METHODS, MODELS, METAPHORICAL AIDS

Background

Scale vs. Pattern-level control in development

Heliconius wing as a computer screen

Regulatory Genes and Computer Screens

Tool Box Part 2. Windows, Shutters, and Walls: the wing patterns of the MCS Clade

What about the silvaniforms, the "S" in MCS ?

Occurrence of natural introgression in Heliconius

Without doubt, the types of hybridization events herein described for synthetic hybrid zones do occur in nature (see Ackery and Smites 1976) and more recently Mallet et al. (1999) and Linares, Salazar and Gilbert, in prep.). It is straightforward to map obvious melpomene-cydno hybrids occasionally captured by collectors. It is a more difficult challenge to locate, gain access to, and document currently active areas of introgression, given the extent of ongoing political turmoil in key neotropical regions. Even then, the outcomes of introgression may be difficult to interpret without detailed study of both wing patterns and molecular traits.

From our observations of greenhouse hybridizations it would appear that the FW red gene and certain window/shutter schemes cross the melpomene-cydno boundary almost exclusively via initial mating of male melpomene with female cydno. In natural hybrid populations, recombinants from backcrosses to cydno which accidentally hit an adaptive peak for Müllerian protection and eventually come to characterize a race or species will: (1) possess cydno mitochondrial DNA, (2) possess the larger size of cydno, typically 5-7% larger in terms of wing span than melpomene, and (3) show pattern elements of melpomene such as red FW bands (e.g., the half red FW [Figure 4.lower center] of H. heurippa), all black FW (e.g., the FW of H. cydno weymeri f. gustavi, [Figure 7, lower left, b]) or the ventral HW closed forceps pattern described above (Figure 7 top, d, e, f, and h). H. cydno races/populations which have made adaptive steps in color pattern by this introgressive process may prove problematic for phylogenies based on mitochondrial DNA. Naturally occuring hybrid individuals originating in this manner will often appear sufficiently distinct to be described as new species. This senario, along with the systematist's null assumption novel molecular and morphlogical trait combinations arise by new mutation, not by interspecific introgression, is the likely explanation for a recently described MCS species, H. tristero (Brower 1996a), which I will refer to below as H. (cydno) tristero. to reflect my opinion of its likely status.

If the F1 [cydno x melpomene] males in the same hybrid zone backcross to pure melpomene females (e.g., Figure 4 lower right) the segregates will have both red forewings (the red gene being epistatic) and possess melpomene mtDNA. A phylogeny based on mtDNA would place such hybrids with melpomene and, if sufficiently distinct, they might be described as a new race like H. melpomene mocoa, an assumed mimic of H. (c.) tristero (Brower 1996a). Because it is straightforward to produce wing pattern/mtDNA combinations closely resembling Brower's new taxa, even using Costa Rican cydno and melpomene stocks in synthetic hybrid zones, it is likely that the type specimens were generated in an area which favors occasional melpomene X cydno hybridization followed by occasional backcrosses by F1 males to pure females of both melpomene and cydno. Thus, in the case of (cydno) tristero and melpomene mocoa, ), we probably have a case of melpomene FW red patterns shared by common descent from F1 fathers in an area of natural hybridization, rather than mimicry between cydno and melpomene (which is not known to exist ).

A diagnostic clue to introgression is provided by Brower's (1996a) figures of the ventral FWs of these newly described taxa. First, H. (c.) tristero, in addition to the red band on the dorsal FW shows white scales ventral to the dorsal red as seen in Browerą's (1996a Figure 7B). As mentioned previously, this is an exclusively melpomene trait. I obtain cydno similar to H. (c.) tristero only after repeated backcrossing to pure cydno females by successive generations of hybrid males which display melpomene FW red. Eventually the FW red is isolated in hybrids that are sufficiently "cydno" genomically to overcome Haldane's rule. At that point it is possible to cross male and female hybrids (a pseudo F2 produced from crossing hybrid males and females once genomic compatibility is achieved with respect to the filtered traits) and observe the segregation of the recessive "white below red" FW trait of melpomene in nearly pure cydno broods (e.g., Figure 2a and b). Certainly, based on these considerations, the type specimen of tristero as as described is not a simple first generation backcross and would indicate prolonged contact between pure cydno females and hybrid males as outlined above.

Other cydno group species and races discussed above which are good candidates to have been produced by hybridization have definable geographical ranges based on a century of collections. The lack of older collections of these new forms described by Brower suggests that he is describing the products of a relatively recent hybrid zone, possibly promoted by habitat changes as in the case of Colombian H. cydno (Linares 1989, 1997a). In the end, I doubt that both of these new taxa will coexist long as Müllerian partners of erato because of the problems of reproductive interference explained in the beginning of this paper. This is another reason to suspect that cydno tristero, melpomene mocoa, and H. erato dingus represent a transient mimetic triangle maintained by introgression. If and when cydno mitochondria go extinct in the H. c. tristero system, H. m. mocoa may live up to its new status as a race tracking H. erato. Perhaps H. c. w. f. gustavi (Figure 7, lower left b) flew briefly with a parallel melpomene "race" (which shared a common hybrid male ancestor) before the latter became extinct in the Cali region of Colombia, where Mauricio Linares studies a cydno hybrid zone.

We have seen that striking and novel phenotypes can be generated in hybrid zones between such genomically close taxa as cydno galanthus and (cydno) pachinus. Yet between species the tracks of introgression can be subtle, yet unmistakable once recognized. Thus, further understanding of the developmental genetic tool box for MCS pattern along the lines described in this chapter is required to search properly for and detect the possible mimetic steps which have been enabled by this process in natural populations. That interspecific gene exchange can account for rather subtle mimetic differences between nearby races of a Heliconius species was first shown by measuring elements of wing pattern in greenhouse populations of H. cydno introgressed with H. melpomene and then by discriminant function analysis to compare experimental hybrids with wild H. cydno weymeri and cydnides (Linares, 1989).

So far, all comparisons of experimentally produced hybrids, natural hybrids and naturally occuring mimetic patterns of various MCS races suggest that major evolutionary advances in wing pattern during the history of the Heliconius MCS clade has been based upon variation arising from introgression rather than de novo from mutation.