New phylogenetic proposal for Procellariiformes (Aves) based on morphological data

Abstract

The phylogenetic relationships within Procellariiformes remain poorly resolved and subject to ongoing debate. Although extant species are traditionally classified into four families (Diomedeidae, Procellariidae -sometimes including Pelecanoididae-, Hydrobatidae, and Oceanitidae), recent hypotheses have relied predominantly on molecular data and often lack comprehensive anatomical sampling. Emphasizing the importance of anatomical studies, in this era dominated by molecular phylogenies, we conducted a detailed morphological analysis based on a dataset comprising 526 characters and 34 species. Phylogenetic analyses were performed under maximum parsimony, unconstrained and with molecular constraints, with Traditional Search and both equal and implied weighting schemes with varying concavity values. Synapomorphy mapping, absolute Bremer support, and Jackknife analyses were also performed. The unconstrained and equal weighting analysis yielded a single most parsimonious tree (length = 2286, CI = 0.383, RI = 0.602), while implied weighting produced topological shifts and increased tree lengths. Several groups wererecovered as monophyletic (unambiguous apomorphies and synapomorphies in parentheses):Procellariiformes (69), Diomedeidae (34), Pelecanoididae (13), Oceanitidae (24), Hydrobatidae (27); and Oceanitoidea (Hydrobatidae + Oceanitidae) (37). In contrast, Procellariidae was not recovered in its traditional sense. However, the clade Procellarii (18), containing all non-diomedeid taxa, with Macronectes giganteus as the most basal member, is proposed. Ardenna and Puffinus were not recovered as monophyletic in any of the analyses. A high proportion of synapomorphies was shared between the unconstrained and constrained analyses (>50%). Morphological characters were taxonomically informative, including characters derived from elements often underestimated in previous studies such as vertebrae, radius, ulna, and fibula.