The initiation of metamorphosis causes the cessation of the larval growth period which determines the final body size of adult insects. Because larval growth is roughly exponential, differences in timing the initiation of metamorphosis can cause large differences body size. Although many of the processes involved in metamorphosis have been well characterized, little is known about how the timing of the initiation of metamorphosis is determined.

Using different strains from Tribolium castaneum, Tribolium freemani, and Manduca sexta and varied nutritional conditions, I was able to document the existence of a threshold size, which determines when the larva becomes competent to metamorphose. Threshold size, however, does not dictate the exact timing of initiation. The exact timing for the initiation of metamorphosis is determined by a pulse of the molting hormone, ecdysone, but only after threshold size has been reached. Ecdysone pulses before the larva attains threshold size only cause the larva to molt to another larval instar. These results indicate the timing of metamorphosis initiation is controlled by two factors: (1) attainment of threshold size, at which the larva becomes competent to initiate metamorphosis and (2) the timing of an ecdysone pulse after attaining threshold size.

I hypothesize the attainment of threshold size, and therefore competence to metamorphose, is mediated by the effect of changing juvenile hormone concentrations caused by the increase in size of the larva. While the larval body grows nearly exponentially, the corpora allata, which secretes juvenile hormone, grows very little if at all. The difference in relative growth causes juvenile hormone concentrations to gradually become diluted. When juvenile hormone concentrations fall below a threshold, changes in protein-protein binding occur that can cause changes in signaling networks and ultimately gene expression. These changes make the larva competent for metamorphosis.

I have demonstrated that only threshold size is consistently correlated with body size; other growth parameters such as growth rate, duration of instars, or number of instars do not consistently correlate with variation in body size. Using the black mutant strain of M. sexta I have shown that lower juvenile hormone titers correlate with lower threshold sizes. My hypothesis is consistent with the large body of literature indicating the involvement of juvenile hormone. I also hypothesize that the diversity of metamorphosis types in holometabolous insects can be explained by heterochronic shifts in the timing of threshold size and other developmental events related to metamorphosis. The heterochronic shifts affect not only the morphology of organs, but can also affect the overall phenotypic response of the larva to changes in the environment. The different phenotypic responses among species may make the more or less suited for certain types of niches.