GoKawiilIn the originally published version of our article, the endocranial landmark configuration from the enantiornithine bird Navaornis hestiae included nine misplaced patch semilandmarks resulting from a sharp corner in the three-dimensional model, which inadvertently shifted those semilandmarks to the next solid surface along the landmark’s normal trajectory. Here, we amend this problem by repositioning these misplaced coordinates using the software Stratovan Checkpoint (the state-of-the-art licensed and maintained version of the software used in our original manuscript: IDAV Landmark Editor) and upload the corrected dataset to the link provided in the Data availability section in the original article. We produced an updated principal components plot summarizing patterns of endocranial shape with the corrected dataset (Fig. 1). Patterns are unchanged regarding the ‘intermediate’ morphology of N. hestiae along PC1, the main axis of endocranial shape variation. However, with the corrected dataset, N. hestiae falls within the crown bird dispersion along PC2, whereas in the initially published article, N. hestiae fell just outside crown bird variation along the same axis. The amount of shape variation explained by each principal component in the corrected dataset varies slightly from the values in the original article; however, our interpretations of avian endocranial evolution are unaltered by this relatively minor change. To further reassess our initial interpretations, we calculated total shape distances (Procrustes distances) from Navaornis hestiae to all the taxa included in our sample (Fig. 2). The six closest species in endocranial shape include the stem bird Archaeopteryx, the unnamed troodontid IGM 100/1126, and crown birds such as Ptilinopus (Columbiformes), Cariama (Cariamiformes), Crypturellus (Palaeognathae, Tinamiformes) and Caloenas (Columbiformes). This assortment of taxa highlights the combination of plesiomorphic and crown bird-like traits in the endocranial morphology of N. hestiae, as originally interpreted. We thank Jesús Marugán-Lobón and Jingmai O’Connor, whose queries helped us identify issues with the original dataset. We also thank Patrick O’Connor and three anonymous reviewers who helped us improve the text of this correction notice.
Fig. 1: Navaornis and endocranial evolution in birds and non-avian dinosaurs. Full size image (a) Three-dimensional principal component (PC) morphospace (PC1 versus PC2, and PC1 versus PC3) of cranial endocasts from crown birds, Archaeopteryx, Navaornis and relevant non-avian taxa. Principal component plots are scaled to the same ratio, so distances are the same on the x axis and y axis. Along PC1, Navaornis falls in an intermediate position between stemward taxa such as Archaeopteryx and crown birds, whereas it falls within the range of crown bird variation along PC2 and PC3. (b) Shape changes associated with scores representing the 5th and 95th quantiles from each of the three main PCs of endocranial shape. Plus (+) and minus (−) symbols indicate whether shape warps are associated with positive or negative scores along each axis. Landmarks are colour-coded for the main brain regions. Brown arrows in b depict the foramen magnum orientation. Percentages in parentheses in a and b indicate the proportion of total shape variance explained by each of the principal components. This figure represents a correction of the PCA plots in Fig. 4b and Extended Data Fig. 10d from the original article.