PhD Defence Devara Adiningrat | Modeling Old-Growth Forests and Soil Microbial Diversity Using Multisource Remote Sensing Data and eDNA

Modeling Old-Growth Forests and Soil Microbial Diversity Using Multisource Remote Sensing Data and eDNA

Devara Adiningrat is a PhD student in the Department of Natural Resources. (Co)Promotors are prof.dr. A.K. Skidmore and dr. M. Schlund from the Faculty of Geo-Information Science and Earth Observation (ITC) and prof.dr. T. Wang from the Faculty of Geo-Information Science and Earth Observation (ITC) and School of Geospatial Engineering and Science, Sun Yat-sen University. 

Biodiversity loss and forest degradation are accelerating globally, increasing the need for monitoring systems that capture ecological complexity across spatial scales. Old-growth forests are particularly important, as they support high biodiversity and regulate key ecosystem processes, including soil microbial dynamics. However, belowground biodiversity remains difficult to observe and map. This dissertation addresses this challenge by testing whether forest structural attributes derived from remote sensing can serve as proxies for modeling soil microbial diversity (as indicated by eDNA profiles) and enable spatial upscaling.

The research integrates airborne LiDAR, multispectral satellite data, and environmental DNA (eDNA) to link aboveground forest structure with belowground biodiversity. Results show that forest structural complexity, particularly canopy heterogeneity and vertical structure captured by LiDAR, is a strong indicator of old-growth conditions and a reliable proxy for soil microbial alpha diversity. In contrast, beta diversity is less explained by structure alone, indicating additional ecological controls.

Comparisons across data sources demonstrate that LiDAR provides more robust and transferable information than multispectral data for identifying structurally complex temperate forests. The integration of remote sensing and eDNA enables spatial modeling of soil microbial diversity across landscapes, although model performance remains context-dependent. Overall, this research provides a scalable framework for linking forest structure to belowground biodiversity, supporting improved ecosystem assessment and conservation planning.