Phylogenetic methods in biology use hierarchical clustering algorithms to trace evolutionary history using heritable genetic data. Stellar phylogenies similarly aim to reconstruct the chemical history of galaxies using heritable information in the form of chemical abundances passed on between generations of stars. Previous studies using the distance based method of the Neighbour-Joining(NJ) algorithm have uncovered a hierarchical relationship between chemical abundances and stellar age (Jofre et al., 2017, Jackson et al.,2021, De Brito Silva in prep) suggesting a non-random phylogenetic signal exists within chemical abundance data. To use tree building algorithms the underlying structure of the data needs to be itself tree-like, a property known as additivity. In biological systems numerous processes influence the additivity of genetic data however it is unclear how different chemical evolutionary events might influence chemical abundance data. Using analytical models and ‘open-box’ FlexCE simulations we simulated chemical abundances from galaxies with various star formation histories and explored the additive structure of these datasets. We identified the NJ algorithm, though generally robust, was sensitive to changes in additivity caused by starburst events and identified a combination of abundances that minimises this effect. Finally, we present a method of assessing the additivity of chemical abundance data.