The term ‘habitat fragmentation’ is frequently associated with the biologically-destructive activities of human development, but an important evolutionary hypothesis posits that much of the biodiversity we see today was generated by episodic, natural habitat fragmentation. This hypothesis suggests that fragmentation can serve as a ‘crucible of evolution’ through the amplifying feedbacks of colonization, extinction, adaptation, and speciation. Interrogating the generality of this hypothesis requires measuring the repercussions of fragmentation at intra- and interspecific levels across entire communities. We use DNA metabarcoding to capture these repercussions from the scales of intraspecific differentiation to community composition in a megadiverse, ecologically foundational group, arthropods, using a natural habitat fragmentation experiment on patches of wet forest isolated by contemporary Hawaiian lava flows (kīpuka). We find a pronounced effect of area in kīpuka cores, where the taxonomic richness supported by a kīpuka scales with its size. Kīpuka cores exhibit higher intra- and interspecific turnover over space than continuous forest. Additionally, open lava, kīpuka edges, and the cores of small kīpuka (which are essentially entirely “edge”) host lower richness, are more biologically homogeneous, and have higher proportions of non-native taxa than kīpuka cores. Our work shows how habitat fragmentation isolates entire communities of habitat specialists, paving the way for genetic differentiation. Parsing the extent to which differentiation in kīpuka is driven by local adaptation versus drift provides a promising future avenue for understanding how fragmentation, and the different isolated communities created through this process, may lead to speciation in this system.

Our current understanding of ecological and evolutionary processes underlying island biodiversity is heavily shaped by empirical data from plants and birds, although arthropods comprise the overwhelming majority of known animal species. This is due to inherent problems with obtaining high-quality arthropod data. Novel high throughput sequencing approaches are now emerging as powerful tools to overcome such limitations, and thus comprehensively address existing shortfalls in arthropod biodiversity data. Here we explore how, as a community, we might most effectively exploit these tools for comprehensive and comparable inventory and monitoring of insular arthropod biodiversity. We first review the strengths, limitations and potential synergies among existing approaches of high throughput barcode sequencing. We consider how this can be complemented with deep learning approaches applied to image analysis to study arthropod biodiversity. We then explore how these approaches can be implemented within the framework of an island Genomic Observatories Network (iGON) for the advancement of fundamental and applied understanding of island biodiversity. To this end, we identify seven island biology themes at the interface of ecology, evolution and conservation biology, within which collective and harmonised efforts in HTS arthropod inventory could yield significant advances in island biodiversity research.