The following paper is the result of making a study of the vegetation of a selected region along Lake Erie at Turkey Point, by observations, field notes and collections of botanical specimens together with information and records acquired by examination of provincial institutional and private herbaria and by a survey of the literature. The field work was carried out during the summer of 1938. / Thesis / Master of Arts (MA)

The following paper is the result of making a study of the vegetation of a selected region along Lake Erie at Turkey Point, by observations, field notes and collections of botanical specimens together with information and records acquired by examination of provincial institutional and private herbaria and by a survey of the literature. The field work was carried out during the summer of 1938. / Thesis / Master of Arts (MA)

Climate change and extreme weather events have impacted global forest ecosystems’ ability to sequester atmospheric carbon dioxide. In this study, the temporal and spatial dynamics of soil CO2 efflux or soil respiration (Rs) was measured in a temperate coniferous (TP74) and a deciduous forest (TPD) over a six-year period (2014 to 2019). Analysis of Rs trends showed a strong positive correlation with soil temperature (Ts) and soil moisture (SM) at TPD and TP74 causing large pulses of Rs. The average annual temperature sensitivity (Q10) was found to be 2.06 for TPD and 1.87 for TP74. Coherence analysis for both sites from 2017 to 2019 showed that in extreme weather events, TP74’s carbon pool was less stable than that of TPD. Dynamics of Rs at both forest sites was further analyzed using thirteen different Rs models (e.g. Ts only, SM only, Ts and SM models, neural network) to evaluate their performance in simulating observed patterns of soil CO2 effluxes. As compared to other models, the Gaussian – Gamma model consistently reproduced observed dynamics of Rs where on average 70% of variability in Rs was explained. This study showed that Ts and SM are key determinants of Rs in both forests. Models that incorporate the influence of SM on Rs and were able to better simulate Rs dynamics as compared to Ts only models. Results also suggest that coherence analysis can be utilized to understand temporal variations in Rs. The knowledge of environmental drivers of Rs can be used to determine the impact of climate change and extreme weather events on Rs and assist in developing ecosystem models. / Thesis / Master of Science (MSc)

<p>Modern advances in biometeorological monitoring technology have improved the capacity for measuring ecosystem exchanges of mass, energy and scalars (such as CO₂). Translating these measurements into robust and accurate scientific information (and ultimately, understanding) requires careful assessment of operations throughout the biometeorological data life cycle. In response, this research analyzed and optimized aspects of data collection, management and filtering for an ecosystem exchange measurement program over an age-sequence of temperate white pine forests.</p> <p>A comprehensive data workflow and management system (DWMS) was developed and implemented to support the entire data life cycle for all past, present and future measurement operations in our research group, and meet the needs of a collaborative, student-led data management environment. Best practices for biometeorological data management were introduced and used as standards to assess system performance.</p> <p>Roving eddy covariance (rEC) systems were examined as a means of producing reliable time-integrated carbon exchange estimates at multiple sites, by rotating an EC system in a resource-mindful approach. When used with an optimal gap-filling model and rEC rotation schedule (2 sites with 15-day rotations), the results suggested its viability, as annual NEE estimate uncertainties ranged between 35 and 63% of the annual NEE flux magnitude at our study sites – even though approximately 70% of half-hours were filled.</p> <p>Lastly, a data-driven approach was used to investigate the effects of different friction velocity and footprint filtering methods on time-integrated carbon exchange estimates at our fetch-limited forests. Though predicted flux source areas varied considerably between footprint models, our objective analyses identified the model (Kljun et al., 2004) and within-fetch requirement (80%) that optimized reliability and representativeness of carbon exchange estimates. Applying this footprint model decreased annual NEE by 31 to 129% (59 to 207 g C m^-2 y^-1) relative to no footprint application, and highlighted the importance of objective analyses of EC flux filtering methods.</p> / Doctor of Philosophy (PhD)