Chapter 5: Paper 4

2.4.1Problem statement¶

Research shows that Australia is a global hotspot for endemism in plants, and that Tasmania is particularly rich in paleoendemic species Cai et al., 2023. However, Australia ranks among the worst globally when it comes to adequate threat assessments for endemic flora Gallagher et al., 2023. Slow growth rate and poor dispersal are characteristics shared by many of Tasmania’s iconic paleoendemic plant species that evolved in cool and stable environments Mokany et al., 2017. Consequently, they all share the same sensitivity to fire due to their inability to recover between fire events Bliss et al., 2021. The intensity and frequency of wildfires is increasing in Tasmania and around the world Bowman et al., 2022Feng et al., 2025 and tolerable fire intervals may already have been exceeded for some taxa Leonard, 2021. Genetic diversity metrics decreased significantly with an index of fire history for Athrotaxis Worth et al., 2017.

The IUCN Red List of Threatened Species is a global framework that aims to assess and monitor the extinction risk of biota IUCN, 2025. The assessment uses five primary criteria (most with sub-criteria), where taxa must meet one or more of the primary criteria to be categorised as threatened. The criteria include population trends, geographic range characteristics, population size, and quantitative analyses of extinction risk IUCN Standards and Petitions Committee, 2024. While the IUCN Red List criteria are comprehensive, there are many compromises and limitations which are unavoidable for a project of this scope. For example, using future projections as justification of vulnerability is encouraged but not mandatory if these criteria are fulfilled using historical evidence. Critically, it is the responsibility of the assessor to judge suitability of the population data (with approval by peer review) used in the evaluation. These flexibilities facilitate assessments in the absence of data -- which is common for rare and enigmatic taxa -- but risks masking vulnerabilities in taxa with long generation times and a lagged response. The result may be an (unintentional) misrepresentation of the true vulnerability of the species, which can have real world consequences for the conservation efforts Possingham et al., 2002.

Regarding the most recent IUCN Red List assessments for our target taxa Table 2.4.1. The habitat data used for the three conifers was generated between 24 to 8 years prior to the assessment in 2012/13 and does not use future projections of habitat suitability. Additionally, the refer to the resilience of these taxa on account of their ability to resprout, and the significant habitat coverage by protected areas (e.g. Tasmanian Wilderness World Heritage Area). The Nothofagus spp. assessments are more recent and comprehensive: exploiting species occurrence databases and research modelling habitat suitability in future climates. Importantly, the conifers were assessed prior to the bushfire events in 2015-2016 and all assessments were made prior to the 2019/2020 bushfires Bowman et al., 2021Leonard, 2021.

The aim of this chapter is not to critique the IUCN Red List or revise the vulnerability category of these taxa. Rather, these IUCN assessments are presented as examples to illustrate that even major conservation organisations like the IUCN are working with and drawing conclusions from incomplete data. Predictably, research shows that “garbage in, garbage out” is true for conservation planning and the “garbage out” is in the form of poor conservation outcomes Biggs et al., 2011Friess & Webb, 2011Wilson et al., 2005.

2.4.2Proposed methods¶

The objective of this chapter is to uncover vulnerabilities of sensitive vegetation on a local scale, demonstrate the importance of local scale dynamics, and the limitations of coarse vegetation mapping for this purpose. To achieve this, we will develop species distribution models (SDMs) using MaxEnt Phillips et al., 2006 to predict habitat suitability for both mature trees and recruitment habitat (i.e. seedling nurseries) under current and projected climate scenarios. The SDMs will include environmental predictors related to climate, fire, topography, and geology. We create three SDMs each using a different occurrence dataset.

• SDM1: DL-derived occurrences

• SDM2: TASVEG vegetation communities Department of Natural Resources and Environment Tasmania, 2020

• SDM3: ALA occurrences Atlas of Living Australia, 2025

We will assess vulnerability by implementing a spatial multi-criteria decision analysis (MCDA) using the variable weights generated by the MaxEnt SDM Rodríguez-Merino et al., 2020. Habitat connectivity analysis will be performed using circuit theory approaches implemented in Circuitscape McRae et al., 2008 to assess dispersal corridors between suitable mature tree habitat and potential recruitment areas. Model performance will be evaluated using area under the curve (AUC) metrics, and spatial cross-validation techniques to account for spatial autocorrelation in occurrence data. Results will be validated against independent field survey data collected in the study area.

We will use the resulting vulnerability maps and the associated variable weightings for each of the three SDMs to characterise the threats to the local population and how these may inform short- and long-range management strategies. The characterisations will enable us to identify the strengths and limitations of the three different datasets as a basis for informing management practices. Acknowledging that the results of this analysis cannot be generalised beyond the target subpopulation, we will draw comparisons between our findings and existing vulnerability evaluations at the population scale (e.g. IUCN Red List) and how automated inventories can add value.

This research will require collaborations with key Tasmanian institutions to ensure access to high-quality datasets and local expertise. The University of Tasmania Climate Futures Research Group will provide climate projections specifically developed for Tasmania, including temperature and precipitation scenarios under multiple representative concentration pathways. These collaborations will ensure access to the most recent climate modelling outputs at appropriate spatial resolutions for a study-site-scale suitability model. Natural Resources and Environment Tasmania (NRE Tas) will contribute expert ecological knowledge on the target taxa, particularly regarding species-specific habitat requirements, fire tolerance thresholds, and observed recruitment patterns across different landscape contexts. Additionally, NRE Tas will provide access to the most current TASVEG mapping data. The Tasmania Fire Service (TFS) will assist with bushfire triage parameters, such as accessibility for preventative and active firefighting.

2.4.3Key innovations¶

This chapter presents three key innovations in local-scale vulnerability assessment for endemic Tasmanian flora. First, we demonstrate a novel comparative framework for evaluating the relative strengths and limitations of different occurrence datasets (automated population inventory, manual community-level mapping, and citizen science records) in vulnerability modelling, providing practical guidance for conservation practitioners working with heterogeneous data sources. Second, we integrate habitat connectivity analysis with multi-criteria vulnerability assessment to identify not only where species are most at risk, but also which populations serve as critical source habitats and dispersal corridors for long-term persistence. Third, we develop a methodology for translating local-scale vulnerability assessments into actionable management recommendations by incorporating operational constraints such as firefighting accessibility and protected area boundaries.