LAUSR

Morphology has always had a central role in the natural sciences and study of the phenotypical diversity of an organism is pivotal to understanding patterns and processes of the living world. It is through their phenotype that organisms interact with the surrounding environment and with each other and, ultimately, phenotypes are the results of evolution (Wagner, 2001a). In the context of systematics, morphological features were the source of data underpinning taxonomic and most evolutionary hypotheses in the pre-Hennigian era, as well as during the advent of cladistics, before the use of molecular data became widespread. Difficulties in interpreting morphological data and conceptualizing characters (as reflected in the debate on what characters are; see Wagner, 2001b), in contrast to the relative ease in generating large molecular data sets, have led some to propose that morphology should have a limited role in phylogenetics and should only be evaluated in light of DNA-based trees (Scotland, Olmstead & Bennett, 2003). Recent advances, however, are improving our ability to assess comparative morphology. Initiatives such as MorphoBank (O’Leary & Kaufman, 2011) allow collaborative work via the internet for the scoring of morphological data and building data matrices (e.g. O’Leary et al., 2013). The use of ontologies for comparing semantic description of phenotypes is also a technical advance for the study of morphology (Vogt, Bartolomaeus, & Giribet, 2010; Deans, Yoder, & Balhoff, 2012; Deans et al., 2015). By having a database of annotated phenotypes, one could automatically extract phylogenetic characters (e.g. Dececchi et al., 2015) and study morphological evolution (e.g. Ramírez & Michalik, 2014). Adoption of a semantic approach could also be a step towards resolution of some problems associated with character formulation (Vogt, 2017). Also, the use of morphological data adds dimensions to studies of biological diversity and establishes a link for the use of both extant and extinct taxa for phylogenetic inference (Jenner, 2004; Wiens, 2004). Of course, and as with any other type of data, including molecular data (Liu et al., 2010), morphology is prone to convergent evolution and must be used with care. Nonetheless, it still has an important role to play in phylogenetics. In the phylogenomic era (Giribet, 2010, 2015), there are examples of concordance between genomic data and morphology, even when such concordance was previously rebutted by small molecular datasets (e.g. Stephens et al., 2015). In recent years, we have seen a large accumulation of molecular phylogenic analyses for Fabaceae (e.g. Luckow et al., 2003; Wojciechowski, Lavin & Sanderson, 2004; McMahon & Sanderson, 2006; Bruneau et al., 2008; Simon et al., 2009; Cardoso et al., 2012; Manzanilla & Bruneau, 2012; Cardoso et al., 2013), the third largest family of flowering plants (Lewis et al., 2005; LPWG, 2017). These studies confirmed the monophyly of Fabaceae, but of the three traditional subfamilies, only Faboideae and Mimosoideae were supported as natural groups, with Caesalpinioideae being polyphyletic. This conclusion was initially highlighted by cladistic analysis of DNA (Doyle, 1995) and morphology (Chappill, 1995) and further supported by combined analysis of the two (Herendeen, Bruneau & Lewis, 2003). Although it was clear that the classification of Fabaceae had to be updated, how to do this was *Corresponding author. E-mail: [email protected]