Biodiversity conservation has long pitted “natural” landscapes against human-managed landscapes. However, biodiversity has long interacted with human-shaped landscapes, starting with Indigenous management thousands of years ago. As the impacts of climate change and biodiversity loss take hold on our ecosystems, the need to see the whole landscape as a partner in biodiversity conservation, rather than nature reserves as the only solution, is imperative. Agricultural landscapes are particularly noted for their ability to bolster biodiversity while producing food, though this largely depends on their management and level of homogenization. Heterogeneous agricultural landscapes over space and time can sustain native biodiversity, even on a scale outpacing non-managed landscapes. Agricultural landscapes do not form independently but are shaped by the hands of farmers and communities. Therefore, understanding the social factors that shape farmer and community decision-making is crucial, along with examining how to ethically and responsibly conduct research with communities.

In this dissertation, I utilize a socio-ecological framework to examine if and how Kānaka ʻŌiwi (Indigenous Hawaiian) agroecosystems can conserve native arthropod biodiversity. First, I describe a framework to identify what mechanisms shape the ability of agroecosystems globally to conserve different taxonomic groups. Second, I present sociological findings based on interviews showing that Indigenous and non-Indigenous farmers have other motives for engaging in agriculture, with ramifications on crop diversity and community impact. Next, I utilize empirical findings based on DNA metabarcoding and Next Generation Sequencing data to showcase how arthropod alpha and beta diversity trends shift between simplified and diversified farms within an agroforestry system on Hawaiʻi island. I then examine the research process I undertook for this chapter through an Indigenous Data Sovereignty framework. Lastly, capitalizing on advances in DNA metabarcoding, Next Generation Sequencing, and information theory, I construct bipartite networks between predators (spiders) and their prey to illustrate how the structure and stability of networks change depending on agricultural landscape composition.