Poor Recruitment is Changing the Structure and Species Composition of an old-growth Hemlock-hardwood forest

Abstract

Anthropogenic factors such as elevated deer populations, invasive earthworms, or climate change may alter old-growth forests of the Upper Midwest region of the United States. We examined demographic trends of woody species across all size classes for a period of 35 years in a late-successional forest dominated by hemlock (Tsuga canadensis), sugar maple (Acer saccharum), and yellow birch (Betula alleghaniensis) in Michigan's Upper Peninsula using two sets of permanent plots. For the duration of the study period, species that were less-preferred white-tailed deer (Odocoileus virginianus) forage, especially sugar maple, comprised a much higher fraction of all seedlings and saplings compared to their fraction of overstory trees. The density of small sugar maple declined across the study period, but no other species became more abundant, creating a more open forest understory. By the most recent census, preferred species for deer browse had been nearly eliminated from the understory, and declines in unpreferred species such as sugar maple were also apparent. We found small changes in temperature (<0.5-1°C rise in minimum and maximum temperatures depending on season) and precipitation (±28 mm depending on season) and little evidence of invasive earthworms impacts. Our results suggest that a sustained elevated deer density is shifting the structure and composition of this old-growth forest. A demographic model showed that if current recruitment, growth, and mortality rates were to continue for 500 years the forest would reach a new equilibrium with virtually no hemlock or yellow birch remaining.Tree coring, or increment boring, has been a common research tool for foresters, ecologists, and climatologists for over a century. Despite its widespread use, there has been very little research into the effect of this practice on the growth and mortality of trees. Using data from two of the permanent plots, we compared the growth and mortality rates of cored trees to a similar set of uncored trees for 16-18 years. While there might have been some slight bias in selecting trees for coring, it is unlikely to have affected our overall results. Cored trees did not differ in their mortality rate from uncored trees and had only minute differences in growth, either when considered collectively or when looking at species individually.