Finding a needle in a heath stack: A strategy to optimize the detection of a rare marsupial on the brink of local extinction

Abstract

Anthropogenic pressures are reducing species distributions and driving extinctions within ecosystems. As the urgency to address biodiversity loss intensifies, the disappearance of species from their ecological niches poses increasing challenges for effective conservation management. This is particularly true for elusive species that have remained undetected for extended periods. Here we use limited locational data to build species distribution models (SDMs) for a threatened marsupial, the long-nosed potoroo (Potorous tridactylus trisulcatus). In Gariwerd (Grampians National Park) the species has been rarely observed following the significant landscape impacts of multiple megafires and drought events. We developed a SDM from sparse historical data (n = 17 records) to guide targeted camera trapping (2021–2024). Despite data limitations, the SDM facilitated the discovery of previously unknown extant populations. We iteratively refined our models through incorporating new field data, tracking parameter changes, and improving model fit. Final predictions identified suitable but currently unoccupied habitat that may be important for enhancing population resilience. We also performed connectivity analysis on habitat patches containing potoroos to estimate connectivity between extant populations and inform conservation planning. Our findings demonstrate the utility of preliminary SDM's, built on extremely limited data, as a starting point for managing rare and threatened species, enabling more targeted interventions to address threatening processes. Furthermore, our approach identified currently unoccupied suitable habitat critical for enhancing population resilience. These findings demonstrate the utility of SDMs, even with limited data, to direct surveys and identify key areas for proactive conservation strategies, including threat management and potential translocations, for data-deficient and declining species globally. Despite the utility of our models, we emphasise the critical importance of field validation and advocate for its inclusion as an essential step in distributional models.

Paper suitability: biological conservation

The escalating biodiversity crisis, driven by anthropogenic pressures, necessitates innovative approaches to conservation management, particularly for elusive species where distributional knowledge is often severely limited. The challenge of determining species persistence in the face of potential extinction is a critical impediment to effective conservation action. In this study, we address this challenge by demonstrating a novel methodology that leverages species distribution models built from extremely sparse historical presence records to strategically guide survey efforts and successfully identify extant populations of rare species. We further illustrate how incorporating newly acquired presence data refines these models, enabling the identification of not only current strongholds but also critical, high-quality habitat that remains unoccupied. By highlighting the potential of this approach to pinpoint both surviving populations and key areas for future conservation interventions, including threat management actions and reintroduction, this research offers a valuable framework for proactive and evidence-based conservation planning in landscapes facing significant biodiversity loss, a core concern for the readership of Biological Conservation.