Development and evaluation of NatureServe’s multi-metric ecological integrity assessment method for wetland ecosystems

• Published in: Ecological indicators 2019-09, Vol.104, p.764-775
• Main Authors: Faber-Langendoen, Don, Lemly, Joanna, Nichols, William, Rocchio, Joe, Walz, Kathleen, Smyth, Regan
• Format: Article
• Language: English
• Subjects: Discriminatory power; Ecological integrity; Metrics; Rapid assessment; Redundancy; Terrestrial; Wetland
• ISSN: 1470-160X; 1872-7034
• DOI: 10.1016/j.ecolind.2019.04.025

•Rapid assessment methods provide an intermediate scale of assessment that is cost-effective, efficient, and ecologically meaningful.•We apply a rapid ecological integrity assessment method across multiple watersheds, and regions in the United States.•Our multi-metric approach evaluates multiple ecological factors, including landscape context (landscape, buffer) and on-site condition (vegetation, hydrology, soil).•Twelve metrics were minimally redundant and have good discriminatory power in response to a gradient of stressors.•Landscape and buffer metrics provide important insights into the on-site condition of wetlands. Many ecological monitoring and assessment programs include rapid assessment methods that employ indicators or metrics to track the degree of divergence of ecosystem condition from reference conditions. Although these rapid assessment methods use a combination of metrics to rate overall ecological condition, they rarely include tests of either the merits of the component metrics being assessed or the method of aggregating the metrics into an overall rating. We used a conceptual model of ecological integrity for wetlands and field data to select and test 15 rapid assessment indicators (using specific metrics) across a spectrum of major ecological factors or MEFs (landscape, buffer, vegetation, hydrology, soil). We applied these metrics to 220 wetland sites across six states (Colorado, Indiana, Michigan, New Hampshire, New Jersey, and Washington), using two assessment area (AA) approaches: 106 sites used 0.5 ha point-based AAs; 114 sites used variable-sized polygon-based AAs. We statistically tested metric ratings and factor scores for their discriminatory power (DP) in relation to a stressor index using the Kruskal-Wallis test, and for redundancy using Spearman rank correlation and scattergrams. Of the 15 metrics, 12 had good or strong DP and were not redundant. Across all metrics, only two pairs (vegetation pair and buffer pair) were strongly correlated. The soil metric had the lowest DP, but it was among the least redundant of any metric. The DP of buffer metrics was lower for point-based approaches than for polygon-based approaches because the buffer for point-based AAs often included additional wetland area. Aggregating individual metrics into MEF scores (e.g., vegetation, hydrology, soil), primary factors scores (Landscape Context, on-site Condition) and overall ecological integrity ratings, either maintained or improved the interpretab