Removing invasive species from islands is a proven conservation intervention to prevent the extinction of native taxa. Despite growing evidence that the removal of invasive species from islands also functions as a nature-based solution for climate resilience through the recovery of native vegetation, < 20% of the 1,500 invasive species removal programs on islands incorporate long-term monitoring to quantify the effects of these interventions. Aggravated by the geographic remoteness of islands, the lack of impact assessment poses a significant challenge to maximizing conservation and climate benefits for these globally imperiled ecosystems. To guide future conservation interventions on islands, my collaborators at Island Conservation, terraPulse, HWS, SUNY ESF, and I developed a globally consistent and scalable framework to evaluate vegetative changes using remotely sensed indices from the NASA Earth Science program. Using five study islands encompassing multiple ecoregions with documented vegetation changes stemming from conservation intervention, we explored the spatially explicit changes in the extent of tree cover, vegetation productivity (NDVI, EVI, and MSAVI2), and vegetation water stress (NDMI) over a 40-year period (1984 – 2023). We tested for effects of invasive species removals and native species repatriations using an interrupted time series analysis and explored how such effects varied across landcover types and elevation. We then validated trends in remote sensing data against in situ vegetation measurements, generating important insights into the validity of our methodology for natural resource managers to complement or improve restoration trajectories. Vegetative indices varied in space and time, with significant spatiotemporal changes reflecting native forest regeneration and tortoise repatriation programs, wildfires, and the removals of invasive herbivores, predatory rats, and copra plantations. We are currently developing a Shiny application for natural resource managers to examine spatiotemporal changes in remotely sensed indices. These tool will help natural resource managers identify and prioritize areas for future conservation intervention, inform intra-island reintroduction site selection, and determine where to implement in situ monitoring to maximize efficiency and adaptive management capabilities. 

The theory of island biogeography is a central tenet of ecology, explaining that biological communities are modulated by area and isolation. While island biogeographical principles have been used post hoc to explain distributions of taxa, opportunities to observe biogeographic forcing on faunal community assembly in situ are scarce, as the generation of novel island community assemblages is rare and requires extensive timeframes to study. However, recovery of islands after invasive species eradication offers a unique opportunity to observe biogeographic forcing on community assembly at realistic spatiotemporal scales. Using data from Aotearoa New Zealand, the world's leader in invasive species eradications, my coauthors and I sought to understand the relative role of abiotic (geography) and biotic (invasive species depredation) factors in shaping seabird community assemblages. We amassed one of the largest long-term invasive mammal datasets, documenting seabird breeding presence, invasive species arrival and subsequent eradications, and restoration programs for >500 islands, with observations from 1664 – 2020. We found strong evidence that seabird assemblages are consistent with patterns of island biogeography, but these trends were disrupted by invasive mammals, with legacy effects detectable long after mammal eradication. 

We examine seabird and vegetative changes at Pokohinu (Burgess) Island, predator-free for almost 30 years, of the Mokohinau Islands. In 2009, we documented seabird distributions, plant diversity, and soil depth across the island and performed smaller surveys in 2014 and 2019. We found an overall 62.5% increase in seabird burrow densities with a strong spatial overlap of species distributions. Diversity of plant species also significantly increased, possibly due to the increased abundance of seed-dispersing birds and the simultaneous expansion of invasive plant species. In addition, we performed burrow and vegetation surveys at other Mokohinau Islands that lack same extent of past human disturbance as Pokohinu, which is a retired lighthouse and farm. Seabird densities between islands were comparable but plant communities lacked many of the invasive species found at Pokohinu. Overall, our results suggest that vegetative cover might not be limiting nesting habitat availability within the Mokohinaus, as seabirds are able to utilize areas dominated by dense vegetation. We emphasize the importance of understanding vegetative succession during recovery periods which may help explain recolonization rates for more specialist species.