Endocasts and brain volume of dogs
Data files
Apr 19, 2023 version files 21.13 KB
-
dog_brain_volumes.xlsx
-
README.md
Abstract
Domestication is a well-known example of the relaxation of environmentally-based cognitive selection that leads to reductions in brain size. However, little is known about how brain size evolves after domestication and whether subsequent directional/artificial selection can compensate for domestication effects. The first animal to be domesticated was the dog, and recent directional breeding generated the extensive phenotypic variation among breeds we observe today. Here we use a novel endocranial dataset based on high-resolution CT scans to estimate brain size in 159 dog breeds and analyze how relative brain size varies across breeds in relation to functional selection, longevity, and litter size. In our analyses, we controlled for potential confounding factors such as common descent, gene flow, body size, and skull shape. We found that dogs have consistently smaller relative brain size than wolves supporting the domestication effect, but breeds that are more distantly related to wolves have relatively larger brains than breeds that are more closely related to wolves. Neither functional category, skull shape, longevity, nor litter size was associated with relative brain size, which implies that selection for performing specific tasks, morphology, and life history do not necessarily influence brain size evolution in domesticated species.
Methods
We processed the collection of dog skulls that is maintained at the Department of Anatomy, Cell and Developmental Biology, Eötvös Loránd University (Budapest, Hungary). This private collection (owned by TC) is composed of specimens that have been obtained mostly in the last 10 years by the appropriate preparation of the heads of deceased dogs (which were donated post-mortem), from which the soft materials have been removed a priori. TC systematically collected the prepared skulls with the aim of having both male and female samples from as many breeds as possible. Breed identity was usually verified upon the collection of cadavers/skulls, given that these materials originate from known dog breeders. Alternatively, we checked the appropriate breed certificates/chips for pedigree. Currently, the collection consists of 383 individual skulls (including males, females and unknown sexes) from 146 breeds. We selected 172 skulls (38 females, 83 males and 50 unknown sexes) across all breeds represented in the collection for subsequent CT scan analysis (see Supplementary Material, Table S1). Skulls were selected from adult individuals, which we verified using morphological characteristics (i.e., the presence of permanent teeth, as dogs should replace all baby teeth before 6-7 months of age).
The selected skulls were transferred to the Diagnostic and Oncoradiology Centre in Kaposvár (Hungary) for CT scanning. We used a Siemens Somatom Definition AS+ CT machine (Siemens, Erlangen, Germany) to digitalize the skulls with high resolution (170 mAs, 140 kV, pixel size 0.323 × 0.322 mm, slice thickness 0.6 mm, with a v80u bone kernel). The resulting DICOM image series were imported into the 3D Slicer software (freeware, www.slicer.org), and using its segmentation and modelling tools, the endocranial volumes (=endocast) were reconstructed (see details in Czeibert et al. 2020). These endocasts reflect the surface morphology of the brain in such detail that external blood vessels and differences in gyrification can be observed (Figure 1). In parallel, we calculated the volume of the endocasts for the analysis (Czeibert et al. 2020) in this study.
We also extracted additional data on brain volumes from the literature for some dog breeds.
Usage notes
Microsoft Excel, R