Understanding thermal adaptation is a priority if we wish to understand and mitigate the impacts of climate change. Organisms' current tolerances and their ability to adapt or move to new areas will determine whether they survive warming environments or not. There is currently a relative paucity of data on thermal adaptation in tropical as compared to temperate species. Tropical insects make up a large proportion of the Earth's biodiversity, but little is currently known about their ability to respond to climate change. We propose to use a well-studied group of tropical ectotherms, the Heliconius butterflies, to assess thermal adaptation across altitudinal gradients, to determine the contributions of genetic and environmental variation to these traits, and to identify the underlying genetic loci. Within Heliconius there are several instances of subspecies that also show structuring by altitude. Using population genomic sequence data and the reference genome for H. melpomene we recently showed that three subspecies pairs of H. melpomene and H. erato across altitudinal gradients exhibit divergence in regions of the genome that contain candidate genes for thermal adaptation such as metabolic enzymes and heat shock proteins. We will characterise the ecophysiology of these species, through characterisation of the physiological differences found between populations and species. Within Ecuador these species have parallel altitudinal clines on the east and west side of the Andes allowing us to assess the replicability of any trends we identify. Temperature can also be an important factor in delimiting the ecological niche and so in driving divergence and speciation. Research on the Heliconius system has largely focussed on the role of colour patterns in driving divergence between populations and species, but habitat differences are also present and thought to contribute to divergence. For example, the species pair H. melpomene and H. cydno on the western side of the Andes, which have become a classic system in speciation research, show divergent altitudinal ranges. The phenotypic differences driving divergence in habitat use have yet to be investigated, but physiological adaptations to temperature seem highly likely. Therefore, we will experimentally test for thermal adaptation differences between the sister clades H. melpomene and H. cydno/timareta, to assess the importance of temperature in determining altitudinal range and species distributions in these species. We will use the extensive genomic resources available for Heliconius to perform genome scans and detailed association mapping analyses to identify loci responsible for thermal adaptation within H. melpomene. Identifying these loci will then allow us to address the question of whether genes involved in temperature adaptation show evidence for either introgression or divergent selection between species. Sharing of temperature adaptation genes between species could allow rapid adaptation to novel environments, while divergence would suggest thermal adaptation is important in maintaining species identities. Research on Heliconius has flourished in recent years leading to many insights into the process of divergence and speciation in the genome. However, the ecological characters investigated have remained largely restricted to colour pattern. This project will be a major step towards establishing Heliconius as a more comprehensive model system for ecological genetics, making use of the existing knowledge, genomic resources and techniques available in this system to investigate broader ecological issues in the tropics.

1) The big questions Convergent evolution is a natural experiment in repeated evolution of similar phenotypes, offering unique insights into the evolutionary process. When similar patterns evolve in different lineages, to what extent are the same molecular mechanisms deployed? Are the same regulatory changes co-opted into generating convergent phenotypes? How can evolutionary change at a single locus regulate complex patterning changes during rapid evolution? What are the precise genetic changes necessary for the evolution of a new developmental pattern? Heliconius butterflies are an excellent system to address these questions. 2) The background Many tropical butterflies have mimetic wing patterns to warn predators of their toxicity, and these have become an excellent system in which to understand the molecular basis for convergence and diversification, and make the link between natural selection in the wild and evolution in the genome. Here we will study the molecular basis for pattern convergence in tropical Heliconius butterflies. Genetic mapping and gene expression experiments have identified a simple system of three genetic loci that control the complex diversification in wing patterning seen in Heliconius. One of these loci regulates yellow pattern elements and the same genomic locus is also involved in wing patterning of both the peppered moth, Biston betularia, and the butterfly Bicyclus anynana, suggesting an ancient shared patterning system in butterflies and moths. Patterns of expression and genetic data from natural populations, suggest that cortex is the functional gene at this locus, although expression data also point to involvement of other linked genes. We will apply recently developed CRISPR/Cas9 methods for gene knockouts to investigate the molecular basis for pattern convergence between species of Heliconius butterflies. 3) Objectives and expected results We will test the frequency with which similar patterns evolving in mimetic butterflies use the same genes. At a closer resolution, we will also test whether those genes are controlled by the same regulatory switches to turn them on and off, when they control similar patterns. This will test for the repeatability of evolution in different genetic backgrounds. These experiments will involve developing and applying novel gene editing techniques to study patterning in these butterflies, which will set the standard for evolutionary studies in the future. The remarkable patterns of mimicry in these butterflies have long been considered an exemplar of evolution by natural selection, and this project will offer unique new insights into the molecular mechanisms that produce such strikingly similar patterns in so many different species.

Convergent evolution, the independent acquisition of similar traits in multiple lineages in response to the same selective pressures, is ubiquitous, facilitating adaptation and diversification across the tree of life. Therefore, understanding the genetic mechanisms by which convergence occurs is critical if we are to understand adaptations that already exist, and the predictability of evolution in response to common selection pressures. We propose to study mimetic convergence across the Lepidoptera using high-throughput sequencing and gene expression analyses to address a major challenge in this field: the contributions of different genetic mechanisms to convergence across evolutionary timescales. This will be the first genetic analysis of convergence for a trait evolving under the same selective force over 2-110 million years of evolution and will uncover the genetic landscape of convergence across evolutionary time. The genetic changes causing convergence can be categorized as divergent genetic mechanisms, parallel evolution, or collateral evolution. We hypothesize that these three processes act at different evolutionary time scales. Most recent understanding of convergent evolution has focused on parallel and collateral evolution among closely related species. We lack studies that investigate the genetic basis of convergence over a range divergence times (from recent to deep time) for a single trait under the same selective force. Only by considering convergence among lineages that split anywhere from a few million to 100 million years ago, or more, can we understand the overall frequency distribution of the genetic mechanisms of convergence. The relative contributions of the three genetic mechanisms will impact on the tempo and direction of evolutionary convergence. For example, interspecific hybridization can greatly facilitate convergence among closely-related species, yet its contribution to convergence is largely unknown. We also lack knowledge of the genetic basis of deep time convergence. An important unanswered question is whether convergence between distant lineages is difficult to evolve. Alternatively, is convergence aided by the existence of conserved genetic architectures and developmental pathways, which may facilitate parallel evolution even after 100 million years of separation? We propose to tackle these fundamental questions about the genetic mechanisms of convergence by exploiting a unique system in the Lepidoptera in which multiple species have converged on the same defensive wing colour patterns across a wide range of evolutionary timescales (2-110 million years). We will use a combination of fieldwork, gene expression analysis and the latest high-throughput sequencing technologies to identify and verify genes responsible for convergence in multiple butterfly and moth species. These data will allow us to assess the relative contributions of divergent genetic mechanisms, parallel and collateral evolution to convergence among 18 species of butterflies and moths representing 2-110 million years of evolution, and will allow us for the first time to visualize the genetic landscape of convergent evolution for a single trait evolving under the same selective force across a wide evolutionary timescale.