In addition to increasing erosion, many of the world's low-lying coastal regions will be exposed to potential inundation. Deltas that are deteriorating as a result of sediment starvation, subsidence, and other stresses are particularly susceptible to accelerated inundation, shoreline recession, wetland deterioration, and interior land loss (Biljsma et al., 1996; Day et al., 1997).
River deltas are among the most valuable, heavily populated, and vulnerable coastal systems in the world. Deltas develop where rivers deposit more sediment at the shore than can be carried away by waves. Deltas are particularly at risk from climate changepartly because of natural processes and partly because of human-induced stresses. Deltaic deposits naturally dewater and compact as a result of sedimentary loading. When compaction is combined with isostatic loading or other tectonic effects, rates of subsidence can reach 20 mm yr-1 (Alam, 1996). Human activities such as draining for agricultural development; levee building to prevent flooding; and channelization, damming, and diking of rivers to impede sediment transfers have made deltas more vulnerable to sea-level rise. Examples of sediment starvation include the Rhone and Ebro deltas (Jimenez and Sanchez-Arcilla, 1997) and polder projects in the Ganges-Brahmaputra (Jelgersma, 1996). Sediment transport by the Nile, Indus, and Ebro Rivers has been reduced by 95% and in the Mississippi by half in the past 200 years, mostly since 1950 (Day et al., 1997). Further stress has been caused by subsurface fluid withdrawals and draining of wetland soils. In the Bangkok area of the Chao Phraya delta, groundwater extraction during 1960-1994 increased average relative sea-level rise by 17 mm yr-1 (Sabhasri and Suwarnarat, 1996). Similar severe land subsidence has been experienced in the Old Huange and Changjiang deltas of China (Chen, 1998; Chen and Stanley, 1998). In the latter case, groundwater removal was curtailed, leading to a reduction in subsidence rates.
Where local rates of subsidence and relative sea-level rise are not balanced by sediment accumulation, flooding and marine processes will dominate. Indeed, Sanchez-Arcilla and Jimenez (1997) suggest that in the case of largely regulated deltas, the main impacts of climate change will be marine-related because impacts related to catchment areas will be severely damped by river regulation and management policies. In such cases, significant land loss on the outer delta can result from wave erosion; prominent examples include the Nile (Stanley and Warne, 1998), Mackenzie (Shaw et al., 1998b), and Ganges (Umitsu, 1997). In South America, large portions of the Amazon, Orinoco, and Paraná/Plata deltas will be affected if sea-level rise accelerates as projected (Canziani et al., 1998). If vertical accretion rates resulting from sediment delivery and in situ organic matter production do not keep pace with sea-level rise, waterlogging of wetland soils will lead to death of emergent vegetation, a rapid loss of elevation because of decomposition of the belowground root mass, and, ultimately, submergence and erosion of the substrate (Cahoon and Lynch, 1997).
In some situations, saltwater intrusion into freshwater aquifers also is a potentially major problem, as demonstrated by a three-scenario climate change and sea-level rise model study of the Nile delta (Sherif and Singh, 1999). In other places, saltwater intrusion is already taking place (Mulrenna and Woodroffe, 1998). In the Yangtze delta, one consequence of saltwater incursion will be that during dry seasons shortages of freshwater for agriculture are likely to be more pronounced and agricultural yields seriously reduced particularly around Shanghai (Chen and Zong, 1999).
An estimate by Nicholls et al. (1999) suggests that by the 2080s, sea-level rise could cause the loss of as much as 22% of the world's coastal wetlands. Although there would be significant regional variations (Michener et al., 1997), such losses would reinforce other adverse trends of wetland loss resulting primarily from direct human actionestimated by DETR (1999) to be about 40% of 1990 values by the 2080s. Stabilization scenarios developed by DETR (1999) show a large reduction in wetland lossesto 6-7%, compared with unmitigated emissions (13%). Two main types of tidal wetlandmangrove forest and salt marshare considered here, although serious impacts on other coastal vegetation types, including subtidal seagrasses, can be expected (Short and Neckles, 1999).
Mangrove forests often are associated with tropical and subtropical deltas, but they also occur in low- to mid-latitude lagoon and estuary margins, fringing shorelines from Bermuda in the north to northern New Zealand in the south. Mangroves have important ecological and socioeconomic functions, particularly in relation to seafood production, as a source of wood products, as nutrient sinks, and for shoreline protection (Rönnbäck, 1999). Moreover, different kinds of mangrove provide different goods and services (Ewel et al., 1998). The function and conservation status of mangroves has been considered in special issues of two journals, introduced by Field et al. (1998) and Saenger (1998).
Many mangrove forests are being exploited and some are being destroyed, reducing resilience to accommodate future sea-level rise. In Thailand, 50% of mangrove has been lost in the past 35 years (Aksornkoae, 1993); yet with greatly increased sediment supply to the coastal zone in some places, mangrove colonization has expanded seaward in suitable habitats (Panapitukkul et al., 1998). As for other shore types, this example emphasizes the importance of sediment flux in determining mangrove response to sea-level rise. Ellison and Stoddart (1991), Ellison (1993), and Parkinson et al. (1994) suggest that mangrove accretion in low- and high-island settings with low sediment supply may not be able to keep up with future sea-level rise, whereas Snedaker et al. (1994) suggest that low-island mangroves may be able to accommodate much higher rates of sea-level rise. This ability may depend on stand composition and status (e.g., Ewel et al., 1998; Farnsworth, 1998) and other factors, such as tidal range and sediment supply (Woodroffe, 1995, 1999; Miyagi et al., 1999). In some protected coastal settings, inundation of low-lying coastal land may promote progressive expansion of mangroves with sea-level rise (Richmond et al., 1997). In contrast, Alleng (1998) predicts the complete collapse of a mangrove wetland in Jamaica under rapid sea-level rise.
The response of tidal marshes to sea-level rise is similarly affected by organic and inorganic sediment supply and the nature of the backshore environment (Brinson et al., 1995; Nuttle et al., 1997). In general, tidal marsh accretion tracks sea-level rise and fluctuations in the rate of sea-level rise (e.g., van de Plassche et al., 1998, 1999; Varekamp and Thomas, 1998). Marsh accretion also reflects marsh growth effects (Varekamp et al., 1999). Bricker-Urso et al. (1989) estimate a maximum sustainable accretion rate of 16 mm yr-1 in salt marshes of Rhode Island (assuming that vertical accretion rates are controlled mainly by in situ production of organic matter); this rate is an order of magnitude higher than rates reported by others. Orson et al. (1998) also emphasizes the effects of variability between marsh species types.
Temporal and spatial variability in rates of relative sea-level rise also is important. Stumpf and Haines (1998) report rates of >10 mm yr-1 in the Gulf of Mexico over several years, where the long-term mean rate of relative sea-level rise is 2 mm yr-1 or less. Forbes et al. (1997b) report multi-year fluctuations in sea-level rise at Halifax, Nova Scotia, of as much as 10 mm yr-1 and occasionally higher, superimposed on a long-term mean of 3.6 mm yr-1. Thus, short-term fluctuations in sea-level rise may approach the maximum limit of accretion, although the drowning of marsh surfaces is unlikely to be a major concern. Higher local rates of sea-level change have been recorded over the past 100 years or so in a few places, one of which is the Caspian Sea. Here, the response of riparian vegetation to the sea-level fall in the early part of the 20th century was a rapid seaward progression of vegetation. This progression ceased with the rise in Caspian sea level averaging 120 mm yr-1 from 1978 to 1996, but it did not result in a similar rapid regression of vegetation. Instead, the vegetation consolidated its position, which has been partly explained by the wide flooding tolerance of the major emergent plant species, with floating vegetation increasing in extent with more favorable (higher water level) conditions (Baldina et al., 1999).
In some areas, the current rate of marsh elevation gain is insufficient to offset relative sea-level rise. For instance, model results from a wetland elevation model designed to predict the effect of an increasing rate of sea-level rise on wetland sustainability in Venice Lagoon revealed that for a 0.48-m rise in the next 100 years, only one site could maintain its elevation relative to sea level; for a 0.15-m rise, seven sites remained stable (Day et al., 1999).
Maintenance of productive marsh area also depends on horizontal controls discussed by Nuttle et al. (1997) and Cahoon et al. (1998). For example, in settings with sufficient sediment influx, the wetland may expand toward the estuary, while also expanding landward if the backshore slope is sufficiently low and not backed by fixed infrastructure (Brinson et al., 1995). If sediment supply is low, however, marsh front erosion may occur (Dionne, 1986).
Although determining the threshold for such erosion is difficult, this erosion is regarded as a negative impact on many wetlands, particularly those constrained by artificial structures on the landward side. Nicholls and Branson (1998) use the term "coastal squeeze" to describe the progressive loss and inundation of coastal habitats and natural features located between coastal defenses and rising sea levels. They believe that intertidal habitats will continue to disappear progressively, with adverse consequences for coastal biological productivity, biodiversity, and amenity value. Where sediment influx is insufficient to sustain progradation, there is potential for significant loss of coastal wetlands. After considering two "what if" climate change scenarios, Mortsch (1998) found that key wetlands around the Great Lakes of Canada-USA are at riskparticularly those that are impeded from adapting to the new water-level conditions by artificial structures or geomorphic conditions.
Other reports in this collection