Research Projects

Climate change is driving dramatic shifts in biodiversity, impacting ecosystems and human livelihoods. Among its many consequences, it poses the greatest threat to global food security, as changes in temperature and rainfall patterns affect crop yields directly. Moreover, climate change also indirectly disrupts food production by affecting insect diversity—key agents of pollination for over 80% of food crops. Despite this growing crisis, our understanding of how climate change influences insect populations remains surprisingly limited. Certain species may be less vulnerable to extinction if they can rapidly evolve their climatic niches, i.e., the range of environmental conditions in which they can survive and reproduce. Behavioral, morphological, or physiological adaptations could enable some species to cope with rising temperatures. However, the rate of climatic niche evolution varies significantly across species and clades, with the underlying drivers remaining poorly understood. Furthermore, no studies have yet explored climatic-niche evolution in diverse insect assemblages, leaving a critical gap in our understanding of biodiversity resilience. This project aims to address two key questions: (i) what is the rate of climatic-niche evolution in insects, and (ii) how do different morphological adaptations shape species’ climatic tolerances? To answer these, I will integrate (a) intensive field sampling across a steep and broad climatic gradient along the Eastern Himalayas in Arunachal Pradesh, (b) a time-calibrated species-level phylogeny, and (c) a neural network architecture. Moths (Lepidoptera) are an ideal model system for this study due to their immense species and morphological diversity (e.g., variations in size, shape, and coloration), sensitivity to temperature and vegetation, stable taxonomy, ease of identification, and vital ecological roles as herbivores, pollinators, and prey. For the first question, I will reconstruct the phylogeny of sampled Lepidoptera species by grafting species onto a time-calibrated, family-level backbone tree using the software PartitionFinder2. Contemporary climatic niches will be calculated from weather sensors, and ancestral values will be reconstructed with phylogenetic generalized models. To address the second question, I will extract image features—geometrical, spectral, and textural—using a ResNet-101 architecture. Discriminant Analysis will be used to assess how effectively these features predict species’ climatic niches. High-resolution images of individual moths will serve as input for this pipeline. By integrating the findings across a diverse set of families and along a comprehensive climatic gradient, this research will provide critical insights into the climate sensitivity of insects. The results will establish a benchmark for understanding global species responses to climate change, and inform mitigation strategies for conservation of vulnerable species.