Abstract
Forest development following major disturbance is thought to follow a fairly repeatable temporal pattern. An initial cohort of trees establishes relatively rapidly (stand initiation), new establishment is precluded for an extended period (stem exclusion), and finally, new individuals again begin to establish, creating new age-classes in the forest understory (understory reinitiation), eventually leading to an uneven-aged condition. The current study was designed to assess the generality of this developmental pattern at the landscape level and gain insight into the possible mechanisms controlling stem exclusion and understory reinitiation in even-aged forests. Research was conducted within two bigtooth aspen (Populusgrandidentata Michx.) dominated landscapes in northern Lower Michigan having similar physical site characteristics, overstory compositions, and disturbance histories. The objectives for the study included (i) assessing the repeatability of development patterns within and between the two landscapes and (ii) exploring relationships between the timing of understory reinitiation and overstory growth characteristics and seed availability. Stem analysis was used to reconstruct establishment and growth histories of surviving stems in mature forest on replicate plots within each landscape. The age distributions of surviving individuals of all tree species in both forests reflected a developmental pattern characterized by rapid initial cohort establishment lasting 5–15 years, stem exclusion lasting 25–35 years, and understory reinitiation. The duration and timing of the developmental stages on the replicate plots were similar both within and between the two landscapes. There were, however, a small number of plots in landscape 2 that had substantially reduced stem exclusion lengths, relative to the remaining plots in both forests. Variation in the timing of understory reinitiation in landscape 2 was related to characteristics of remnant eastern white pine (Pinusstrobus L.) seed trees. The stem exclusion period was shorter, or almost nonexistent, on plots close to several large seed trees because of early, low-frequency establishment of white pine in the understory. Substantial increases in the frequency of understory establishment on all plots in landscape 2, as well as all new understory establishment on plots in landscape 1, were often associated with radial growth increases in overstory stems. The radial growth increases presumably reflected an increase in resource availability, possibly occurring in response to a concentrated wave of natural thinning within the bigtooth aspen dominated overstory. These results suggests that the timing of understory reinitiation may have been influenced by variation in seed availability interacting with changes in resource availability in the forest understory. A limited amount of new establishment occurred relatively early in forest development on plots that experienced abundant early seed rain, yet the bulk of new establishment occurred only after an increase in resource availability in the understory. The general patterns of forest development described in this study were similar both within and between landscapes. Similar descriptive studies are needed to address the repeatability of development patterns at the landscape level within other forest types. Additionally, experimental studies are needed to unequivocally identify the mechanisms controlling stem exclusion and understory reinitiation in even-aged forests.