Modeling Early Regeneration Processes in Mixed Species Forests of Maine: Germinant Emergence and Early Survival
- Wilson, Geoffrey F.
The Graduate School, University of Maine
Successful regeneration of a tree species is a function of both the abiotic environment and the surrounding biota. These environment components, in turn, influence the regeneration microenvironment both indirectly through physical modification and directly through competitive effects. In the coming century, the climate is expected to change at an unusually rapid rate, yet the implications of this change on forest composition are at present hypothetical. Most of Maine's forests are periodically harvested, and natural regeneration, as opposed to ditect seeding or planting, is by far the most common reforestation method. Understanding early regeneration processes, including both abiotic and biotic influences on germination and early survival, is crucial for anticipating the future composition of Maine's forests. To date early regeneration processes in these forests have received little research attention. The objective of this thesis is to investigate the relationship between germinant dynamics and the biotic environment, including the tree overstory, understory, herb layer, and forest floor.
Influences of the surrounding vegetation and substate on the density of newly emerged germinants in a variety of silviculturally manipulated stands were analyzed using Poisson regression. Species modeled included Abies balsamea, Tsuga canadensis, Acer rubrum, Betula sp., Picea sp., and Thuja occidentalis. Results indicated dominant influences of the canopy on all of the species; in many cases the models showed multiple canopy influences. The influences of the herb layer vegetation and substrate availability were minor compared with the roles of the canopy. Most germinants had emerged by mid-July in 1996.
One and two season survival of germinates was also analyzed in relation to pre-existing vegetation and growing substrate. Analysis was performed using binomial regression. As with the germinate density models, survival models also indicated a dominant influence of the canopy, and again the canopy showed multiple effects. The substrate on which a germinate was growing was significant in describing its survival in three of the species. This effect may vary from year to year, possibly due to interaction with climatic factors.
Description of the overstory was conducted on two different spatial scales in order to detemrine which was more appropriate for characterizing the regeneration microenvironment. A 200-m2 plot centered on each 4-m2 regen plot was superior to a O.08-hectare plot for characterizing a particular microsite, yet the latter was adequate for characterizing gross overstory conditions influencing the germinant density and survival in the regen plots.
Use of Poisson and binomial regression was successful in elucidating some effects the surrounding biota exert on early regeneration processes over the short-term. However, long-term monitoring of these processes will be necessary before a predictive ability will emerge, particularly given the extreme annual variation in biotic processes such as seed production and weather events such as drought and late frosts.
You must be logged in to post a comment.