Multiple origins of secondary temporal fenestrae and orbitozygomatic junctions in birds

Abstract Orbitozygomatic junctions (bony connections between the postorbital and zygomatic processes) and secondary temporal fenestrae have long been known to occur in a few avian species, but no comprehensive study of this phenomenon has ever been published. Having surveyed all non‐passerine and most passerine families, we established that the orbitozygomatic junction evolved 18–20 times independently in Cracidae, Phasianoidea (Odontophoridae and Phasianidae), Gastornis , Columbidae (three times), Pteroclidae ( Syrrhaptes ), Aptornis , Thinocoridae ( Thinocorus ), Scolopacidae (possibly twice), Ciconiidae (twice), Brachypteraciidae ( Uratelornis ), Picidae ( Picus spp.), Psittacidae ( Melopsittacus ), Cacatuidae, and Alaudidae (twice). The junction arises in evolution as a result of either elongation of the two processes that meet at angles or the appearance of a bony cross‐bridge in place of a ligament or aponeurosis. In the first case, the junction is initially non‐adaptive, as indicated by its extreme variation (e.g., in Cracinae and Ciconiidae), and may or may not prove functional as an exaptation. Whenever adaptive, the junction supports an expansion of the adductor mandibulae externus (primarily its pars media). In addition, a rostral extension of the tympanic wing has come to cross the temporal fossa in Strigidae (at least twice) and probably Podargus . Altogether, secondary bony connections across the temporal fossa evolved independently at least 21–23 times in neornithine birds.