Conservation priorities for functionally unique and specialized terrestrial vertebrates threatened by biological invasions

Abstract

Invasive non‐native species (INS) continue to pose a significant threat to biodiversity, including native population declines, which can ultimately disrupt ecosystem processes. Although there is growing evidence of the impacts of INS on functional diversity, most of the existing approaches to prioritization of species for conservation still focus on taxonomic diversity, neglecting the ecological role of species. We developed the functionally unique, specialized, and endangered by invasive non‐native species (FUSE INS) score to fill this gap by combining functional irreplaceability (i.e., uniqueness and specialization) of species with their extinction risk due to INS. We calculated this score for 3642 terrestrial vertebrates exposed to INS by assessing how INS affected them based on the IUCN Red List and by evaluating their specialization and uniqueness in a multidimensional functional space. Thirty‐eight percent of native species were both at high extinction risk because of INS and functionally unique and specialized, making them priority species for INS impact mitigation. Priority species of amphibians concentrated in Central America and Madagascar and of lizards in the Caribbean islands, northern Australia, New Zealand, and New Caledonia. Priority bird and mammal species were more widespread (birds, mostly in coastal areas, on Pacific islands, and in northern India and New Zealand; mammals, in southwestern Europe, Central Africa, East Africa, Southern Africa, Southeast Asia, and eastern Australia). Seventy‐eight species were also highly irreplaceable but not yet threatened by INS, suggesting that preventive conservation measures may help protect these species. For the 50 birds of the highest priority, 64% required conservation actions to mitigate the INS threat. The FUSE INS score can be used to help prioritize indigenous species representing large amounts of functional diversity. Incorporating functional diversity into the conservation prioritization of species and associated areas is key to accurately reducing and mitigating the impacts of INS on native biodiversity.

INTRODUCTION

According to the International Union for Conservation of Nature (IUCN) Red List, 333 contemporary extinctions have occurred worldwide due to biological invasions and nearly 7000 species are currently at risk of extinction or extirpation due to the effects of invasive non‐native species (INS) (defined by Pyšek et al. [2020]). It is expected that 2301 species will go extinct within the next 50 years (IUCN, 2022). The consequences of biological invasions go far beyond a simple number of declining species; they represent profound changes in the functional diversity of vertebrate communities (Bellard et al., 2021). For instance, Sayol et al. (2021) found that non‐native bird species lead to the functional homogenization of insular avian communities. The number of introduced bird species is often equal to or higher than the number of extinct birds on islands; however, they do not seem to compensate for the functional loss due to the extinction of functionally distinct birds (such as large‐bodied birds with low flight ability and animal‐ or nectar‐based diets [Sayol et al., 2020; Soares et al., 2022]). INS threaten 10% of the taxonomic and 50% of the functional diversity of insular amphibians (Marino et al., 2022). Thus, focusing on taxonomic diversity greatly underestimates the impacts of INS, especially regarding the maintenance of ecosystem functions and services (Egerer et al., 2018). Therefore, other dimensions of diversity need to be accounted for in the establishment of conservation priorities to ensure they cover the complete picture of INS impacts on native biota.

Taxonomic diversity has been for decades the main indicator of the state of biodiversity (Purvis et al., 2019); thus, most conservation initiatives have relied on this dimension (e.g., global biodiversity hotspots [Myers et al., 2000]). This practice disregards species’ functional role and evolutionary history (Bonn et al., 2002; Rodrigues & Gaston, 2002). Moreover, taxonomic diversity is an unreliable indicator of other diversity dimensions, such as phylogenetic and functional diversity (Brum et al., 2017); therefore, it is even more important to integrate them into conservation prioritizations (Diamond & Roy, 2023; Gaüzère et al., 2022).

New metrics have been developed to better integrate the multidimensionality of diversity, accounting for species’ evolutionary history or functional diversity, in conservation measures (Cui et al., 2024; Gumbs et al., 2023; Hidasi‐Neto et al., 2015; Isaac et al., 2007; Pimiento et al., 2020). For example, the evolutionarily distinct and globally endangered (EDGE) metric, which accounts for species’ evolutionary distinctiveness and extinction risk, has been applied to many taxa and integrated into conservation schemes (Gumbs et al., 2018). In parallel, Pimiento et al. (2020) developed the “functionally unique, specialized, and endangered” (FUSE) metric and applied it to marine megafauna to identify species of critical importance for functional diversity and at high extinction risk. Nevertheless, the integration of functional features into conservation policies remains anecdotal, especially the FUSE metric that has, to our knowledge, never been applied to any other taxon besides marine megafauna and elasmobranchs (i.e., sharks, rays, and skates) (Pimiento et al., 2023). Although conservation measures require identifying the threats species face, most prioritization metrics do not account for the specificity of threats. For instance, INS have caused the extinction of native species with unique functional traits, some of which no longer exist in the current fauna (e.g., birds: Matthews et al., 2022). Once identified, INS need to be controlled or eradicated to save native species from extinction. It is thus crucial to define a species conservation prioritization strategy that takes into account the traits of species and their uniqueness and specialization among their taxonomic group from a threat perspective.

We developed the FUSE INS score to identify species that are both endangered by INS and functionally unique and specialized, meaning they are irreplaceable in terms of functional strategies in the global functional space. Taking advantage of recent improvements in the FUSE and EDGE frameworks (Gouhier & Pillai, 2020; Griffin et al., 2020; Gumbs et al., 2023), we implemented a new method to calculate the FUSE INS score of species and applied it to almost all (not crocodiles, turtles, or snakes) extant terrestrial vertebrates (n = 27,841 species). We relied on 2 major improvements that consider the uncertainty associated with the extinction probability of assessed species, recently developed by Gumbs et al. (2023) (EDGE2) and the severity of the impact of INS on native species, based on the IUCN Red List.

To date, all the major international institutions agree that biological invasion represents one of the main direct drivers of biodiversity loss (Bellard et al., 2022). In December 2022, COP15 of the Convention on Biological Diversity adopted a specific target for reducing and mitigating the impacts of INS on native biodiversity (Secretariat of the United Nations Convention on Biological Diversity, 2021). To achieve this goal, it is essential to identify the species and areas most vulnerable to INS impacts (Essl et al., 2020), not only by considering the imperiled taxonomic diversity but also by accounting for the functional value of species and ecosystems. In this context, this index dedicated to species affected by INS will help identify and prioritize functionally irreplaceable vertebrates at high risk of extinction due to INS and identify and monitor species (and associated sites) that are not yet threatened by INS but have a high functional irreplaceability, representing an opportunity for the future of biodiversity. Preserving species that present exceptional traits will directly contribute to the conservation of global functional diversity. Because INS involve specific measures of prevention, control, and eradication relative to habitat loss or overexploitation, which require other types of conservation measures, we focused only on INS threat.

METHODS

Rationale

Inspired by the FUSE approach, we built the FUSE INS score to focus on species that have a high extinction risk due to INS and are of particular importance for functional diversity (Pimiento et al., 2020). We took advantage of the recent development of the EDGE2 metric to quantify the uncertainty associated with the species’ probability of extinction. The FUSE INS score of a given species is the product of its extinction risk due to INS (P _INS) and its functional irreplaceability accounting for all species in the taxonomic group. Species that are strongly threatened by INS and functionally irreplaceable will have a high FUSE INS score, whereas species that are less affected by INS or less exceptional in terms of ecological characteristics will have a lower score. Specifically, we defined the FUSE INS score of a given species as the log‐transformed combination of its extinction probability due to INS threat (P _INS) and functional irreplaceability, captured by functional specialization (FSp) and functional uniqueness (FUn), as follows (Figure 1):

$$\mathrm{FUSE}\mathrm{INS}=\ln (1+P_{\mathrm{INS}}\times \mathrm{FSp}+P_{\mathrm{INS}}\times \mathrm{FUn}).$$ (1)

FIGURE 1.

Schematic workflow of the FUSE INS (functionally unique, specialized, and endangered by invasive non‐native species) score calculation: (a) estimated extinction probabilities (P _INS) randomly selected within the range of probability values of an International Union for Conservation of Nature (IUCN) Red List category and the INS impact magnitude (details on ranges of probability values in Appendix S2); (b) functional specialization (FSp) and functional uniqueness (FUn) measured in a functional space of n dimensions, adapted to each taxonomic group, and normalized (range 0–1); and (c) 1000 FUSE INS scores computed as the combination of P _INS, FSp, and FUn for each species and species lists built based on the median FUSE INS score of each taxonomic group and the conservation status of species (types of lists described in “Final metric and priority lists”).

Species threats and conservation status

We characterized species threats, conservation status, and conservation actions with the advanced research tool from the IUCN Red List of Threatened Species (IUCN, 2022). First, we identified the terrestrial vertebrates exposed to INS as all species associated with the threats “invasive non‐native/alien species/diseases” or “problematic species/diseases of unknown origin” (the latter, only when we could determine the non‐native origin of the named problematic species). This process identified 1895 amphibians, 816 birds, 508 lizards, and 423 mammals affected by INS (3642 total species). Vertebrates not exposed to INS (n = 24,199) received a P _INS of 0 and thus a FUSE INS of 0. We also collected information regarding the other threats these species face. We then extracted information about the scope and severity of the INS threat specifically to quantify the INS impact magnitude. Species with >50% of their total population experiencing a significant decline (i.e., slow, rapid, or very rapid) due to INS were classified as sustaining a high impact from INS, whereas species with <50% of their population experiencing fluctuations or no declines were classified as experiencing a low impact. Species that had no information for the INS impact magnitude were classified as NA.

Second, we collected the IUCN Red List category (i.e., global conservation status) of species affected by INS: least concern (LC), near threatened (NT), vulnerable (VU), endangered (EN), critically endangered (CR), or data deficient (DD). Finally, we extracted the conservation actions needed (which indicate the conservation actions or measures needed for the concerned animal), based on category “invasive/problematic species control” from the Conservation Actions Classification Scheme (version 2.0), as conservation actions for INS management.

Extinction probability of species associated with INS threat

Following the EDGE2 protocol, we incorporated uncertainty in extinction probability (Gumbs et al., 2023). Specifically, we computed a smoothed distribution of extinction probabilities as in Gumbs et al. (2023). The median probability of extinction was 0.06 for LC, 0.12 for NT, 0.24 for VU, 0.49 for EN, and 0.97 for CR species (Mooers et al., 2008; Appendices S1 & S2). Thus, each IUCN Red List category had an associated range of extinction probabilities (e.g., 0.0001–0.09 for LC species and 0.7–0.9999 for CR species) (ranges of all categories in Appendix S1). The size of probability ranges differed among the categories, introducing a higher variance in extinction probability for intermediate categories. A species at very high (or low) extinction risk is more likely to become extinct (or not), with very low uncertainty, as illustrated by a low variance in P _INS. In contrast, a species at moderate extinction risk could become extinct but could also thrive for more than 50 years. In such cases, the associated uncertainty is higher at those intermediate risk of extinctions, illustrated by a larger range of P _INS. This method for deriving P _INS presents several advantages: extinction probabilities are continuous, avoiding a strong bias in the final score due to a coarse categorization of species extinction risk and allowing a smooth changing of categories; some uncertainty is attributed to each probability; and the amount of uncertainty depends on the IUCN Red List category (Gumbs et al., 2023).

We further divided each IUCN Red List category into 2 subcategories based on the INS impact magnitude. We then drew 1000 values of P _INS for each species by randomly selecting 1000 P _INS in the range corresponding to the species’ subcategory, considering both conservation status and INS impact magnitude (Figure 1a; Appendix S2). For instance, if a species is CR with a high INS impact magnitude, P _INS would be randomly picked from 0.874 to 0.9999, whereas for a CR species with a low impact magnitude, it would be picked from 0.7 to 0.874. For NA species regarding the INS impact magnitude, P _INS values were selected in the whole range of extinction probabilities of the species’ conservation status. For DD species, P _INS was selected in the whole range of the 10 categories (0.0001–0.9999), regardless of conservation status and INS impact magnitude.

Functional irreplaceability definition

The functional value of species was measured in a functional space, which was defined as a multidimensional space containing all species from a taxonomic group, where axes are a combination of species traits. The functional irreplaceability represents the degree of specialization and uniqueness of species in the functional space (Pimiento et al., 2020). Functional specialization (FSp) accounts for species’ distance from the average strategies of the taxonomic group, and functional uniqueness (FUn) measures the level of isolation of each species in the functional space of its respective taxon (Mouillot et al., 2013).

Trait data and functional spaces

We combined a global list of 27,841 terrestrial vertebrates (6492 amphibians, 10,943 birds, 5505 mammals, and 4901 lizards), for which we had trait information from Marino et al. (2022) for amphibians, lizards, and mammals and from Marino and Bellard (2023) for birds. Because reptile traits were poorly reported for major groups, such as snakes, turtles, and crocodiles, we focused on lizard species for this taxon. Traits were related to species morphology, life history, and ecology, which together reflect species functional strategies (Marino & Bellard, 2023; Marino et al., 2022). We ended up with a total of 4 traits for amphibians, 10 traits for birds, and 5 traits for both lizards and mammals (details on trait definition and sources are in Appendix S3). These traits are commonly used in studies evaluating functional diversity and summarizing the effects of species on ecological processes and their responses to environmental change (Bellard et al., 2021; Cooke et al., 2019; Marino et al., 2022).

We built a multidimensional functional space for each taxonomic group separately with a 2‐step approach that allowed us to consider various types of variables (e.g., continuous, ordinal, factorial). First, we computed pairwise trait‐based distances between species with the Gower (1971) dissimilarity index. Second, we calculated species coordinates in the functional space by applying a principal coordinate analysis (PCoA) on the distance matrix to reduce the number of dimensions. The trait‐based distance matrix was obtained with the funct.dist() function from the mFD package, and the PCoA was computed with the pcoa() function from the ape package (Magneville et al., 2021; Paradis & Schliep, 2019). Then, we evaluated the quality of spaces with an increasing number of principal components to build them. Quality was assessed by the root mean square deviation (rmsd) between trait‐based distances and distances in the PCoA‐based space, and the best functional space was obtained for the lowest rmsd score (Maire et al., 2015). The best functional spaces contained 3 dimensions for amphibians (rmsd = 0.070), 7 dimensions for birds (rmsd = 0.032), and 5 dimensions for lizards (rmsd = 0.071) and mammals (rmsd = 0.069) (Appendix S4). The cumulated variance explained by the final functional spaces ranged between 69% and 74% (Appendix S4).

Functional specialization and uniqueness

Finally, for a given species i, we calculated FSp as the Euclidean distance between i and the n‐dimensional functional space centroid and FUn as the sum of the Euclidean distances between i and its 5 nearest neighbors (Figure 1b). When combined, FSp and FUn represent the functional irreplaceability of a given species within its own taxonomic group (Pimiento et al., 2020). Both metrics were normalized from 0 to 1. For FSp, 0 represented the most central species in the functional space, whereas 1 was the most distant species from the centroid (Appendix S5). For FUn, 0 represented the least isolated species (given the distance from its 5 nearest neighbors), and 1 was the most isolated species (Appendix S6). The 2 metrics were not correlated because they were complementary for all taxonomic groups (Appendix S7).

Final metric and priority lists

We calculated the FUSE INS score for each species with Equation (1) and used the values of P _INS, FSp, and FUn calculated previously. Because all 3 metrics ranged from 0 to 1, the FUSE INS score ranged from 0 to ln(3)≈1.1. We also calculated FUSE INS scores with 2 alternative formulas to quantify the sensitivity of this index to different formulations (Appendix S8). Each species received 1000 FUSE INS scores, one for each of the 1000 generated P _INS, from which we calculated the median score per species and its standard deviation. Furthermore, for each taxonomic group, we calculated the median score based on all species of the taxonomic group to classify species in priority lists based on different conservation objectives (Gumbs et al., 2023) (Figure 1c): core, research, watch, and borderline lists. The core list contained threatened species with a CR, EN, or VU conservation status that had more than 95% of their scores above the median score of all species of their respective group. Species in this list were of high functional value and would require an urgent and adapted conservation response to prevent their extinction. The research list contained DD species with more than 95% of their scores above the median score of their respective group. This list would help target species that are of high functional value but have not yet received enough research to have a proper conservation evaluation. The watch list contained nonthreatened species with an NT or LC conservation status that had more than 95% of their scores above the median score of their respective group. Species in this list were of high functional value but not immediately prone to extinction; however, they would benefit from proactive conservation strategies. Finally, the borderline list contained threatened species with a CR, EN, or VU conservation status that had 80–95% of their scores above the median score of their respective group. Species in this list had a high extinction risk but were not as functionally exceptional as the species from the core list. However, if species in the core list became extinct, they would be the next species on the verge of extinction harboring the most functionally distinct features.

We derived the 4 lists for each taxon (i.e., amphibians, birds, lizards, and mammals). We also analyzed the lists’ sensitivity to the 80% and 95% thresholds by building the lists based on thresholds of 75% (or 85%) and 90% (or 97.5%) (results in Appendix S9).

Species distribution ranges and richness maps

We collected the distribution range of amphibians and mammals from the IUCN Red List, of birds from Birdlife, and of lizards from the GARD database (BirdLife International & Handbook of the Birds of the World, 2020; IUCN, 2022; Roll et al., 2017). We considered only range polygons where the species were considered extant, native, or reintroduced and resident or breeding. We then extracted the species’ potential presence in hexagonal cells 110‐km long (dimension of 1° cells at the equator) to derive maps of species richness. At the cell level, we evaluated the correlation between the species richness of FUSE INS species and the sum of FUSE INS scores of all present species. Because the 2 metrics were highly correlated (Pearson's correlation coefficient >0.80 for all groups) (Appendix S10), we displayed only the maps based on the species richness of FUSE INS species.

All analyses were conducted with R 4.2.2 (R Core Team, 2022). Data and codes used are archived in Zenodo and linked to a GitHub public repository: https://zenodo.org/doi/10.5281/zenodo.10716188.

RESULTS

We calculated the FUSE INS score of 3642 terrestrial vertebrates listed as associated with the INS threat on the IUCN Red List (the scores for all 3642 species are available from Zenodo [https://zenodo.org/doi/10.5281/zenodo.10716188]). The scores ranged from 0 to 0.992, thus never reaching the maximum possible value. The highest score was for kakapo (Strigops habroptila), a large, flightless, endemic, ground‐nesting, nocturnal parrot from New Zealand. Median scores for each group were low overall: 0.05 (SD 0.13) for amphibians, 0.06 (SD 0.11 and 0.13) for birds and lizards, and 0.04 (SD 0.09) for mammals. The FUSE INS scores calculated with alternative formulas were highly correlated with the main formula (Pearson's correlation coefficient >0.99 for all taxonomic groups and formulas) (Appendix S8) and resulted in the same ranking of species.

We identified 1378 species in the FUSE INS core list: 684 amphibians, 343 birds, 199 lizards, and 152 mammals (Figure 2). These represented 38% of all species exposed to INS (complete lists with species scores are in Appendix S15). We did not identify any species belonging to the FUSE INS research list (i.e., species with high FUSE INS score but of DD status; 211 out of 3642 species were DD). However, 78 species were listed on the FUSE INS watch list (nonthreatened species with high FUSE INS scores), and the large majority was classified as NT. Forty‐three species were assigned to the borderline list, including 5 critically endangered species. The thresholds used for defining the 4 lists had a negligible effect on the species included on the lists, except for the FUSE INS research list for amphibians, which contained 39 species with lower thresholds (i.e., 70% and 90% [Appendix S9]). In all taxonomic groups, most species across the 3 lists were associated with only one INS. Except for amphibians, which were mostly associated with 2 fungi (Batrachochytrium dendrobatidis and Batrachochytrium salamandrivorans), invasive mammals stood out as the most commonly identified INS group across all lists and taxa. However, for 29% of species, INS were referenced as unspecified species.

FIGURE 2.

Number of species in the core, watch, and borderline lists (lists described in “Final metric and priority lists”) (circles, all species in the lists for each taxon; numbers inside circles, number of species; CR, critically endangered; EN, endangered; VU, vulnerable; NT, near threatened; LC, least concern).

Throughout all taxonomic groups and lists, most species tended to be associated with other threats besides INS (details in Appendix S11). Seventy‐eight species were threatened by INS alone (i.e., not threatened by any other pressure): 11 amphibians, 19 birds, 26 lizards, and 8 mammals on the core list; one amphibian, 3 birds, 4 lizards, and one mammal on the watch list; and 3 birds and 2 lizards on the borderline list (Appendix S12). On average, species on the core list were associated with more than 4 threats, including the INS threat (mean [SD] = 4.1 [1.8]), except lizards that were on average associated with 3 threats. Biological resource use and agriculture and aquaculture were among the most common other threats associated with amphibians, birds, and mammals. Lizards were threatened mostly by agriculture and aquaculture, followed by natural system modifications. Birds and mammals were the only groups associated with climate change and severe weather.

Overall, information on INS impact magnitude was lacking. Information was missing for 88% of amphibians, 67% of mammals, 53% of lizards, and 21% of birds on the core list. Because birds had more information than other taxa, we focused on the 50 birds on the core list of the highest conservation priority to illustrate associated threats and conservation actions needed based on the IUCN Red List classification (Figure 3). More than half of them were threatened only by one INS. Invasive mammals affected 56% of the species, followed by invasive birds (22%), plants (12%), and other taxa to a lesser extent. Almost all of these species were threatened by other threats besides INS (92%). The most commonly identified threats were biological resource use (70%) and climate change and severe weather (56%), followed by agriculture and aquaculture (48%), natural system modifications (38%), and pollution (34%). All 50 species were cited as in need of conservation action (i.e., land and water protection or management) by the IUCN. Specifically, 64% of the 50 birds of the highest conservation priority were associated with unimplemented INS control.

FIGURE 3.

Functionally unique, specialized, and endangered by invasive non‐native species (FUSE INS) scores for the 50 birds of the highest conservation priority on the core list (i.e., species of high functional value and at high extinction risk due to INS) (bars, median score; error bars, standard deviation of the 1000 scores computed for each species; CR, critically endangered; EN, endangered; NA, missing impact information) (left) and other threats besides INS to species and conservation actions needed for INS management (green, biological resource use; brown, habitat loss and degradation [combined residential and commercial development, agriculture and aquaculture, energy production and mining, transportation and service corridors, human intrusion and disturbance, and natural system modification]; orange, climate change and extreme weather; red, pollution; gray, other threat types; black, conservation actions needed; filled square, existence of a threat or conservation action needed related to that species) (right).

Birds and mammals on the core list were widespread, whereas amphibians and lizards were regionally clustered (Figure 4). We detected hotspots of core‐list species in Central America (especially Guatemala, Costa Rica, northern Ecuador, and northern Colombia) and eastern Madagascar for amphibians and in the Caribbean islands, northern Australia, New Zealand, and New Caledonia for lizards. For birds, areas with the highest density of core‐list species were the coasts of the Southern Hemisphere, Pacific islands (including Hawaii and Galapagos), northern India, and New Zealand. For mammals, species on the core list occurred mostly in the East African Rift, Zambia, Botswana, the Pyrenees, Madagascar, Himalayas, Southeast Asia, and eastern Australia. Species on the watch list occurred all over the world for birds and mammals, but amphibians and lizards were restricted to a few areas (e.g., northern Andes, Madagascar, New Zealand) (Appendix S13). In contrast, all species on the borderline list, regardless of taxa, were restricted to areas, such as Madagascar or western Pacific islands, with species with small range sizes (Appendix S14).

FIGURE 4.

Global maps of species on the core list (i.e., species of high functional value and at high extinction risk due to INS) for each taxon (values, taxonomic richness of species ranging from 1 to 22 for amphibians and birds and from 1 to 7 for lizards and mammals; gray, no species on the core list).

DISCUSSION

Across terrestrial vertebrates, we identified more than 1300 species with high functional irreplaceability at high risk of extinction, partly because of INS (including 78 with INS as the only threat). On the core lists, amphibians represented most of the species with high FUSE INS scores, followed by birds, lizards, and mammals. In general, amphibians have the highest amount of taxonomic and functional diversity threatened by biological invasions (Bellard et al., 2016b; Marino et al., 2022), which our FUSE INS scores confirmed. Overall, the high number of species on these priority lists reinforces the vulnerability of these taxonomic groups to biological invasions (Bellard et al., 2016a). Moreover, it highlights the importance of considering species traits when evaluating the INS threat because many of their impacts jeopardize more than taxonomic richness alone. The FUSE INS scores can be used to identify species and areas of high conservation priority relative to the threat of INS. This approach also has the potential to be used for the other conservation goals listed, along with warnings and recommendations, in Table 1.

TABLE 1.

Warnings, recommendations, and ways to adapt the functionally unique, specialized, and endangered by invasive non‐native species (FUSE INS) score to set conservation goals.

Score construction
Warning	Traits used for functional space construction must encompass species’ ecological niches to better represent their role in the ecosystem. Thus, in general, the number of traits to consider must be >3.
Warning	Use at least one continuous trait to avoid a patchy distribution of species in the functional space.
Recommendation	Data on species extinction risks, threats, or traits are regularly updated. The score can be recalculated to match the most updated information.
Possible adaptation	For taxa with fewer available traits at a global scale, such as plants or arthropods, the score can be calculated on a taxonomic or geographic subset. In that case, regional International Union for Conservation of Nature (IUCN) assessments can be used instead of the global IUCN Red List.
Possible adaptation	Criteria for establishing priority lists are fully adaptable: the threshold of 95% of score values above the median can be adjusted to fit the project's specific objectives.

Score interpretation and priority lists
Warning	Absolute value of the index is not informative per SE. Use the ranking of species to obtain priorities based on the conservation objectives.
Recommendation	Report other threats associated with species on the lists to manage the INS threat.
Recommendation	Consider providing proactive conservation guidance for species in the watch list a and not restraining conservation efforts to core list b species.
Recommendation	Score ranks species based on their functional irreplaceability, but one could incorporate species’ evolutionary history. The score can thus be combined and compared with other metrics, such as the EDGE score adapted to INS threat.

^a Species of high functional value but not immediately prone to extinction (LC and NT).

^b Species of high functional value and at high extinction risk due to INS (CR, EN, or VU).

Identification of priority species and areas with the FUSE INS score

Assessing the spatial distribution of species from the FUSE INS core lists allowed us to identify areas that concentrated functionally irreplaceable vertebrates at high risk of extinction due to INS. Core‐list birds occurred mostly along coastlines of the Southern Hemisphere but were also present on many Pacific islands and New Zealand. Pacific islands are the object of several conservation initiatives, including marine protected areas (PAs), to protect bird reproduction and ensure population persistence for key species. For instance, important bird and biodiversity areas (IBAs) are defined based on their contribution to global avifauna preservation (Donald et al., 2019) and highlight zones of high irreplaceability in terms of taxonomic richness (Di Marco et al., 2016). These areas could gain importance and receive even more attention if other facets of diversity were added as a criterion, for example with the FUSE INS score, which considers species functional irreplaceability.

We identified some hotspots of core‐list species that, despite the exceptional species they host, are not represented in any IBAs, such as New Zealand or coastlines. An extreme example is the kakapo, the species with the highest FUSE INS score. This species has particular ecological characteristics that are generally associated with a high vulnerability to INS, such as ground‐dwelling foraging, habitat specialism, large size, and plant‐based diet (Marino et al., 2022). As the largest parrot in the world and the only one that does not fly, this species is certainly exceptional in terms of functional strategies at the global scale. Coastal birds are also distinct regarding functional traits. For instance, the New Zealand storm‐petrel (Fregetta maoriana), Guadalupe storm‐petrel (Hydrobates macrodactylus), and MacGillivray's prion (Pachyptila macgillivrayi) are seabirds that nest on islands and, together with the kakapo, are among the 50 birds of the highest conservation priority on the core list that are threatened only by INS (i.e., not by other threats according to the IUCN Red List).

Apart from birds, most species on core lists had no information regarding their INS impact magnitude; up to 88% of amphibian species lacked information. This large deficiency is likely to reduce the importance of the INS impact compared with the conservation status itself in the final calculations of the FUSE INS score. Despite the limited information, we still found that amphibians represented half of the species on the core lists. Amphibians affected by INS are mostly threatened by the 2 chytrid fungi responsible for the chytridiomycosis disease, which has caused declines in more than 500 amphibian species worldwide (Scheele et al., 2019). The chytrid is widespread throughout the world but has the greatest effect on large‐sized and range‐restricted amphibians from the tropical climates in the Americas. Therefore, it was not surprising to find that amphibians on the core list were restricted to zones in Central America and the northern Andes.

Lizards on the core list were almost all from northern Australia or native to islands (oceanic islands in the Caribbean and Pacific, as well as New Zealand and New Caledonia), of which 81% were insular endemics. These findings are consistent with the facts that insular endemic species are more prone to extinction than continental species and threatened predominantly by INS (Leclerc et al., 2018). Moreover, endemic species also tend to have unique characteristics due to their isolated evolutionary histories, such as gigantism or dwarfism, inability or poor ability to fly, and reduced clutch size (Fernández‐Palacios et al., 2021; Whittaker et al., 2017).

Finally, we found that the hotspots of core‐list mammals were localized in different areas compared with the global hotspots of mammals sensitive to INS obtained with taxonomic diversity alone (Bellard et al., 2016b). In line with previous findings, eastern Australian mammals contributed significantly to the core list. However, new areas, such as the Pyrenees, Eastern and Southern Africa, and the Himalayas, were revealed to be important based on the FUSE INS score. Moreover, our score corroborates some priority areas identified to preserve the global phylogenetic diversity of mammals in Amazonia, Central Africa, Eastern Africa, Southern Africa, Madagascar, and Southeast Asia (Robuchon et al., 2021). Yet, the importance of areas such as southwestern Europe or eastern Australia was specific to the INS threat and the functional dimension of diversity; thus, they were excluded in previous conservation prioritization strategies that are highly biased toward taxonomic diversity (Llorente‐Culebras et al., 2023).

Using the FUSE INS score for conservation guidance

Once identified based on an FUSE INS score, priority species and areas can be the object of conservation measures dedicated to mitigating INS threat. Identifying which INS are threatening each species would allow direct targeting of INS for management actions. Information about associated INS was unavailable or incomplete for most native species on the IUCN Red List, even for birds, which have the most comprehensive data. In those cases, we suggest complementing the IUCN Red List with other sources, such as the Global Invasive Species Database (https://www.iucngisd.org/gisd/), to identify the associated INS and then implementing an appropriated conservation action targeting the given INS.

For specific predators (e.g., introduced rodents and cats) or habitat disruptors (e.g., introduced herbivores or plants), eradication campaigns can benefit core‐list species, especially if they are endemic species. Although costly to implement, eradications can yield real benefits for native diversity when the context of INS presence and its integration into native ecosystems is properly identified (Philippe‐Lesaffre et al., 2023). However, this control measure is mainly effective in isolated systems, such as islands, where INS are the main threat to the species and their numbers are low (maximum 2–3 species), making them easier to eradicate (Glen et al., 2013). Our results showed that almost half of the birds and the majority of reptiles from the core lists were insular endemics, offering high hopes for the potential use of this strategy (Barbraud et al., 2021; Medina et al., 2011). For instance, the Tristan albatross (Diomedea dabbenena), ranking third in the bird core list, would highly benefit from a mice eradication program on Gough Island, an important reproductive area where mice threaten their eggs (Wanless et al., 2009). An eradication would also benefit MacGillivray's prion, which also inhabits this island (Dilley et al., 2015).

Eradication is only possible when the INS is a visible organism; therefore, it cannot be applied to introduced diseases. Moreover, pathogens, such as chytrids, pose a major management challenge because traditional area‐based conservation methods, such as habitat protection, are ineffective. Consequently, new strategies must be designed, such as the translocation of populations, which requires sufficient knowledge to ensure its efficacy in the long term, especially in the context of chytrid disease (Scheele et al., 2021).

Species with high FUSE INS scores were also threatened by other pressures, predominantly biological resource use and agriculture or aquaculture. There was, on average, a higher number of threats associated with the core‐list species than with endangered species in general (Berglund et al., 2013; Capdevila et al., 2022a; Leclerc et al., 2018). The recognition of threat co‐occurrences and interactions through networks could greatly facilitate the decision‐making processes of practitioners by prioritizing conservation actions linked to one or more threats (Geary et al., 2019). Besides, removing INS pressure for core‐list species could increase the ability of these species to recover from other global threats (Capdevila et al., 2022b).

Making the most of FUSE INS watch and borderline lists

The FUSE INS watch list contained species of high functional value (i.e., functionally unique or specialized) that are affected by INS but have a low global extinction risk. This list therefore included species with strategies not frequently associated with other species that seem to resist (or to be less exposed to) INS or other threats. In contrast, species on the borderline list are threatened and have a high extinction risk but are less functionally exceptional relative to core‐list species. However, if the species on the core list become extinct, these are the next species on the verge of extinction that harbor functionally distinct features and thus represent great amounts of functional diversity.

Species on both the watch and borderline lists may therefore need proactive conservation strategies to prevent them from being more exposed to INS threats and consequently facing an increased extinction risk. In contrast, core‐list species require reactive conservation measures following an INS threat already in place. In fact, protecting the watch‐list species can be a less expensive option to guaranteeing the preservation of functional diversity. Applying less costly conservation actions, such as habitat protection or monitoring, to protect functionally distinct species that still maintain viable populations and large distribution areas can provide high benefits at low cost. By simply overlapping the zones where these species occur (Appendices S13 & S14) with the current PA coverage, one can identify the extent to which a PA could contribute to the protection of these species (Daru et al., 2019). Moreover, one could determine zones where a small increase in PA size would have large benefits by encompassing significant regions of these species’ distribution ranges. Although current PAs do not fully address species conservation needs, at least for mammals (Williams et al., 2022), these are still an efficient way to protect native species from INS and ensure a high level of resilience (Liu et al., 2020). We emphasize the urgent need to improve the network of PAs by making them adaptable to the different threats and conservation priorities identified.

Applications of the FUSE INS score

The FUSE INS score represents a promising approach to inform conservation priorities regarding the threat of INS on a macroecological scale because it allows the identification of functionally irreplaceable species at high risk of extinction due to INS. The list of warnings and recommendations in Table 1 can help one make the most of the FUSE INS score in conservation planning (Table 1). By definition, the FUSE INS score depends on the probability of extinction due to INS and on the traits used to calculate the functional spaces and indices. Therefore, those 2 components need to be carefully defined and might be adjusted based on the final conservation objective. For instance, if the study is global, one might use a small amount of broad traits that have large taxonomic coverage. However, if the study is conducted at a smaller taxonomic or spatial extent, the traits can be selected to better represent the ecological or functional role of species within a particular ecosystem.

Although defined within a global perspective, the FUSE INS score and lists can be useful to prioritize functionally irreplaceable species and associated areas at a national or state‐island scale. Moreover, they can be complemented with other prioritization tools, such as the EDGE score, for comparing or reaching a consensus list across the several facets of diversity (i.e., taxonomic, phylogenetic, and functional). Finally, this index can be constructed to prioritize species at local scales simply by creating a functional space that reflects the local species community. The FUSE INS score is an adaptable index that can be applied to all taxonomic groups with sufficient trait data and downscaled to focus on smaller spatial scales to meet the project's conservation objectives.

Supporting information

Supporting Information

Supporting Information

Marino, C. , Soares, F. C. , & Bellard, C. (2025). Conservation priorities for functionally unique and specialized terrestrial vertebrates threatened by biological invasions. Conservation Biology, 39, e14401. 10.1111/cobi.14401