DISCUSSION AND CONCLUSIONS (continued)

8. How did the tool box originate?

Almost certainly, what I have been metaphorically calling "the tool box" involved an innovation with respect to how and when in development scale cell fates are determined such that changes in few regulatory genes might have large qualitative effects on wing pattern. The question of what precisely these innovations might have been may be revealed by investigation of wing pattern development in Philaethria , a heliconiine which displays open windows of green wing membrane characterized by highly reduced scales. However, within the "windows" on both FW and HW, these paddle-like structures are basically type I (yellow/white) scales (personal obs.). Thus, Philaethria wings show important traits one might expect of a proto-Heliconius wing pattern, and, with its closest relatives, Dryas and Dryadula, constitutes an outgroup to Heliconius (Penz 1999). Dryas and Dryadula show type III scales over all window regions and would therefore provide useful comparisons for any developmental studies of Philaethria. The evolution of genes that switched ground plan of the windows from type III to type I scales in Philaethria may represent the first steps in creating the tool box shared by Eueides, Neruda, Laparus, and Heliconius, all genera which separate from other heliconiiti with respect to potential for evolving wing pattern variety under mimetic selection. Does the genus Heliconius represent an additional adaptive radiation based on some further improvement in the contents of the pattern tool box? I think not. I have argued elsewhere that pollen feeding represents the crucial adaptive zone for this genus (Gilbert 1991) and that this innovation led to unpalatibility and selected for adult behavioral traits which in turn led to population structures that promoted niche diversification.

Early in the history of Heliconius the pattern tool box might have gradually diversified through mutation. However, if we are indeed dealing with a simple set of on/off switches that govern a cascade of regulatory genes in a few sectors of the wing, and those at the bottom are in charge of only three scale types, the major ways that species and races now differ may have been quickly accumulated among early evolving species of the genus. At that point we can imagine that hybridization and introgression would have begun to accelerate in importance and quickly would have replaced mutation as the proximate generator for novel pattern genotypes in Heliconius.

The recent history of diversification of the genus Heliconius does I think, reflect two intertwined adaptive zones. The first of these is based on the habit of pupal mating and its indirect but positive effects on wing pattern diversification (as detailed for the ESS clade above). The second and to me most amazing adaptive zone, reflected in an adaptive radiation of intra-generic mimicry by the melpomene/cydno group and of ithomiine mimicry by cydno/silvaniforms of the MCS clade, is based on an upgrade of the window/shutter system (it is now more flexible, and there is a shutter color option) which further accelerated the rate at which hybridization and introgression can facilitate rapid evolution of wing pattern in the tracking of comodels. For lack of access, I have ignored the so-called "primitive species" of Heliconius (e.g., Brower 1994b), all from South America. This clade may be important in testing hypotheses developed in this chapter.