Butterflies have evolved different colour patterns on their dorsal and ventral wing surfaces to serve different signalling functions, yet the developmental mechanisms controlling surface-specific patterning are still unknown. Here, we mutate both copies of the transcription factor apterous in Bicyclus anynana butterflies using CRISPR/Cas9 and show that apterous A, expressed dorsally, functions both as a repressor and modifier of ventral wing colour patterns, as well as a promoter of dorsal sexual ornaments in males. We propose that the surface-specific diversification of wing patterns in butterflies proceeded via the co-option of apterous A or its downstream effectors into various gene regulatory networks involved in the differentiation of discrete wing traits. Further, interactions between apterous and sex-specific factors such as doublesex may have contributed to the origin of sexually dimorphic surface-specific patterns. Finally, we discuss the evolution of eyespot number diversity in the family Nymphalidae within the context of developmental constraints due to apterous regulation.

Supplementary Figure 1: Majority-rule consensus bayesian phylogenetic tree of 53 species of the Bicyclus genus with Hallelesis as the outgroup. The tree was generated in MrBayes v3.2.7a. Numbers at the node indicate posterior probabilities. Supplementary Figure 2: Notation of wing sectors and veins on Bicyclus anynanawings. The names of the sectors (within the wing) and veins (outside the wing) used to reference the different locations on the wing disc is shown.

SummaryDirect regeneration of shoot buds in vitro is an important technique in plant genetic manipulation. We describe the isolation and functional characterization of a novel MADS box cDNA (PkMADS1) from Paulownia kawakamii leaf explants undergoing adventitious shoot regeneration. mRNA gel blot analysis con®rmed the expression of PkMADS1 in the shoot-forming cultures, but no signal was observed in the callus-forming cultures. PkMADS1 transcripts were also detected in shoot apices, but not in root apices, initial leaf explants or the¯ower. In situ hybridization revealed that its expression was restricted to developing shoot primordia in the excised leaf cultures, suggesting a role for this gene in adventitious shoot formation. Transgenic Paulownia plants over-expressing the PkMADS1 gene showed some changes in phenotype, such as axillary shoot formation. In the antisense transformants, shoots were stunted and had altered phyllotaxy, and, in some lines, the shoot apical meristem appeared to have been used up early during shoot development. Leaf explants from the antisense transgenic plants showed a tenfold decrease in shoot regeneration compared with explants from sense transformants or wild-type. Our results show that PkMADS1 is a regulator of shoot morphogenesis.

Species co-occurrence in ecological communities is thought to be influenced by multiple ecological and evolutionary processes, especially colonization and competition. However, effects of other interspecific interactions and evolutionary relationships are less explored. We examined evolutionary histories of community members and roles of mutualistic and parasitic interactions (Müllerian and Batesian mimicry, respectively) in the assembly of mimetic butterfly communities called mimicry rings in tropical forests of the Western Ghats, India. We found that Müllerian mimics were phylogenetically clustered, sharing aposematic signals due to common ancestry. On the other hand, Batesian mimics joined mimicry rings through convergent evolution and random phylogenetic assembly. Since the Western Ghats are a habitat island, we compared species diversity and composition in its mimicry rings with those of habitat mainland to test effects of biogeographic connectivity. The Western Ghats consisted of fewer mimicry rings and an overall smaller number of aposematic species and mimics compared to habitat mainland. The depauperate mimicry rings in the Western Ghats could have resulted from stochastic processes, reflecting their long temporal and spatial isolation and trickling colonization by the mimetic butterfly communities. These results highlight how evolutionary history, biogeographic isolation, and stochastic colonization influence the evolutionary assembly and diversity of ecological communities.