We summarize three existing conceptual frameworks for forest regeneration in northeastern North America, and suggest that none consider both a range of disturbance characteristics and a range of forest conditions at the time of disturbance. We offer a more general conceptual model, within which the existing models can be seen as special cases. We propose that the abundance of characteristic seed-bank, pioneer species, such as Prunus pensylvanica L.f. and Rubus spp. (often allegheniensis, hispidus, or odoratus), is dependent on propagule availability, which in turn is determined by forest age and size. Specifically, following disturbance, large tracts of forest and older forests (> ca. 125 yr) are predicted to have very low densities of the above pioneers.
As a result, population, community, and ecosystem parameters of the regenerating forest may be substantially different than in the familiar cases of regeneration in secondary forests. Indeed, the presettlement forest of much of northeastern North America may have experienced a notable scarcity of pioneers after disturbances, in areas far enough inland for hurricanes to be unimportant.
Our hypothesis makes predictions of seed bank abundance that are well supported in a variety of forest types; we also provide support for our hypothesis with data on regeneration following catastrophic windthrow in Allegheny National Forest, Pennsylvania, and elsewhere. Finally, our hypothesis also predicts that the potential regeneration in much of the secondary forest of northeastern North America should profoundly shift as stands age from roughly 100 to 130 years.