BACKGROUND

In this study, Rooney and colleagues (2000) examined whether deer browse might be a factor affecting the regeneration of eastern hemlock (Tsuga canadensis). A long-lived, shade-tolerant conifer occurring on moist, acidic soils, eastern hemlock once dominated much of the forest in upper Wisconsin and Michigan. Heavily cut over by the 1920s, hemlock-northern hardwood forests were replaced by second-growth hardwoods and have yet to rebound. Today, hemlock remains a common community element in the few remaining primary forests and has partially rebounded in some lowland and riparian areas. Hemlock remains uncommon, however, in previously logged, upland stands and occupies only 0.5% of the upland landscape in the northern Great Lakes region.

Restoration of hemlock-dominated landscape elements is being considered by federal land management agencies but significant barriers exist. Even in forest stands where hemlock does compose a substantial portion of the canopy, hemlock seedlings and saplings are conspicuously absent. A variety of hypotheses could explain this mystery. Seedling establishment may be limited by the availability of appropriate microsite conditions, such as moss beds, nurse logs, or bare mineral soil. Or, moisture availability, soil characteristics, or other habitat characteristics may limit hemlock seedlings. Because they grow slowly and provide winter browse, hemlock seedlings are also sensitive to herbivory by white-tailed deer, as demonstrated in several other studies.

To help solve this mystery, Rooney and colleagues studied 100 hemlock stands in northern Wisconsin and western upper Michigan. The sites included land in county, state, and national forests, national lakeshore, Indian reservations, and private ownership. These entities managed their forests for different goals and took different approaches to deer management. The stands selected varied in hemlock density. The researchers sought to understand factors responsible for this variation. In other words, what factors explained why hemlock were regenerating in some sites and not in others?

At each study site, researchers counted the number of hemlocks in each of four size classes (seedlings, small saplings, medium saplings, large saplings) and collected data on several factors suspected behind poor hemlock regeneration. These were seed input, leaf litter type, leaf litter depth, light availability, habitat type, and deer browsing intensity. They then conducted statistical analyses to assess which of these factors contributed to variation in hemlock density between sites.