Prioritising naturalised plant species for threat assessment: developing a decision tool for managers: final report

Abstract Naturalised, but not yet invasive, plants pose a latent threat to Australia’s biodiversity. Approximately 10% of the almost 30,000 non-native plant species introduced to Australia have formed self-sustaining or naturalised populations in the landscape. Whilst some of these naturalised plants have become highly invasive where they pose significant threats to biodiversity, primary production and human health (e.g. the Weeds of National Significance), the majority have yet to become invasive and as such may be described as ‘sleeper weeds’. As these species have yet to become invasive, their management is often piecemeal unless they have been assessed as having a high weed risk potential. Given the time and resources required to undertake each weed risk assessment, many of these plants have not been assessed at all. The potential for changing climate regimes to create more favourable conditions for these plants, and thereby increase their invasive potential, is yet to be assessed. The central aim of this research was to assess the current extent of environmentally suitable habitat for a suite of naturalised, but not yet invasive non-native plants within Australia and to evaluate how projected changes in climate may alter these patterns in the coming decades. We used species distribution modelling – a recognised method for building spatial projections of suitable habitat based on correlations between known occurrences of species and environmental variables – as the basis for assessing the threat posed by 292 naturalised, but not yet invasive plants under current and future climates. Individual species profiles were created by compiling key trait data, observation records and maps of current habitat suitability and projected change in suitability across Australia. These profiles will serve as the basis to evaluate how the habitat suitability for each species is likely to change in the future under different scenarios. As a second level of assessment, we also constructed vulnerability maps of both current and future time periods by overlaying binary maps for all 292 species. In doing so we identified ‘hotspots’ of climatically suitable habitat for large numbers of naturalised plants in the Australian landscape, and provided an assessment of the risks posed on a state-wide and national scale. Overall, the southerly coastal areas and Tasmania have the highest risk of invasibility, under both current and modelled future (2035) climates under high (RCP 8.5) emission scenarios based on species numbers. Finally, we developed a point-based prioritisation scheme, based on: (1) gridded observations per 100,000 km2 (2) habitat suitability of observations (3) area of habitat suitability (4) area of highly suitable habitat and (5) minimum distance between gridded observations and highly suitable habitat. Using this assessment process we classified 4% of Australia-wide species as having a high risk of invasion, 83% as having a medium risk, and 13% as having low risk, under current climate conditions. However, the percentage of high-risk species when assessed at a state and territory scale differed widely between states. Under a future scenario (RCP 8.5 2035) at a national scale we classified 3% of the species as having a high risk of invasibility, 81% as having a medium risk, and 16% as having low risk. It is envisaged that this prioritisation approach for determining weed management priorities for naturalised plants, will be the basis for a tool for allocating economic and human resources for on-the-ground actions now and in the future in light of climate change. For identified high risk species, long-term management programs and allocation of resources will be required. This information provides vital baseline data for prioritising which naturalised plants to target for weed risk assessments, as well as active management/intervention (e.g. containment and/or eradication) now and in the future. The outputs generated from this research will be freely available to end-users via a web-based portal. This portal is a decision-support tool that provides end-users with the ability to interrogate individual profiles for 292 species and interactively map emerging weed threats for regions or management units of interest. Additionally, there is species-level habitat suitability information categorised by: the Collaborative Australian Protected Area Database (CAPAD); Local Government Areas (LGA); wetlands of international significance (RAMSAR); Natural Resource Management regions (NRM) and Interim Biogeographic Regionalisation for Australia (IBRA 7) areas. Our integration of modelling, spatial analysis and species trait information provides a comprehensive assessment of which naturalised plants to target as Australia’s climate changes. Such assessments provide significant economic benefits by targeting control to high priority naturalised, but not yet invasive plants before they become significant problem weeds.