| Eolambia | |
|---|---|
 | |
| Skull reconstruction (A) and life restoration (B) | |
Scientific classification | |
| Kingdom: | Animalia |
| Phylum: | Chordata |
| Class: | Reptilia |
| Clade: | Dinosauria |
| Clade: | †Ornithischia |
| Clade: | †Ornithopoda |
| Superfamily: | †Hadrosauroidea |
| Clade: | †Hadrosauromorpha |
| Genus: | †Eolambia Kirkland,1998 |
| Type species | |
| †Eolambia caroljonesa Kirkland, 1998 | |
Eolambia (meaning "dawnlambeosaurine") is agenus of herbivoroushadrosauroiddinosaur from the earlyLate Cretaceous of theUnited States. It contains a single species,E. caroljonesa, named bypaleontologistJames Kirkland in 1998. Thetype specimen ofEolambia was discovered by Carole and Ramal Jones in 1993; the species name honors Carole. Since then, hundreds of bones have been discovered from both adults and juveniles, representing nearly every element of the skeleton. All of the specimens have thus far been found inEmery County,Utah, in a layer of rock known as the Mussentuchit Member of theCedar Mountain Formation.
Measuring up to 6 meters (20 ft) long,Eolambia is a large member of its group. While it closely approaches the Asian hadrosauroidsEquijubus,Probactrosaurus, andChoyrodon, in traits of the skull,vertebrae, and limbs, it may actually be more closely related to the North AmericanProtohadros. This grouping, based on the straightness of thequadrate bone andscapula, would represent an isolated,endemicradiation of hadrosauroids. Despite resemblinghadrosaurids – lambeosaurine hadrosaurids in particular – in several features, leading to its initial identification as one of them, these similarities have been rejected as either entirelyconvergent or misinterpreted.
Eolambia would have lived in a forested environment at the edge of lakes in a humidfloodplain environment, feeding ongymnosperms,ferns, andflowering plants. The water levels in the lakes changed over time with cyclical wet and dry spells caused by theprecession of the Earth, reflected by alternating bands in the sediments of the Mussentuchit Member. As a juvenile,Eolambia would have been preyed upon by largecrocodylomorphs residing in the lake waters. With increasing age, however, they became impervious to the crocodylomorphs, and mature individuals (at least eight to nine years in age) were preyed on by largetheropods such as theneovenatoridSiats.
In 1979Peter Galton andJames A. Jensen described a fragmentary right femur,BYU 2000, belonging to ahadrosaurian dinosaur discovered in sediments belonging to theCedar Mountain Formation inArches National Park,Utah. Though poor material, it was important for it (alongside a second North American femur described in the paper) was the first hadrosaur specimen from theLower Cretaceous anywhere in the world. Galton and Jensen hypothesized more complete remains of a hadrosaur may be found from the formation in the future.[1] Various hadrosauroid teeth had also been found in quarries of small vertebrates in the western region of theSan Rafael Swell, nearCastle Dale inEmery County, Utah; they were described in 1991 by J. Michael Parrish.[2] Subsequently, in 1993, Carole Jones and her husband Ramal Jones discovered fragmentary bones in a fossil site located in the northwestern region of the Swell. They brought the site to the attention of Donald Burge, director of the institution then called theCollege of Eastern Utah Prehistoric Museum (CEUM). The site, which is formally known as CEUM Locality 42em366v, would subsequently be named Carol's Site (sic) in her honour. The fossils, stored under the specimen number CEUM 9758, represent the partial remains of an adult hadrosauroid, including parts of the skull, vertebrae, ischium, and leg. CEUM 5212, a partial skull and forelimb from an adult, was found nearby in CEUM Locality 42em369v.[3] CEUM 8786, a left femur from an adult, was discovered later in Carol's Site, and was not described until 2012.[4]
At the same time, theOklahoma Museum of Natural History (OMNH) had recovered remains belonging to the same hadrosauroid from excavations in the southwestern region of the Swell.[3] Specimens discovered by the OMNH initially consisted of six individuals from five localities: two juvenile skeletons,[5] including vertebrae, scapulae, an ulna, an ilium and ischium, and hindlimb elements from locality OMNH v237; a partial juvenile skeleton, including parts of the skull, forelimbs, and hindlimbs along with a dorsal vertebra from OMNH v824; OMNH 27749, a sacrum and ischium from OMNH v696; OMNH 24389, an ischium from OMNH v214; and OMNH 32812, a partial skeleton including a scapula, two caudal vertebrae, and other unexcavated elements from OMNH v866. The OMNH also made further discoveries of the hadrosauroid's teeth in various quarries. Richard Cifelli of the OMNH allowedJames Kirkland to study the collected specimens.[3] Kirkland went on to preliminarily describe the specimens in 1994 in a conference abstract at the annual meeting of theSociety of Vertebrate Paleontology.[6]
In a 1998 paper published in the bulletin of theNew Mexico Museum of Natural History and Science, Kirkland formally named the specimens as a newgenus,Eolambia, with thetype and only species beingEolambia caroljonesa. The generic name combines theGreek prefixeos/ἠώς ("dawn", "morning", implying "early") with the suffixlambia (derived fromLambeosaurus, which is in turn named afterCanadianpaleontologistLawrence Lambe). In all, the genus name means "dawn (or early)lambeosaurine", in reference to its supposed position as the most basal lambeosaurine. Meanwhile, thespecific name honors Carole Jones. The nameEolambia was suggested bypaleoartistMichael Skrepnick; it replaced the informal name "Eohadrosaurus caroljonesi", which was used by Kirkland before the 1998 paper.[3] However, the epithetcaroljonesa, following theICZN, is incorrect and in this case the correct spelling would becaroljonesae.[7]
Additional excavations since 1998 have revealed additional skeletal elements belonging toEolambia. It is now the most completely-known iguanodont from between theCenomanian andSantonian stages. Nearly every single skeletal element belonging to the taxon is represented by multiple fossils across existing collections. At least twelve individuals, including at least eight juveniles,[8] were discovered in a quarry south of the town ofEmery, including parts from both the skull and the rest of the body. The quarry is known as the Cifelli #2 ''Eolambia caroljonesa'' Quarry, or more formally CEUM locality 42em432v. Another quarry south of the town of Emery – the Willow Springs 8 quarry, or CEUM locality 42Em576v – preserves at least four juveniles. Material from these quarries was described in 2012 by Kirkland along with Andrew McDonald, John Bird, and Peter Dodson, who also provided an updateddiagnosis forEolambia based on this material.[4]
One specimen discovered at the Cifelli #2 quarry, a right dentary catalogued as CEUM 34447, is unusual compared to otherEolambia specimens, including both adults and juveniles. While the ratio between the mid-point depth and overall length of the dentaries in other individuals generally ranges from 0.19 to 0.24, the same value in CEUM 34447 is 0.31, making it unusually deep. The dentary also does not deepen substantially at the front end, and thus lacks one of the distinguishing traits ofEolambia dentaries. However, the other distinguishing trait, which is the expansion of the front end of the coronoid process, is present in the specimen. Although it is possible that this individual represents an unusualEolambia specimen, McDonald and colleagues cautioned that it may represent a distinct hadrosauroid that differs in themorphology of the dentary fromEolambia (albeit one that is not suggested by any of the other materials at the quarry). Thus, they regarded it as an indeterminate basal hadrosauroid.[4]
TheField Museum of Natural History (FMNH) also conducted excavations south of the town of Emery from 2009 to 2010, with permits from theUtah Geological Survey. These excavations were conducted in a site discovered by Akiko Shinya in 2008, FMNH locality UT080821-1, which has been named Akiko's Site in her honor. During the excavations, 167 disarticulated but closely associated elements were collected in two plaster jackets. One of these plaster jackets, containing elements from the hip and vertebrae, was given the specimen number FMNH PR 3847; the specimen was subsequently described by McDonald along with Terry Gates, Lindsay Zanno, and Peter Makovicky in 2017. Another locality discovered by the FMNH is the Triple Peak quarry, FMNH locality UT130904-2, which contains a bonebed of over 400 juvenile bones that belong to at least four individuals.[9]
All of the localities thatEolambia were discovered in belong to the Mussentuchit Member of theCedar Mountain Formation. Predominant rocks in the Mussentuchit consist of interleaved light gray to gray layers ofmuddysandstone andmudstone. Deposits representing two separateriver channel systems are also present in the Mussentuchit. The lower of these systems consists of four layers, respectively of muddy sandstone, fine-grained sandstone,silty mudstone, and fine-grained sandstone, with the latter two layers bearing plant debris. The Cifelli #2 Quarry is associated with this system. The upper system consists of greenish fine-grained sandstone. Three layers ofbentoniteclay layers, representingvolcanic ash deposits, are also present in the Mussentuchit, being respectively whitish-gray, yellow, and tan in color. The middle of these layers cuts through the Cifelli #2 Quarry.[8] Bentonite is also present within the sandstone itself throughout the Mussentuchit.[9] In the Cifelli #2 quarry, the bones themselves are blackened, having been replaced bycalcite,quartz,pyrite, andcarbonate-hydroxyapatite.[8]
Palynology (the study of pollen),[10][11][12] as well as studies of fossilmegaflora and invertebrates,[13] had initially suggested that the uppermost Cedar Mountain Formation – including the Mussentuchit – dated to theAlbian stage.[3] A 1997argon-argon dating ofsanidine crystals recovered from a volcanic ash layer in the upper Cedar Mountain Formation by Cifelli and colleagues indicated an age of 97 ± 0.1 million years, which was then considered to be just after the boundary between the Albian and Cenomanian stages. This is similar to the early Cenomanian age reported for the overlyingDakota Formation.[14] Subsequently, in 2007, James Garrison and colleagues dated the volcanic ash layer passing through the Cifelli #2 Quarry – which is likely the same layer dated previously by Cifelli and colleagues – to 96.7 ± 0.5 million years ago. At an approximate level, the Cifelli and Garrison estimates agree with each other, thus indicating a Cenomanian age forEolambia. Garrison and colleagues also re-analyzed the pollen assemblage from localities throughout the Mussentuchit, concluding that the assemblage spans the Albian-Cenomanian boundary.[8]
Eolambia is a largehadrosauroid. Initial estimates placed the length of its skull at 1 meter (3 ft 3 in) in length,[3] although this was due to a disproportionately long snout that was later corrected by the discovery of additional material.[4] In 2016,Gregory S. Paul estimated a body length of 6 meters (20 ft) and a weight of 1 metric ton (0.98 long tons; 1.1 short tons) forEolambia,[15] which agrees with a prior body length estimate of 6.1 meters (20 ft) byThomas R. Holtz Jr. in 2012.[16] Earlier, in 2008, an adult specimen was estimated as having a length of 5.2 meters (17 ft) and a height at the hip of 2 meters (6 ft 7 in).[8]
The crestless skull ofEolambia has a similar overall shape to those ofEquijubus andProbactrosaurus. The front of the snout is highly roughened, being punctuated by manyforamina (openings). At the tip of eachpremaxilla, there are two tooth-like structures known as denticles, which is also seen in its closest relativeProtohadros. Further back, the rear portion of the lower branch of the premaxilla abruptly projects upwards, closing off the nostril at the rear as inProbactrosaurus,Protohadros, and other hadrosauroids. This part joins with the two finger-like processes of themaxilla, which is similar toProtohadros. The body of the maxilla itself does not bear a recess or any indication of anantorbital fenestra, likeEquijubus,Protohadros, and other hadrosauroids. One of the characteristics used to distinguishEolambia is the concave profile of the tooth row of the maxilla when viewed from the side, which is likeEquijubus,Probactrosaurus, and several otherhadrosauriforms but unlikeProtohadros.[4]
LikeProbactrosaurus and other hadrosauroids, the back of the maxilla connects to thejugal – which borders the bottom of the eye socket andinfratemporal fenestra – through a finger-like projection that fits into a recess. The bottom margin of the jugal bears a strong flange beneath the level of the infratemporal fenestra; this is also seen inEquijubus,Probactrosaurus,Protohadros, and several other hadrosauroids. Connecting to the jugal from above is thepostorbital, which has a roughened surface where it borders the eye sockets (likeProtohadros), but the side of the bone is otherwise smooth. At the back of the skull, thequadrate articulates with thesquamosal with a joint that is D-shaped when viewed from the top. The left and right squamosals would have contacted each other extensively, being only separated at the back by a small process of theparietal. Thesupraoccipital bone, which forms the top portion of the back of the skull, is flat and nearly vertical, as is the case inProbactrosaurus and other hadrosauroids.[4]\
As with the premaxilla, thepredentary ofEolambia bore denticles. There is a prominent dorsomedial process, a tab-like structure also seen inProbactrosaurus and other hadrosauriforms. Several additional tab-like denticles were present on either side of the dorsomedial process, which are likewise present inProbactrosaurus. The predentary is joined at the back by thedentary, which constitutes most of the lower jaw. There is a short recess, ordiastema, between the articulation of the predentary with the dentary and the first tooth position on the dentary, which is observed inEquijubus,Probactrosaurus, and other hadrosauroids. The front of the dentary characteristically deepens, as inProtohadros,Ouranosaurus, andBactrosaurus. Two bulges are present on the outer surface of the dentary, one of them representing thecoronoid process as inProbactrosaurus and other hadrosauroids. LikeProtohadros and several other hadrosauriforms but unlikeProbactrosaurus, only the front end of this process is thickened. As inEquijubus,Probactrosaurus, and various otheriguanodonts, a small foramen is present on the side of thesurangular, which is located behind the dentary.[4]
Similar to other iguanodonts, the teeth ofEolambia are arranged in tightly-spaced and interlocking rows. At any given time, each of the 32 maxillary tooth sockets holds three teeth, while each of the 30 dentary tooth sockets holds four teeth. Out of these, two of the teeth in each socket are replacement teeth like those ofProbactrosaurus; the others are active teeth. Every active tooth has onewear facet. LikeProbactrosaurus,Gongpoquansaurus,Protohadros, and other hadrosauroids, each maxillary tooth crown has only one ridge, which is slightly offset towards the midline of the mouth. Meanwhile, each dentarytooth crown characteristically bears a primary ridge, and an accessory ridge closer to the midline of the mouth, a condition which is also present inProtohadros and other hadrosauroids. All of the crowns also bear small, denticle-like serrations on the front and rear edges, which is also seen inProbactrosaurus.[4]
LikeEquijubus,Probactrosaurus,Gongpoquansaurus, and otherstyracosternans,Eolambia hascervical vertebrae which are veryopisthocoelous, meaning that their front ends are strongly convex while their back ends are strongly concave. Unusually, the front end of the third cervical is set slightly higher than the back end, which is not seen in any other cervical vertebrae. Vertebrae from further back in the neck have a greater angle between the elongatearticular processes known as the postzygapophyses, and also more elongatetransverse processes. On the transverse processes, there are further articular processes – the rounded parapophyses and rod-like diapophyses – are located. In the rearmost cervicals, there is also a deep depression separating the postzygapophyses, and theneural spines have steeply-angled front margins, which has the effect of creating prominent spike-like projections.[4][9]
The first fewdorsal vertebrae are similar to the cervical vertebrae, but have taller and more prong-like neural spines. The rest are amphiplatyan, meaning that they are flat at both ends. They also have postzygapophyses which are less elongate and more pedestal-like, in addition to taller and more rectangular neural spines. The parapophyses, which are depressions instead of projections, have moved off from the transverse processes to theneural arch, between the transverse processes and the articular processes known as the prezygapophyses at the front of the vertebrae. However, they move back onto the base of the transverse processes in the last few dorsals.[4] In the third or fourth dorsal, the parapophysis is located very close to thesuture between the neural spine and thecentrum, which is unlike the other dorsals but similar to the first few dorsals inEdmontosaurus. Also likeEdmontosaurus, the median ridge separating the prezygapophyses become more pronounced in the rear dorsals.[9]
Among theEolambia specimens found to date, the best-preservedsacrum includes seven vertebrae. Given that this individual is immature, and hadrosaurs increase their sacral vertebra count with age, adults may have had more sacrals.[3] The caudal (tail) vertebrae areamphicoelous, having both the front and back ends being concave. They have pedestal-like prezygapophyses and tab-like postzygapophyses, the latter of which are separated by a depression. The prezygapophyses lengthen to become stalks and the postzygapophyses shrink in the middle and rear caudals. Furthermore, the neural spines transition from rectangular to strongly curved (concave in front and convex behind), the centra become more elongated, and the transverse processes disappear as well.[4][9]
LikeProbactrosaurus, thescapular blade ofEolambia is nearly rectangular, with the sides of the blade being straight and meeting the end of the blade at an almost-right angle. Thehumerus is bowed towards the midline, and the deltopectoral crest on the humerus is prominently thickened. A knob is present on the outer surface of the joint with theradius on the humerus. Both theulna and the radius are relatively straight bones. The top of the ulna bears three "prongs", namely theolecranon process and two additional triangular flanges. Meanwhile, the radius bears a flattened inner surface at the top end, and a distinct platform on the inner surface at the bottom end, both for articulating with the ulna. The hand has five digits. Among themetacarpal bones of the hand, the second is curved, the third is straight, and the fourth is strongly curved. LikeIguanodon,Probactrosaurus, and otherbasal iguanodonts, the claw on the first digit is distinctively conical. The second and third claws are hoof-like, with the third claw bearing roughened flanges on its sides – a trait also shared with other iguanodonts.[4]
The preacetabular (i.e. located in front of thehip socket) process of theilium bears a horizontal "boot" which forms an obtuse angle with the bottom of the process. Additionally, the top edge of the bone projects outwards to form a rim located above the ilium-ischium articulation on the bottom edge of the bone. These traits are also seen inProbactrosaurus and other iguanodonts.[4] The hip socket itself is quite large relative to the ilium for an iguanodont.[5] A distinctive flange is present along the top surface of the ilium, a trait which is shared with an unnamed hadrosauroid from theWoodbine Formation. The postacetabular (i.e. located behind the hip socket) process of the ilium tapers at its rear with no break in its top margin. On thepubis, the forward-projecting process expands towards the front, likeProbactrosaurus,Gongpoquansaurus, and other iguanodonts. The shaft of the ischium is characteristically straight in adults (juveniles have a curved shaft, which is more common among iguanodonts), and the bottom of the shaft bears a compressed "boot" that is expanded forwards.[4][9]
Like the humerus, thefemur is bowed, but outwards. However, characteristically, the bottom half of the femur is straight. Thefourth trochanter, located halfway down the femur, is vertically tall and triangular, similar toProbactrosaurus,Gongpoquansaurus, and other iguanodonts but dissimilar to more basalornithopods. Like the ulna and radius, thetibia andfibula have straight shafts, and they articulate with each other; a roughened projection on the bottom of the fibula fits between two bulb-likecondyles located just behind thecnemial crest on the bottom end of the tibia. The top of the tibia is rough, indicating the presence of a cap ofcartilage between the femur and tibia. The foot has four digits. Like the metacarpals, the secondmetatarsal is curved, the third is straight, and the fourth is strongly curved. Thephalanges of the foot are very stout, with the terminal phalanges near the end of each digit being particularly stout. As with the second and third claws on the hand, the second, third, and fourth claws of the foot are uniformly broad, flat, and hoof-like.[4]
In Kirkland's initial description ofEolambia, he considered it to be a member of theHadrosauridae, as defined byDavid B. Weishampel,David B. Norman, and Dan Grigorescu in 1993. Weishampel and colleagues used seven unifying characteristics to define the Hadrosauridae: the upward expansion of the ascending process of the maxilla; the absence of the paraquadrate foramen, which separates the quadrate andquadratojugal; the location of theangular on the inner surface of the lower jaw; the absence of the surangular foramen on the surangular; the narrow teeth of the maxilla; the presence of three or more teeth in each dentary tooth position; and the reduction of the top margin of the scapular blade.[17] The first, fifth, sixth, and seventh of these traits were recognized inEolambia, with the rest being unknown due to missing material. Kirkland further assignedEolambia to the Euhadrosauria, defined by Weishampel and colleagues to include the common ancestor ofHadrosaurinae (now the Saurolophinae) and Lambeosaurinae – the two primary branches of hadrosaurids – and all of its descendants.[17] This assignment was based on five characteristics, all present inEolambia: the presence of denticles on the premaxilla; the quadrate's narrow joint with the lower jaw; the narrow dentary teeth; the presence of a single edge, or carina, on each dentary tooth; and the angled deltopectoral crest on the humerus.[3]
Within the Hadrosauridae, Kirkland further consideredEolambia to either be a basal member of the Lambeosaurinae, or thesister group of Lambeosaurinae. He identified five characteristics shared with the Lambeosaurinae, as defined across various studies:[17][18][19][20] the absence of a foramen on the premaxilla; the (at least partial) enclosure of the nostril by the premaxilla; the development of a shelf on the maxilla; the very tall neural spines of the caudal vertebrae; the robustness of the humerus; and the large "boot" of the ischium in adults. Kirkland found that two additional traits separated the crestlessEolambia from other, crested lambeosaurines, which were considered as being related to the development of the crest: the elevation of thenasal cavity above the eye socket, and the shortening of the parietal. IfEolambia were to be recognized as a lambeosaurine, then the Lambeosaurinae would have to be redefined to exclude those two traits. Alternatively, ifEolambia was instead the sister group of the Lambeosaurinae, then it would represent a morphology in hadrosaurids that is close to the divergence between hadrosaurines and lambeosaurines. This possibility is supported by the presence of a groove on the bottom of the sacrum ofEolambia, which was recognized by Kirkland as a defining trait of hadrosaurines[19][20] – albeit oneconvergently present in theAnkylosauria andCeratopsia.[3]
To support the lambeosaurine affinities ofEolambia, Kirkland conducted aphylogenetic analysis; thetree recovered by his analysis, which accordingly placesEolambia as the most basal lambeosaurine, is reproduced below. Despite the evidence he cited in support of this position, Kirkland also noted a seemingly inconsistent trait. As observed in the juvenile ilium he initially described, the antitrochanter – a projection on the ilium that limits thegreater trochanter of the femur – is more poorly developed than would be expected for a basal hadrosaurid. Michael Brett-Surman noted that adults may have had a better-developed antitrochanter. On the other hand, thederived lambeosaurineHypacrosaurus has a well-developed antitrochanter even at the hatchling stage. Kirkland noted the possibility that basal hadrosaurids developed their antitrochanter later in life than derived hadrosaurids such asHypacrosaurus.[21] He thus remarked that adultEolambia specimens would resolve this issue.[3]
In 2001, Jason Head re-evaluated the phylogenetic position ofEolambia based on specimens stored at the OMNH. He recognized Kirkland's identification of supposed lambeosaurine distinguishing traits inEolambia as either erroneous or dubious. As demonstrated by Head, there is actually a foramen on the premaxilla; the premaxilla only borders the nostril on the bottom edge; there are two finger-like processes at the front of the maxilla in addition to the shelf; high neural spines of the caudal vertebrae also occur in the non-lambeosaurinesOuranosaurus andBactrosaurus; and the expansion of the tip of the ischium also occurs inCamptosaurus,Iguanodon,Ouranosaurus,Gilmoreosaurus, andBactrosaurus, indicating that the expansion isplesiomorphic for iguanodonts and was secondarily lost in hadrosaurines.[20] While M.L. Casanovas and colleagues in 1999 characterized lambeosaurines as having an "expanded foot" at the tip of the ischium, as opposed to a "club-like" condition,[22] Head argued thatCorythosaurus andLambeosaurus[23] have ischial tips no more expanded than those ofMantellisaurus orOuranosaurus.[5][24]
Head also demonstrated that the teeth ofEolambia were broader and more asymmetrical than Kirkland had assumed; he also noted that subadults only possessed two teeth per tooth position, unlike hadrosaurids,[5] although adult specimens in fact have the three or more tooth positions as originally demonstrated by Kirkland.[4] Head further showed thatEolambia possessed seven sacral vertebrae, unlike eight or more in derived hadrosaurids, and that its antitrochanter was actually more poorly developed than those ofOuranosaurus andAltirhinus.[25] Examination of the OMNH specimens further demonstrated that the postacetabular region of the ilium did not form a distinct process, and the articulations at the bottom of the femur are not very prominent, both of which separateEolambia from more derived iguanodonts such as hadrosaurids. Thus, Head concluded thatEolambia was abasal member of the Hadrosauroidea, outside of the Hadrosauridae, a conclusion supported by a phylogenetic analysis finding it to be a non-hadrosaurid hadrosauroid closely related toProbactrosaurus.[5]
In 2004, Norman includedEolambia for the first time in a comprehensive phylogenetic analysis of iguanodonts, as part of a chapter focusing on iguanodonts in the second edition ofThe Dinosauria. He found thatEolambia was the sister group toAltirhinus, with both of them being more basal thanProtohadros,Probactrosaurus, or the Hadrosauridae.[26] Within the same volume, Horner, Weishampel, andCatherine Forster conducted a separate phylogenetic analysis of hadrosaurids, likewise includingEolambia. They foundEolambia again as a non-hadrosaurid hadrosauroid, albeit as the sister group toProtohadros. Although Head had specifically redefined the Hadrosauridae, based on shared characteristics, to includeProtohadros,[5][27] Horner and colleagues adapted a taxon-based definition that excludedProtohadros and thusEolambia. They also identified additional characteristics differentiatingEolambia from hadrosaurids: there are coarse denticles on the teeth of the dentary, and the coronoid process is weakly expanded.[28]
Variance in recovered phylogenetic positions forEolambia persisted in the following years. In the 2009 description ofLevnesovia,Hans-Dieter Sues and Alexander Averianov found thatProtohadros occupied an intermediate position relative toAltirhinus andProbactrosaurus, being the sister group ofFukuisaurus.[29] Meanwhile, in accordance with the results of Horner and colleagues, Albert Prieto-Márquez andMark Norell found in 2010 thatEolambia was instead the sister group ofProtohadros. He also noted features ofEolambia that were convergent upon hadrosaurids: the presence of a single tooth carina; the nearly-square outer corner of the predentary; and the midpoint of the quadratojugal notch (which articulates with the quadratojugal) being located roughly halfway up the surface of the quadrate (being located less than 60% of the bone's height down from the top).[30] In 2012, Holtz classifiedEolambia as a primitive member of theStyracosterna along withAltirhinus.[16]
In 2012, McDonald conducted a phylogenetic analysis of iguanodonts incorporating data from newEolambia specimens. Contrary to previous authors, he found instead thatEolambia was the sister group ofProbactrosaurus. This close relationship was based on two shared characteristics: the quadrate being straight save for the top end, which is curved backwards; and the top and bottom margins of the scapular blade being nearly parallel.[4] Wenjie Zheng and colleagues addedJintasaurus to the group in 2014,[31] while José Gasca and colleagues recovered the original group in 2015.[32] However, McDonald extensively revised his phylogenetic dataset further in preparation for the 2017 description of the FMNH specimens. As with Horner and colleagues, as well as Prieto-Márquez and Norell, McDonald and colleagues found thatEolambia was the sister group ofProtohadros, with both of them being placed as basalhadrosauromorphs (results reproduced below).[9] Using the 2012 version of the McDonald dataset, Francisco Verdú and colleagues independently recovered a similar result,[33][34] withEolambia being closely related toProtohadros andJeyawati; however, McDonald and colleagues found the latter to be more derived in 2017.[9]
In 2018, Terry Gates and colleagues described the new iguanodontian genusChoyrodon, from the same locality asAltirhinus, considered in the past a relative ofEolambia. Their phylogenetic analysis, based on the matrix of Norman (2015), foundChoyrodon to be the sister taxon ofEolambia.Eolambia's commonly recovered relativeProtohadros was not present in this matrix. Three characteristics supported this sister relationship: a trapezoidally-shapedocciput, and two shared traits of the teeth. The placement of these genera relative to other hadrosauriformes depended on whether the antorbital fenestra was coded as present or absent inChoyrodon; the juvenile type specimen has one, but the possibility that adults would have a closed one could not be ruled out.[35]
Eggshells possibly belonging toEolambia have been discovered at various localities alongside isolated teeth. They have reticular (net-like) surface patterns, and have a thickness of 2 millimeters (0.079 in).[3]
Being known from a variety of juvenile and adult specimens, the changesEolambia underwent as it grew are well-documented. The number of teeth in the maxilla increased with age, from 23 in a juvenile to 33 in an adult. A similar increase occurred in thedentary teeth, from 18–22 in juveniles to 25–30 in adults. While juveniles lack secondary ridges on the crowns of their dentary teeth, faint secondary ridges are present in adults.[4] The size of the "boot" of the ischium also increased with age, as documented by Kirkland.[3] Changes may also have occurred in the straightness of the shaft of the ischium, as documented by specimens discovered by the FMNH.[9]
Results from ahistological analysis performed on a rib from specimen FMNH PR 3847 were reported by McDonald and colleagues in 2017. Extensivebone remodeling has occurred in the internal matrix of the rib. Themedullary cavity, which houses themarrow, is marked by the presence of many cavities, orlacunae. On the outer surface of the bone,osteons have almost completely replaced the interior of the bone; the layered growth of the bone is more visible on the interior of the bone, where fivelines of arrested growth (LAGs) are visible. A sixth LAG is possibly present within the medullary region, but is relatively limited in extent, which makes this identification questionable.[9]
Since LAGs can be obliterated by bone remodeling, this represents an underestimate of the true number of LAGs that had actually developed; McDonald and colleagues estimated three or four missing LAGs by examining the spacing between the visible LAGs. This indicates that the animal was eight to nine years old when it died. The spacing between the LAGs decreases significantly towards the outer surface, indicating that growth was slowing, but the lack of the external fundamental system (a series of tightly-spaced LAGs) indicates that growth had not yet ceased. This is consistent with the unfused sutures between the centra and neural spines in the cervical and dorsal vertebrae, and the unfused sutures between the sacral centra and their corresponding ribs.[36] Overall, the growth rate ofEolambia appears to have been comparable to that ofHypacrosaurus andMaiasaura, which reached maturity after seven or eight years.[9][37][38]
Sedimentological data from the Mussentuchit Member indicates thatEolambia lived on a poorly-drainedfloodplain, around a system of large, perennial lakes. The lakes fluctuated between high and low water levels inMilankovitch cycles of roughly 21,000 years, caused by the Earth'saxial precession. These cycles is reflected by the alternating layers of muddy sandstone and mudstone in present-day rock deposits. During dry periods, the exposed lakebed formed a broad beach devoid of plants. Thecracked mud of the lakebed producedmud clasts, while organic debris and scavenged carcasses were scattered around the water's edge as the lake progressively retreated. Meanwhile, during periods in which thesea level rose, small rivers invaded the floodplain, whichreworked and displaced the debris that had been deposited during the dry season.[8]
Most of the fossils discovered in the Mussentuchit are scattered and disarticulated, with intact skeletons being relatively rare. Although the lakes did not actively destroy bones, it was not particularly conducive to good preservation either due to the long duration of the burial process. Before burial, scavengers could have disrupted the skeletal material, which is evident in the severedEolambia tibiae bearing tooth marks discovered in the Cifelli #2 quarry. The regression of the lake could have occasionally createdbogs, but they would not have been permanent enough to entrap and preserve larger vertebrates. Action by water currents seems to have played a role in the deposition ofEolambia bones in the Cifelli #2 quarry, where the bones are largely deposited at 28°, 69°, 93°, 131° and 161°azimuth (i.e. relative to north). These directions would have been parallel or sub-parallel to the edges of the lake, indicating orientation by lake currents, except for the 69° and 93° azimuths, which probably represent river currents.[8]
The abundant plant debris, combined with the absence ofsink (playa) deposits, indicates that the local climate was relatively humid during the deposition of the Mussentuchit. This is in marked contrast to earlier deposits in the area, which arecalcareous (chalky), indicating periods of aridity lasting up to 10,000 years.Paleoclimatic reconstructions support these interpretations, with the encroachment of the Mowry Sea from the north (a process which would eventually form theWestern Interior Seaway) resulting in arid (Mid-latitude Continental Interior) climates being replaced by more humid (Mid-latitudeEast Coast) climates.[39] With a paleolatitude of 46° N, the Mussentuchit would have had a warm mean surface temperature of 40–45 °C (104–113 °F), about 6–8 °C (43–46 °F) warmer than the present day, according togeneral circulation models.[8][40]
Crocodylomorphs are abundant in the Mussentuchit deposits, being represented largely by teeth. The largest known teeth have been referred to thePholidosauridae;[a] they are broad with rounded bases, and some of them bear longitudinal ridges.[9] These large crocodilians would have been capable of preying upon at least juvenileEolambia at the water's edge. The abundance of juvenileEolambia specimens in the lakeside fossil material supports this hypothesis; largerEolambia would have been invulnerable to crocodilians.[8] Smaller crocodilian teeth are also present, with two distinct morphotypes: an ovoid form bearing ridges, which has been referred to two different species ofBernissartia;[8][a] and a compressed, triangular form bearing heavy ornamentation, which has been attributed to theAtoposauridae.[14][42][a] However, given the uniform small size of these teeth, and the co-occurrence of these morphotypes in some crocodilians,[41] they may belong to the same species.[9]
LargerEolambia would have been vulnerable[8] to the largeneovenatoridtheropodSiats,[43] which was discovered by expeditions from the FMNH.[9] A variety of smallermaniraptoran theropods are also present in the Mussentuchit, which would have functioned as scavengers; their remains include teeth attributed to theDromaeosauridae,Paronychodon, andRichardoestesia.[42] As for herbivorous dinosaurs, thenodosauridAnimantarx was discovered in Carol's Site, near the type specimen ofEolambia.[44] Teeth belonging to more basal ornithopods have also been discovered, with similar teeth having been referred to thePachycephalosauria or to toothed birds by Cifelli in 1999.[8][42] More complete remains referred to two different ornithopods – anorodromine and a basal iguanodont – remain unpublished.[45] A giantoviraptorosaur has also been discovered,[46] alongside teeth from ceratopsians andsauropods.[9][42]
Fish, which may have been preyed upon by smaller crocodilians, are represented by teeth and scales. They include amyliobatiform ray referred to the genusBaibisha;[42] an indeterminate member of theNeopterygii; asemionotid referred to the genusLepidotes; agar related toAtractosteus;[9] apycnodontid referred to the genusStephanodus; and an indeterminatebowfin. Indeterminatecaudatans (salamanders) and mammals are also present. As for plants, pollen samples indicate a tiered forest surrounding the lakes of the Mussentuchit, containing acanopy oftree ferns,araucarias, andconifers, accompanying an undergrowth offerns, smallgymnosperms, andflowering plants. Algae would also have been present in the waters of the lakes.[11][47] These plants would have provided an abundant source of food forEolambia.[8]
Kirkland, Cifelli, and colleagues noted that the fauna of the Mussentuchit – iguanodonts, pachycephalosaurs, and ceratopsians – bears strong similarities to contemporary Asian faunas. They proposed thatEolambia was part of an influx of Asian dinosaurs into North America during the Cenomanian, which supplanted the earlier low-diversity native fauna.[3][14] This hypothesis is supported by the close relationship betweenEolambia and eitherProbactrosaurus orFukuisaurus, which have respectively been recovered by the phylogenies of Head and Sues & Averianov.[4][5][29]
However, the results of Horner and colleagues, Prieto-Márquez, and McDonald and colleagues, which consider the North AmericanProtohadros to be the closest relative ofEolambia, contradict this hypothesis.[9][28][30][48] This alternative interpretation suggests that the common ancestor ofEolambia,Protohadros, and more derived hadrosauroids was not exclusively Asian, but instead distributed across Asia and North America.[4] McDonald and colleagues further suggested thatEolambia andProtohadros represent a localized group of hadrosauroidsendemic to North America, much like other groups of styracosternans throughout the Cretaceous. Such contemporary groups included a central Asian group ofShuangmiaosaurus,Zhanghenglong, andPlesiohadros.[9]
