Global mean sea levels are expected to rise up to 1.9m by 2100 and will change the distribution and community structure of low-lying coastal ecosystems due to flooding, erosion, and salt-water intrusion. Although habitats will be inundated, ecosystems have the potential to shift inland and endemic species may persist if conditions are favorable. Multiple-stressors including invasive species and destructive land-use practices have caused widespread coastal aquatic ecosystem degradation over the last century. Predictions of ecosystem migration due to sea level rise need to account for these stressors, which, even at low levels, may reduce the resilience of these ecosystems. The goal of this dissertation was to predict potential consequences of future sea level rise on groundwater-fed anchialine pool ecosystems in Hawaii by comparing scenarios of inundation with current patterns of habitat, introduced species, and land use.

First, high resolution sea level rise models that incorporate groundwater were developed for the west coast of the island of Hawaii. Ecological surveys were also conducted to identify habitat condition and the occurrence of native and non-native species at 398 pools located along 280 km of the coastal corridor. Statistical models were used to determine the current relationship between native and non-native species occurrence and various physical and biological parameters. Stable isotope analysis was used to examine food web relationships within pools. Geospatial analyses were used to predict future pool inundation, pool creation, and non-native species dispersal under various sea level rise scenarios. Results showed that sea level models incorporating groundwater levels were up to 37% better at detecting known anchialine pools than corresponding models without groundwater levels (Chapter 1). Pool surveys showed that two dominant endemic shrimp Halocaridina rubra and Metabetaeus lohena were located across the region in a wide range of habitats, but that introduced fishes (tilapia, poeciliids) were present in ~ 25 % of pools. Statistical models showed that introduced fishes had a strong negative effect on the occurrence of H. rubra and M. lohena, while benthic silt cover and adjacent development also had significant negative relationships with shrimp occurrence (Chapter 2). Geospatial models indicate that current habitats will be increasingly inundated by marine waters, but that new habitat will emerge in the landscape in low lying open space (Chapter 3). Because of high subsurface hydrologic connectivity, many endemic species are likely to populate these new habitats by moving through the coastal aquifer. However, in some areas, rising sea levels will provide surface connectivity between pools currently infested with introduced fishes (Chapter 3). Additionally, results suggest that tilapia are a bigger threat to endemic anchialine species and ecosystem health and should be targeted for removal before poeciliids (Chapter 4). This study demonstrates an interdisciplinary approach to examining ecosystem risk due to sea level rise. Successful conservation of coastal aquatic biodiversity will continue to require current restoration efforts along with protection of potential habitat sites under future climate conditions.