Most mammalian embryos develop for weeks to months inside the mother’s body. This developmental strategy of giving birth to live young –known as viviparity– both protects the developing offspring and allows the mother to continuously supply nutrients over an extended period. Apart from monotremes, or egg-laying mammals, all other mammals give birth to live young, making studies of their early development difficult. Collection of prenatal materials for species other than established models is extremely challenging, both practically and ethically, due to factors such as the body size, habitat, reproductive biology, and conservation status.

Natural history museums around the world offer a simple, ethical means of studying mammalian development because their collections consist of hundreds to thousands of historically (and opportunistically) collected prenatal specimens for a wide range of mammals. By using non-invasive imaging technologies, such as micro-computed tomography (micro-CT), researchers can image the developing skeleton without harming these precious specimens and digitally share these data with the scientific community through public repositories.

This study was inspired when we discovered that the wet collections of the Harvard Museum of Comparative Zoology (MCZ) Mammalogy Department contain hundreds of mammalian embryos preserved in jars of ethanol which had not been previously used in research. Many were collected over 100 years ago.

To conduct our research, we used micro-CT to study the developmental sequence of bone formation (ossification sequence) in the mammalian order Carnivora. Species included the Steller sea lion, small Indian mongoose, lion, gray fox, brown bear, and small Indian civet. We chose to focus on carnivorans because previous research on carnivoran ossification sequences was largely limited to two domestic species: dogs and cats. Furthermore, the MCZ collection included singleton embryos from a wide taxonomic range of species, as well as numerous Steller sea lion embryos of different developmental stages. This diversity allowed us to compare both within and between species, as well as among animals living in terrestrial versus aquatic environments.

After CT-scanning, we digitally reconstructed the developing skeleton of each specimen and identified which bones were present. Based on these data, we could determine the order of bone formation by comparing the bones observed with those expected based on the adult skeleton. For example, if the femur was present in one specimen while the tibia was absent, we infer that the femur ossifies before the tibia in this species. Using this logic, we constructed ossification sequences for the skull and postcranial bones of each species. We then used an analysis method called Parsimov-based genetic inference to identify heterochronic changes in ossification sequence between species and to reconstruct ancestral ossification sequences throughout the carnivoran tree. Briefly, this method treats ossification sequence as a single complex character and generates trees to identify the minimum number of heterochronic shifts necessary to explain the data.

Carnivores are divided into two major lineages, the dog-like ‘caniforms’ and cat-like ‘feliforms.’ As luck would have it, each of these lineages includes a popular domestic species. We found that both the dog and cat are good models for the ossification sequence of their respective linages. This is a helpful finding since dogs and cats are much easier to study than other species, making them very useful models in both veterinary science as well as carnivoran biology in general. Pinniped species, however, were the exception; seals and sea lions showed differences in the timing and order of ossification of different bones – i.e.  skeletal heterochronies – when compared both to each other as well as to the terrestrial carnivorans. This finding matches the high rates of heterochrony observed in the sutures of the crania of adult pinnipeds and may be related to adaptations to aquatic locomotion.

The micro-CT scans generated through this study are now publicly available via the Morphosource repository and we hope these data will facilitate the inclusion of embryonic data in future studies of carnivoran biology. We encourage other researchers to explore the collections available at their local natural history museum to gain critical insights into development and evolution.

We would like to thank the Genes, Organisms, and Ecosystems Research Experiences for Undergraduates (GEO REU) program for making this research possible, our reviewers for helpful comments on this manuscript, and the editor for inviting us to write this post.