Evaluation of the Basis and Effectiveness of Habitat Assessments in Wetland Functional Assessment Methods

Abstract

I studied the basis and effectiveness of wetland assessment methods in providing habitat assessments. While it is well understood that wetlands and riparian areas provide important ecological functions and habitat for a wide variety of wildlife species, much is still to be learned about providing meaningful, accurate and repeatable methods for assessing them. I examined and evaluated four assessment methods to determine their accuracy and usefulness in assessing a site's provision of habitat. One hypothesis I tested is that if the assessment methods studied provide an accurate assessment of wetland functions, then the resulting site scores for the methods should be correlated. The second hypothesis is that there is a correlation between the site scores and an independent measure of function, specifically the number of riparian-associated bird and butterfly species observed at each site. Biological and physical data collected from 47 riparian sites in California's Central Valley were used to calculate site scores using Habitat Assessment Technique (HAT), Rocky Mountain Riparian Hydrogeomorphic (HGM), Southern California Riparian Model, and Reference Wetland assessment methods. The rankings of these site scores were also calculated for each method. Correlation coefficients (r) were calculated between the site scores of the four methods, as well as between the site scores and the numbers of riparian-associated bird and butterfly species for each plot. The site scores were mostly uncorrelated. Only one statistically significant correlation was demonstrated between the site scores for the Southern California Riparian Model and Reference Wetland methods (df = 46, r = 0.46, p = 0.00103, with Bonferroni correction). With Bonferroni corrections (p < 0.00625), the site scores were also uncorrelated with the numbers of riparian-associated bird and butterfly species. Without Bonferroni corrections, only two statistically significant correlations were demonstrated: between the number of riparian-associated bird species and the HAT score (df = 46, r = 0.37, p = 0.0095) and the number of riparian-associated butterfly species and the Reference Wetland score (df = 46, r = 0.38, p = 0.0092). I rejected both original hypotheses, which demonstrated that the assessment tools currently available do not consistently produce relatively precise, or reproducible results. Possible reasons for these problems include attempting to assess a function that is too broadly defined, inappropriately or subjectively selected variables, subjectively assigning values to variables, or inappropriately selecting reference sites. The existing attempts at assessing wetland or riparian function are important steps in the right direction toward assessment of wetland and riparian sites and achievement of "no net loss," but functional assessment must be considered a work in progress.