Considerable attention has focused on the effects of global climate change on biodiversity, but few analyses and no broad assessments have evaluated effects of sea-level rise on biodiversity. Menon et al. calculated the total area lost for all terrestrial ecoregions using new maps of marine intrusion under scenarios of 1 and 6 m sea-level rise. Areal losses for particular ecoregions ranged from nil to complete. Marine intrusion is a global phenomenon, mostly affecting Southeast Asia and nearby islands, eastern North America, northeastern South America, and western Alaska. Assuming that fauna respond to reduced ecoregions in a predictable manner, the authors estimated likely numbers of extinctions caused by sea-level rise. They found that northeastern South America is most susceptible to marine-intrusion-caused extinctions, although anticipated extinctions in smaller numbers will be scattered worldwide. This assessment is the first global analysis of sea-level rise impacts on terrestrial biodiversity, complementing recent estimates of losses owing to changing climatic conditions. —Michelle Schulte
Menon, S., Soberón, J., Xingong, L., Peterson A.T., 2010. Preliminary global assessment of terrestrial biodiversity consequences of sea-level rise mediated by climate change. Biodiversity Conservation 19, 1599–1609.
Past studies indicated that the rate of future melting of polar ice sheets and related sea-level rise could be faster than widely thought, resulting in a sea-level rise of 4–6 m by 2100. Menon et al. used geographic information systems (GIS) to delineate potential inundation areas resulting from projected sea-level rise of 1 and 6 m. In this analysis, cells below a projected sea-level rise that connect to the ocean and are not presently inland water are designated as inundation cells. The authors used the Terrestrial Ecoregions GIS Database and the Terrestrial Ecoregions Base Global Dataset as a source of geospatial data showing the global extent of ecoregions, as well as providing data on numbers of endemic species in each ecoregion. Menon et al. used values for strict endemic species and near-endemic species across all 827 terrestrial ecoregions in this analysis. They then converted the vector-format terrestrial ecoregions coverage into a grid, so as to estimate the area lost from marine intrusion by overlaying it with the 1 and 6 m inundation scenarios grids and performing raster map algebra.
A decrease in the area of an ecoregion can be used to estimate biodiversity losses under certain sets of assumptions. Past studies have employed the relationship between the numbers of species present and area under consideration (species–area relationship, or SAR) to calculate future extinctions. The SAR is a steady-state relationship between number of species (S) and area (A) of the form S = cAz, where cand z are constants estimated empirically. If the present number of species Snow is existing in an area Anow, which is reduced to Afuture, and if c and z remain constant, then the number of species will eventually decrease to a new steady state Sfuture = Snow (Afuture/Anow)z. The authors estimated z in two different ways: as the overall SAR across all ecoregions globally, and SARs for 3 latitudinal bands (polar, temperate, tropical). The authors calculated Sfuture for each ecosystem under the general z and the latitude-specific z, and estimated confidence intervals for each Sfuture.
Globally, 0.7% of global land was inundated and therefore lost under 1 m of sea-level rise, and 1.5% of global land area under 6 m of sea-level rise. Proportional losses in ecoregions ranged from 0 to 100%. The most affected ecoregions were Southeast Asia and associated islands, northeastern South America, eastern North America, and western Alaska. Even under a 1 m sea-level rise scenario, 21 ecoregions are expected to lose >50% of their land area, which include 8 mangrove-dominated ecoregions, lowland forest and scrub on 8 islands, and 5 low-lying continental areas. Thus, sea-level rise manifested as marine intrusions is expected to greatly affect terrestrial ecoregions.
For the global SAR fitting, z was estimated at 0.124 ± 0.015 s.e., although the overall fit was not particularly tight. Out of a total of 18,628 endemic or near-endemic species in single ecoregions, this single SAR yielded a calculated loss of 117 ± 27 species for the 1 m sea-level rise scenario, and 221 ± 51 species for the 6 m scenario. Splitting SAR regressions into polar, temperate, and tropical subsets, important regional differences were observed. The slope of the SAR (z) was highest in tropical regions, and lowest in polar regions. Also, these SAR differences translated into different rates of estimated species loss under the 1 and 6 m scenarios: 0 of 35 polar species under both scenarios; 10 and 30 out of 3,117 species in temperate regions; and 170 and 307 out of 15,476 species in tropical regions. Overall, with region-specific z estimates, global species losses sum to 181 ± 23 species under the 1 m scenario and 337 ± 44 species under the 6 m scenario, out of 18,628 current species.
Past studies have criticized on a number of grounds the use of the linear relationship of species and area to estimate future extinctions. Menon et al., however, controlled for the possible errors given the data limitations in resolution of area and in the taxon, range, and fragmentation of species. Certainly, both the marine-intrusion and the biodiversity distribution summaries could be improved significantly. For the marine-intrusion scenario, improvements are needed in the horizontal (from 1 km to 10 m resolution) and vertical (<1 m) resolutions. Also, moving from crude ecoregion-based summaries to actual species-specific distributional information would improve the estimates of the biodiversity distribution. Finally, because some species, such as keystone species, may play more critical roles in maintaining communities than others, categorizing the individual species as to their relative ‘importance’ in community structuring will clarify the magnitude of secondary effects.
Overall, the authors present a valid preliminary assessment of likely biodiversity consequences of sea-level rise and marine intrusion caused by climate change. The most realistic scenario of the two that were explored is a rise of 1 m by 2100, although the 6 m scenario is still very possible given the effects of glacial calving and ice-sheet loss. This analysis does not account for second-order effects on biodiversity caused by humans affected by rising sea levels, such as migrations and land use shifts, which may cause yet more negative effects on natural systems.