New insights into drivers and passengers of tree architecture
Cumulative thesis
Date of Examination:2023-05-02
Date of issue:2023-06-27
Advisor:Prof. Dr. Dominik Seidel
Referee:Prof. Dr. Holger Kreft
Referee:Prof. Dr. Peter Annighöfer
Files in this item
Name:Final.PhD.Dissertation_Dorji2023.pdf
Size:6.20Mb
Format:PDF
Description:PhD Dissertation for Publication as eDiss
Abstract
English
Individual tree architecture and species composition perform a critical role in many ecological processes and resources a forest offers, such as wood value, biodiversity, and ecosystem stability. The structure and dynamics of a forest are determined by its indi-vidual trees' architecture and growth patterns. In turn, the interaction of ecological pa-rameters and genomic structural components contributes to the architecture and growth of trees. However, understanding how tree structure develops and adapts in response to various factors, such as competition, drought stress, sunlight angle, and resource utilization, is still scarce. Several theories and models exist in advocating and disseminating the concerns and issues about this discipline. However, many aspects are still unknown due to the scarcity of data. Therefore, the scope and aim of this dis-sertation are to look into the drivers and passengers of tree architecture from empirical evidence. We quantitatively analyzed the three-dimensional architecture of the trees using LiDAR (light detection and ranging) and fractal geometry approaches. The combination of Li-DAR technology and fractal analysis has made it possible to give a holistic, in-depth analysis of tree architecture. Hence providing an avenue to empirically draw links be-tween the trees’ architecture (and the complexity of this architecture) and several eco-system functions, which were not possible in the past. In this thesis, we present three independent papers (chapters 2 to 4) related to exploring the drivers and passengers of tree architecture as follows. In the first study, we explored the intricate 3D structure of 24 beech (Fagus sylvatica L.) trees that grew amid various levels of competitive pressure using 3D LiDAR data from the German Biodiversity Exploratories. We developed robust quantitative structure models (QSMs) of each tree to understand their branching patterns. The box-dimension (Db) method from fractal analysis was used to quantify the architectural complexity and self-similarity of the trees. The findings showed that the competition appears to signifi-cantly influence the branching structure of trees, as demonstrated by the strong re-sponses exhibited by various tree architectural measures. A new metric presented here, the ‘Db-Intercept’ (intercept of the regression used to derive the box-dimension), showed the most robust response to competition, with a correlation coefficient of -0.78. In the second study, we sought to determine whether (i) latitudinal adaptations of crown shape result from distinctive solar inclination angles at a species' origin? (ii) structural variances in trees are associated with seed dispersal methods? and (iii) tree structural complexity is linked with tree growth performance? We scanned 473 trees using MLS (mobile laser scanning) to obtain 3D data for each tree. The arboretum's en-vironmental conditions were the same for all the tree species being investigated, alt-hough coming from different latitudinal regions. Then, applying fractal analysis and the box-dimension method, the tree's structural complexity was quantified. Also, the topological measurement of a tree's top-heaviness (Rel.Hmaxarea) was derived. We observed that trees from higher latitudes had significantly less top-heavy geometry than those from lower latitudes. Therefore, to some extent, a tree species' crown form appears to be influenced by solar elevation angles at the species' origin. Additionally, we revealed that tree species with wind-dispersed seeds had higher tree architectural complexity than those with seeds dispersed by animals (p < 0.001). Furthermore, tree structural complexity was positively associated with the trees' radial growth increment (p < 0.001). In the third study, we used terrestrial laser scanning (TLS) to scan 71 trees of 19 spe-cies and generated 3D attributes of each tree. We constructed QSMs to characterize their branching patterns. Additionally, the box-dimension approach from fractal analy-sis was used to assess the overall structural complexity of the trees. The pressures in-ducing 12%, 50%, and 88% losses of stem hydraulic conductance (P12, P50, P88) of all the concerned trees were measured. Our findings revealed that the tree's structural com-plexity (Db) relates significantly to xylem safety (p < 0.001). The branching geometry also showed significant results relating to xylem pressure (p < 0.01). We further ob-served a close relationship between specific hydraulic conductivity (Ks) of the branches and Db, while the hydraulically-weighted vessel diameter was also related to Db at mar-ginal significance. Finally, we also conclude that using 3D data from LiDAR in combination with geomet-rical analysis, including fractal analysis, is a promising tool to investigate tree architec-ture relating it to ecosystem functionality.
Keywords: Keywords: Tree architecture, LiDAR, Fractal analysis, Box-dimension, Competition, Seed dispersal strategy, Sunlight angle, Tree growth, Climate Change, Xylem pressure, Hydraulic vulnerability.