Troodon

1860 illustration of the T. formosus holotype tooth

The name was originally spelled Troödon (with a diaeresis) by Joseph Leidy in 1856, which was officially amended to its current status by Sauvage in 1876.[2] The type specimen of Troodon has caused problems with classification, as the entire genus is based only on a single tooth from the Judith River Formation. Troodon has historically been a highly unstable classification and has been the subject of numerous conflicting synonymies with similar theropod specimens.[3]

The Troodon tooth was originally classified as a "lacertilian" (lizard) by Leidy, but reassigned as a megalosaurid dinosaur by Franz Nopcsa von Felső-Szilvás in 1901 (Megalosauridae having historically been a wastebin taxon for most carnivorous dinosaurs). In 1924, Gilmore suggested that the tooth belonged to the herbivorous pachycephalosaur Stegoceras, and that Stegoceras was in fact a junior synonym of Troodon (the similarity of troodontid teeth to those of herbivorous dinosaurs continues to lead many paleontologists to believe that these animals were omnivores). The classification of Troodon as a pachycephalosaur was followed for many years, during which the family Pachycephalosauridae was known as Troodontidae. In 1945, Charles Mortram Sternberg rejected the possibility that Troodon was a pachycephalosaur due to its stronger similarity to the teeth of other carnivorous dinosaurs. With Troodon now classified as a theropod, the family Troodontidae could no longer be used for the dome-headed dinosaurs, so Sternberg named a new family for them, Pachycephalosauridae.[4]

Comparison of troodontid teeth; A is the T. formosus holotype

The first specimens assigned to Troodon that were not teeth were both found by Sternberg in the early 1930s, in the Dinosaur Park Formation of Alberta. The first was named Stenonychosaurus inequalis by Sternberg in 1932, based on a foot, fragments of a hand, and some tail vertebrae. A remarkable feature of these remains was the enlarged claw on the second toe, which is now recognized as characteristic of early paravians. Sternberg initially classified Stenonychosaurus as a member of the family Coeluridae. The second, a partial lower jaw bone, was described by Gilmore (1932) as a new species of lizard which he named Polyodontosaurus grandis. Later, in 1951, Sternberg recognized P. grandis as a possible synonym of Troodon, and speculated that since Stenonychosaurus had a "very peculiar pes" and Troodon "equally unusual teeth", they may be closely related. Unfortunately, no comparable specimens were available at that time to test the idea. In a recent revision of the material by van der Reest & Currie, Polyodontosaurus was determined to be a nomen dubium, not fit for synonymy with other taxa.[5]

A more complete skeleton of Stenonychosaurus was described by Dale Russell in 1969 from the Dinosaur Park Formation, which eventually formed the scientific foundation for a famous life-sized sculpture of Stenonychosaurus accompanied by its fictional, humanoid descendant, the "dinosauroid".[6] Stenonychosaurus became a well-known theropod in the 1980s, when the feet and braincase were described in more detail. Along with Saurornithoides, it formed the family Saurornithoididae. Based on differences in tooth structure, and the extremely fragmentary nature of the original Troodon formosus specimens, saurornithoidids were thought to be close relatives while Troodon was considered a dubious possible relative of the family. Phil Currie, reviewing the pertinent specimens in 1987, showed that supposed differences in tooth and jaw structure among troodontids and saurornithoidids were based on age and position of the tooth in the jaw, rather than a difference in species. He reclassified Stenonychosaurus inequalis as well as Polyodontosaurus grandis and Pectinodon bakkeri as junior synonyms of Troodon formosus. Currie also made Saurornithoididae a junior synonym of Troodontidae.[7] In 1988, Gregory S. Paul went farther and included Saurornithoides mongoliensis in the genus Troodon as T. mongoliensis,[8] but this reclassification, along with many other unilateral synonymizations of well known genera, was not adopted by other researchers. Currie's classification of all North American troodontid material in the single species Troodon formosus became widely adopted by other paleontologists, and all of the specimens once called Stenonychosaurus were referred to as Troodon in the scientific literature through the early 21st century.

Teeth from South Dakota assigned to T. formosus, with a US dime coin for scale, Children's Museum of Indianapolis

However, the concept that all Late Cretaceous North American troodontids belong to one species began to be questioned soon after Currie's 1987 paper was published, including by Currie himself. Currie and colleagues (1990) noted that, while they believed the Judith River troodontids were all T. formosus, troodontid fossils from other formations, such as the Hell Creek Formation and Lance Formation, might belong to different species. In 1991, George Olshevsky assigned the Lance formation fossils, which had first been named Pectinodon bakkeri but later synonymized with Troodon formosus to the species Troodon bakkeri, and several other researchers (including Currie) have reverted to keeping the Dinosaur Park Formation fossils separate as Troodon inequalis (now Stenonychosaurus inequalis).[9]

In 2011, Zanno and colleagues reviewed the convoluted history of troodontid classification in Late Cretaceous North America. They followed Longrich (2008) in treating Pectinodon bakkeri as a valid genus, and noted that it is likely the numerous Late Cretaceous specimens currently assigned to Troodon formosus almost certainly represent numerous new genera, but that a more thorough review of the specimens is required. Because the holotype of T. formosus is a single tooth, this renders Troodon a nomen dubium.[3]

In 2017, Evans and colleagues further discussed the undiagnostic nature of the holotype of Troodon formosus and suggested that Stenonychosaurus be used for troodontid skeletal material from the Dinosaur Park Formation.[10] Later in the same year, Aaron J. van der Reest and Currie came to a similar conclusion as Evans and colleagues, and also split much of the material assigned to Stenonychosaurus into a new genus: Latenivenatrix.[5] In 2018, Varricchio and colleagues disagreed with Evans and colleagues, citing that Stenonychosaurus had not been used in the thirty years since Currie and colleagues synonymized it with Troodon, and they indicated that "Troodon formosus remains the proper name for this taxon".[11] This conclusion by Varricchio was agreed upon by Sellés and colleagues in their 2021 description of Tamarro.[12] Varricchio's comments were later addressed by Cullen and colleagues in their 2021 review of Dinosaur Park Formation biodiversity, where they noted that while Stenonychosaurus has indeed not been used for 30 years, Currie's original hypothesis of subjective synonymy (based on tooth and jaw morphology) was never directly tested, and given that later research found that teeth were not diagnostic below the family level in troodontids, Currie's original hypothesis is therefore not supported by the available data, regardless of the amount of time since it was originally proposed.[13] They suggested that the description of more complete skeletal material (i.e. containing dental, frontal, and postcranial elements) that can be tied to the holotype could allow the direct testing of the synonymy hypothesis, but re-affirmed that for now, given the lack of supporting evidence, the synonymy of Troodon and Stenonychosaurus cannot be maintained, and that merely remaining untested for 30 years is not sufficient justification to accept a proposed lumping of taxa lacking overlapping diagnostic materials.[13] However, Varricchio and other still insist on their naming method.