The paradox of polymorphic mimicry in H. doris
The strong selective forces that drive Müllerian mimicry are predicted to result in monomorphism among mimicking species, yet as in H. doris , there are many examples of polymorphic mimicry in nature. Our study sheds some light on how this paradox may be achieved. Our results suggest that pFDS can vary at regional scales, and is constrained to knowledgeable predator communities which are savvy to the aposematic forms found only in their local ecosystem (Chouteau et al., 2016; Langham, 2004). For example, over the relatively short distance of ~30km, we found significant differences in the attacks on native red morphs of H. doris , with significantly less attacks occurring at the sites where other red co-mimics are present. This suggests the predator community knowledge was quite distinct at the different sites and corresponds to reports of Jacamars having rather narrow home ranges (Chai, 1986). However, this begs the question of “how do the red H. doris persist in areas lacking red co-mimics?”, as we would expect the lack of co-mimics to result in higher predation and eventual removal of the red morph from the population.
A possible explanation lies in the dispersal behaviour from nearby populations where red co-mimics are present and the red H. dorismorphs have greater protection. Other Heliconius species such asH. erato and H. melpomene have an estimated dispersal range of only ~2.5-5km (Mallet, 1986a; Mallet et al., 1990), as a result of their “trap-line” behaviors as adults (Young & Montgomery, 2020). However, it has been suggested that Heliconius doris may disperse much larger distances immediately post pupal eclosion, which could reduce chances of sib-competition and sib-matings (Mallet, 1999).Heliconius doris females are known to gather in groups and lay eggs on single plants, often even the same leaf, which we observed first-hand in French Guiana. This results in a mass of gregarious larvae that will often fully consume all leaves and tendrils on thePassiflora host. After consumption, an individual host plant can require several years to reach a size sufficient to host another population of H. doris eggs. It would then likely benefit newly eclosed females to disperse larger distances than otherHeliconius species that tend to oviposit much fewer eggs in close proximity. Therefore, it is possible that group egg laying, and relatively greater dispersal in H. doris could drive a “mismatch” of warning colors in the distribution of Heliconiusco-mimetic species, as seen in French Guiana. This dispersal-based hypothesis would result in sink populations for H. doris morphs, where the red co-mimics are lacking, that are continuously replenished from source populations where red morphs have greater protection. It is difficult to understand how this could be an evolutionarily stable strategy and dispersal data for H. doris are lacking to support such a source-sink model for the presence of red H. doris morphs where the co-mimics are absent.
Another important aspect that could explain the distribution of red morphs and polymorphic mimicry in H. doris is the genetic basis for the color variation. In H. numata, polymorphic color patterns result from allelic changes at a single locus, P(Joron et al., 2006). More specifically, the different color patterns result from varying combinations of chromosomal inversions across the P locus (Joron et al., 2011). The color pattern variation is maintained in local populations through disassortative mating (Maisonneuve et al., 2021), a form of negative frequency-dependent selection where rare morphs are preferred mates resulting in offspring of variable colorations. Since the color pattern differences are controlled by a single locus, and the different alleles cannot recombine due to the inverted orientations (Jay et al., 2021), disassortative mating will keep producing color pattern variation in perpetuity. In H. numata , each of the different color patterns also correspond to local co-mimics, and different morphs appear to share similar predation pressures (Chouteau et al., 2016). We propose that a similar system may have evolved in H. doris , with non-recombining alleles at a single locus controlling color pattern variation coupled with disassortative mating as such a system would result in distinct red and blue morphs in each generation across theH. doris range. Currently, there is no data for the inheritance of color patterns or mate preference in H. doris , which would be vital for determining how polymorphic mimicry is maintained in the species.