Forecasting ecological impacts of sea-level rise on coastal conservation areas in India

1 Bombay Natural History Society, Opp. Lion Gate, Shaheed Bhagat Singh Road, Mumbai, Maharashtra 400001, India 1 Current Address: National Wildlife Research Center, PO Box 1086, Taif, Saudi Arabia 2 Department of Biology, Grand Valley State University, Allendale, Michigan 49401 USA 3 Department of Geography, University of Kansas, Lawrence, Kansas 66045 USA 4 Biodiversity Institute, University of Kansas, Lawrence, Kansas 66045 USA 1 mzafarul.islam@gmail.com (corresponding author), 2 menons@gvsu.edu, 3 lixi@ku.edu, 4 town@ku.edu


INTrOduCTION
Several recent studies have accumulated empirical evidence of climate change effects on the distribution and diversity of species and ecosystems.Examples include poleward and upward elevational range shifts in butterfly (Parmesan et al. 1999;Parmesan & Yohe 2003) and mammal species (Moritz et al. 2008;Tingley et al. 2009), mistimed reproduction in bird species (Visser et al. 1998) and advanced spring greenup and other changes in plant phenology (Miller-Rushing & Primack 2008;Vitasse et al. 2009).Such empirical evidence is complemented by predictive modeling efforts based on linkage of ecological niche models with general circulation model (GCM) outputs (e.g., Erasmus et al. 2002;Peterson et al. 2002Peterson et al. , 2005;;Thomas et al. 2004a;Araújo et al. 2005;Anciães & Peterson 2006), which anticipate similar poleward and upward shifts, with significant range losses when species' dispersal potential is constrained by geographic factors.
Indirect climate change-associated phenomena such as sea-level rise are also beginning to receive attention.Major causes of rising sea levels include thermal expansion of the ocean, mountain glacier melting, and discharge ice from ice sheets (Dyurgerov & Meier 1997).Accelerating discharge of glacial ice due to ice sheet melt and tidal and storm surges are expected to exacerbate the situation further.Projections of sealevel rise can vary dramatically owing, at least in part, to the complexity of the factors contributing to this phenomenon.Carter et al. (2007) and IPCC (2007) offered a conservative estimate of anticipated sea-level rise on the order of 0.5-1.0m, while other estimates are much higher on the order of 4-6 m (Bindschadler 1998;Thomas et al. 2004b;Rignot & Kanagaratnam 2006).
Rising sea levels and associated marine intrusion into terrestrial environments are expected to be among the most serious consequences of climate change.Projected estimates of 1-6 m rise of sea levels are likely to have catastrophic consequences for biodiversity and humans.The human and economic consequences of sea-level rise have received some attention in recent years (Titus 1990;Mimura 1999;Hitz & Smith 2004;Bosello et al. 2007).A few analyses have addressed various aspects of the biodiversity consequences of sea-level change: threatened and endangered species' habitat destruction in the southeastern US (Daniels et al. 1993), potential losses of intertidal habitat for shorebirds (Galbraith et al. 2001), potential effects in a mangrove ecosystem (Gopal & Chauhan 2006), likely effects on a single endangered species (LaFever et al. 2007) and ecosystem adaptation to rising sea levels (McKee et al. 2007).
A first-pass global assessment of biodiversity consequences of the sea-level rise (Menon et al. 2010), provided rough estimates of areal losses of ecoregions and species extinctions due to marine intrusion: several ecoregions were projected to lose more than half of their present-day land area even under a 1m sea-level rise.Recent events, such as the disappearance of New Moore Island in the Bay of Bengal (BBC News 2010) have further underscored the urgency of this issue.Acknowledging the significance of this topic, the Survey of India announced, in March 2010, a plan to map the hazard line along India's coastlines.
Here, we present an overview of potential consequences of 1m and 6m sea-level rise for coastal conservation areas on the Indian subcontinent.Specific objectives of this study were to explore (i) areal losses due to marine intrusion in coastal biogeographic zones and ecoregions; (ii) likely impacts of marine intrusion on coastal protected areas (PAs) and Important Bird Areas (IBAs), which provide critical habitat for species; and (iii) adaptation and mitigation strategies for minimizing such impacts.

METhOdS data Sources
We used the output of the study generated by Li et al. (2009).This method implemented several steps in a GIS raster analysis framework to select and designate as 'inundated areas' those cells that (i) are below a projected sea level rise, (ii) are connected to the ocean, and (iii) are not part of existing inland water bodies.The sea-level rise scenarios generated by this method are an improvement on previous estimates (Dasgupta et al. 2007;LaFever et al. 2007) which tended to overpredict potential inundated areas.
GIS data sets of conservation areas were obtained from various sources.Biogeographic zones (areas with shared biological and geographic characteristics) were delineated from maps prepared by Rodgers & Panwar (1988) and Rodgers et al. (2002).Ecoregion data were obtained from the Terrestrial Ecoregions GIS Database (Olson et al. 2001).Protected areas maps were obtained from the World Database on Protected Areas (IUCN and UNEP-WCMC 2010).Finally, point location data for conservation areas were obtained from field surveys, published records and cross-checked with gazetteers of India (Islam & Rahmani 2004;Islam & Rahmani 2008).

Estimates of areal loss and biodiversity impacts
We converted the GIS layers of biogeographic zones, ecoregions and protected areas into equal-area grids (we used Albers projection because it preserves area measurements).After assuring that all grids were on a common projection and grid resolution, we estimated areal loss resulting from marine intrusion by overlaying them with the 1m and 6m inundation grids.Gridsquare resolution was 822m on a side.Data on globally threatened birds were collected from field surveys, published records, BirdLife International, and Indian Bird Conservation Network (IBCN) partners, and were validated by comparison with summaries from regional IBA workshops across the country in 2001 and 2002 (Islam & Rahmani 2004).Digital data on biogeographic zones were obtained from BirdLife International (1998), Rodgers & Panwar (1988), Rodgers et al. (2000), and Champion & Seth (1968).

rESuLTS
Total areal loss due to marine intrusion into coastal areas of the Indian subcontinent is estimated at approximately 13,973km 2 (3%) and 60,497km 2 (14%) of the land area under the 1m and 6m sea-level rise scenarios, respectively (Table 1).Inundation of biogeographic zones ranged from 0-18% under 1m sea-level rise and 0-56% under 6m sea-level rise.As  expected, "Coasts" and "Islands" are the zones predicted to be most severely impacted by potential marine intrusion."Coasts" are predicted to undergo 12% and 56% inundation under the two sea-level rise projections, and "Islands" are predicted to undergo 18% and 23% inundation (Table 1).Marine intrusion is predicted to affect 18 of the 48 ecoregions in India (Table 2).Under the 1m sea-level rise scenario, estimates of ecoregion inundation ranged from 19% to 59%.Under the 6m sea-level rise scenario, estimates of ecoregion inundation ranged from 27-58 % (Table 2).Under the 1m sea-level rise scenario, one ecoregion (Godavari-Krishna mangroves) is predicted to lose more than a quarter of its area and another (Sunderbans mangroves) is predicted to lose more than half of its area.Under the 6m sea-level rise scenario, three ecoregions (Sunderbans freshwater swamp forests, Andaman Islands rain forests, and Maldives-Lakshadweep-Chagos Archipelago tropical moist forest) are predicted to lose more than a quarter of their land areas, and three more (Sunderbans mangroves, Godavari-Krishna mangroves, and Rann of Kutch seasonal salt marsh) are predicted to lose more than half of their land area.
A closer examination of a sample of 22 coastal conservation areas (Table 3) indicates that nine will be spared effects of marine intrusion under 1m sea-level rise, but only one will be spared under a 6m sea-level rise scenario.Of those impacted by marine intrusion, the extent of predicted inundation is quite variable, ranging from 1-95 % and 2-100 % under 1m and 6m sealevel rise, respectively (Table 3).Seven protected areas (Bhitarkanika, Chilka Lake, Point Calimere, Interview Island, Lothian Island, Sajnakhali, and Pulicat Lake) are expected to experience >50% inundation under 1m sea-level rise, and an additional four protected areas (Kachchh Desert, Velavadar, Pulicat, and Nal Sarovar), join this list under 6m sea-level rise.Images 1-4 depict the extent of predicted marine intrusion in some of these protected areas and their surroundings.

dISCuSSION
As in the preliminary assessment of global areal losses of ecoregions (Menon et al. 2010), the increased losses under the 6m scenario is clear in this analysis.Several coastal ecoregions and conservation areas are predicted to lose over half of their land areas to marine intrusion, particularly under the 6m sea-level rise scenario.Coastal zones have high biological productivity and support large number of birds and other taxa including mangroves.Populations of several species are expected to be displaced as a result of changes in the timing and magnitude of coastal biological productivity due to climate change (Both et al. 2006).
Most IBAs in coastal areas support species of global conservation concern, including some endemic and several breeding populations of threatened species.
The Andaman Islands support eight endemic bird species, and an additional four restricted-range species that they share with the Nicobar Islands.One endemic species, the Narcondam Hornbill (Aceros narcondami), is globally threatened and confined to the tiny island of Narcondam (<7km 2 ) in the northern part of the island group.The status of Narcondam Hornbill needs to be assessed quickly, given the very small size of the island and its potential vulnerability.In addition, Andaman Teal (Anas albogularis), endemic to the Andamans, is scarce, has recently declined, and is likely to be affected by sealevel rise because of its coastal distribution.
Similarly five bird species are endemic to the Nicobar Islands.Three of the endemic species are globally threatened: Nicobar Sparrowhawk (Accipiter butleri), Nicobar Megapode (Megapodius nicobariensis) and Nicobar Bulbul (Hypsipetes nicobariensis); of these, Nicobar Megapode is of particular concern because its greatest concentrations are found in coastal forests.The other two endemic species, South Nicobar Serpenteagle (Spilornis klossi) and Nicobar Parakeet (Psittacula caniceps) are both Near Threatened and confined to the southern islands, and might also be affected.
Hundreds of species of flora and fauna, including globally threatened species, depend upon low-lying coastal ecosystems for their survival.Indeed estuarine habitats and mudflats are sensitive to changes in both their upstream watersheds and the off-lying oceans.Several important biodiversity areas including in particular protected and non-protected IBAs in the Bay of Bengal and Arabian Sea, have already been stressed by numerous anthropogenic impacts as well as by invasive species (Islam & Rahmani 2004;Islam & Rahmani 2008), perhaps making them less resilient to change from the outset.
Coastal areas are subject to a range of intrinsic and extrinsic factors that could inhibit the ability of populations to adapt to climate change (Crick 2004).The prediction of future coastal evolution is hindered by the lack of standard methodology or agreement about the types of data and approaches required (Boesch et al. 2000).Factors that may impede tracking coastal changes include anthropogenic transformations, disjunct potential areas for species migration, and barriers to dispersal; where possible, ecosystems may respond by shifting inland, rather than with losses.
However, we note that protected areas, which generally have legal descriptions as part of their decrees, do not shift so easily, which may leave coastal areas inundated and key habitats unprotected.The inundation of mangroves is likely to result in a shift in species composition, with the deepest mangroves dying out.The methods used in our analyses are clear and quantitative with explicit assumptions.The relative susceptibility of different coastal environments to sea-level rise may be quantified at regional to national scales (Gornitz et al. 1994) using basic data on coastal geomorphology, rate of sea-level rise, and past shoreline evolution.More detailed coastal and marine geological data is needed to permit a comprehensive assessment of the susceptibility of the Indian subcontinent to sealevel rise.

Figure 1a .
Figure 1a.Important Bird Areas (IBAs) directly affected by a 1m sea-level rise, shown as solid blue circles.Other IBAs are shown as red circles and inundated areas appear in gray.

Figure 1b .
Figure 1b.Important Bird Areas (IBAs) directly affected by 6m sea-level rise, shown as solid blue circles.Other IBAs are shown as red circles and inundated areas appear in gray.
Images 1 & 2. Selected conservation areas (protected areas and IBAs) impacted by sea-level rise (1) haliday Island, Lothian Island, Narendrapur, and Sajnakhali Wildlife Sanctuary, (2) Chilka Lake [Key: These are true color satellite images, in which green represents vegetation, brown represents bare ground and dark blue is water.In the insets, the area inundated by 1m sea level rise is indicated in white and the area inundated by 6m sea level rise is shown in light blue.]