Fossilized embryos are widespread but the record is temporally and taxonomically biased

Philip C. J. Donoghue, Artem Kouchinsky, Dieter Waloszek, Stefan Bengtson, Xi‐ping Dong, Anatoly K. Val'kov, John A. Cunningham, John E. Repetski

We report new discoveries of embryos and egg capsules from the Lower Cambrian of Siberia, Middle Cambrian of Australia and Lower Ordovician of North America. Together with existing records, embryos have now been recorded from four of the seven continents. However, the new discoveries highlight secular and systematic biases in the fossil record of embryonic stages. The temporal window within which the embryos and egg capsules are found is of relatively short duration; it ends in the Early Ordovician and is roughly coincident with that of typical "Orsten"-type faunas. The reduced occurrence of such fossils has been attributed to reducing levels of phosphate in marine waters during the early Paleozoic, but may also be owing to the increasing depth of sediment mixing by infaunal metazoans. Furthermore, most records younger than the earliest Cambrian are of a single kind-large eggs and embryos of the priapulid-like scalidophoran Markuelia. We explore alternative explanations for the low taxonomic diversity of embryos recovered thus far, including sampling, size, anatomy, ecology, and environment, concluding that the preponderance of Markuelia embryos is due to its precocious development of cuticle at an embryonic stage, predisposing it to preservation through action as a substrate on which microbially mediated precipitation of authigenic calcium phosphate may occur. The fossil record of embryos may be limited to a late Neoproterozoic to early Ordovician snapshot that is subject to dramatic systematic bias. Together, these biases must be considered seriously in attempts to use the fossil record to arbitrate between hypotheses of developmental and life history evolution implicated in the origin of metazoan clades.