Vegetation-based metrics of biotic integrity for assessing the ecological condition of wetlands of Kentucky.

摘要

Over the last two centuries, wetland acreage across the world has significantly declined due to human disturbances. It has been estimated that Kentucky has lost over 80% of its wetland area. In response to these losses occurring across the United States, the Clean Water Act was passed to halt this dramatic decline and to restore the ecological integrity of waters of the United States. To enforce the Clean Water Act, a number of ecological assessment techniques have been developed to quantify the ecological quality of the waters of the United States. Kentucky recently adopted a rapid method for assessing the ecological condition of wetlands, but there is no standardized means to rigorously assess the ecological quality of its wetlands. Indices of biotic integrity (IBI) represent such a rigorous method and has become one of the most common approaches for intensive ecological assessment. IBIs evaluate the ecological condition of a site based on indicator organisms that reflect current and past anthropogenic disturbances of the area. In this study I report on initial efforts to develop a vegetation-based IBI for Kentucky (KY VIBI). Ohio has a state-wide applicable vegetation-based IBI (VIBI) for wetlands that has undergone multiple iterations of testing and refinement over more than 10 years. Due to the geographic and vegetative similarities between Ohio and Kentucky, Ohio's VIBI (OH VIBI) was used as a model to begin developing a state-wide applicable vegetation-based IBI for Kentucky.;A unique approach was used to begin the process of developing the KY VIBI. I developed a set of candidate metrics that included unmodified and slightly modified OH VIBI metrics, unmodified metrics from a VIBI study conducted in Colorado, and newly hypothesized metrics based on similar studies and my own professional knowledge of the plant communities of wetlands in Kentucky. The candidate metrics were tested for their response to disturbance indices using correlation analysis with data obtained from 68 wetland sites in Kentucky. Since metric response is expected to vary along a disturbance gradient, the resultant ecological condition of a site can be evaluated based on a core set of metrics that are related to anthropogenic disturbance. Sites were distributed across wetland types (emergent, forest, and shrub), as well as across the three major river basins (Green River, Kentucky River, and Upper Cumberland River). The disturbance indices were created by combining the non-biological submetrics of a newly developed rapid assessment method, the Kentucky Wetland Rapid Assessment Method (KY-WRAM), and the landscape disturbance index (LDI). The KY-WRAM and LDI were statistically combined using Principal Components Analysis (PCA) to create new disturbances indices. Combining these two separate measures of anthropogenic disturbance in a PCA resulted in better metric correlation compared to using either the KY-WRAM or LDI individually. The first two PC axes explained 48.35% and 13.47% of the total variation, respectively, and so those two axes were retained for comparison to the candidate KY VIBI metrics. Loading scores of variables were relatively strongly weighted on just the first or second axis of the PCA, suggesting good, simple structure in the PCA.;A list of the best ten candidate metrics were selected for each wetland type (emergent, shrub, and forest) based on their correlation and apparent response to disturbance, along with using best professional judgment. Many of the best metrics are related to invasiveness, tolerance, and floristic quality scores (e.g. Mean Coefficient of Conservatism for all Species, % Non-native Species, % Intolerant). Although the OH VIBI provided a sound methodological foundation for developing a VIBI for Kentucky, the results presented here suggest that many of the metrics in the current OH VIBI do not accurately reflect the biological effects of disturbance in Kentucky's wetlands, at least for the three river basins sampled for this study. The ten best candidate metrics will need to be further tested and evaluated for performance in other basins of Kentucky to ensure complete and proper calibration. Also, the individual metrics will need to be scaled, combined, and possibly weighted to create an index. By selecting and utilizing a different set of metrics with stronger association to disturbance we can more accurately describe wetland quality in the state of Kentucky.