Congratulations to Molly Ng who defended her dissertation on Friday, June5, 2020

Advisor: Selena Smith

Plant response to climate is valuable in understanding many aspects of evolution and ecology, including climate and biogeochemical processes. The close relationship between plants and climate has been extensively studied in eudicot angiosperms (woody flowering plants), showing links between climate and leaf shape. However angiosperms only represent a small portion of plant life on land, ~125 of 425 Ma (millions of years). In contrast, non-angiosperm approaches tend to be qualitative or semi-quantitative, limiting implications that could potentially be derived. Exploring other methods, like plant anatomy, which has been shown to reflect their growing environment, has potential to expand our knowledge of non-angiosperm response to climate. 

In this dissertation, I focus on conifers within Cupressaceae because of their diverse morphology and habitats, and extensive fossil record. I investigate anatomical links to climate across gradients in the modern and apply these data to answer questions about their evolution and ecology. I focus on Metasequoia glyptostroboides Hu and Cheng (Cupressaceae), the dawn redwood, a deciduous conifer that naturally inhabits a small valley in central China.  Metasequoia has an extensive fossil record (~100 Ma) across the Northern Hemisphere from mid- to high-latitudes. I collected leaves from  Metasequoia glyptostroboides and their close relatives, Sequoia sempervirens and Taxodium distichum, across their natural and cultivated ranges across North America, Asia, and Europe. I made leaf cross-sections, and measured the following traits: cross sectional area, vascular bundle area, resin canal area, leaf width and thickness.

In Chapter 2, I focused on how modern leaf anatomy within M. glyptostroboides relates to climate across a gradient. I use Canonical Correlation Analysis (CCA), a multivariate ordination method, to link anatomy and 19 Bioclim climate variables. I found leaf width and cross-sectional area were associated with cold-season precipitation, vascular bundle area with warm-season precipitation, leaf thickness with mild cold-season temperatures and mean annual temperature, and resin canal area with daily temperature fluctuations and mild cold-season temperatures. To test how this relationship held across time, I applied it to the fossil record by measuring the same anatomical traits in Metasequoia milleri, an anatomically preserved fossil from the Princeton Chert Locality (Allenby Formation, BC, Canada) from the early Eocene (~55 Ma). These climate estimates were compared to previous reconstructed climates of nearby fossil localities. My estimates were within these previously independently derived estimates of climate, demonstrating how anatomy links to climate and that these relationships held over geologic time. 

Whether leaf anatomical links to climate I found in Chapter 2 holds across other conifers was investigated in Chapter 3. I included closely related taxa S. sempervirens and T. distichum and added physiological proxies related to photosynthesis (carbon/nitrogen content; δ13C to calculate photosynthetic discrimination (Δleaf), water use efficiency (WUE), and internal cellular:atmospheric pCO2 (ci/ca)). These data were combined with 19 Bioclim variables and I found relationships were different across and within species. Most significantly, I found \textit{Metasequoia} and \textit{Sequoia} separately showed a strong association between cross-sectional area and precipitation of the driest quarter, suggesting a shared or conserved climatic response. Further, I was able to show that while anatomy between taxa are distinct, their physiology converge, implying different strategies to achieve a similar physiology.

Chapter 4 uses the modern trait-climate relationships I established in Chapter 3 to test whether past traits could be reconstructed using climate. I combine modern anatomy-climate relationships of M. glyptostroboides, published fossil localities of M. occidentalis and M. sp., and early Eocene (~50 Ma) climate simulations at 3x and 6x pre-industrial CO2 to create, test, and apply generalized linear model with fossil leaf measurements. Generally, cross-sectional area, leaf width, and leaf thickness decrease with increased CO2, while resin canal area increase. These responses to warming climate most likely reflect cooling strategies to offload heat. Leaf width match 6x reconstructed traits, showing promising results for applications to test evolutionary hypotheses. 

By studying conifer adaptations, or traits, we can deepen our understanding of conifer evolution. This thesis demonstrated how conifer leaves were linked to climate and how I applied these relationships to uncover evolutionary implications within conivers. Conifer leaves are reflective of their growing environment and expanding research into conifer response to climate has valuable significance for understanding evolutionary history within conifers.