2000) revisited the sites of Egler's study and found that, during the past 40 years, a landward shift in the bands had occurred. In particular, vegetation in a zone termed the "white zone" (band 4 in Egler 1952) had shifted from a mixture of C. jamaicense and R. mangle to pure R. mangle. The authors attributed this change to sea level rise and water management practices that have reduced freshwater flow to the southern Everglades. Landward movement of mangroves into freshwater marsh in other areas of the southwestern Everglades has been identified through aerial photographs (Smith III et al. 2013).
Much of what has been documented concerning the spatial and temporal patterns of vegetation along the marine to terrestrial transect in southern Florida can be explained in terms of familiar ecological concepts. Higher tolerance to salinity favors mangrove species dominance at the marine end of this transect, while competition favors dominance by glycophytic species at the terrestrial (freshwater) end of the transect. Gradual rise in sea level and anthropogenic decreases in freshwater flow cause salinity intrusion and a landward shift of the vegetation bands, with halophytic mangroves replacing glycophytic freshwater marsh and hardwood hammocks. Other factors, such as changes in fire frequency, may also be affecting the ecotones between some of these bands (Egler 1952).
An additional factor, however, may also play a role in the dynamics of the shifting ecotones: positive feedbacks between a given vegetation type and its local soil conditions. Sternberg et al. (2007) hypothesized that feedback effects of two vegetation types, mangroves and hardwood hammocks, on local soil salinity help maintain the sharp ecotone frequently observed between them. Although mangrove vegetation is outcompeted by glycophytic vegetation under less saline conditions inland, mangroves, once established, can influence vadose zone salinity in their favor. A suite of salt tolerance mechanisms in mangroves allows them to transpire during periods of higher salinity, despite having rather conservative rates of water use overall. This facilitates higher evapotranspiration rates than would occur with the presence of only glycophytic vegetation. Mangrove roots access to water allows saline groundwater to move up in the vadose zone, further increasing its salinity. Thus, mangroves create locally high soil salinity conditions, while neighboring glycophytic vegetation, which downregulates evapotranspiration during the dry season, maintains locally low soil salinity. Each vegetation type creates soil conditions favoring itself over the other type. The resulting sharp change in soil salinity from one vegetation type to the other is confirmed by observations (Sternberg et al. 2007; Saha et al. 2011). This is an example of the more general "vegetation switch" mechanism (Agnew et al. 1993), which has been used to explain sharp ecotones between differing vegetation types, such as forest-grassland, forest-mire, Alpine treelines, etc. The vegetation switch can be thought of as a form of vegetation pattern generation or self-organization, that is, it arises through internal interactions to create structure (e.g., Temmerman et al. 2005, 2007; Kirwan and Murray 2007).
There are several implications of the existence of such a vegetation switch for ecotones in general and the halophyte- glycophyte ecotone in particular. First, as noted above, this positive feedback mechanism should act to sharpen ecotones between the vegetation types. Second, because each vegetation type has a positive effect on its local soil conditions that favor itself, the spatial position of the ecotone should be fairly resilient to small disturbances. For example, a minor disturbance that changes soil salinity by only a small amount might be counteracted by the local vegetation. The third implication is that, although the ecotone between the two vegetation types is sharp, its exact location on the marine-terrestrial transect may not be deterministic. The underlying abiotic conditions at a given point along a segment of the transect may be suitable for either vegetation type to thrive, but, once established at some point on that segment, one of the vegetation types can influence those conditions such that it can exclude the other type. The segment of the transect along which either vegetation type could exist in the absence of the other is called a region of bistability, and the two vegetation types that could each dominate the region are called alternative stable states (Scheffer 1997; Beisner et al. 2003; Folke et al. 2004). The exact location of the ecotone may depend on initial conditions. A fourth implication of the vegetation switch hypothesis is that a sufficiently large disturbance may be able to overcome the resilience of the ecotone in the region of bistability, causing a shift, called a regime shift, from the existing stable vegetation state on one side of the ecotone to the alternative stable vegetation state. For a regime shift to occur, however, the disturbance would have to exceed some threshold in strength and duration of the change in environmental conditions and be accompanied by input of propagules of the alternative vegetation type. But once a regime shift occurs, it may be difficult to reverse due to hysteresis. A fifth implication is that a mixture of the alternative vegetation types is unstable, and the system will move toward dominance of one or the other type, or possibly toward segregation of the two types on either side of a sharp ecotone. This transition may be slow, however, depending on the rates of competitive exclusion.
The existence of a vegetation switch mechanism in the mangrove-freshwater marsh ecotone zone changes the sort of dynamics that can occur when external conditions change. For example, a gradual rise in sea level might not lead to gradual landward movement of vegetation bands, as the vegetation types are self-maintaining and the ecotones would thus be resilient to small changes in sea level. The switch mechanism suggests that ecotones may be stable much of the time but be punctuated by large jumps in space when major disturbances occur.
The hypothesis of Sternberg et al. (2007), specifically the sharp mangrove-hardwood hammock ecotone, has been supported through simulation modeling (Teh et al. 2008; Jiang et al. 2012a), consistent with the idea of a vegetation switch mechanism. If the other implications of the switch mechanism also apply to the transition between halophytic and glycophytic vegetation, then we expect there to be segments of the transition that are bistable and that could undergo regime shifts from their current vegetation state to an alternative vegetation, if sufficiently large disturbances occurred. In particular, a storm surge with large overwash of saline water might shift a region currently covered with glycophytic vegetation to undergo a shift to halophytic vegetation. Ocean water intrusion through storm surges may affect large areas on a short time scale (e.g., cause a short-term salinity increase in the soil and groundwater of an inundated area). Whether such storm-related pulses lead to long-term effects on vegetation depends on many specific factors, such as the physiological and competitive properties of local vegetation, precipitation, overland freshwater flow, elevation gradient, depth and salinity of groundwater, and sufficient input of seedlings of the alternative vegetation type (Jiang et al. 2012a; White and Falkland 2010). If these conditions are satisfied, the possibility of a regime shift over a broad area affected by a storm surge exists, which is why storm surges are of special interest in relation to mangrove migration and other vegetation changes (Ross et al. 2009; Teh et al. 2008). There have been some relevant studies on transitions from glycophytic to halophytic vegetation types due to salinity inputs. For example, Baldwin and Mendelssohn (1998) studied the effects of a pulse of salinity and inundation coupled with clipping of aboveground vegetation on two adjoining plant communities, Spartina patens and Sagittaria lancifolia. The study concluded that the vegetation might shift to a salt-tolerant or flood-tolerant species, depending on the level of flooding and salinity at the time of disturbance. Also see Person and Ruess (2003) and Steyer et al. (2010) for relevant studies.
The mangrove-freshwater marsh (Cladium) ecotone is often very sharp, which suggests that positive feedback effects similar to those at the mangrove-hardwood hammock boundary, and thus a switch mechanism may occur. The question we ask here is whether regime shift through salinity overwash is a feasible mechanism for explaining some of the past inland movement of the boundary in the coastal Everglades of southern Florida, or for influencing future movement. The hypothesis of a regime shift requires two primary prerequisites: an increase in salinities and an invasion of mangrove propagules following a storm surge, or alternatively a prior presence of propagules. Through an analysis of the effects of increased salinity and mangrove propagules with a model based on a regime shift theory, we demonstrate how shifts resulting in long-term changes in the mangrove-marsh ecotone could occur after storm surge events and inquire whether, on the basis of empirical data, they are likely to occur in a particular region of the Everglades affected by a hurricane. First, coastal water level and salinity data from 2000-2010 are used as a hydrological baseline and to observe possible trends in increased coastal groundwater salinity. Second, vegetation dynamics are simulated based on hydrology data from Hurricane Wilma (Oct 2005). Third, long-term vegetation changes under given scenarios of salinity intrusion and the density of mangrove propagules transported to freshwater marsh sites by hypothetical storm surge are simulated.