This study was conducted in the Afram Headwaters, Tain Tributaries Block II and Worobong South Forest Reserves. Satellite record of fire incidence for the country over 11 years (1997 to 2007) was modelled via binary logistic regression analysis, and correlations between fire incidence and the correlates of fire used to explain the observed trends. Fire incidence was found to be correlated with multiple variables which probably covary. Rainfall, vegetation type and geology showed the strongest correlations with fire incidence. Recurrent fire has impacted negatively on forest structure, ground cover biomass and species composition in two forest reserves, but more marked in the wetter Worobong South Forest Reserve than the drier Tain II Forest Reserve. Basal area has reduced from 40 m²ha^-1 in least-degraded to <1 m²ha^-1 in heavily-degraded stand in Worobong South Forest Reserve, along with tree density, whereas canopy openness has increased from 6% in least-degraded to 83% in heavily-degraded forest. In Tain II Forest Reserve, however, the heavily-degraded forest has lost close to 50% of its maximum value in terms of basal area, tree density and canopy cover, all in approximately 20 years. Early-burning, with maximum seedling height growth rates of 130 cm yr^-1 and 40 cm yr^-1 for Worobong South and Tain II Forest Reserves respectively, might help control the fires, and restore forest canopy in about 10 to 20 years if regularly maintained, but must be accompanied by green firebreaks. Complete protection from fire (during convalescence), on the other hand, would take between five and 15 years to restore forest canopy, but at huge resource cost. Implications for sustainable forest management are discussed.

Forests play an integral role in regulating the exchange of carbon between the atmosphere and the terrestrial biosphere. These ecosystems only cover for about 30% of the land surface, forests account for almost half of the annual carbon uptake. The amount of carbon sequestered by forest ecosystems is largely dependent upon favorable climate conditions that promote increases in growth. Under the lowest emissions scenario, the United States is projected to undergo an almost 2˚C increase in temperature by the end of the century and it is important that we assess the contemporary climate-growth relationships of multiple forest types to better evaluate the stability and persistence of this vital carbon sink. Tree rings have been used to assess forest response to macroclimate conditions, but often the trees sampled for these analyses are only the most dominant individuals in the forest. This excludes individuals found in the understory of complex forest systems, such as those in the temperate forests of the eastern United States, and climate-growth relationships calculated from only dominant individuals may not be representative of the entire stand. Recent studies have shown that structural complexity of the forest canopy can significantly alter the microclimate conditions at which understory trees grow. Altered growth responses of the understory trees could increase the resilience of complex forest systems to the changes in temperature that are expected by the end of the century. Here, I apply dendrochronology methods to quantify climate-growth relationships of canopy strata in temperate forests of the eastern United States. Many different forest types are found in this region and have been the focus of numerous species-specific studies on climate growth relationships. However, the integrated response of co-occurring species within canopy units is not often investigated, despite measures of productivity being an integration of ecosystem processes. I present research that investigates the differential climate sensitivities of canopy strata, and I present a means to more accurately represent biomass estimates calculated from tree-ring data. The first study quantifies the climate sensitivities of different canopy strata from five temperate forests in the eastern US. We used a generalized additive model (GAM) to assess the influence that growing season mean temperature, growing season precipitation, and tree size have on dominant (uppermost), intermediate, and understory (lowermost) canopy strata. We found that differential climate sensitivities do exist between canopy strata, causing each canopy class to respond to extreme climate conditions in a different manner. For example, during the hottest and wettest years dominant and intermediate trees show slight increases in growth, whereas understory trees show significant decreases in growth. These results suggest that the climate and competitive environments created within stratified canopy layers may provide an added degree of ecosystem stability in the face of changing climate conditions. The second study assesses the spatial coherence of climate-growth relationships between canopy layers from the eastern temperate forest region. We collected increment cores from sites in Missouri, Indiana, Ohio, Michigan, Massachusetts, and Maine and we found that site groupings were relatively consistent between canopy layers. Dominant and intermediate trees showed a strong correlation with temperature that also coincides with the forest types and species distributions that are observed across the region. However, understory trees show stronger relationships with precipitation. Sites from the northeast US and Michigan displayed muted climate relationships, likely due to having both coniferous and hardwood species present. The midwest sites, composed of mostly hardwood species, showed relatively strong, negative temperature relationships in the dominant and intermediate canopy layers, but understory trees displayed strong positive relationships with temperature. These results suggest that although macroclimate conditions influence species distributions and affect the dominant trees, understory trees are likely responding to microclimate conditions. This also suggests that regions with increased functional diversity and complex canopy structure may be better buffered against changing climate conditions. Finally, we identify four main sources of uncertainty in estimating aboveground biomass from tree-ring data. Tree rings are being used more frequently to estimate the annual uptake of biomass by forested ecosystems. However, these calculations require several steps and assumptions that affect the overall accuracy of the biomass estimates. The error range around tree-ring estimates of aboveground biomass is seldom reported. We illustrate how increment upscaling, allometric, stand density, and mortality uncertainties can affect biomass estimates from a well-studied site in the Valles Caldera in northern New Mexico. We found that dominant sources of uncertainty change depending upon whether cumulative or incremental biomass is calculated. At the cumulative level, choice of allometric equation and tree mortality estimates dominate the uncertainty, whereas inter-annual variability in the tree-ring record dominates incremental biomass estimates. Despite the calculations that are required to translate linear ring-width measurements into biomass quantities, the underlying climate-growth relationships recorded within the tree rings are not significantly altered. Tree-rings provide a means for non-destructively quantifying the aboveground biomass in a forest and reporting the accompanying uncertainties will facilitate more accurate comparisons between disparate forest types.

Thesis (Ph. D.)--University of Washington, 2004. / Vita. Includes bibliographical references (leaves 165-194).

Thesis (M.S.)--Oregon State University, 2008. / Printout. Includes bibliographical references (leaves 45-51). Also available on the World Wide Web.

Thesis (M. Sc.)--Carleton University, 2003. / Includes bibliographical references (p. 28-34). Also available in electronic format on the Internet.

Thesis (Ph. D.)--University of Tennessee, Knoxville, 2007. / Title from title page screen (viewed on June 10, 2008). Thesis advisor: Henri D. Grissino-Mayer. Vita. Includes bibliographical references.

Thesis (Ph. D.)--University of Tennessee, Knoxville, 2004. / Title from title page screen (viewed Sept. 20, 2004). Thesis advisor: Wayne K. Clatterbuck. Document formatted into pages (xiii, 174 p. : ill.). Vita. Includes bibliographical references (p. 144-152).

Thesis (Ph.D.)--Aberdeen University, 2008. / Title from web page (viewed on July 20, 2009). Includes bibliographical references.

Thesis (M. Sc.)--University of Canterbury, 2009. / Typescript (photocopy). "May 2009." Includes bibliographical references (leaves 98-115). Also available via the World Wide Web.

Thesis (M.S.)--Oregon State University, 2006. / Printout. Includes bibliographical references (leaves 127-136). Also available on the World Wide Web.