Wetland-nonwetland transition zones in the Missouri River Basin of the north-central United States were studied in seven wetlands ranging in size, permanence, and salinity to develop an efficient sampling methodology that utilized vegetation data to delineate wetland boundaries. One hundred and sixty species of vascular plants occurred in twenty sample transects. Sampling methodology suggested for use in the study area (200-km radius of Sioux Falls, South Dakota) is a combination of the belt transect method (contiquous quadrats) to estimate cover by species and a cover board to measure vertical structure. The methods require a combined sampling time of about 9 min/m of transect, corresponding to a sampling time of 3.2 hr for a transect of average length. A streamlined methodology was also devised whereby sampling time could be cut in half. General upper and lower borders of the transition zone were determined from direct gradient analysis graphs. Specific borders were determined from the occurrence of compositional dichotomies displayed in ordination models.
Available at: http://works.bepress.com/terry_sharik/67/