original plans following unanticipated experimental challenges, new technological developments and/or serendipitous scientific observations. Aim 1. We made and validated a variety of tools in D. sechellia for labelling specific populations of neurons, for tracing neuronal architecture with fluorescent reporters, for physiological measurements of neuronal activity using calcium sensors, and for artificial neuronal activation with light-sensitive ion channels. We also generated and validated more general genetic tools for efficient integration of new transgenes into the genome. Aim 2. We surveyed a variety of behaviours of D. sechellia (e.g. chemosensory, visual, mechanosensory, oviposition, circadian as well as interactions with parasitoids and microbes) in comparison with those of D. melanogaster – and another close cousin, D. simulans – leading us to identify both conservation and divergence between these species. Notably, our study of circadian behaviours led us to discover a loss in plasticity of D. sechellia, in that it cannot adapt its circadian rhythms to changing day-length, which we linked to regulatory differences in a key circadian neuropeptide gene. Neuronal circuit comparisons focussed principally on the olfactory system, where functional significance can be more easily interpreted and investigated, and we characterised changes in cell number and projection patterns at several levels of this sensory system. The original proposal to survey neuronal gene expression patterns across these drosophilids by bulk RNA-sequencing of nervous system tissues was superseded by single cell-resolution transcriptomics approaches, allowing us to generate the first cross-species atlases for the whole central brain in any species; comparisons of these atlases pointed to multiple examples of D. sechellia-specific cell type representation and gene expression patterns, notably in glial populations. Aim 3. our efforts to map the genetic basis of OSN population expansions in D. sechellia revealed a surprisingly complex genetic architecture underlying these phenotypic differences, which unfortunately prevents pinpointing of the responsible genes. Greater progress was made in our understanding of the functional significance of these neuronal population increases, through comparative neurophysiological analyses in D. sechellia and D. melanogaster. We found that increased sensory pooling of OSNs results in reduced post-synaptic depression in their partner projection neurons in D. sechellia compared to D. melanogaster, enabling more persistent transmission of repeated or prolonged olfactory signals by these interneurons to higher brain centres. This phenomenon reveals a striking synergy with the changes in sensory sensitivity conferred by olfactory receptor tuning evolution. Taken together, this project has generated tools, datasets and results to reveal many new insights into how and why nervous systems have evolved. Moreover, it lays the foundation for a wealth of future research on other divergent behaviours and brain regions (as well as other phenotypes) of these drosophilid species, defining principles and mechanisms of relevance for all animals.