Avisaurids were among the largest and last enantiornitheans to have lived, although they are also among the most poorly preserved. The majority of them are known primarily from fossilized tarsometatarsal bones, the part of a bird's leg formed by fused metatarsals. As a result, members of this family are distinguished from other enantiornitheans exclusively by features of the tarsometatarsal and pedal phalanges. Unlike in some prehistoric birds, avisaurid tarsometatarsals were not completely fused, with the distal parts of the metatarsals being separate from each other. The proximal half of metatarsal III is convex from the front. The inside edge of this bone's trochlea has a bony tab which points downward, known as a plantar projection. The innermost bone of the tarsometatarsus, metatarsal I, is small, laterally compressed, and J-shaped from the side. It is connected to a reversed hallux sporting a very large and curved claw. Chiappe and Calvo found that the Avisauridae shared adaptations of the foot — including a fully reversed and distally placed hallux with a large claw — that indicated the ability to perch in trees. They argued that an arboreal habit was most likely for all of the Avisauridae.
History and classification
Avisauridae was erected as a family by Brett-Surman and Paul in 1985. At that time the family consisted of a few fossils that they believed belonged to small non-avian dinosaurs. They doubted that these fossils belonged to birds due to the presence of several features of the tarsometatarsus. In Avisaurus, only the proximal parts of the metatarsals were fused, the proximal part of metatarsal III was wide, and the hypotarsus was poorly developed. However, Chiappe later reassigned the Avisauridae to the class Aves and the subclassEnantiornithes in 1992. He noted that the features used to exclude avisaurids from birds are in fact present in some early birds such as Archaeopteryx, as well as various Cretaceous bird groups. Avisaurids also had a thin metatarsal IV and a bony knob on the front of metatarsal II for the insertion of M. tibialis cranialis, both believed to be enantiornithean features. Chiappe defined the family as the common ancestor of Neuquenornis volans and Avisaurus archibaldiplus all its descendants. In 2008, the family was given a broader definition courtesy of Cau and Arduini. They redefined the group as Avisaurus archibaldi and all genera more closely related to it than to either Longipteryx, Gobipteryx or Sinornis. Matt Martyniuk gave the name Avisauroidea to this group, although the erection of that name has been criticized by Cau. Under this broader definition, several other enantiornitheans, such as Enantiophoenix, would qualify as members of the family. Nevertheless, enantiornithean taxonomy is notably difficult to resolve, and some analyses on enantiornitheans have not resolved the family. However, this may be due to such analyses focusing on early Cretaceous enantiornitheans rather than fragmentary late Cretaceous taxa, such as most avisaurids. The following is a cladogram based on Cau and Arduini :
