A bird egg can look absurdly large for the body that produced it. In deep time, the puzzle becomes even stranger: the largest known bird eggs — laid by the extinct elephant bird of Madagascar, bigger than even an ostrich's — exceed the largest eggs found from non-avian dinosaurs, even though many of those dinosaurs dwarfed every bird that ever lived.

A new study by researchers at the American Museum of Natural History (AMNH) and Princeton University, published in Royal Society Open Science, offers an explanation. Stephanie Lechki (Princeton) and Roger Benson (AMNH) analyzed reproductive and anatomical data from mammals, birds, and reptiles together in a single evolutionary framework and found a consistent link: species with relatively larger brains tend to produce fewer, larger offspring — whether that investment takes the form of a larger egg or a larger newborn.

The logic, the authors argue, is energetic. Growing a large brain is expensive, and that cost must be paid up front in each individual offspring. "At first glance, this seems counterintuitive," said Lechki. "Many non-avian dinosaurs were enormous, yet even the biggest dinosaur eggs were smaller than the largest bird eggs. Our results suggest that the answer lies in brain evolution. As birds evolved larger brains, they also evolved larger offspring, which required larger eggs."

The pattern also gives new context to famous fossils, including the nesting oviraptorosaurs found by AMNH expeditions to the Gobi Desert in the 1990s. A troodontid egg clutch from the Gobi, for instance, held eggs about half the size they would be in a bird of equivalent size. "This work places those remarkable fossils into a much larger macroevolutionary picture," Benson said. "If larger brains required larger offspring, and larger offspring required larger eggs, then other aspects of anatomy and behavior may have evolved in response."

Those knock-on effects could include a wider pelvic canal for laying, more open and aerated nest structures for incubation, and greater parental investment after hatching. The researchers caution that some details of extinct reproduction — such as how many times a year an animal bred — rarely leave fossil evidence, and they plan to test whether the pattern holds among living species that reproduce multiple times per year.