Quantifying Uphill-Changes in Understory Density in Broadleaf Forests on Hillsides on Different Substrates in the Netherlands Using Lidar Point Cloud Data – Assessing the Effect of Edaphic Factors on Understory Vegetation Development

By Roland van Vulpen

Abstract
Growth of understory vegetation in broadleaf forests is limited by light conditions and edaphic factors. Forests on slopes are faced with different degrees of canopy openness depending on slope angle. Researching changes in understory in these forests can provide an insight in the impact of reduced light limitation. Airborne Laser Scanning (ALS) LiDAR has proven a useful tool to assess and quantify forest structure. Point cloud data collected for the third iteration of the Dutch mission for the national digital elevation model (AHN3) was used to derive metrics on the understory layer and canopy of forests. For this research project 19 forests on slopes with similar range in slope angle from 5 different geological areas of the Netherlands and from the four typical different broadleaf dominant forest types were compared in terms of understory density on different slope angles. Understory in this case was defined based on a height definition: returns from 0-5m were considered understory and above this canopy. The understory layer itself was subdivided into a herb and a shrub layer (0-0.5 and 0.5-5m respectively). Point cloud data on all forests was used to derive a slope map of the areas fit onto seven slope sector categories. By comparing trends in derived metrics understory density it was found that for all forest types the increase in slope angle led to an increase in density (returns per cell) of 15% for all forest types. Aggregating the forest types based on the carbon to nitrogen ratio (C/N-ratio) and their dominant tree species showed clear differences in slope angle sectors where understory density increased most comparatively. Changes in the variability of the height of the canopy compared to the changes in the herb and shrub layer. For the different forest types it followed that the shrub layer would follow an inverse trajectory to the variability in the canopy height: trends towards lower variability resulted in an increase in growth in understory. Soil pH however did not prove to be as effective a tool to create relevant groups for comparison. Overall for most soil types the resulting forest type showed to be the most relevant indicator of forest structure and thereby forest and understory responses to the uphill gradient in light conditions. Tree species showed to be a more complex tool as some species were found in different environmental conditions. Validation with field visits and the inclusion of two validation plots amongst the 19 forests that were assessed provided relevant input to the final assessment. Management (felling) in forests helped create more gaps that was beneficial to more understory development than expected, which affected some of the results and lead to the decision to exclude one forest. Limitations in terms of flight lines and spatial heterogeneity might have caused more flattened trend lines than might have been expected in reality.