Background/Question/Methods Vegetation structural complexity is an ecosystem morphological trait complementary to others like vegetation height or cover, which together complete the spectrum of ecosystem structure Essential Biodiversity Variables (EBVs). However, there is much confusion around the definition and indicators employed to measure the structural complexity of ecosystems, with authors speaking of ‘structural diversity’, ‘roughness’, ‘horizontal and vertical structure’, etc. The most popular approach follows the early works of McArthur (1961) who calculated Foliage Height Diversity (FHD) and related it to diversity of bird species. However, soon Lovejoy’s work in the Amazon (1972) and other authors reported concerns when not finding similar relationships. As technologies advanced and Light Dectection and Range (LiDAR) made available very detail characterizations of the vertical profile, the most popular approach was to follow up with FHD. However, the early works by Valbuena et al. (2012) developed mathematical demonstrations that FHD is essentially flawed, showing that while McArthur stratification into three strata could show good relationships, the method fails once stretched into LIDAR finer resolutions (strata number dependency). This presentation shows that entropy and variance are interrelated components in defining ecosystem structural complexity, with a robust mathematical framework that can allow determining either while avoiding FHD’s methodological flaws.
Results/Conclusions A unified framework for the characterization of ecosystem structure by measuring either entropy – Simpson or Shannon indices –, or variability – variance or the Gini coefficient (GC) – is presented. It shows how maximum entropy can be flagged up from values of the GC. This mathematical framework provides unified means for determining maximum entropy in the 3D space of information provided by LiDAR. Results, from several study areas such as the Biological Dynamics of Forest Fragments Project (BDFFP), show the relationships between measures of entropy and variability, and with other ecosystem structure traits. Also, using data from the Santa Rosa National Park Environmental Monitoring Super Site we evaluate linkages between ecological diversity and entropy as a function of ecological succession. In the AMAZECO project we also investigated common workflows for the derivation of ecosystem morphological traits using satellite-borne and airborne LIDAR. The former allowed to develop an ecosystem structure EBV product covering the entire of the Brazilian amazon using data from the Global Ecosystem Dynamics Investigation (GEDI) mission plus extensive collections of circa 1,000 airborne LIDAR transects. This investigation opens the door to reporting of ecosystem structure EBVs at national and global scales from cross-platform LIDAR datasets.
