The sustainability of the northern hardwood forest is threatened by disturbances and perturbations including chronic air pollution, invasive pests, and rare catastrophic events. Hubbard Brook Experimental Forest (HBEF) in New Hampshire is a prime example of a forest experiencing multiple environmental stressors with serious but subtle impacts. Long-term studies at HBEF have documented a decline in forest biomass accumulation that seems to be the result of reduced growth of the dominant tree species: sugar maple (Acer saccharum Marsh.) and beech (Fagus grandifolia Ehrh.). Previous studies have linked soil calcium depletion resulting from acidic deposition to reduced health in sugar maple. In addition, the spread of the exotic scale insect, Cryptococcus fagisuga Lind., that produces bark cankering has reduced growth rates in beech trees. Each of these perturbations may influence species competitive hierarchies. My dissertation focuses on teasing apart these complex competitive interactions in relation to tree growth and examines the role of chronic acid deposition and an ice storm on demographic processes that affect forest productivity. In addition, it synthesizes best practices for evaluating changes in forest dynamics through tagged-tree inventories.

In Chapter 1, I conducted a neighborhood analysis to examine the nature of competition in influencing tree growth and evaluate the dominance of species under current perturbations. A dominant competitor can be defined either in terms of its ability to suppress other individuals (competitive effect) or its ability to avoid being suppressed (competitive response). Using neighborhood models, I quantified the species-specific competitive effects and responses to determine the competitive hierarchy of species in their respective communities (northern hardwood, hemlock, and fir-birch). I predicted late-successional tree species to be at the top of the competitive hierarchy. I used spatially-explicit demography plots, growth rings as a measure of tree growth, and likelihood methods to parameterize and compare a variety of growth models with various neighborhood sizes. These models made different assumptions about the effect of competing neighbors: no competitive effects (i.e., the null model), species-equivalent competitive effects, and species-specific competitive effects. The results demonstrate that the radii of tree neighborhoods varied from 4 to 22 m depending on the target species and community type. Competition was important in these forests and that species-composition of target tree's neighborhood clearly influenced growth in addition to simple crowding. Results from the neighborhood analysis suggest there is no evidence of competitive dominance of late-successional species sugar maple and beech but for different reasons. In the northern hardwood community, sugar maple was very sensitive to competition while beech had low predicted growth rates. In the hemlock community, both species had lower growth rates than its competitors. In contrast, mid-successional shade-intolerant yellow birch exhibited strong competitive effects on its neighbors and had only an intermediate level of sensitivity to competition. In the high elevation fir-birch community, red spruce growth experienced the greatest decline in growth from competition from neighbors, but the absolute values were still higher than for competing species. The success of mid-successional shade-intolerant species and the lack of dominance in the late-successional species suggest that the competitive hierarchy expected at HBEF has been reversed, possibly due to the impact of calcium depletion and beech bark disease.

In Chapter 2, I examined forest dynamics of HBEF in the context of the hierarchical response framework (HRF). HRF identifies a hierarchy of mechanisms leading to ecological change as ecosystems are exposed to press disturbance: individual-level responses, species-reordering, and species immigration/loss within the ecosystem. An experimental calcium amendment to watershed 1 (W1), compared to reference watershed 6 (W6), provided the opportunity to study tree response to chronic acid deposition and potential interactions with a pulse disturbance, an ice storm that occurred in 1998. First, I used a comparative demographic analysis on the paired watershed to quantify the initial individual-level response to chronic acid deposition. Next, I tracked the recovery of the forest after the ice storm to assess whether species-reordering has subsequently occurred in response to exposure to chronic acid deposition. Results from the study provided support for the first two stages of the HRF but there was no evidence of species immigration or loss at HBEF. Evidence for individual-level responses included increased sugar maple growth in response to the Ca amendment. In W6, increased beech growth and recruitment resulted in watershed-level biomass gains. The pattern suggests that strong interspecific competition between sugar maple and beech is driving forest response to disturbance, resulting in species reorganization. Moreover, the increase in beech dominance observed under chronic acid deposition was accelerated by the pulse disturbance. The implications of these results in the context of HRF are that increased pulse disturbances acting on the backdrop of a press disturbance will continue leading to species-reordering, resulting in a more heavily beech dominated forest and potential loss of sugar maple from the northern hardwood forest.

Results reported in Chapters 1 and 2 relied on measuring the demographic responses of individual trees. These tagged-tree inventories not only provide a baseline for community composition, but also insight into species-specific changes through time. In Chapter 3, I identified and summarized potential sources of error at each step of a basic tagged-tree inventory. These steps include keeping track of trees, measuring tree diameters, identifying trees to species, and determining tree vigor. Each source of error affects various demographic components and in different directions and magnitudes. I presented best practices from experiences at HBEF that will help minimize the occurrence of these errors. Recommendations for managing successful long-term tagged-tree inventory monitoring include: (1) Develop robust field procedures with solid initial training and season-long check; (2) Use available data to measure extent of error, to inform field procedures, and to make corrections to the database; (3) Consistently review practices to detect challenges and improve efficiency.