Tongli Wang

Greenhouse CO₂ enrichment has been widely applied in commercial production and investigated in numerous experiments worldwide, yet findings vary across dispersed studies, and its yield benefits have not been systematically quantified and interpreted at the global scale. This dissertation quantifies the greenhouse CO₂ fertilization effect, explains variation across crop functional types and harvested organs, and identifies where CO₂ enrichment is most effective under current and future climates. In Chapter 2, I compiled a global dataset of greenhouse experiments (454 observations from 147 studies) and applied meta-analysis using log response ratios (lnRR) to estimate yield gains and benchmark greenhouse CO₂ enrichment against Free-Air CO₂ Enrichment (FACE). Across comparable CO₂ increments of 115–300 ppm (ppm, parts per million by volume (μmol mol⁻¹)), greenhouse CO₂ enrichment increases yields by ~28% on average and delivers ~1.4 times larger response than FACE. Dose-response analyses suggest that yield gains peak at CO₂ concentration increment of 800–1,200 ppm. Chapter 3 identifies differences in responses among crops with harvestable organs, showing that below-ground crops (roots and tubers) exhibit roughly double the yield response of above-ground crops, which may be attributed to stronger sink capacity and carbon storage of below-ground crops. In Chapter 4, I evaluate climatic and soil controls on enrichment performance and train Random Forest models (R² around 0.7) using CMIP6 GCM ensemble to map global suitability and project changes under SSP126 and SSP585. The maps suggest a poleward shift in suitability for greenhouse CO₂ enrichment under future scenarios. Overall, this thesis integrates meta-analysis, greenhouse–FACE benchmarking, yield-CO₂ response surfaces, and machine-learning suitability mapping to provide evidence-based guidance for climate-smart greenhouse CO₂ enrichment under ongoing climate change.

View record

Forest ecosystems cover almost a third of the earth's land area and are vital for providing ecological services, such as supplying habitats, cleaning the air, and mitigating climate change. However, climate change is rapidly altering the environmental conditions necessary for the survival and productivity of forest tree species. My PhD research project is focused on using ecological niche models (ENMs) to assess the impacts of climate change on forest ecosystems and develop climate-smart forestry strategies. In the second chapter, I conducted a comprehensive review of recent advancements in ENMs, including advances in machine learning algorithms, the data sources being used, and their applications in forestry. This review explored how ENMs are applied at various ecological scales and how they advance our understanding of the impacts of climate change on forest tree species’ suitable habitats. Chapter 3 developed an innovative approach by incorporating soil variables and interspecific competition into climate niche models (CNMs), using an endangered species Chamaecyparis formosensis (red cypress) as a case study, to predict climate change impacts and suitable habitats more accurately. Chapter 4 is focused on examining common drivers for species niche distribution, and identifying biodiversity hotspots through building niche models for 100 key tree species in China. The impact of climate change on the distributions of suitable habitats of the species and the shift of biodiversity hotspots were also assessed under various future climate scenarios. In Chapter 5, I developed a site-based climate-smart tree species selection tool to help foresters identify species best suited to future climate conditions at any given specific planting site in China, aiming to enhance the effectiveness of forestation efforts and improve the ecological resilience of plantations. Finally, Chapter 6 synthesized the major findings of the research, highlighting the implications of my study for forest management and ecosystem restoration in the context of climate change. The study filled critical knowledge gaps and proposed a pathway for adaptive forest management, including species selection, assisted migration, and conservation, to mitigate climate change's impacts on forest ecosystems.

View record

Species climate niche models (CNMs) have been widely used for assessing climate change impact and developing adaptation strategies for forestry. However, CNMs built based on species occurrence data reflect only species’ realized niche, which may overestimate the suitable habitat loss of existing forests, underestimate assisted migration potential, and are unable to quantify productivity. To address these deficiencies, my research objectives were to explore modeling approaches aimed at predicting forest species fundamental niche and productivity based on species provenance trials and occurrence data. In chapter 2, a universal response function (URF) was developed for lodgepole pine (Pinus contorta Dougl. ex Loud.), through comparison and optimization of the best existing modeling approaches, to predict the species’ fundamental climate niche and productivity. This model explained 80 % of the variation among provenance and test sites, and the prediction of fundamental climate niche was validated with global observations with 94.6 % agreement. While this approach is considered ideal, it may not be applicable for many forest tree species due to the lack of comprehensive provenance trials. In the chapter 3, I built a fundamental climate niche model using widely available species occurrence data with lodgepole pine and Douglas-fir (Pseudotsuga menziesii Franco.). I identified a new cut-off threshold of 0.3 and extended the CNMs in predicting realized climate niches to fundamental climate niches, the result presented greater niche gain (up to 187 %) and reduced habitat loss (up to 80 %) by 2050s under a moderate climate change scenario. Similarly in Chapter 4, I investigated the potential to extend occurrence-based CNMs to predict species productivity using lodgepole pine and Douglas-fir as the template species for their comprehensive range-wide occurrence data and availability of site productivity data. The CNMs were optimized through a series of steps, achieves R2 above 0.9 in reflecting measured site productivity and validated with R2 above 0.7 using independent datasets for each species. My research provided crucial tools for evaluating climate change’s impact on species suitable habitat distribution and productivity and holds potential for informed forest management decisions, including conservation and assisted migration aimed at maximizing future productivity and carbon sequestration.

View record