95th ESA Annual Meeting (August 1 -- 6, 2010)

COS 82-7 - Persistence of early-successional tree species in a late-successional hemlock-hardwood forest landscape

Thursday, August 5, 2010: 10:10 AM
336, David L Lawrence Convention Center
Robert T. Fahey, The Morton Arboretum, Lisle, IL, Craig G. Lorimer, Department of Forest and Wildlife Ecology, University of Wisconsin-Madison, Madison, WI and David J. Mladenoff, Department of Forest & Wildlife Ecology, University of Wisconsin-Madison, Madison, WI
Background/Question/Methods

Establishment of early-successional species by catastrophic disturbances can help maintain landscape-level species diversity.  However, early-successional species may lead a tenuous existence in landscapes where such disturbances are infrequent relative to the species’ life-span.  In such landscapes, early-successional trees may persist as fugitives or may rely on marginal habitats (such as rock outcrops, inclusions of sandy soil, and lake, wetland, and river edges) as refuges to provide a persistent seed source. The primary objective of this study was to investigate the fine-scale distribution of early-successional tree species in relation to habitat features and environmental variables in the pre-settlement hemlock-northern hardwood forest landscape of northern Wisconsin. We also aimed to relate the distribution of early-successional tree species to possible modes of persistence for these species in this late-successional landscape.  We used witness tree data from U.S. Public Land Survey records and modern GIS data on environmental variables to determine habitat and species associations of the primary early-successional species found in the Great Lakes region hemlock-hardwood forest: aspen (Populus spp.), paper birch (Betula papyrifera), and jack, red and white pine (Pinus banksiana, resinosa, and strobus).  Of particular interest was white pine, which is a restoration priority in the region.

Results/Conclusions

The primary early-successional species accounted for ~12% of the total witness tree sample. Each early-successional species was more common in potential refuge habitats (which made up ~44% of the landscape) than the overall landscape.  In addition to refuge habitats, the occurrence of white pine as a witness tree was strongly related to proximity to large sandy outwash plains and broad-scale variation in climate.  White pine also occurred more often in upland mesic habitats with loamy soils than would be expected based on the modern landscape.  Across all habitats, there was a strong association between white pine occurrence and that of other fire-adapted species.  Early-successional species made up a larger proportion of the witness tree sample than in similar forests in northeastern North America, which could be explained by the abundance of potential refuges in this landscape (especially the high density of lake and wetland edges).  Our results illustrate the potential importance of these refuges in mesic landscapes, but also show that early-successional species frequently occurred outside these refuges in upland mesic habitats.  The finding that white pine occurred across a wide habitat range in the pre-settlement forest supports efforts to restore this species to mesic hemlock-hardwood forests.