Assessing the condition of Hawaiian coastal wetlands using a multi-scaled approach

Abstract

The U.S. Environmental Protection Agency (EPA) advises using a multi-scaled approach to evaluating the integrity of wetland ecosystems that includes intensive field surveys (Level I), rapid on-site assessments (Level II), and remote or landscape-scale assessments (Level III). Although wetland condition assessment procedures have been developed, validated, and calibrated in the continental U.S., they have not yet been fully developed or field-tested with wetlands in the Pacific Islands. To my knowledge, this project was the first to field test rapid assessment methods (RAMs) in Hawaiian coastal wetlands. Three on-site RAMs, developed for coastal wetlands in the continental U.S. (Florida Wetland Rapid Assessment Procedure, California Rapid Assessment Method, and the Hawaiʻi Hydrogeomorphic Method), were field tested in 27 coastal Hawaiian wetlands. In addition, three different indicators of landscape condition (landscape development intensity, road density, and forest cover) were used in remote (Level I) condition assessments of the wetland buffers (100m and 1,000m radii) and watersheds. Based on the assumption that condition assessments collected at different levels of detail should provide consistent information, the results from the remote and rapid assessments were compared with detailed (Level III) field data collected on soil and water quality during prior surveys (2007-2009). Results showed that landscape indicators (development intensity and road density) calculated from readily available GIS data, particularly for the 1,000m buffers and watershed basins of each wetland, were significantly correlated with wetland soil parameters (i.e. bulk density [BD], pH, soil organic matter [SOM], total nitrogen [TN], and extractable phosphorus [ExP]) and with water quality parameters (total dissolved nitrogen [TDN], total phosphorus [TP], and δ15N levels in wetland plant tissue). In addition the high degree of correlation between landscape development intensity (LDI) and road densities (in watersheds; r = 0.95, p < 0.01) suggests that both provide comparable indicators of anthropogenic stressors in Hawaiian landscapes that vary from natural, mixed agriculture, to urban. The scores from the rapid assessment methods (RAMs), however, were somewhat weakly correlated with soil parameters (i.e. BD, total carbon [TC], SOM, and TN, ExP), yet strongly correlated with the δ15N levels in wetland plant tissue, an indicator of human and animal waste. The correlations between RAM scores and landscape indicators (LDI scores and road density), calculated for the wetland buffers (1,000m) and watersheds, were also strong. Among the rapid assessment methods, CRAM showed the strongest correlations with Level I and Level III data, which suggests that although CRAM was developed for condition assessments along the Pacific West coastline (from Mexico to Oregon), with modifications to account for local conditions, the method could be applied to assessing the condition of Hawaiian coastal wetlands.