Implications of beak morphology for the evolutionary paleoecology of the megaherbivorous dinosaurs from the Dinosaur Park Formation (upper Campanian) of Alberta, Canada

https://doi.org/10.1016/j.palaeo.2013.11.014Get rights and content

Highlights

  • Megaherbivores from the Dinosaur Park Formation differ broadly in their beak morphologies and inferred feeding styles.

  • Beak morphology does not differ between confamilial taxa.

  • Different feeding styles were necessary but insufficient to sustain megaherbivore diversity.

  • The distribution of beak morphologies was maintained for at least 1.5 Myr.

Abstract

Using the megaherbivorous dinosaur assemblage from the Dinosaur Park Formation as a model, linear and geometric morphometrics are applied to examine the degree to which different feeding styles—as reflected by beak morphology—facilitated the coexistence of these animals on the Late Cretaceous island continent of Laramidia. Our findings indicate that megaherbivorous dinosaurs occupied a spectrum of feeding habits. The wide, square beaks of the ankylosaurs suggest that these animals were bulk-feeders that consumed more fibrous herbage than traditionally assumed. Conversely, the narrow, square beaks of the ceratopsids evoke concentrate feeders, although the large body sizes and sophisticated dental batteries of these animals suggest a diet of forbs and low-growing scrub, akin to the feeding strategy of the black rhinoceros (Diceros bicornis). Both nodosaurids and hadrosaurids had beaks of intermediate size and shape, suggesting that these were mixed feeders that consumed a diversity of plant types of variable nutritional quality. Contrary to previous suggestions, there is little evidence for different feeding styles within the aforementioned families. Feeding styles were evolutionarily stable, and lend further support to the contention that the fossil assemblage of the Belly River Group constitutes a chronofauna.

Introduction

During the Late Cretaceous, the epeiric Western Interior Seaway divided North America into two. The resulting western landmass—called Laramidia (Archibald, 1996)—was diminutive in size, with an estimated area of just 4–7 million km2 (Lehman, 1997, Sampson and Loewen, 2010). Despite its small size, Laramidia supported a rich variety of dinosaurs, particularly megaherbivorous forms whose diversity is unrivaled by any modern mammal community (Lehman, 1987, Lehman, 1997, Lehman, 2001).

For various reasons (summarized in Mallon et al., 2013), experts have long wondered how so many large, diverse herbivores could coexist on such a small landmass. Two basic explanations have been proposed: (1) megaherbivores coexisted because dietary resources were non-limiting (Ostrom, 1964a, Farlow et al., 1995, Lehman, 1997, Sampson, 2009), and (2) megaherbivore coexistence was achieved as a result of dietary niche partitioning (Coe et al., 1987, Lehman, 2001, Sander et al., 2010). The latter claim has been the subject of recent attention (Mallon et al., 2012, Mallon and Anderson, in press, Mallon et al., 2013), and is the continued focus of this study. More specifically, the hypothesis that megaherbivore coexistence was mediated primarily by different feeding styles is tested. It is predicted that coexisting species should be distinguishable by beak morphology, which influences the ability of an individual to crop food at coarse or fine spatial scales. The megaherbivore assemblage of the Dinosaur Park Formation (DPF) is used as a study model, which has been justified elsewhere (Mallon et al., 2012, Mallon and Anderson, in press, Mallon et al., 2013).

Plants typically differ in both nutritional quality and the way they are distributed throughout the environment. For example, grasses are of low nutritional quality and are homogeneously distributed throughout the environment. By contrast, dicotyledonous (dicot) plants are commonly of higher nutritional quality and are heterogeneously distributed throughout the environment. In this way, the most nutritious food items are variably dispersed in a substrate of readily-available, but low-nutrient fodder. This pattern is fractal in nature (Jarman, 1974, Lucas, 2004); for example, even within a single shrub, the most nutritious fruits and shoots occur within a matrix of less nutritious leaves and twigs.

If natural selection has acted to allow vertebrate herbivores to forage optimally (MacArthur and Pianka, 1966, Charnov, 1976) on different plants, they should differ in the size and shape of their rostra, which are employed in the cropping of vegetation (Hanley, 1982). Because grasses are of low nutritional quality, yet uniformly distributed across the substrate, natural selection would be expected to produce animals with wide, square rostra that allow them to increase intake rates to offset the low nutritional quality of grass. By contrast, animals that specialize on dicot browse should have narrow, pointed rostra that allow them to concentrate on the most nutritious forbs and shoots distributed patchily throughout the environment. Rostral morphology is likewise related to body size; larger taxa can tolerate lower quality foodstuffs as a result of their lower mass-specific metabolic requirements and prolonged gut retention times (Owen-Smith, 1988), and are therefore expected to have the widest, squarest rostra in the absence of evolutionary constraints.

Because different rostral morphologies are required to feed on different plants or plant parts, it might be expected that sympatric herbivore species facilitate coexistence by evolving disparate rostral sizes and shapes, thereby reducing competition for food. In fact, sympatric bovids from the African Serengeti exhibit a panoply of rostral morphologies that allow them to specialize on different plant tissues (Gordon and Illius, 1988, Janis and Ehrhardt, 1988, Janis, 1990, Solounias and Moelleken, 1993, Spencer, 1995, Dompierre and Churcher, 1996, Mendoza et al., 2002, Fraser and Theodor, 2011). The gerenuk (Litocranius walleri) possesses a particularly narrow, pointed rostrum, even for its small size, allowing it to concentrate on dicot forage. At the other extreme, the African buffalo (Syncerus caffer) possesses a wide, square rostrum that allows it to ingest vast quantities of grass. There are also a variety of intermediate feeders, including gemsbok (Oryx gazella), common eland (Taurotragus oryx), and hartebeest (Alcelaphus buselaphus), which have intermediate-shaped rostra that enable them to feed on both grass and dicot browse. Similar adaptations to different feeding styles among ungulates have been noted in cervids (Gordon and Illius, 1988, Janis and Ehrhardt, 1988, Solounias and Moelleken, 1993, Dompierre and Churcher, 1996, Fraser and Theodor, 2011), camelids (Dompierre and Churcher, 1996), and megaherbivorous rhinocerotids (Janis and Ehrhardt, 1988, Owen-Smith, 1988, Janis, 1990).

It follows that the megaherbivorous ankylosaurs, ceratopsids, and hadrosaurids from the DPF might also have differed in their feeding styles. However, given that Cretaceous grasses were relatively rare (Prasad et al., 2005), the browser–grazer continuum typically used to classify mammals cannot be extended to dinosaurs. Dodson et al. (2004, p. 511) suggested the terms “generalized (nonselective) and specialized (selective) feeders”, instead, citing the narrow-beaked ceratopsids as an example of the latter. However, it must be noted that animals with wide rostra may also feed selectively, typically specializing on a low-nutrient diet requiring high rates of intake (e.g., grazing ungulates requiring > 90% grass in their diet; Hofmann and Stewart, 1972). Therefore, the nomenclature of Hofmann and Stewart (1972) is offered as a more profitable alternative. They distinguished between bulk feeders, concentrate feeders, and intermediate feeders. Bulk feeders are adapted to eating low nutrient roughage, and possess wide rostra as a result. At the opposite end of the spectrum, concentrate feeders select only the most nutritious plant parts using their relatively narrow rostra. Intermediate feeders fall between these two extremes, consuming a range of plant types, and having intermediate snout shapes accordingly.

The paleoenvironments of the DPF varied, and likely contained a mix of open and closed habitats that would have been suitable for dinosaurs exploiting these different feeding styles (Krassilov, 1981, Crane, 1987, Upchurch and Wolfe, 1987, Wolfe and Upchurch, 1987, Carrano et al., 1999, Braman and Koppelhus, 2005). Differences in beak shape are also thought to have facilitated the coexistence of ankylosaurids and nodosaurids (Carpenter, 1982, Carpenter, 1997a, Carpenter, 1997b, Carpenter, 2004), centrosaurines and chasmosaurines (Lull, 1933), and hadrosaurines and lambeosaurines (Dodson, 1983, Bakker, 1986, Carrano et al., 1999, Whitlock, 2011), but no formal test of the hypothesis that beak shape differs within and between all these groups has been offered to date. The present study addresses this challenge using a combination of linear and geometric morphometrics.

Section snippets

Specimens examined

The dataset used here (Fig. 1; Appendix 1) comprises 72 specimens spanning 12 genera from the megaherbivorous dinosaur clades Ankylosauria, Ceratopsidae, and Hadrosauridae. The ankylosaurids Dyoplosaurus (Parks, 1924, Arbour et al., 2009) and Scolosaurus (Nopcsa, 1928, Penkalski and Blows, 2013), and the ceratopsid Spinops (Farke et al., 2011), are not included because suitable material is lacking (the provenances of the last two genera are also unknown).

Caveats

A few caveats must first be considered

Configuration of morphospace

The shape PCA (Fig. 2) produced two principal components (PCs) that cumulatively account for ~ 96% of the total shape variance (Table 1), indicating that most of the variation in the dataset can be accounted for by just a couple of simple shape deformations. Positive scores along PC 1 correspond to elongate, pointed beaks, whereas negative scores correspond to stout, square beaks. Positive scores along PC 2 correspond to ‘spoon shaped’ beaks that constrict caudally, whereas negative scores

Ankylosauridae

The relatively wide, square beaks of Euoplocephalus suggest that these animals were bulk feeders, specializing on fibrous plant tissues with comparatively low nutritional values. This, in combination with their low feeding heights (< 1 m; Mallon et al., 2013) and ventrally deflected beaks (Mallon and Anderson, 2013), is reminiscent of the condition seen in grazing ungulates. The most fibrous plants of the Late Cretaceous (having > 60% neutral detergent fiber) included araucarian and podocarpaceous

Conclusions

This study demonstrates that, among the megaherbivorous dinosaurs from the DPF, beak shape discriminates taxa better than size. The ankylosaurid Euoplocephalus possess the widest, squarest beaks, suggestive of a fibrous diet typical of bulk feeders. Conversely, ceratopsids have the narrowest, most triangular beaks, otherwise observed among concentrate feeders that usually specialize on high protein, low fiber plant tissues. However, given their large body size and specialized shearing

Acknowledgments

Thanks to P. Dodson, K. Ruckstuhl, M. Ryan, J. Theodor, and F. Therrien for commenting on early drafts of the manuscript. Two anonymous reviewers offered additional constructive criticisms, and F. Surlyk provided valuable editorial assistance. D. Fraser provided stimulating discussion, and Ø. Hammer, D. Sheets, and M. Zelditch offered valuable technical assistance. C. Mehling (AMNH); W. Simpson (FMNH); P. Barrett and S. Chapman (NHMUK); M. Currie, A. McDonald, and K. Shepherd (CMN); B. Iwama,

References (127)

  • J.D. Archibald

    Dinosaur Extinction and the End of an Era: What the Fossils Say

    (1996)
  • R.T. Bakker

    Dinosaur feeding behaviour and the origin of flowering plants

    Nature

    (1978)
  • R.T. Bakker

    The Dinosaur Heresies: New Theories Unlocking the Mystery of the Dinosaurs and their Extinction

    (1986)
  • D.J. Beerling

    Modelling palaeophotosynthesis: Late Cretaceous to present

    Phil. Trans. R. Soc. B

    (1994)
  • D.J. Beerling

    Global terrestrial productivity in the Mesozoic era

    Geol. Soc. London Spec. Publ.

    (2000)
  • A.K. Behrensmeyer et al.

    Paleoenvironmental contexts and taphonomic modes

  • P. Béland et al.

    Paleoecology of Dinosaur Provincial Park (Cretaceous), Alberta, interpreted from the distribution of articulated vertebrate remains

    Can. J. Earth Sci.

    (1978)
  • F. Bookstein et al.

    Comparing frontal cranial profiles in archaic and modern Homo by morphometric analysis

    Anat. Rec.

    (1999)
  • D.R. Braman et al.

    Campanian palynomorphs

  • C.M. Brown et al.

    Evidence for taphonomic size bias in the Dinosaur Park Formation (Campanian, Alberta), a model Mesozoic terrestrial alluvial–paralic system

    Palaeogeogr. Palaeoclimatol. Palaeoecol.

    (2012)
  • R.J. Butler et al.

    The phylogeny of the ornithischian dinosaurs

    J. Syst. Palaeontol.

    (2008)
  • K. Carpenter

    Skeletal and dermal armor reconstructions of Euoplocephalus tutus (Ornithischia: Ankylosauridae) from the Late Cretaceous Oldman Formation of Alberta

    Can. J. Earth Sci.

    (1982)
  • K. Carpenter

    Ankylosaur systematics: example using Panoplosaurus and Edmontonia (Ankylosauria: Nodosauridae)

  • K. Carpenter

    Ankylosauria

  • K. Carpenter

    Ankylosaurs

  • K. Carpenter

    Redescription of Ankylosaurus magniventris Brown 1908 (Ankylosauridae) from the Upper Cretaceous of the Western Interior of North America

    Can. J. Earth Sci.

    (2004)
  • M.T. Carrano et al.

    Hadrosaurs as ungulate parallels: lost lifestyles and deficient data

    Acta Palaeontol. Pol.

    (1999)
  • K. Chin

    The paleobiological implications of herbivorous dinosaur coprolites from the Upper Cretaceous Two Medicine Formation of Montana: why eat wood?

    Palaios

    (2007)
  • K. Chin et al.

    Dinosaurs, dung beetles, and conifers: participants in a Cretaceous food web

    Palaios

    (1996)
  • M.J. Coe et al.

    Dinosaurs and land plants

  • P.R. Crane

    Vegetational consequences of the angiosperm diversification

  • P.J. Currie et al.

    Stomach contents of a hadrosaur from the Dinosaur Park Formation (Campanian, Upper Cretaceous) of Alberta, Canada

  • S.M. Decherd

    Primary Productivity and Forage Quality of Ginkgo biloba in Response to Elevated Carbon Dioxide and Oxygen — An Experimental Approach to Mid-Mesozoic Paleoecology

    (2006)
  • W.A. DiMichele et al.

    Long-term stasis in ecological assemblages: evidence from the fossil record

    Annu. Rev. Ecol. Evol. Syst.

    (2004)
  • P. Dodson

    Taxonomic implications of relative growth in lambeosaurine hadrosaurs

    Syst. Biol.

    (1975)
  • P. Dodson

    A faunal review of the Judith River (Oldman) Formation, Dinosaur Provincial Park, Alberta

    Mosasaur

    (1983)
  • P. Dodson et al.

    Ceratopsidae

  • H. Dompierre et al.

    Premaxillary shape as an indicator of the diet of seven extinct late Cenozoic New World camels

    J. Vertebr. Paleontol.

    (1996)
  • D.A. Eberth

    The geology

  • G.M. Erickson et al.

    Complex dental structure and wear biomechanics in hadrosaurid dinosaurs

    Science

    (2012)
  • D.C. Evans

    Ontogeny and Evolution of Lambeosaurine Dinosaurs (Ornithischia: Hadrosauridae)

    (2007)
  • A.A. Farke et al.

    A new centrosaurine from the Late Cretaceous of Alberta, Canada, and the evolution of parietal ornamentation in horned dinosaurs

    Acta Palaeontol. Pol.

    (2011)
  • A.A. Farke et al.

    Ontogeny in the tube-crested dinosaur Parasaurolophus (Hadrosauridae) and heterochrony in hadrosaurids

    Peer J.

    (2013)
  • J.O. Farlow

    A consideration of the trophic dynamics of a Late Cretaceous large-dinosaur community (Oldman Formation)

    Ecology

    (1976)
  • J.O. Farlow et al.

    Dinosaur biology

    Annu. Rev. Ecol. Syst.

    (1995)
  • D. Fraser et al.

    Anterior dentary shape as an indicator of diet in ruminant artiodactyls

    J. Vertebr. Paleontol.

    (2011)
  • P.M. Galton

    The cheeks of ornithischian dinosaurs

    Lethaia

    (1973)
  • P.M. Galton

    Herbivorous adaptations of Late Triassic and Early Jurassic dinosaurs

  • C.T. Gee

    Dietary options for the sauropod dinosaurs from an integrated botanical and paleobotanical perspective

  • I.J. Gordon et al.

    Incisor arcade structure and diet selection in ruminants

    Funct. Ecol.

    (1988)
  • Cited by (28)

    • The Albian vegetation of central Alberta as a food source for the nodosaurid Borealopelta markmitchelli

      2023, Palaeogeography, Palaeoclimatology, Palaeoecology
      Citation Excerpt :

      Indeed, it is commonly held that most megaherbivorous dinosaurs could not have been preferential feeders, given the metabolic requirements of such a large body size, and must, therefore, have been generalist herbivores (Wing and Tiffney, 1987; Taggart et al., 1997). What is known of the dietary ecology of dinosaur megaherbivores is largely based on indirect evidence, including dental morphology and microwear, beak morphology, jaw biomechanics, nutritional quality or palatability of extant nearest plant relatives, and known macrofloral associations (e.g., Hummel et al., 2008; Sander et al., 2010; Mallon and Anderson, 2014a, 2014b; Mallon, 2019; Ősi et al., 2022). If dinosaur megaherbivores were specialists or had dietary preferences, as opposed to a generalist feeding behaviour, these preferences would have influenced their roles as ecosystem engineers on Cretaceous landscapes (Bakker, 1978; Wing and Tiffney, 1987; Tiffney, 1992; Taggart et al., 1997; Brown et al., 2020).

    • Beak morphology and limb proportions as adaptations of hadrosaurid foraging ecology

      2023, Cretaceous Research
      Citation Excerpt :

      On the other hand, Mallon and Anderson (2013) found that hadrosaurine beaks are transversely narrower and less ventrally deflected than the lambeosaurines, using linear measurements that represent skull ecomorphologies. More recently, Mallon and Anderson (2014) performed two-dimensional geometric morphometrics on the hadrosaurid beak shape and recovered no statistically significant difference between hadrosaurines and lambeosaurines. These differences are likely to originate in the different analytical approaches in each study and thus are not directly comparable.

    • Stable isotope record of Triceratops from a mass accumulation (Lance Formation, Wyoming, USA) provides insights into Triceratops behaviour and ecology

      2022, Palaeogeography, Palaeoclimatology, Palaeoecology
      Citation Excerpt :

      It has been suggested that gymnosperms have more positive δ13C values than angiosperms (by 2–3 ‰ (Hare and Lavergne, 2021)), and recent studies have indicated that small but consistent offsets in δ13C between different plant types may vary strongly depending both on the plant species and prevailing climatic and environmental conditions (e.g., up to 2–3 ‰ for halophytes (Wei et al., 2008; Sheldon et al., 2020; Hare and Lavergne, 2021)). Although specific studies on Triceratops feeding are lacking, it is proposed that ceratopsians were able to efficiently process fibrous plant material such as gymnosperms using their sharp beak and tooth batteries (Ostrom, 1966; Mallon and Anderson, 2014). A diet shifted more towards gymnosperms and/or halophytes may have led to higher δ13C values.

    View all citing articles on Scopus
    View full text