Ancient therizinosaur fossil unearthed in Britain, the oldest of its kind

"Previous research had suggested that the maniraptorans were around in the Middle Jurassic, but the actual fossil evidence was patchy and disputed."
Nergis Firtina
Late Cretaceous therizinosaurid Therizinosaurus cheloniformis.
Late Cretaceous therizinosaurid Therizinosaurus cheloniformis.

Wikimedia Commons 

By examining mystery teeth, machine learning models have been used to identify the oldest therizinosaur fossils discovered to date in Britain. For the first time, therizinosaurs have been recognized from fossils discovered in the UK thanks to their toothy remnants, which were discovered in Oxfordshire, Gloucestershire, and Dorset.

Researchers from the Natural History Museum and Birkbeck College trained computer models to recognize the mystery teeth using cutting-edge machine learning techniques, which pushed the origin of some of the group's members back by about 30 million years, according to NHM.

"Previous research had suggested that the maniraptorans were around in the Middle Jurassic, but the actual fossil evidence was patchy and disputed. Along with fossils found elsewhere, this research suggests the group had already achieved global distribution by this time," said Simon Wills, a Ph.D. student at the Natural History Museum who led the research.

"The teeth we analyzed include what are currently the only troodontid and therizinosaur fossils ever recorded from the UK and are the oldest evidence of these dinosaurs anywhere in the world," he added.

Machine-learning models can recognize isolated teeth

Therizinosaur, a sizable herbivorous dinosaur from the late Cretaceous, was distinguished by its long claw bones that resembled scissor blades. These prehistoric species were included in the most recent Jurassic World movie because of its distinctive appearance.

While earlier research has attempted to categorize isolated teeth using several statistical techniques, it hasn't always succeeded. The current study's authors have been trying to enhance this after demonstrating that machine-learning models can recognize isolated teeth from known species with high accuracy.

"The use of machine learning in vertebrate paleontology is still in its infancy, although its usage is growing," Simon says.

"The main drawback is the need for a comprehensive training dataset for the models to learn from. In our study we are fortunate that there is already a relatively large dataset of dinosaur tooth measurements available that we could use to train the models," Simon adds.

The study was published in Papers in Palaeontology.

Study abstract:

Papers in Palaeontology includes papers that document the diversity of past life and its distribution in time and space. As a sister publication to Palaeontology, its focus is on descriptive research, including descriptions of new taxa, systematic revisions of higher taxa, detailed palaeoecological, biostratigraphical and biogeographical documentation, and descriptions of significant floras and faunas from specific localities or regions. Submissions to Papers in Palaeontology should consider the broader implications and present the key questions and conclusions of the work in a wider context. In general, descriptions of single taxa, or local case studies primarily of interest to taxonomic or regional specialists, are likely too narrow in scope to be considered for the journal. Papers submitted to either journal will be considered for both, and authors will be advised if the editors consider the alternative to be more appropriate.