From Leaves to Ecosystems: Resolving the Responses of Evergreen Forests to Seasonality

Abstract

To predict and mitigate the effects of global climate change requires a more comprehensive understanding of the processes that control the exchange of carbon, water, and energy between the biosphere and the atmosphere. Examining how plant metabolic processes respond to seasonal variation in climatic variables can help parametrize and validate model representation of controls on carbon exchange in vegetation models. Plant responses to the seasonality of climatic variables include not only instantaneous responses, but also changes in plant status due to phenology (the field of science concerned with cyclic and seasonal natural phenomena in plants). In this dissertation I first broadly explore whether we are adequately measuring, and successfully modelling, the effects of phenology on carbon allocation and acquisition in plants. Then I focus on the role of phenology and seasonality in driving two distinct evergreen forest systems: a broadleaf tropical forest in the Brazilian Amazon, and subalpine coniferous forest in the Rocky Mountains. In a the tropical forest I test the hypothesis that leaf phenology drives an increase in gross ecosystem production during the dry season that has been observed at the ecosystem scale (from eddy covariance). I test this hypothesis at the individual tree scale by investigating age-dependent leaf physiology and seasonal shifts in leaf demography for five broadleaf canopy trees representing abundant species in the Tapajós National Forest (near Santarém, Brazil). The results of this study suggest that aggregated whole-branch Vcmax (the maximum carboxylation rate of Rubisco, Vcmax, weighted by branch leaf demography) increases during the dry season, with a magnitude consistent with increases in ecosystem-scale photosynthetic capacity observed from flux towers. In the subalpine coniferous forest, I examined the seasonal climate determinants of net ecosystem production and evapotranspiration by analyzing a 15-year eddy covariance time-series from a subalpine forest using an ensemble of artificial neural networks at the daytime/nighttime time-step. This study shows the relative rankings of climate drivers and driver-response relationships directly from the eddy covariance dataset with minimal a priori assumptions, and found that air temperature emerged as the determinant of a seasonal 'switch' in net ecosystem productivity from winter to spring. This dissertation advances our knowledge of the complex interplay between phenological responses and direct climate responses of evergreen forests from the scale of leaves to the scale of ecosystems.