Date of Award

Winter 2006

Project Type


Program or Major

Natural Resources and Earth Systems Sciences

Degree Name

Doctor of Philosophy

First Advisor

George C Hurtt


Forests are structurally diverse, but these structures derive from the same processes of disturbance and growth. Understanding forest structure can help unlock the history, function, and future of a forested ecosystem. Components of forest structure include tree size distributions, foliage distributions and variation in canopy density, and coarse woody debris (coarse necromass). Tropical rainforests are structurally the most complex of all ecosystems. In addition to having high biological diversity, Amazon forests are marked by complex vegetation dynamics and a diverse forest stand structures, which play an important role in the interactions of water and carbon between the biosphere and atmosphere. Two aspects of forest structure in Amazonia are examined in this thesis, canopy geometry and coarse necromass. A crown delineation algorithm was developed that uses high resolution satellite image data. This algorithm was applied to two forests with field based data on forest structure and then applied at seven locations across the Amazon basin. The algorithm provided forest structure based on crown geometry across vast areas the where field based studies would be prohibitive due to cost and time. Coarse necromass dynamics were studied through a combination of field work using novel techniques that measured necromass density, volume, and calculated mass for fallen and standing coarse necromass stocks at two tropical forested sites. For both sites, the effect that reduced impact logging (RIL) had on coarse necromass pools was found to generate 50% more coarse necromass. Density and void space estimates were found to not be significantly different between sites. Standing dead and fallen coarse necromass were found to be proportionally related across sites and forest types. The production of necromass at one site, Tapajos, was examined over a 4.5 year period, providing an estimate of necromass cycling. RIL and undisturbed forests were found to similar coarse necromass production. Mortality rates used to estimate coarse necromass production tend to underestimate the amount by about 55%. Finally a review of current literature dealing with coarse necromass dynamics in tropical forests was conducted.