Evolution and Development of Tendrils in Bignonieae (Lamiales, Bignoniaceae)1

The tribe Bignonieae includes all Neotropical lianescent Bignoniaceae. The leaves of Bignonieae are generally 2- or 3-foliolate, with terminal leaflets modified into a tendril. These tendrils have varied morphologies and are thought to have been involved in the diversification of Bignonieae. Little, however, is still known about the biology and evolution of tendrils. This study investigated the evolution and development of tendril types in Bignonieae in order to further understand how changes in leaf morphogenesis led to current patterns of variation in tendril morphology. For that, we investigated the ontogeny of 11 species representing a wide diversity of tendril types (i.e., simple, trifid, and multifid) and used a recently published phylogeny of Bignonieae as the basis to reconstruct patterns of evolution for tendril types. For those analyses, we used maximum likelihood (ML) and maximum parsimony (MP) approaches, with both ACCTRAN and DELTRAN optimization schemes implemented in the latter. Ancestral character state reconstructions of tendril type suggest that the ancestral condition for the whole tribe (Core Bignonieae plus Perianthomega Bureau ex Baill.) is a lack of tendrils, whereas parsimony reconstructions indicate an ambiguous ancestral condition. However, all reconstructions suggest that trifid tendrils represent the ancestral condition for the Core Bignonieae. Other tendril types evolved subsequently through a series of developmental changes. Furthermore, tendril ontogenetic studies provided key information for the resolution of major ambiguities in the ancestral state reconstructions of tendril type in Bignonieae. For instance, in Bignonia callistegioides Cham. and B. prieurei DC. (simple tendrils), no traces of remnant lateral branches from a trifid-tendrilled ancestor were found, corroborating the ACCTRAN hypothesis of a simple tendril condition for the ancestor of that lineage. In Tanaecium pyramidatum (Rich.) L. G. Lohmann (trifid tendrils), on the other hand, we detected a pattern and rate of tendril differentiation that initially followed the same pattern seen in taxa with simple tendrils (i.e., Cuspidaria DC. and Fridericia Mart.), with a developmental delay relative to other trifid-tendrilled species, favoring a simple-tendrilled ancestor hypothesis and supporting the ML and ACCTRAN optimizations. The interpretation of the ontogenetic data in light of a robust phylogenetic framework led to specific hypotheses about the evolutionary processes and respective changes in gene regulation that may have led to the current tendril morphologies found in Bignonieae. In particular, we suggest that tendril evolution involved heterochrony and hypothesize that changes in the expression of genes that are associated with compound leaf development may have led to the diversity of tendril morphology currently observed in Bignonieae.