henhaobio
Sign in

bioRxivPaleontologyPreprint

Developmental variation in pterygoid segmentation clarifies patterns of avian bony palate evolution

Annabel K. HuntiD, Juan BenitoiD, Olivia PlateauiD, Adam Dawid UrantówkaiD, Daniel J. FieldiD

PDFDOI

Abstract

The morphology of the palate has long constituted the primary basis for differentiating between the two deepest clades of crown group birds, Neognathae and Palaeognathae. However, published literature on the bird palate is dominated by classical anatomical descriptions pre-dating the advent of contemporary three-dimensional imaging techniques, hindering our understanding of bird palate disparity and development. Pterygoid segmentation, the process by which the rostral portion of the pterygoid separates and fuses with the palatine during ontogeny in neognathous birds, remains a poorly understood aspect of avian cranial development despite giving rise to an important component of the cranial kinetic system. Here, we use micro-computed tomography to explore ontogenetic change of the palate during the process of pterygoid segmentation across an unprecedentedly broad taxonomic sample of immature and mature birds. With this large dataset, we present direct evidence for the full process of post-hatching pterygoid segmentation being exclusively restricted to the major avian subclade Neoaves. Further, our investigation helps refine several hypotheses of homology related to the pterygoid-palatine contact in crown birds and their close stem-group relatives: we hypothesise that the rostral projection of the pterygoid observed in Anseres (and potentially Anhimidae and Megapodiidae) is homologous with the hemipterygoid of neoavians, and that the discrete and unfused hemipterygoids of some crownward stem-birds suggest that a discrete hemipterygoid originated prior to the origin of the process of pterygoid segmentation. However, whether pterygoid segmentation represents a synapomorphy of Neornithes, Neognathae, or Neoaves remains ambiguous in light of elusive data from the fossil record and early stages of cranial development. Overall, our study clarifies the process of pterygoid segmentation in crown birds and raises new questions regarding the major morphological modifications that have characterized the evolutionary history of the avian bony palate.

In-depth summary

Claude Opus 5.5 · xhigh effort · 8 Nov 2026 · from the full text of version 2 (22 Sept 2026) · CC BY 4.0

Micro-computed tomography of 69 immature and mature bird specimens finds complete post-hatching pterygoid segmentation (hemipterygoid detaches from pterygoid, fuses to palatine) only in Neoaves, not Galloanserae or palaeognaths.

The palate has long been the main trait separating Palaeognathae from Neognathae, yet pterygoid segmentation, which helps form the mobile pterygoid-palatine joint of neognaths, is known mostly from classical descriptions drawn in ventral or lateral view that obscure the dorsally placed hemipterygoid. This study uses three-dimensional μCT reconstructions across a broad ontogenetic and taxonomic sample to establish which crown-bird clades actually undergo segmentation.

Key results

  1. ShownAmong extant birds, the complete post-hatching segmentation sequence was observed only in Neoaves: a distinct hemipterygoid forms, separates from the pterygoid body and fuses to the caudal palatine.

    All 22 immature neoavian specimens showed evidence of a hemipterygoid (attached to the pterygoid, discrete, or fusing to the palatine), and it was merged into the palatine in all 14 adults. No transient discrete hemipterygoid or segmentation-consistent ontogenetic change in palatine and pterygoid shape was found in any surveyed palaeognath, anseriform or galliform. · NEOAVES; 4. DISCUSSION; Anseriformes; Mature

  2. SuggestedNeoavian immatures can be ordered into five hypothetical stages, from an unsegmented rostral pterygoid projection to complete fusion with the palatine. A fully detached hemipterygoid (Stage 3) was rare, so many taxa may start fusion before detachment is complete.

    Stage 1 (no suture): immature Tauraco, Musophaga, Otidiphaps, Pteroglossus, Ara, Lorius. Stage 2 (sutured but still attached): e.g. Apus, Columba, Spheniscus, Morus, Nycticorax, Vultur. Stage 3: only one immature Gallirallus australis. Stage 4 (sutured to the palatine): Recurvirostra and Egretta, the latter with tentative signs of incomplete detachment that scan resolution could not settle. · STAGE 1 - UNSEGMENTED HEMIPTERYGOID (figs. 21…; STAGE 2 - PARTIALLY SEGMENTED HEMIPTERYGOID…; STAGE 4 – HEMIPTERYGOID PARTIALLY FUSED TO THE…; Neoaves

  3. SuggestedIn most neoavians the adult pterygoid process of the palatine is a composite: the hemipterygoid supplies its dorsal part (sometimes a separate dorsal condyle) and the palatine its ventral part. The neoavian joint is therefore only partly homologous with the joints of other neognaths.

    In Stage 2 immatures the hemipterygoid sits in a shallow medial depression along the dorsal palatine crest, a depression that adults lack; the ventral condyle is formed by the palatine alone. The Stage 3 Gallirallus hemipterygoid matches the dorsal part of the adult pterygoid process, and an adult Opisthocomus retains an incompletely fused suture where the immature overlap lies. · STAGE 2 - PARTIALLY SEGMENTED HEMIPTERYGOID…; STAGE 3 - FULLY SEGMENTED HEMIPTERYGOID (figs…; Neoaves

  4. SuggestedThe anatid rostral pterygoid projection, which persists into adulthood without segmenting, is proposed as homologous with the Stage 1 neoavian hemipterygoid, and the authors recommend calling it a 'hemipterygoid process'. Parts of the rostral projections of megapodes and Anhimidae may be equivalent.

    No distinct hemipterygoid appeared in any immature or adult anatid. The projection spans ~1/4 (Cygnus, Anser) to ~1/3 (Dendrocygna, Aythya, Spatula, Malacorhynchus, Anas) of immature pterygoid length and changes from blade-like to roughly cylindrical in adults; its shape and position match Stage 1 neoavians. Non-megapodiid galliforms had no rostral projection at any post-hatching stage. · Immature; Anseriformes; Potential homology of the anseriform rostral…; Galliformes

  5. SuggestedDiscrete, unfused hemipterygoids reported in stem ornithurines (Ichthyornithes, Hesperornithes) lead the authors to argue that the hemipterygoid predates pterygoid segmentation. They favour segmentation as a Neoaves synapomorphy but say alternative scenarios remain viable.

    Ichthyornithine and hesperornithine hemipterygoids are described as very large relative to neoavian ones and unfused to pterygoid or palatine. Stem anseriforms Nettapterornis and Presbyornis, examined here, have short, robust rostral pterygoid projections like those of megapodes. The authors read these as hemipterygoid processes, implying hemipterygoid-pterygoid fusion is plesiomorphic for Galloanserae. · FOSSILS AND THE ORIGIN OF PTERYGOID SEGMENTATION; 5. CONCLUSIONS

Methods, data and reproducibility

Palatines, pterygoids and any hemipterygoids were manually segmented in VGSTUDIOMAX from micro-computed tomography (μCT) scans of late-stage embryos, hatchlings, juveniles and mature birds. These were compared within immature-mature pairings across Palaeognathae, Galloanserae and Neoaves, and neoavian immatures were assigned to five hypothetical stages of segmentation.

  • Extant bird specimens scanned by μCT, covering 50 species-level taxa in 21 ordinal-level subclades of Palaeognathae, Galloanserae and Neoaves · birds (Neornithes) · n = 69 specimens (43 immature, 26 comparatively mature)
  • Palaeognath specimens (Tinamidae, Casuariidae, Apterygidae, Struthionidae) · birds (Palaeognathae) · n = six immature, four comparatively mature
  • Anseriform specimens (Anhimidae, Anatidae) · birds (Anseriformes) · n = eight immature, three adult
  • Galliform specimens (Megapodiidae, Numididae, Phasianidae) · birds (Galliformes) · n = seven immature, five adult
  • Neoavian specimens, including immature-adult pairs from most major neoavian subclades · birds (Neoaves) · n = 22 immature, 14 adult
  • Subadult subfossil holotype of the extinct Upland Moa · Megalapteryx didinus (extinct moa) · n = holotype specimen
  • Pterygoids of fossil total-group anseriforms Nettapterornis and Presbyornis · fossil total-group Anseriformes · n = two fossils · NHMUK PVA 5922 (Nettapterornis); USNM PAL 299846 (Presbyornis)
  • Additional immature and mature specimens examined visually for comparison (supplementary table 2) · birds (Neornithes)
Methods & tools
micro-computed tomography (μCT), VGSTUDIOMAX 3.4.5 (VolumeGraphics), VGSTUDIOMAX 2023.2 (VolumeGraphics), MorphoSource (planned data deposit)
Reagents & environment
VGSTUDIOMAX version 3.4.5 and version 2023.2 (VolumeGraphics), used for manual segmentation
Reproducibility gaps
CT data and 3D models are to be deposited on MorphoSource upon acceptance (DOI not yet assigned). The specimen list is in supplementary tables 1 and 2, which are not in the provided text. No code is mentioned; segmentation was manual.
Limitations (6)
  • The study cannot say whether segmentation arose as a neornithine, neognath or neoavian synapomorphy; settling this needs fossil and early embryonic data. · stated by the authors
  • Only late embryos and post-hatching birds were sampled. An early embryonic hemipterygoid that fuses straight to the palatine, as reported for Gallus by Jollie (1957), would be missed, so pre-hatching imaging is needed. · stated by the authors
  • The pterygoid-hemipterygoid suture was hard to resolve in lower-resolution scans and weakly ossified specimens. Scan resolution also left open whether the immature Egretta hemipterygoid was still attached to the pterygoid. · stated by the authors
  • The proposed homologies (anatid rostral projection with the neoavian hemipterygoid; kiwi medial pterygoid process; anhimid and megapode projections) are flagged by the authors as needing developmental testing. · stated by the authors
  • Ontogenetic series are cross-sectional and often pair different species (e.g. immature Macronectes giganteus with mature M. halli, immature Aythya affinis with mature A. ferina), so stage order is inferred rather than tracked within a lineage. · noted in this summary
  • Descriptions are qualitative, with few specimens per taxon and stage (Stage 3 rests on one Gallirallus). No morphometric or statistical analysis supports the clade-level contrasts. · noted in this summary

Read this if you study avian skull development, cranial kinesis or palate homology and want 3D μCT documentation of how the hemipterygoid forms, detaches and fuses across bird clades and ontogenetic stages.

Citations
1

Reproducibility ledger

Nobody has reported trying to reproduce this paper yet. Did you replicate an experiment or run the code? Your report saves the next lab a week.

Sign in to file a reproducibility report.