PS 51-61 - Ecological drivers of plant community composition in the Seney Wilderness Area, Upper Michigan

Thursday, August 10, 2017
Exhibit Hall, Oregon Convention Center


Alex J. Graeff, Northern Michigan University; Alan J. Rebertus, Northern Michigan University; R. Gregory Corace III, U.S. Fish and Wildlife Service; Matthew J. Van Grinsven, Northern Michigan Univeristy


Patterned fens are peatlands characterized by alternating hummocks (strings) and hollows (flarks), but often contain other landforms. Several studies of patterned fens have described vegetation and landform patterns in relation to water chemistry. However, ecoregional differences in geomorphology, vegetation, landforms, and water chemistry give way to a variety of unique peatlands. This variation necessitates studies in a variety of physiogeographic settings. The Seney Wilderness Area’s (SWA) peatland is the second-southernmost patterned fen in North America and contains unique patterning associated with extinct dunes, yet has received little attention. To our knowledge, no study has previously co-examined geomorphology, vegetation, and water chemistry among landform types in the SWA. The objectives of this study were to (1) characterize plant community composition of fen landforms and (2) examine the relationships between plant community composition and environmental variables. We estimated percent cover of vascular plant species in 298 relevé-style plots across six fen landform types: peripheral water tracks, featureless water tracks, swamp forests, strings, flarks, and ponded sedge lawns. At each plot, we measured peat depth and depth to water table (DWT) in addition to pH, specific conductance (μS/cm), and Ca2+ (mg/L) for both surface and subsurface water samples.


Non-metric multidimensional scaling (NMS) isolated four significantly distinct vegetation groupings (P<0.01) including: (1) peripheral water tracks, (2) swamp forests, (3) ponded sedge lawns and flarks, and (4) featureless water tracks and strings. Significant differences (P<0.05) were detected among landforms for peat depth, DWT, surface pH, subsurface pH, surface conductivity, subsurface conductivity, surface Ca2+, and subsurface Ca2+ variables. Canonical correspondence analysis (CCA) identified DWT, peat depth, surface conductivity, subsurface Ca2+, and subsurface pH as the most important variables (r2>0.4) related to community composition. Variation in water chemistry and physical characteristics among fen landforms types was strongly related to plant community composition. This study reveals that plant communities, and their physical and chemical properties, differ not only across fen landform types, but also within landform types across the landscape. Further study may reveal other forces, such as spatial characteristics, driving plant community variation in the large fen landscape.