Abstract

The Mesozoic and Cenozoic track record of large birds (avian theropods) with footprint lengths (FL) > 10.0 cm is quite limited, whereas small tracks (FL < 10.0 cm) are abundant from the Early Cretaceous onwards. This large versus small threshold value is consistent with the track record of extant birds among which only ˜10% are large, and so is scaled appropriately to Class Aves. The proportion of large pachydactylous (thick‐toed) or robust non‐avian theropod ichnotaxa reported from the Mesozoic considerably exceeds the few large leptodactylous (thin‐toed) or gracile forms such as Archaeornithipus and Magnoavipes, named, albeit controversially, to suggest avian affinity. The Cenozoic record of large avian tracks is quite different, with large pachydactylous tracks limited to a few Eocene‐Oligocene (Palaeogene) ichnogenera (Ornithoformipes and Rivavipes), and rare unnamed Neogene and Holocene dinornithiform tracks. Thus, large flightless species are less well represented by tracks than body fossils. This suggests probable preservational bias in favour of waterbird or shorebird‐like tracks in shoreline facies. However, the Cenozoic record of large leptodactylous tracks attributable to cranes, herons and related trackmakers is quite extensive, mostly confined to the Neogene and formally named to suggest representatives of extant avian clades capable of flight. These distribution patterns apparently reflect the dominance of large pachydactylous non‐avian and avian theropods in the Mesozoic and early Palaeogene, and contrast with the rise of large, as well ongoing diversification of smaller leptodactylous Aequornithes in the Neogene. These distribution patterns in space and time likely reflect changing selection pressures as today's modern avifauna evolved. □ - bird, Cenozoic, footprints, ichnospecies, ichnotaxabase.