Hesperornithes

Hesperornithes is an extinct and highly specialised clade of Cretaceous toothed birds. Hesperornithean birds, apparently limited to former aquatic habitats in the Northern Hemisphere, include genera such as Hesperornis, Parahesperornis, Baptornis, Enaliornis, and Potamornis, all strong-swimming predatory waterbirds. Many, if not all, species were completely flightless. The largest known hesperornithine, described in 1999 and named Canadaga arctica, may have reached a maximum adult length of over 1.5 metres (5 ft).

Hesperornithean birds were the only Mesozoic birds to colonize the oceans. They became extinct at the Cretaceous–Paleogene boundary, along with enantiornithine birds, all non-avian dinosaurs, and many other mostly reptilian life forms.

Anatomy and ecology
Most of what is known about this group rests on analyses of single species, as few provide sufficiently diverse fossils for analysis. In most cases, what holds true for one hesperornithine also applies to the others, as these birds were quite stereotypical and extreme in their autapomorphies.

Although some of the smaller species might have been able to fly, Hesperornis and Baptornis had only vestigial wings. As in the case of living foot-propelled diving birds, the femur and metatarsus were short, whereas the tibia was long. The legs were also set far back on the body, as in loons, grebes or penguins. Hesperornithids must have been powerful swimmers and divers but extremely ungainly on the land, and probably spent little time ashore except to nest. They were rather long-bodied, and measured about 6 ft long.

Some researchers think that on land they had to slide on their bellies and push with their legs; the hip and knee joints were shaped such that these birds could not move them dorsoventrally, and in a resting position the feet projected sideways from the body, which would have prevented them from walking upright. Whereas Hesperornis had probably lobes of skin on its toes similar to the grebes of today, the toes of Baptornis could not rotate (necessary in lobed feet to reduce drag) and thus were more likely webbed, as in loons. The dense bones of these birds decreased their buoyancy, making diving easier.

The beak was long, probably hooked at the tip, and bore a series of simple, sharp teeth which were set into a longitudinal groove. These probably helped to seize fish, as does the serrated beak of mergansers today; unlike the reptilian teeth of more primitive toothed birds, those of the hesperornithids were unique. They also retained a dinosaur-like joint between the lower jaw bones. It is believed that this allowed them to rotate the back portion of the mandible independently of the front, thus allowing the lower teeth to disengage.

Evolution
Currently, the hesperornithine birds are recognized as a very specialized lineage that is not ancestral or otherwise closely related to the birds of today. Still, their relationship is close enough that they probably diverged from the ancestors of today's birds as late as the earliest Cretaceous.

The discovery that Gansus yumenensis, which lived some time before the main radiation of hesperornithids, was a primitive member of the Ornithurae – that is, birds which had developed the type of pygostyle that is also found in modern birds – suggests that the ancestors of Hesperornithes – which also ornithuran – may have been birds generally similar to Gansus and living some 130–150 mya. Given the tendency of birds that have become flightless to increase in size markedly in just a few million years, the probability of Hesperornithes deriving from an animal at least superficially similar and not too far related to Gansus is quite high. The new fossil material of that species should assist in determining the number of synapomorphies of hesperornithines and other ornithurans, which have until now been muddied by the numerous autapomorphies of the former.

Loss and/or fusion of caudal vertebrae in pygostyle-like structures was a general trend in Cretaceous avian evolution, and a full pygostyle and associated structures may have evolved more than once to similar shapes. The fact that Gansus had non-pneumatized, dense bones, like those of hesperornithines, although it was not a specialised diver, is interesting to note. Similarly, the bone structure of Hesperornis indicates that as opposed to Enantiornithes and in line with other Ornithurae it showed rapid, uninterrupted growth to adult size.

The earliest known hesperornithine is the Early Cretaceous Enaliornis although these birds are somewhat tentatively assigned to this group because of the bad preservation of their remains; the majority of hesperornithine taxa are known from the Late Cretaceous. Small hesperornithine bones are known from the freshwater deposits of the Late Cretaceous of the Judith River Group as well as the Hell Creek and Lance Formations as well as Eurasian sites. These birds were about the size of a cormorant or a loon. Less well-verified is the presence of the lineage around Antarctica, which was a temperate, ice-free region in their time.

Classification
The clade Hesperornithes was originally named as a subclass of Aves by Furbringer in 1888. However, it was generally ignored in the scientific literature in favor of the order-level name Hesperornithiformes, coined one year later. In 2004, Clarke became the first to define the hesperornithean group in terms of phylogenetics. Clarke defined Hesperornithes as all birds closer to Hesperornis than to modern birds, and regarded Hesperornithiformes as a junior synonym, though she did not define the latter name. Clarke also defined the more inclusive group Hesperornithidae as all hesperornitheans closer to Hesperornis than to Baptornis.

These birds were originally combined with Ichthyornis in the paraphyletic "Odontornithes" by Othniel Charles Marsh, in 1873. In 1875, they were separated as Odontolcae. The group was often considered to be allied to loons and grebes, or to the Paleognathae (based on a perceived similarity of the bony palate). These similarities, however, as the more recently determined fact that the osteons of their bones – at least in Hesperornis – were arranged in a pattern similar to that in Neognathae, are today considered to be due to convergent evolution.