Large-scale estimation and mapping of forest biodiversity indicators and ecological threats in the West Usambara montane forests
| dc.date.accessioned | 2024-08-26T07:18:37Z | |
| dc.date.available | 2024-08-26T07:18:37Z | |
| dc.date.issued | 2024-05 | |
| dc.description | MSc. in Forestry | |
| dc.description.abstract | Assessment of the indicators of forest biodiversity, such as tree species diversity, evenness, and richness, is crucial for monitoring and managing forest ecosystems. Quantifying the relationship between environmental factors, such as topography, soil, and climate, and tree species diversity and distribution is also critical for understanding the pattern of the indicators of forest biodiversity and informing conservation efforts. Various studies have therefore, shown that environmental factors can significantly influence tree species diversity and distribution in forest ecosystems, with topography and soil moisture being important predictors of forest structure and species richness. Such factors have also been used to develop models for large scale predictions under different environmental conditions. However, in the recent decades, remote sensing based had been considered to be as one among the tools for facilitating the assessment of tree species diversity and distribution patterns across large areas of dense montane tropical forests. By modelling the relationship between remotely sensed data and tree species diversity, these techniques can help identify priority areas for conservation and guide management strategies. Furthermore, remotely sensed based techniques had useful in assessing ecological threats such as wildfire which essentially play a significant role in shaping the structure and composition of tropical forests. This thesis consists of three manuscripts that assessed indicators of forest biodiversity in the West Usambara montane forests of Tanzania. The first manuscript aimed to determine the role of environmental factors on tree species composition and diversity in the West Usambara montane forests. The field data were collected through a two-phase systematic sampling approach, and environmental data were obtained from USGS, ISRIC, and NCCS for topographic, soil and climate respectively. The second manuscript aimed to assess the potential use of remotely sensed data to model and monitor forest biodiversity in the study area. The study computed field diversity, predicted diversity using GAM and XGBoost models for Sentinel 2 and PlanetScope imagery, and compared the efficiency of the sensors and models. The third manuscript aimed at studying the post fire recovery of forest composition and structure. It assessed the difference in structure and composition between burnt and unburnt areas and lastly burnt area mapping was conducted.The results grouped the 183 identified tree species into three distinct forest communities using cluster analysis. Indicator species analysis identified species significantly associated with each community, like Dombeya burgessiae in the higher elevation community. Environmental data on climate, soil properties, and topography were compiled. Canonical correspondence analysis revealed variables like precipitation, soil nitrogen, and elevation were influential in driving community patterns. The communities differed significantly in diversity and richness, with greater values in lower elevation communities. Species turnover linked to environmental gradients was the primary contributor to beta diversity. Overall, the study highlighted the importance of multi-scale abiotic factors in shaping tropical montane forest communities. The findings have implications for ecological monitoring and conservation efforts in these biodiverse yet threatened ecosystems. An integrated assessment of climate, edaphic, and topographic variables is therefore key to understanding the environmental forces structuring tree communities and diversity in tropical montane systems. Combining field-based approaches and remote sensing techniques provides valuable insights into the factors that influence tree species diversity and distribution in West Usambara Montane forests. The results highlight the strong influence of soil factors such as pH and nitrogen on tree diversity and distribution, while also demonstrating the potential of remote sensing data, particularly PlanetScope data, for assessing and estimating forest biodiversity indicators. The study recommends that conservation efforts should prioritize areas showing low tree species diversity and take into account the influence of environmental factors such as soil properties. The use of remote sensing techniques can facilitate the identification of these priority areas and guide management strategies. Additionally, further research is needed to explore the potential of other remote sensing data sources and models to improve the accuracy and reliability of biodiversity assessments. Forest fires play a significant role in the diversity and composition patterns in the tropical montane forests. It is therefore important to understand the manner through which these ecosystems are affected by forest fires together with their recovery patterns. From this study, it was observed that while fire did not significantly alter overall tree species composition, it did reduce structural parameters like density, basal area, biomass, and diversity compared to unburned forest. However, these structural attributes exhibited recovery and increased with time since fire. Indicator species analysis identified unique taxa in burned areas. A 14-variable model integrating spectral, textural, and vegetation indices was used to map ~ 1430 ha of burned forest. The study demonstrated the resilience of tropical montane forest composition to fire disturbance, although more work is needed to fully understand post-fire dynamics and long-term recovery. Overall, the study highlights the utility of integrating field measurements and satellite data for assessing fire impacts in tropical ecosystems. | |
| dc.description.sponsorship | The Eastern Arc Mountain Conservation Endowment Fund (EAMCEF), Tanzania Forest Fund (TaFF), and RMCRD/GMES and Africa grant | |
| dc.identifier.uri | https://www.suaire.sua.ac.tz/handle/123456789/6388 | |
| dc.language.iso | en | |
| dc.publisher | Sokoine University of Agriculture | |
| dc.subject | Large-Scale Estimation | |
| dc.subject | Forest Biodiversity Indicators | |
| dc.subject | Ecological Threats | |
| dc.subject | Usambara Montane Forests | |
| dc.title | Large-scale estimation and mapping of forest biodiversity indicators and ecological threats in the West Usambara montane forests | |
| dc.type | Thesis |