The idea of DAI is to disentangle architectures of trees with a strong apical dominance, i.e. with a central main stem growing more strongly than other side axes (such as in most conifers, for instance), from those having a spread out branching pattern with similar axes and no obvious main stem.
The higher the index the more dominant the principal axis. DAI is computed based on the indices A0 which is an index of the probability of being the principal axis for each path of the tree (see function Compute_A0
).
More information are given in Martin-Ducup et al 2020