Structural Complexity and Biomass of Mountain Forests in the Annapurna Region, Himalaya
Doctoral thesis
Date of Examination:2025-09-10
Date of issue:2026-01-06
Advisor:Prof. Dr. Dirk Hölscher
Referee:Prof. Dr. Dirk Hölscher
Referee:Prof. Dr. Dominik Seidel
Sponsor:DAAD
Files in this item
Name:eddis_forest_structure_himalaya_smita_das.pdf
Size:3.37Mb
Format:PDF
Abstract
English
Heterogeneous environmental conditions characterize mountain forests and can be highly dynamic due to climatic and socio-economic changes. Despite their importance, our understanding of forests in the Himalaya is limited, partly because these forests are difficult to access. This study aimed to advance the scientific understanding of mountain forest structures and above-ground biomass in the Annapurna region of Nepal. The study was carried out on 69 plots, covering forests across a broad range of elevation (1200 - 3800 m asl), precipitation (900 - 3700 mm yr⁻¹) and slope (2 - 42˚). The forests on the study plots comprised evergreen needle-leaved tree species such as Abies spectabilis, Cupressus torulosa and Pinus wallichiana, evergreen broad-leaved tree species such as Daphniphyllum himalense and Rhododendron spp., and deciduous broad-leaved tree species such as Alnus nepalensis, Acer spp. and Betula utilis. We employed ground-based mobile laser scanning to capture tree architectural characteristics and to quantify forest and tree structural complexity. Study 1 compared relative canopy height and aboveground biomass derived from ground-based and spaceborne laser scanning. Spaceborne data originated from the Global Ecosystem Dynamics Investigation (GEDI). Relative canopy height estimates from both approaches were moderately correlated, with a correlation coefficient of r = 0.42 for RH98, though GEDI consistently produced higher values. Aboveground biomass from ground-based scans ranged from 28 to 260 Mg ha⁻¹, while GEDI-derived estimates were on average 65% higher and moderately correlated with ground-based values (r = 0.51). A significant discrepancy was observed in plant functional type classifications used in the GEDI biomass estimation, which often did not match field observations. This led to an average biomass overestimation of 18%. Unlike studies in other mountainous regions, our findings showed only a slight slope effect. Overall, this study highlights both the potential of spaceborne GEDI data for forest monitoring in remote mountain regions and the need for further refinement to improve its accuracy. Study 2 addressed the drivers of forest stand structural complexity. It was based on ground-based mobile laser scanning. The explanatory variables used strongly influenced forest stand structural complexity (adjusted R² = 0.60), with significant contributions by the number of trees, maximum height of the forests, species diversity, north-facing aspect, soil pH and forest disturbance. In contrast, climatic factors such as precipitation and temperature showed no detectable effect. It has, however, to be noted that the study plots were located below the tree line ecotone. Study 3 focused on the relationship between tree architecture, environmental conditions and tree structural complexity. It was further examined whether and how tree structural complexity translates into forest stand structural complexity. The study covered 546 trees on 14 undisturbed study plots. It was found that tree structural complexity, expressed as Db, was lowest for the needle-leaved Pinus wallichiana and highest for the broad-leaved Daphniphyllum himalense. A high share of variation in Db was explained by tree architecture. In multivariate models, tree height, crown radius and crown length explained more than 60% of the observed variation in Db. Stem density of the plot accounted for 19% of the variation in Db and there was no influence of tree diversity. Precipitation explained 13% of the observed variation in tree Db, but elevation and slope did not have significant influences. As expected, tree height declined with elevation, but small trees often exhibited high Db values. The standard deviation of tree-level Db within a plot explained 47% of the variation in stand-level complexity among plots, surpassing the maximum tree-level Db. This suggests that the sole removal of small or large trees would reduce stand-level complexity by 36%. We conclude that in the Himalayan forests, species identity and tree architecture play a significant role in determining tree structural complexity, while environmental factors have a smaller role. Furthermore, structural variation among the trees within a plot plays a crucial role in the structural complexity at the stand level. Overall, these studies provide insights into forest and tree structures in the Annapurna range across environmental gradients and thereby may contribute to the good management of these forests.
Keywords: Mountain Forest; Forest structure; LiDAR; Biomass; Tree architecture; Himalaya
