Reply to: Triassic sauropodomorph eggshell might not be soft

In the accompanying Comment, Choi et al. dispute morphological, histological, geochemical and biomechanical evidence for a soft-shelled egg in Mussaurus, an Early Jurassic sauropodomorph, based on three main assertions, namely that: (1) the carbonate signal associated with Mussaurus eggshell could represent original biomineral, despite its chemical and petrographic resemblance to the limy sediment matrix; (2) the endogenous N-, O- and S-heterocyclic polymers, which generate the Mussaurus egg morphology and encode the biomineralization signal, might represent sedimentary kerogen (refuted in figure 2a of ref.) or graphitic carbon, which cannot possibly form under the pressure and temperature conditions in question; (3) birefringence, which would indicate eggshell calcite, may be obscured in the thin section by organic matter (contra the previous claim suggesting only graphite might be preserved). Choi et al. disagree with our analyses of the Mussaurus eggshell while ignoring the biomechanical evidence and compound maps presented in Norell et al. and accepting identical analyses for the Protoceratops eggshell. Choi et al. did not perform new analyses, but scored the Mussaurus eggshell as hard on the basis of comparisons with hard-shelled but unrelated non-dinosaurian eggshell (Aenigmaoolithus) from a dissimilar age, region and volcanic setting. They re-ran our ancestral state reconstruction without time scaling and obtained an ambiguous 52% probability of hard-shelled eggs in the ancestral dinosaur, in contrast to our inference of an ancestrally soft egg.