Constructing multimetric indices and testing ability of landscape metrics to assess condition of freshwater wetlands in the Northeastern US. (July 2016)

• Record Type: Journal Article
• Authors: Miller, Kathryn M.; Mitchell, Brian R.; McGill, Brian J.
• Is Part Of: Ecological indicators. Volume 66(2016)
• Page Start: 143
• Page End: 152
• Publication Date: 2016-07
• Subjects: Multimetric indices -- Freshwater wetland condition -- Landscape context -- Wetland ecological integrity -- National Wetland Condition Assessment -- National Park Service Inventory and Monitoring Program; Environmental monitoring -- Periodicals; Environmental management -- Periodicals; Environmental impact analysis -- Periodicals; Environmental risk assessment -- Periodicals; Sustainable development -- Periodicals; 333.71405
• DOI: 10.1016/j.ecolind.2016.01.017
• Languages: English
• ISSNs: 1470-160X

Abstract

Highlights: We built multimetric indices (MMIs) of wetland condition for vegetation, soil, water and algae. We used data from the Environmental Protection Agency's National Wetland Condition Assessment. Vegetation and soil were the best performing MMIs, and included many commonly cited indicators. Adjacent land use and vegetation were strong predictors of soil, water, and algae MMIs. The wetland MMIs we constructed are applicable to a range of wetland types covering 11 eastern US states. Abstract: Using data collected for the Environmental Protection Agency's (EPA) 2011 National Wetland Condition Assessment (NWCA), we developed separate multimetric indices (MMIs) for vegetation, soil, algae taxa, and water to assess condition of freshwater wetlands in the northeastern US. This study represents the first attempt at developing multiple biotic and abiotic MMIs of wetland condition over this large of an area, and is only possible because of the high quality data collected by the NWCA. We chose metrics that distinguished between reference and most disturbed sites, had a signal:noise ratio > 2, and were not strongly correlated with other metrics, latitude, or longitude. The vegetation and soil MMIs were the best performing indices, with good separation between reference and most disturbed sites, and included commonly used condition metrics (e.g., pH and P concentration for soil, and percent cover of exotic species for vegetation). The algae MMI was the weakest index, with considerable overlap between reference and most disturbed sites. For areas smaller than our study, algae taxa may be suitable for wetland MMIs. However, in our study area, many algae taxa followed strong latitudinal or longitudinal gradients, and could not be considered for the algae MMI. Small sample size and several metrics with a high signal:noise ratio were the major limitations of the water MMI. We also examined how well landscape (level 1) and rapid assessment (level 2) metrics predicted MMIs using random forest regression. Agricultural land use surrounding wetlands was an important predictor for all four MMIs, although the soil, algae and water MMI models performed best when intensive (level 3) vegetation metrics were also included in the random forest regression models. Based on these results, we recommend wetland assessment programs employ a combination of landscape and rapid assessment monitoring at many sites, along with level 3 monitoring at a subset of sites. We developed these MMIs to evaluate freshwater wetland condition for a long-term monitoring program in Acadia National Park. These MMIs are also applicable to a range of wetland types covering 11 states in the northeastern United States and can be calculated using a downloadable spreadsheet that calculates and rates each MMI using raw metric values.