THE VEGETATION OF WET MEADOWS IN RELATION TO THEIR LAND-USE SUSAN M. GALATOWITSCH ∗ , DIANE C. WHITED, RICHARD LEHTINEN, JASON HUSVETH and KAREN SCHIK Departments of Horticultural Science and Landscape Architecture, University of Minnesota, 305 Alderman Hall, 1970 Folwell Avenue, St. Paul, Minnesota, 55108, U.S.A. ( ∗ author for correspondence, e-mail: galat001@maroon.tc.umn.edu) (Received 19 January 1997; accepted 15 September 1998) Abstract. Wetland biomonitoring approaches are needed to determine when changes in response to stressors are occurring and to predict the consequences of proposed land-use changes. These approaches require an understanding of shifts in biota that occur in response to land-use, data that are lacking for most kinds of wetlands. Changes in floristic composition corresponding to land- use differences at multiple scales (site to 2500 m radius) were characterized for 40 wet meadows associated with prairie glacial marshes in Minnesota (U.S.A.). In general, guild was more useful than species composition for indicating land-use impacts. Site impacts (stormwater, cultivation) and landscape disturbance (agriculture and urbanization, combined), coincide with a reduction in native graminoid and herbaceous perennial abundance (e.g., Carex lasiocarpa, Calamagrostis canadensis, Spartina pectinata). This vegetation is replaced with annuals (e.g, Bidens cernua, Polygonum pen- sylvanicum) in recently cultivated sites or introduced perennials (e.g., Phalaris arundinacea, Typha angustifolia) and floating aquatics (lemnids) in stormwater impacted wetlands. Ditches also reduce native perennial importance and increase perennials, but only when they are in highly impacted landscapes. Keywords: impact assessment, land-use impacts, Minnesota, wet meadows, wetland vegetation 1. Introduction Since many wetlands are discontinuous in the landscape, existing as ‘biogeograph- ical islands’, and most are in topographically low, depositional locations, they are prone to impacts from external stressors imposed by surrounding land-uses includ- ing altered hydrology, nutrient and pollutant addition, and barriers to movement (Adamus, 1992; WCMC, 1992; Denny, 1994). Assessing the risks of impairment requires that land-use effects on ecosystem structure and function as well as the potential to reverse degradation or loss be predicted. Most well-documented studies of the consequences of anthropogenic stress on aquatic ecosystems are from severe disruptions; yet, most land-use impacts are low-level or incremental (Schindler, 1987). Research in lacustrine and lotic ecosystems suggests that changes in species assemblages (especially phytoplankton, invertebrates) are more sensitive indicators of stress than are measures of ecosystem function (e.g., nutrient cycling, produc- tivity, decomposition) or chemical and physical composition) (e.g., pollutant levels Environmental Monitoring and Assessment 60: 121–144, 2000. © 2000 Kluwer Academic Publishers. Printed in the Netherlands.