Torosaurus

Description

Torosaurus was a plant-eating dinosaur that lived a very long time ago, during the last part of the Late Cretaceous period, about 68 to 66 million years ago. Its fossils have been found across the Western Interior of North America, from Canada to the southern United States.

Torosaurus was a very big dinosaur, measuring between 7.5 and 9 metres (25 and 30 feet) long and weighing about 8 tonnes (18,000 pounds). It had the biggest skulls of any land animal known, with the frilled part of the skull reaching almost 3 metres (9.1 feet) long. Torosaurus looked similar in size to its close relative, the Triceratops, but had a longer frill with big oval holes in it. It also had several small bony points on the back of its frill and did not have the long nose horn that Triceratops prorsus had. Instead, it had a shorter nose horn, like the older form, Triceratops horridus.

Classification debate

In 2010, scientists questioned whether Torosaurus was a real, separate dinosaur. Some studies suggested it might just be the older, fully grown form of Triceratops, where the frill grew longer and developed holes as it matured. But more recent research supports the idea that Torosaurus is its own distinct genus, based on differences in skull shape and the discovery of intermediate forms.

Discovery and species

In 1891, two years after naming Triceratops, two dinosaur skulls with long frills that had holes were found by John Bell Hatcher in Niobrara County in southeastern Wyoming. The scientist who studied them, Othniel Charles Marsh, named them Torosaurus.

The name Torosaurus likely comes from a Greek word meaning "to perforate". This refers to the holes in the frill, which makes it different from the solid frill of Triceratops. Two species of Torosaurus have been identified: T. latus, meaning "the wide one", and T. utahensis.

Fossils of Torosaurus have been found in many places across North America, including Wyoming, Montana, South Dakota, North Dakota, Colorado, Utah, and Saskatchewan. Scientists think Torosaurus specimens are not as common in the fossil record as Triceratops specimens.

Description

Torosaurus was a very large dinosaur, about the same size as the biggest Triceratops. It measured between 7.5 and 9 meters long and weighed between 6 and 11 metric tons. One of its most special features was its huge skull, which could be up to 3 meters long—the largest skull of any land animal known to science.

Scientists have noticed that Torosaurus skulls can look different from each other. Some have big, curved brow horns, while others have shorter, straighter ones. The position of these horns can also vary. The nose horn can be straight and upright or just a small bump. The frill, or bony shield, at the back of the skull also changes in shape and size. Some frills curve upward, while others are flat or heart-shaped. Despite these differences, Torosaurus skulls are very similar to those of Triceratops.

Classification

In 1891, Marsh placed Torosaurus in the Ceratopsidae family of Ceratopsia, a group of herbivorous dinosaurs with parrot-like beaks. They lived in North America and Asia during the Jurassic and Cretaceous Periods.

Torosaurus has, with its long frill, traditionally been placed in Chasmosaurinae. It was seen as a later member of a line coming from Anchiceratops or Arrhinoceratops. It was placed in a different group from Triceratops because of its short frill. But in the 1990s, exact cladistic analysis showed that both were chasmosaurines. Recent studies show a close link between Torosaurus and Triceratops.

Possible synonymy with Triceratops

During the late 2000s and early 2010s, people debated if Torosaurus might be the same as Triceratops. In the Maastrichtian age of Laramidia, two closely related chasmosaurines lived in the same habitat. The only clear difference between them was the shape of the frill. No Torosaurus juveniles are known, but many Triceratops juveniles have been found. Triceratops differs from other chasmosaurines because adults keep a juvenile trait. Adults have short squamosals, a case of paedomorphosis.

In 2009, John Scannella studied dinosaur ontogeny in the Hell Creek Formation of Montana. He thought that Triceratops and Torosaurus could be growth stages of one genus. The Torosaurus specimens would be fully grown individuals of Triceratops. Torosaurus would be a junior synonym of Triceratops.

In 2010, Scanella and Jack Horner, Scannella's teacher at Montana State University, researched growth patterns in thirty-eight skull specimens from the Hell Creek formation. They concluded that Torosaurus is the grown form of Triceratops. Horner said that the frill of ceratopsian skulls is made of metaplastic bone. This bone can change size over time, growing longer or shorter. Significant growth happens even in skulls already known as Triceratops. Horner noted, "where the horn direction is backwards in juveniles and forward in adults". About 50% of all young Triceratops skulls have thin areas in the frill that match the "holes" in Torosaurus skull frills. These areas are surrounded by mature granular bone, suggesting they grow to balance the weight as Triceratops grow longer frills. Horner argued that many dinosaur species might be growth stages of other known species. With old Triceratops, the frill would grow longer, causing it to flatten and widen at the back. At the same time, parietal fenestrae would appear, resulting in the typical chasmosaurine frill shape.

Scannella and Horner said not all data fit their idea easily. One problem was that if Torosaurus were the last growth phase of Triceratops, called the "toromorph phase", Torosaurus fossils should be common, but they are rare. They explained this by higher death rates in young animals and the idea that old animals lived in high places where erosion stopped fossilization. A second problem was the size range of Torosaurus specimens, which seemed to show real Torosaurus young animals. They said the bone structure showed these were fully grown, and the size differences were just individual variation. A third problem was the lack of forms between those with and without parietal holes. These fenestrae are always perfectly shaped. They pointed to specimen USNM 2412, the main example of Nedoceratops, as an example of such a changing form. The problems of this genus would show its early change into a "toromorph". A last problem was the number of osteoderms on the frill edge. With Triceratops, there are usually five epiparietals, including a midline osteoderm. With Torosaurus, there are ten or twelve, with no midline epiparietal. Also the number of episquamosals on the side edge of the frill differs (five with Triceratops, six or seven with Torosaurus). This was explained by the idea that the number of epoccipitals increased during growth. It was also noted that both number and position of the osteoderms change with Triceratops. This is shown by specimen MOR 2923, which has six epiparietals, but no midline one.

Scannella and Horner's ideas have not been widely accepted. Some experts, while seeing the possibility that the "toromorph" idea is right, have said it is not likely. The idea was directly challenged by a 2011 paper by Andrew Farke and a 2012 paper by Nicholas Longrich. In 2011, Farke redescribed the problem Nedoceratops hatcheri as an old or sick individual of its own genus. Scannella and Horner said it was a Triceratops. Farke said the irregular holes in the Nedoceratops frill, far from cutting thin bone, were surrounded by thick swellings. Farke also said several facts were hard to match with the idea of a Triceratops turning into a Torosaurus. In general, the number of epoccipitals in ceratopsians does not increase when the frill grows. Even though the number of episquamosals often changes, there seems to be no link with size because some young already show the highest number. It seems to be a matter of individual variation, not growth. Likewise, with Ceratopia in general, the forming of holes in the frill is not linked to age, as even the youngest individuals often have the parietal fenestrae. Farke explained the thin bone areas on the frill of Triceratops, the supposed start of holes, to be where muscles attach. There would be no steady link between holes and a granular bone structure. Many Triceratops specimens have frills with a deeply veined surface, showing they are old. The bone of their frills would have to get new again and then become granulated for hole forming to start, which Farke thought unlikely. Finally, Farke noted that specimen YPM 1831, despite its huge size, seemed not fully grown, as shown by its unfused joints and smooth bone texture. So, it seemed to be a real Torosaurus young animal.

The same year, Scanella and Horner answered some of Farke's comments. They said that USNM 2412, because of its problems, was not the best example of a changing form, but they said that, apart from swellings, the holes in its frill were also bordered by granular and thinning bone. Considering all the evidence, they thought it much more likely that Nedoceratops was a sick individual of Triceratops than its own genus. They also pointed to Triceratops specimens showing the exact mix of veined, granular, and young striated bone that Farke had said was unlikely. The idea that the thin areas on Triceratops frills were muscle attachment sites was rejected because the bone at these points did not show the rough surface typical for such an attachment. For the difference in the number of epoccipitals, they gave two more reasons. The osteoderm tips of old individuals might have worn away in life so that each osteoderm looked like two. In this way, the normal number of five or six Triceratops epiparietals could have doubled to ten or twelve, exactly the amount seen with the Torosaurus specimens. Or, the Torosaurus latus specimens, found in older layers, might, in a process of anagenesis, show an early stage of Triceratops evolution. The oldest dated specimen, MOR 1122, has twelve epiparietals, while the younger MOR 981 has ten, seeming to show an evolutionary change in which the number of epiparietals slowly decreased.

In 2012, Longrich studied the problem using the idea of falsification. He said the "toromorph" idea suggested three such tests. First, if Torosaurus were the same as Triceratops, their fossils should be found in the same places. In fact, their areas do not fully match. In the far north, no Torosaurus fossils have been found, while from the south, only Torosaurus utahensis is known. However, this could be an artefact of the few Torosaurus remains and poor searching. Longrich said the idea was corroborated by the first test. Second, the idea said all Torosaurus specimens would be adults, while no Triceratops specimens would be very old. Longrich said this last point had not yet been proven. In 2011, Horner had published a study of bone showing that all Triceratops specimens studied had a young bone structure, but the group was too small to prove this for all Triceratops fossils. To test this better, Longrich suggested twenty-four outside skull traits, by which specimens could be checked for their level of skull fusion and their age. Thirty-six specimens were studied using these rules. It showed that fusion usually happened in a certain order, giving more information about their age. Indeed, by these rules, most Torosaurus specimens were very old. However, there were two exceptions. The small individual ANSP 15192 was a relatively young adult, as shown by the lack of fusion of the snout bones. The youngest specimen was YPM 1831, with an unfused snout, epijugal, and occipital condyle. Also, it had lost all its frill osteoderms because they had not yet fused, while the frill edge looked like growing, young bone. On the other hand, Longrich found that ten of the Triceratops skulls studied had reached the same level of growth as the oldest Torosaurus specimens. Longrich said the test of the second prediction did not support the idea. The third prediction was that changing forms could be found between Torosaurus and Triceratops. Longrich thought the claim that the thin areas on Triceratops frills were the start of parietal fenestrae, the strongest proof of a changing phase. However, he said these structures were in different places. The Triceratops areas are partly on the squamosal, while the Torosaurus holes are fully inside the parietal. Also, the areas are bordered by much thicker bone, while the Torosaurus holes are surrounded by thin bone. Longrich said the idea failed regarding the third prediction. Being wrong on two of the three predictions, the idea should be rejected.

Longrich also gave some more reasons against the "toromorph" idea. There are no changing forms known for the number of epiparietals. Also, it is hard to see how their number could have increased, as they took up the full frill edge and would, in a metaplastic growth, simply grow larger and the rest of the frill. The suggested splitting of osteoderms by wear has only been proven with episquamosals, never with epiparietals. Torosaurus has a squamosal that is thickened on the inside and curved on the outside, while the Triceratops squamosal is curved on the inside and flat on top. Changing forms are unknown. The Torosaurus squamosal is also, apart from size, much more stretched out. Longrich noted that when Torosaurus and Triceratops specimens are put together to make one growth series, as Scanella and Horner did, Torosaurus specimens ANSP 15192 and YPM 1831 were different from the regression line because their squamosals were more stretched than could be explained by allometry. Longrich said that while Horner, in his bone study, could only find Triceratops young animals, this was a clue, but offered the other idea that Triceratops was different from its relatives in keeping a young bone structure until old age. On the other hand, bone changing is not a good way to measure maturity because tests show that differences in the force on various bones can greatly change the rate or level of such changing and may make it seem like old bone. Longrich expected that Scanella and Horner would say his second test of their idea was caused by individual variation. But Longrich said the importance of this was small: e.g. the size difference between ANSP 15192 and YPM 1831 was better explained by sexual dimorphism, the first possibly being a young adult female and the second being a young male.

In 2013, Farke and Leonardo Maiorino published morphometric research, a statistical study of the shape space describing the change of the Torosaurus, Triceratops horridus, Triceratops prorsus, and Nedoceratops skulls linked to growth. They concluded that Torosaurus latus skulls kept a different shape from T. horridus and T. prorsus through growth, even when the frill shape was left out. Nedoceratops proved, except for size, not to be a possible changing form between Torosaurus and Triceratops horridus. Farke and Maiorino said that the small number of Torosaurus specimens made these results less certain, but concluded that Torosaurus and Triceratops were different groups, though they allowed for the idea of anagenesis, the several groups forming one single chronospecies line of descent, because of the lack of good rock layer data.

The idea that the Torosaurus latus specimens might show a "toromorph" phase of Triceratops growth has caused the question of whether the second Torosaurus species, Torosaurus utahensis, is a "toromorph" as well. This issue has been made harder by the lack of good fossil material, as most specimens are single bones. T. utahensis was mainly put into Torosaurus because of stretched squamosals, showing a long frill. The number of epiparietals and the size, place, or even existence of parietal fenestrae are unknown. Researchers have said that distinct young Torosaurus have been found from a bone group in the Javelina Formation of Big Bend National Park, saying they are Torosaurus cf. utahensis because of their place near an adult with a typical Torosaurus parietal. Scanella and Horner said that only future finds could solve this problem. They suggested that this group, which extends the Torosaurus range south of that of Triceratops, might be a separate chasmosaurine group or a third Triceratops species. Farke's 2013 morphometric study was not clear on this point, with T. utahensis shape space falling between Triceratops and Torosaurus latus and not well apart from either.

In 2022, Mallon et al. said that two specimens found in Canada's Frenchman and Scollard Formations, EM P16.1. (at Eastend Historical Museum in Saskatchewan) and UALVP 1646 (at the University of Alberta), are young and can be called Torosaurus, showing that it is a real group. The same study also noted that Torosaurus lived during the Late Maastrichtian.