Monozygotic twinning: an evolutionary hypothesis.

Monozygotic twinning is rare within populations yet taxonomically widespread. We explore the evolution of monozygotic twinning by modeling an allele in a newly formed offspring that causes it to undergo mitosis and separation to form one or more clones (twins), potentially in conflict with the parents' best interest. The success of this twinning allele in our haploid models depends on the balance of the benefit of increased frequency in the clutch and the cost of reduced survival resulting from limited parental resources. The trait reaches high frequency in a broad range of plausible conditions but also fails to spread or is kept at low frequency in others when the survival cost is high (e.g., in small clutch sizes). Interestingly, there are two reasonable conditions that predict high frequency of the trait but low visibility: random parental abortion and selection for low penetrance. Thus our models suggest reasons why monozygotic twinning might be rare, or alternatively, be common yet appear rare. In addition, we discuss the implications for sex-linked twinning, dizygotic twinning, and twinning by gametes.