Cranial Crest

The cranial crest of lambeosaurine hadrosaurids is constructed primarily from the premaxillae and nasals. The dorsal process of the premaxilla extends posterodorsally above the skull roof to form the major structural framework of the crest, while the nasal bones contribute to the ventral paired tubes that carry the narial passages. The lateral premaxillary process acts as a laminar covering between the dorsal and ventral tubes (Gates et al., 2021). Internally, the crest houses complex airway structures derived from the nasal vestibule. These include s-loops—sinuous curves of the nasal passage within the premaxilla—a common median chamber dorsal to the braincase, and lateral diverticula that branch from the common median chamber. The prefrontals provide structural support at the base of the crest complex, while the frontonasal platform—a thickened, steeply angled shelf formed by the frontals—underlies and supports the nasal elements as they ascend into the crest.

A landmark 2026 study by Dudgeon, Brown, and Evans (published in The Anatomical Record) used CT scanning of mature specimens of Corythosaurus casuarius, C. intermedius, and Lambeosaurus lambei to provide the most detailed comparison of internal crest anatomy in adult Lambeosaurini to date. Their results showed that the s-loops of L. lambei are far more convoluted than previously recognized, forming exaggerated loops anterior to the orbit and adding a dorsal loop unique to this species. In contrast, Hypacrosaurus altispinus displays a reversed direction of s-loop curvature (medial to lateral rather than lateral to medial) and greatly expanded anterior projections of the lateral diverticula, representing a derived narial morphology among lambeosaurines.

In theropods, crest anatomy is structurally distinct. Dilophosaurus wetherilli bears a pair of thin, high nasolacrimal crests formed by expansions of the nasal and lacrimal bones. These crests are solid, containing no narial passages, and are too structurally fragile for use in physical combat. Oviraptorid crests are formed from dorsal expansions of the premaxillae and nasals, with considerable variation among taxa; Corythoraptor jacobsi convergently evolved a cassowary-like helmet crest (Lü et al., 2017, Scientific Reports).

The acoustic hypothesis for lambeosaurine crests—first rigorously formulated by David B. Weishampel in a 1981 Paleobiology paper—remains the most widely supported functional interpretation. Weishampel treated the nasal passages within the crest as analogous to wind instruments. By modeling the crest as a series of closed and open tubes, he calculated that Parasaurolophus could produce resonant frequencies in the range of approximately 55–720 Hz, with variation between species owing to differences in crest size, shape, and nasal passage length. The shorter-crested P. cyrtocristatus would have produced higher-frequency sounds than the longer-crested P. walkeri or P. tubicen.

In 1997, a collaborative project between Sandia National Laboratories and the New Mexico Museum of Natural History and Science advanced this work dramatically. Using CT scans of an exceptionally preserved Parasaurolophus tubicen skull discovered in 1995 near Farmington, New Mexico, computer scientist Carl Diegert and paleontologist Tom Williamson constructed a three-dimensional digital model of the crest's internal air passages. Their analysis revealed that the internal structure was more complex than the simple trombone-like tubes previously described, containing additional chambers within the crest. By computationally simulating the passage of air through these passages, they produced a low-frequency, resonant bellowing sound. The simulation included variations with and without vocal cords, since it remains uncertain whether hadrosaurs possessed a syrinx or laryngeal vocal apparatus. The reconstructed sound was a deep, trumpeting rumble that could change in pitch, and each individual Parasaurolophus likely produced a voice distinctive enough to be distinguished from conspecifics.

The 2026 study by Dudgeon et al. carries significant implications for acoustic function. Their finding that Lambeosaurus lambei possesses far more convoluted and lengthened nasal passages than Corythosaurus species suggests the potential for lower-frequency sound production in the former. Since Lambeosaurus lambei, Corythosaurus casuarius, and Parasaurolophus walkeri were partially contemporaneous in the Dinosaur Park Formation during the Late Campanian, the significant intergeneric variation in internal crest anatomy may represent selection for differing vocalizations, enabling acoustic niche partitioning in a shared ecosystem.

There is broad scholarly consensus that the primary evolutionary driver of cranial crest elaboration in dinosaurs was socio-sexual display (Dodson, 1975; Hopson, 1975; Horner et al., 2004; Evans et al., 2009). The crests are conspicuously visible, taxonomically diagnostic, and exhibit the kind of exaggerated ornamentation that, in extant animals, is strongly associated with sexual selection. Hone, Naish, and Cuthill (2012) proposed that mutual sexual selection—in which both sexes select mates based on ornamental traits—was a key evolutionary force shaping cranial crests in both pterosaurs and dinosaurs. Knell and Sampson (2011) argued that the species recognition hypothesis alone is insufficient to explain the evolution of elaborate structures, noting that simple, low-cost signals would suffice for species identification, while the exaggerated crests observed in lambeosaurines and ceratopsians are more consistent with the runaway dynamics of sexual selection.

For solid crests (as in Dilophosaurus and oviraptorids), acoustic function is excluded, making visual display the primary interpretation. The paired crests of Dilophosaurus are too thin and fragile for use as weapons, supporting a role in intraspecific visual signaling—species recognition, mate attraction, or status advertisement. A 2024 preprint reviewing the function and evolution of cranial crests in Dilophosaurus concluded that display was the most parsimonious functional interpretation, consistent with the Knell and Sampson (2011) framework.

Cranial crests in lambeosaurines undergo dramatic morphological change through ontogeny. In hatchlings and early juveniles, the crest is small and externally simple, lacking the elaborate superstructure seen in adults. Dodson (1975) documented allometric growth patterns in lambeosaurine crests that showed rapid expansion during the transition from juvenile to adult stages. Evans, Forster, and Reisz (2005) described a juvenile Parasaurolophus from Dinosaur Provincial Park with only the incipient beginnings of a crest, demonstrating that the elongated tube characteristic of adults develops relatively late in ontogeny. This poses taxonomic challenges: juvenile specimens of different species may appear superficially similar due to their undeveloped crests, and conversely, ontogenetic stage differences within a species can be mistaken for interspecific distinctions.

Gates et al. (2021), in their rediagnosis of Parasaurolophus cyrtocristatus, emphasized that crest shape was historically the primary diagnostic criterion for this species, but because crest morphology changes dramatically during growth, this criterion alone is unreliable. They identified non-crest cranial characters (e.g., squamosal and supraoccipital morphology) that provide more ontogenetically stable diagnostic features. Dudgeon et al. (2026) further documented that in Corythosaurus, the anterior projections of the lateral diverticula become progressively reduced during ontogeny as the s-loops expand posterodorsally, suggesting spatial constraints within the growing premaxillae.

In 2014, Phil R. Bell, Federico Fanti, Philip J. Currie, and Victoria M. Arbour reported a mummified specimen of Edmontosaurus regalis from the Wapiti Formation of Alberta, Canada, that preserved a fleshy, non-ossified cranial crest analogous to a rooster's comb (Bell et al., Current Biology, 24(1), 70–75). Edmontosaurus regalis is a member of Hadrosaurinae—the so-called 'non-crested' hadrosaurs that lack the elaborate bony supracranial crests of lambeosaurines. The soft-tissue crest, positioned on the dorsal midline of the skull, was interpreted as a sexually selected display structure, possibly brightly colored in life.

This discovery fundamentally altered understanding of dinosaur ornamentation. It demonstrated that the absence of bony crests does not equate to the absence of cranial display structures, and that many dinosaurs traditionally classified as 'non-crested' may have borne conspicuous soft-tissue adornments that left no skeletal trace. In October 2025, a study published in Science (concerning Edmontosaurus annectens) reported additional mummified specimens preserving a fleshy midline ridge extending along the neck and trunk, further expanding the known extent of soft-tissue ornamentation in hadrosaurines. A 2025 re-evaluation of the original E. regalis specimen (published in The Anatomical Record) provided updated anatomical descriptions and discussed implications for understanding soft-tissue ornamentation across Dinosauria.

Several functional hypotheses for lambeosaurine crests have been proposed and subsequently rejected or demoted. The snorkeling hypothesis (Sternberg, 1935)—that the crest served as an underwater breathing apparatus—was refuted because the crest lacks an external opening at its apex, making it useless for surface respiration. The air-storage hypothesis (also Sternberg) suggested the crest served as an auxiliary air reservoir for prolonged submersion, but the volume of the crest is too small relative to the animal's lung capacity. The defensive weapon hypothesis (Abel, 1924) was rejected because hollow crests are structurally too fragile to withstand combat impacts. The enhanced olfaction hypothesis (Ostrom, 1961, 1962)—proposing that the enlarged nasal cavity increased olfactory surface area—was weakened by Evans et al. (2009), whose endocranial CT analysis demonstrated that the expanded portions of the nasal cavity within the crest correspond to the nasal vestibule, which in extant archosaurs is associated with air conditioning and vocalization rather than olfaction. The thermoregulation hypothesis proposed that the crest's vascularized surface area could assist in heat dissipation or absorption, but the crest's surface area relative to body size is generally considered insufficient for meaningful thermoregulatory impact.

Parasaurolophus: Tube-shaped crest extending posterodorsally beyond the occiput, formed by the dorsal premaxillary process with nasals forming ~80% of the ventral tubes. Three recognized species (P. walkeri, P. tubicen, P. cyrtocristatus) with distinct crest lengths and curvatures.

Corythosaurus: Helmet-shaped or semicircular crest. Well-developed s-loops curve from lateral to medial within the premaxillae; large lobate posterior lateral diverticula (Dudgeon et al., 2026).

Lambeosaurus: Hatchet-shaped crest tilted anteriorly in L. lambei, with the most exaggerated s-loop development known among lambeosaurines, including a unique dorsal loop (Dudgeon et al., 2026). L. magnicristatus possesses an even larger crest relative to skull size.

Hypacrosaurus: Crest intermediate in shape. Uniquely derived internal anatomy with s-loops curving from medial to lateral (opposite direction to Corythosaurus and Lambeosaurus) and greatly expanded anterior lateral diverticulum projections (Evans et al., 2009).

Dilophosaurus: Paired, thin, plate-like nasolacrimal crests along the skull midline. Solid. Function: visual display. A 2026 comparative cranial biomechanics study also suggested a secondary buttressing function during biting.

Oviraptoridae: Variable crests. Oviraptor philoceratops bears a modest crest; Citipati osmolskae has a tall, rounded casque; Corythoraptor jacobsi convergently evolved a cassowary-like crest, suggesting mutual sexual selection as a driving force (Lü et al., 2017).

Guanlong: An early tyrannosauroid from the Late Jurassic bearing a thin midline crest along the nasals, representing one of the earliest known examples of cranial ornamentation in the tyrannosauroid lineage.

Gates et al. (2016, Nature Communications) demonstrated a correlation between bony cranial ornamentation and the evolution of gigantic body size in theropod dinosaurs, finding that lineages with cranial ornaments speciated at significantly higher rates and evolved larger body sizes than non-ornamented lineages. This suggests that socio-sexual signaling structures like crests may have played a direct role in driving macroevolutionary patterns in Dinosauria. The diversity of crest forms among contemporaneous lambeosaurines in the Dinosaur Park Formation—where Corythosaurus, Lambeosaurus, and Parasaurolophus coexisted—may reflect character displacement driven by the need for species-specific visual and acoustic signals in a multi-species ecosystem. The 2026 demonstration of strongly divergent internal crest anatomy among these coeval genera strengthens the hypothesis that selection for differing vocalizations contributed to the remarkable external crest diversity observed in the Late Cretaceous of western North America.