Craniofacial modularity and cranial kinesis evolution in the adaptive radiation of Furnariidae (Aves: Passeriformes)
Data files
Dec 27, 2024 version files 14.64 MB
-
Beak.zip
4.78 MB
-
Furnariidae_cranial_kinesis_Variables.csv
2.80 KB
-
Neurocranium.zip
128.45 KB
-
Phylogenetic_trees.zip
28.98 KB
-
README.md
2.79 KB
-
Skull_Prokinetic_sps.zip
2.72 MB
-
Skull_Proximal_rhynchokinetic_sps.zip
2.13 MB
-
Skull.zip
4.84 MB
Abstract
The role of phenotypic modularity in the evolution of skull morphology in birds has been a subject of debate in recent years. Furnariids (ovenbirds, woodcreepers, and allies), a spectacular avian adaptive radiation, are distinguished in their cranial morphology as the only passerines with two types of cranial kinesis, constituting a great model to test whether the evolution of novelties in cranial kinesis was associated with shifts in patterns of evolutionary modularity and allometry in the avian skull. Our analyses by means of geometric morphometric tools and phylogenetic comparative methods show that the beak and neurocranium of furnariids evolved in a modular fashion and were shaped by the cranial kinesis evolution. Besides, species with prokinesis show a higher degree of modularity and morphological disparity, lower phenotypic rates, as well a higher contribution of allometry in the evolution of the beak morphology than species with proximal rhynchokinesis, suggesting, as observed in several vertebrates, that the functional demands associated with higher degrees of cranial kinesis promoted rapid integration throughout the skull. Prokinetic - robust morphotypes and proximal rhynchokinetic - gracile morphotypes, have repeatedly evolved by evolutionary convergence in both evolutionary modules, which suggests the existence of functional trade-offs and long-standing adaptive optima related to cranial kinesis.
README: Craniofacial modularity and cranial kinesis evolution in the adaptive radiation of Furnariidae (Aves: Passeriformes)
https://doi.org/10.5061/dryad.4tmpg4fmv
Files and variables
File: Furnariidae_cranial_kinesis_Variables.csv
Description: Variables associated with cranial kinesis used in phylogenetic comparative methods.
- Species: Species used in geometric morphometric analyses and PCM
- Kinesis: A binary variable derived from the discretization of the range of values adopted by the kinetic index.
- Indice_Kinetico: A variable constructed from the sum of the states of three variables associated with cranial kinesis.
File: Phylogenetic_trees.zip
Description: The complete phylogenetic tree obtained by Harvey et al. (2020) and the three pruned variations used in the present study.
File: Neurocranium.zip
Description: Landmark coordinates obtained with Stereomorph of all specimens per species. Neurocranium Procrustes coordinates and Centroid Size: a proxy for size (Centroid Size) and Procrustes coordinates with and without allometric variation of species means as used in the PCM.
File: Skull_Prokinetic_sps.zip
*Description: *Landmark and sliding semilandmarks coordinates obtained with Stereomorph of all specimens per prokinetic species (45sps). Skull Procrustes coordinates and Centroid Size: a proxy for size (Centroid Size) and Procrustes coordinates with and without allometric variation of species means as used in the PCM.
File: Skull_Proximal_rhynchokinetic_sps.zip
Description: Landmark and sliding semilandmarks coordinates obtained with Stereomorph of all specimens per proximal rhynchokinetic species (38sps). Skull Procrustes coordinates and Centroid Size: a proxy for size (Centroid Size) and Procrustes coordinates with and without allometric variation of species means as used in the PCM.
File: Skull.zip
Description: Landmark and sliding semilandmarks coordinates obtained with Stereomorph of all specimens per species (83 sps). Skull Procrustes coordinates and Centroid Size: a proxy for size (Centroid Size) and Procrustes coordinates with and without allometric variation of species means as used in the PCM.
File: Beak.zip
Description: Landmark and sliding semilandmarks coordinates obtained with Stereomorph of all specimens per species (83 sps). Beak Procrustes coordinates and Centroid Size: a proxy for size (Centroid Size) and Procrustes coordinates with and without allometric variation of species means as used in the PCM.
Code/software
Free available R packages: Stereomorph, geomorph, ape, phytools, magrittr, morphospace, mvMORPH, Morpho, RRphylo
Access information
NA
Methods
Data sampling
Skull morphology was studied on photographs taken in lateral and dorsal view of 153 museum skeleton specimens representing 106 species, 51 genera (covering 73% of all currently recognized genera; sensu Clements et al., 2022), three subfamilies (i.e., Sclerurinae, Dendrocolaptinae, and Furnariinae), and all monophyletic clades of Furnariidae (i.e., tribes; sensu Harvey et al., 2020) (see Supplementary Table S1 for information on each specimen examined in this study). Although subtle differences in size between males and females have been reported in some traits in certain species of furnariids (e.g., Moreno et al., 2007; Diniz et al., 2016), there is no significant sexual dimorphism reported in the skulls of these birds (Stefanini et al., 2016). Therefore, our sampling did not discriminate between sexes.
Morphological quantification
Variation in skull shape was quantified in 110 specimens representing 83 species (see Supplementary Table S1) using 17 landmarks and 30 curve semilandmarks, digitized on photographs taken in lateral view of each specimen using Stereomorph (v. 1.6.7) (Olsen & Westneat, 2015) (Figure 2; Supplementary Figure S1A). Landmarks were mainly taken from Marugán-Lobón & Buscalioni (2006) and Navalón et al. (2020) (see description of each landmark in Table 1). Since we were particularly interested in capturing the variation associated with the cranial kinesis, we added one landmark to quantify the presence and extension of the rhynchokinetic 'slit-like gap' hinge (hinge E in Zusi, 1984), which is formed in the junction between the lateral and ventral bar (see Figure 1A, B). Moreover, we quantified the contour of the external naris fossa in order to reflect the variation of pseudo-schizorhinal and holorhinal nares linked to proximal rhynchokinesis and prokinesis respectively (Figure 1). In the neurocranium, traits that were related to the pterygoid and quadrate bones were quantified, since these elements belong to the osteological complex involved in cranial kinesis (Bock, 1964). In addition, traits linked to the degree of ossification, such as those related to the fenestrae of the interorbital septum as well as the traits related to the temporal fossa and zygomatic process were quantified to capture the variation associated with the gracile and robust morphotypes which are frequently linked to proximal rhynchokinesis and prokinesis, respectively (Figure 1A).
To avoid introducing unnatural variation (i.e., artifacts of preservation; Adams, 1999; Rhoda et al. 2021b) due to the disarticulation of elements, such as those involved in the multi-bar linkage system (e.g., jugal bar, quadrate, pterygoid, palatal elements, and beak), we only included species represented by complete dry skulls that had no signs of deformation in the geometric morphometric analyses. In passerine bird skulls, this ensures that the angles between the different anatomical regions remain constant.
Curve semilandmarks (Bookstein, 1997) capturing the shape of the naris, tomium, and culmen were slid until minimizing the bending energy (Gunz et al., 2005). Differences in scale, position, and orientation among specimens were removed by means of Generalized Procrustes Analysis (hereafter GPA, Gower, 1975). Total skull size was measured using the log-transformed centroid size (log-Cs) of each specimen prior to GPA.
To obtain species means of shapes and sizes, Procrustes shape coordinates and log-Cs were averaged within each species. The mean Procrustes coordinates retrieved were subjected to a new round of GPA superimposition to obtain the shape coordinates that include allometric variation (hereafter full-shape). In addition, a second set of shape coordinates devoid of allometric variation (hereafter non-allometric shape) was estimated by computing the phylogenetically corrected shape residuals resulting from the linear regression of the full-shape on log-Cs (Klingenberg, 1996; Revell, 2009). To obtain summarized information on the variation of shape that would allow, on the one hand, an exploratory assessment of the data and, on the other hand, the assessment of the evolutionary convergence and the rates of phenotypic evolution, a principal component analysis (PCA) was carried out for the entire skull and on the neurocranium and beak (i.e., rostrum including naris) for both datasets (i.e., full-shape and non-allometric shape).
For the sake of brevity, the description of full-shape variation summarized by the meaningful Principal Components (PCs) of the neurocranium and the beak (i.e., those PCs retained with the Broken Stick method; Jackson, 1993) are provided in Appendix S1 of Supplementary Information. With the aim of graphically representing the variation associated with common ancestry, the morphospaces generated by the first two principal components of the entire skull, neurocranium, and beak for full-shape and non-allometric shape (Figure 3A and Figure 4A, B, C, D; see also Supplementary Figure S2A, B for morphospaces with tips labelled with species names) were depicted following a phylomorphospace approach (Sidlauskas, 2008). All the analyses were performed within the R programming environment (R Core Team, 2022) using the Morpho (v. 2.12) (Schlager, 2017), geomorph (v. 4.0.8) (Baken et al. 2021, Adams et al., 2023), morphospace (v. 0.009) (Milla Carmona, 2022), and phytools (v. 2.3-0) (Revell, 2012) packages.