The feasibility of implementing management for threatened birds in Australia

ABSTRACT

There are many impediments to conserving threatened birds. Some can be overcome through concerted action across multiple species while others require species-specific research and intensive management. We assess the feasibility of managing 202 threats identified in the Action Plan for Australian Birds across 217 Australian threatened bird taxa against five metrics – financial, technical, temporal, political and social. A higher percentage of all threats to all taxa was assessed as having low or very low financial (43%), technical (32%) or temporal (29%) feasibility than political (9%) or social feasibility (7%). The feasibility of managing the most difficult threat to a taxon was low or very low for 52% of the threats identified. Within Australia, the main impediments to addressing most threats are technical and financial, indicating that research and funding should be domestic conservation priorities. Addressing threats linked to habitat destruction/degradation, climate change, invasive species and pollution is limited by the long timescales required to alleviate them, emphasising the urgency of commencing action. Outside Australia, social and political constraints are the main impediments to Australian conservation managers influencing action on habitat loss, species over-exploitation and climate change, but there are also technical, financial and temporal impediments to reducing threats from pollution, invasive species and climate change. Advocacy is therefore also important for addressing threats facing birds outside Australia. Our study provides a platform upon which to build more explicit and efficient management of threatened birds, prioritise efforts in research, funding and advocacy and contribute to reducing losses from Australia’s avifauna.

POLICY HIGHLIGHTS

• Feasibility analysis can identify the principal constraints on taxon recovery and help shape funding threat-based or species-specific priorities.

• In Australia, financial and technical deficits are the principal constraints on threat mitigation, emphasising the need for funding and research.

• Outside Australia, social and political issues are the major problem, requiring diplomacy to effect change.

• Many threats also face temporal constraints, highlighting the need for sustained funding.

KEYWORDS:

• Conservation prioritisation
• ornithology
• threatened species
• conservation decision-making

Introduction

Threatened species recovery involves implementing conservation actions that address the drivers of population decline (Caughley 1994). To achieve this successfully not only requires a sound understanding of the threatening processes at work, but also of how they can best be addressed (Commonwealth of Australia 2022). Where available, information on threat management can be used to prioritise conservation actions that are most likely to have positive impacts on population growth rates (Joseph et al. 2009; Gillespie et al. 2020). Where this information is lacking, systematic analysis can be used to identify knowledge gaps and associated research priorities such that evidence-based actions can be implemented in the near future (Chaplin-Kramer et al. 2022).

The threats facing declining bird species are numerous and diverse in their nature, spatial and temporal extent and specificity. The Action Plan for Australian Birds 2020 identifies 202 threats facing Australia’s threatened birds (Garnett and Baker 2021). These threats range in scale from global issues such as climate change and its associated droughts and wildfires (Reside et al. 2012; Nimmo et al. 2021), to local scale, species-specific threats such as nestling parasitism by insects (Edworthy et al. 2019). This list of threats is non-exhaustive because the nature and magnitude of threats are best identified through targeted research and monitoring programs, which are lacking for many endangered species (Robinson et al. 2018), including in Australia (Verdon et al. 2024). However, the list of threats identified in the action plan currently provides the best available baseline with which to prioritise the management actions, future research and funding required for the conservation of Australia’s avifauna.

A crucial consideration when deciding which conservation actions to prioritise is their feasibility – defined as the ease with which management can successfully be implemented to mitigate the threats identified (Gillespie et al. 2020). Understanding the feasibility of managing various threats makes it possible to target actions appropriately (Carwardine et al. 2019), which is important to avoid wastage of limited conservation resources (Wintle et al. 2019).

In a systematic assessment of priority conservation actions for Australian frogs, Gillespie and colleagues (Gillespie et al. 2020) developed a simple and repeatable means of scoring the feasibility of conservation actions to address threats by disaggregating feasibility into five metrics: social, technical, financial, temporal and political. Social feasibility reflects the level of social acceptance of particular management actions. For example, mitigating threats through culling native or culturally significant species may face greater societal backlash than other management actions such as habitat restoration (Driscoll et al. 2019; Beggs 2022, but see; Zander et al. 2022). Social feasibility also considers whether management actions are ethically justifiable (Coghlan and Cardilini 2022), and can change dramatically over time as public attitudes shift (Niemiec et al. 2022). Technical feasibility relates to the practical challenges associated with the implementation of particular management actions. Technical feasibility of recovery actions may be impacted by the life-history of the species in question, or by the scale at which the threat operates on that species. For example, implementing targeted actions to improve breeding rates of nomadic species that nest in unpredictable locations will be less technically feasible than protecting nests of species with regular, known breeding sites (Stojanovic et al. 2018; Heinsohn et al. 2022). Similarly, invasive species management may technically be more feasible on islands than in mainland areas (Kopf et al. 2017). Low technical feasibility does not mean that no action should be taken but may indicate more research is needed to reduce the risk of failure before action commences. Financial and temporal feasibility consider, respectively, the cost of and time required to address threats. While the need for funds is obvious, the problem with projects that take a long time is that the longer a process of threat reduction takes, the greater the chance it will be disrupted by random events, funding shortages, staff losses etc. Political feasibility is influenced by aspects of the other four criteria, but additionally considers the variable environmental policies of governments (Haas and Ferreira 2018). Political feasibility is particularly important for mobile species (Runge et al. 2014), where addressing threats may require coordinated responses across multiple jurisdictions (Yong et al. 2018; Nandintsetseg et al. 2019). However, political feasibility is also the most unstable and can change at very short notice.

These five feasibility criteria are not mutually exclusive but each, by themselves, can potentially impede conservation action. For example, conservation actions with low technical feasibility will likely cost more, take longer to achieve and may therefore be less favourable politically. Actions that are socially contentious are likely to be less feasible politically, which in turn will reduce their temporal feasibility. The important point is that by identifying the major constraint(s) on threat alleviation, efforts to overcome those constraints can be prioritised, whether that be through research, funding, advocacy or diplomacy, or some combination. At the very least, articulation of feasibility constraints can lead to valuable debate about which approaches should be prioritised and why.

Using a comprehensive database derived from the updated Action Plan for Australian Birds 2020 (Garnett and Baker 2021; Garnett et al. 2024), we aim here to assess patterns in the feasibility of managing threats to all of Australia’s threatened birds. Our goal was to identify the principal constraints on the feasibility of implementing management actions as a guide for focusing research, advocacy, diplomacy and communication over the coming decade.

Methods

Taxonomic and geographic scope

Feasibility was assessed for all of Australia’s bird ultrataxa (monotypic species or subspecies of polytypic species; hereafter ‘taxa’) that had been assessed as threatened or Near Threatened in Garnett and Baker (2021) using the IUCN Red List guidelines. Threats to seabird taxa at sea were assessed separately from threats to the same seabirds nesting on land.

The feasibility for managing threats was assessed and analysed separately for two geographic regions – within Australian territory and outside it. It was assumed that the feasibility of influencing management of threats while taxa are within Australian territory, which includes mainland Australia, oceanic territories (Macquarie, Heard, Christmas, Cocos (Keeling), Norfolk and Lord Howe Islands) and Australia’s Exclusive Economic Zone, will be qualitatively different from when they are outside, as occurs with 54 taxa (32 seabirds, 21 migratory shorebirds and one swift). Outside Australian territory, feasibility was assessed from an Australian perspective. That is, we assessed the probability that an Australian government, non-government organisation or an individual would be able to influence the management of threats in areas not under Australian jurisdiction by means, for example, of membership of an organisation like the Agreement on the Conservation of Albatrosses and Petrels or the East Asian-Australasian Flyway Partnership. Such organisations can attract Australian investment in research (e.g., the Australian government funded study of migratory shorebird hunting in eastern Russia; Gallo-Cajiao et al. 2020) or enhanced by international partnerships (e.g., World Heritage nomination of Yellow Sea wetlands for shorebird conservation (Castagnino Vera 2020).

Threat database compilation and feasibility estimation

In a review of all Australian bird taxa to determine their risk of extinction, accounts of each taxon considered threatened or Near Threatened under the IUCN Red List guidelines were prepared by 1–18 taxon experts (median 4, mean 5.1; Table S1). Each account included a list of threats affecting each taxon identified to the finest scale practically possible (e.g., each invasive species was considered to be a separate threat; Table S2). For analysis, threats were aggregated into seven broad classes (including other) following Kearney et al. (2023) which can be matched against the 13 classes commonly used by the IUCN (2012, Table 1; Table S1). At the same time, data were collected on the impact of every threat on the basis of its timing (i.e., past, ongoing or future), scope (i.e., the proportion of the total population affected) and severity (i.e., the overall population declines likely caused by the threat; IUCN 2022, Table S3, see Garnett et al. 2019; 2024 for details of methodology).

Table 1. Classification of threats facing Australia’s birds. Overarching threatening processes follow Kearney et al. (2023).

Overarching threatening process	IUCN threat class	Description	No. threats			Example threat	Example impacted taxa
			In class	Inside Australia	Outside Australia
Habitat destruction and degradation	1	Development-associated habitat loss	3	27	42	1.2 Habitat loss caused by commercial development	Curlew Sandpiper (outside Australia)
	2	Agriculture-associated habitat loss	13	117	21	2.1 Habitat loss caused by clearing for large scale agriculture	South-eastern Brown Treecreeper
	3	Resource extraction-associated habitat loss	11	19	1	3.2 Habitat loss caused by bauxite mining	Palm Cockatoo
	4	Collisions	3	8	0	4.1 Collisions with wind turbines	Orange-bellied Parrot
	6	Recreation-associated disturbance	5	16	1	6.1 Dog walking	Eastern Hooded Plover
Species over-exploitation	5	Hunting, harvesting and bycatch	17	66	97	5.4 Pelagic longline bycatch	Wandering Albatross
Altered fire regimes	7	Natural system modifications	11	130	20	7.1 Changed fire regime	Carpentarian grasswren
Invasive species and disease	8	Invasive species	115	431	56	8.1 Black rat predation	Norfolk Island Robin
Pollution	9	Pollution	9	18	67	9.2 Oil spills	Matsudaira’s Storm-Petrel
Climate change	11	Climate change impacts	9	231	40	11.1 Sea level rise	Samphire Thornbill
Other	10	Geological events	2	1	2	10.1 Volcanoes	Macaroni Penguin
	12	Genetics, Allee effects and captive breeding	3	13	1	12.1 Lack of genetic variability	Norfolk Island Morepork
	13	Unknown	1	5	0	13.1 Unknown cause of decline	Red Goshawk

*taxa can face multiple threats in each class.

Taxon experts also assessed the overall feasibility of managing the threats to reduce their impact on extinction risk on a subjective five-point scale (very low, low, medium, high, very high). Such broad classes were considered appropriate since assessing feasibility involves assessing the likelihood of future actions occurring under conditions of substantial uncertainty. To understand feasibility more fully, these feasibility assessments were disaggregated into each of five feasibility metrics – social, political, technical, financial or temporal, based broadly on Gillespie et al. (2020; Table 2).

Table 2. Definitions used to score the social, political, technical, financial or temporal feasibility of alleviating threats to Australian threatened birds.

Definition	Feasibility (score)
Almost insurmountable constraints; success highly unlikely (<20%) or will take a very long time (>50 years)	Very low (1)
Substantial constraints; low chance of success (20–50%); may take a long time (20–50 years)	Low (2)
Moderate constraints; success possible (50–80%); may take a moderately long time (5–20 years)	Medium (3)
Some constraints; success probable (80–95%); potentially over short time (1–5 years)	High (4)
No substantive constraints; success highly likely (>95%) in <1 year	Very high (5)

To obtain and check scores for this disaggregation, we formed a judgement after consulting relevant literature and policy documents when these were available (See Table S4 for worked examples for each threat class). Where documentation was lacking, we inferred feasibility based on efforts to conserve similar taxa or alleviate similar threats as well as drawing on the expertise of the taxon experts (Table S1).

For social feasibility, information was often available from national attitudinal surveys on the acceptability of many management options (e.g. Garnett et al. 2018; Zander et al. 2022). For political feasibility, assessments were based on the policies or legislation related to the jurisdiction where each threatened taxon occurred and the extent to which these could influence the impact of the threat. For example, the legislation and policy governing some threats (e.g., urban planning) operates at a local government level, others at a sub-national level (e.g., land use, forestry, bushfire management) and some at a national level (e.g., climate change mitigation). Given national climate change policy in Australia changed soon after the publication of Garnett and Baker (2021), political feasibility scores were adjusted for threats associated with climate change to reflect change in national policy. For technical, financial and temporal feasibility, we consulted recovery plans, conservation advices, relevant methodological literature and the research priorities detailed in taxon texts in Garnett and Baker (2021).

For each feasibility metric, the feasibility scores were estimated independently, i.e., assuming no other form of constraint on implementation exists. For example, some threats are technically difficult to alleviate even if funds are available – technical feasibility was only considered very low if constraints could not be overcome by an injection of funds. Similarly, other threats such as a lack of natural tree hollows available for nesting at a landscape scale for widespread species (Manning et al. 2013) cannot reasonably be reduced for many decades regardless of the money available. We also tailored the feasibility of managing threats to the context in which each threatened bird taxon encountered them – e.g., predatory cats are easier to eradicate from small islands than from extensive landscapes; constraints on habitat clearance are stronger in some jurisdictions than others.

The minimum feasibility score for any of the five metrics scored for each threat (social, political, technical, financial, temporal) was guided by the overall level of feasibility for managing that threat indicated by the experts in their accounts in Garnett and Baker (2021; individual citations in Table S1). Scoring for each metric was initially performed by one author (STG). After the initial scoring, the authorship team checked the feasibility scores for each metric; any discrepancies were discussed until a consensus was reached.

Feasibility metrics

For analysis of each feasibility class, we used the mean score for feasibility across all threats to a taxon because the number of threats to a taxon varied between one and 20. However, for overall feasibility of alleviating any threat to a taxon and allow population recovery, we calculated the mean of the lowest score in any of the five feasibility metrics. While a threat with very low scores for all five metrics is likely to present more challenges than one with, for example, four high feasibility scores and one very low score, the very low scored threat may ultimately constrain progress with any of the others when it comes to the feasibility of abating it.

We also calculated a feasibility score weighted by the impact of that threat (‘scaled feasibility’) because, to choose between threats in which to invest funds for research, management or advocacy requires a trade-off between the feasibility of achieving threat alleviation and the impact of that threat on populations of the taxon concerned. That is, at a taxon level, we assumed that investment in alleviating negligible threats has as little benefit as investing in alleviation of serious threats that are currently intractable (e.g., volcanic eruptions). To balance impact with feasibility, we multiplied the minimum feasibility score for each threat to a taxon by the impact of that threat on the taxon then scaled the result to the maximum impact of any threat to the taxon as follows

$\mathit{Fxwt}=\frac{\mathit{Ft}\ast \mathit{It}}{\mathit{Max}\mathit{Itx}}$

where F_xwt is the weighted feasibility score for threat t to taxon x, F_t is the unweighted feasibility score for threat t, I_t is that threat’s current impact on the taxon and Max I_tx is the maximum impact of any threat to the taxon. The maximum F_xwt score across all threats to a taxon, termed ‘scaled feasibility’, was selected as being the score that best represented the feasibility of alleviating a significant threat to a taxon. The feasibility metric(s) responsible for the scaled feasibility for a taxon was then used to identify the type of action needed to have greatest efficacy in alleviating threats to the taxon.

Data visualisation and analysis

To account for the different political challenges associated with taxa whose range extends outside of Australian territory, the analysis was divided into two sections. Domestic threats impacting taxa within Australia were assessed first, then international threats were assessed separately for taxa that either occur in Australia but breed overseas, or breed in Australia but spend a portion of the lives overseas.

Overall feasibility and mean feasibility for each feasibility class were analysed with Generalised Linear Models (GLM) via the package lme4 v1.1–31 (Bates 2022), to identify patterns in the feasibility of addressing threats based on taxonomic grouping, habitat, life-history, conservation status/effort and geography, as per Garnett et al. (2019). Thus, each taxon contributed a single datapoint to each GLM for each feasibility variable. All GLMs were fitted with a Gaussian distribution. Domestic social and political feasibility were inverse transformed (1/(max (x + 1) – x) to account for the left skew in the distribution of those response measures. The following covariates (Tables S5 and S6) were included in the GLMs (after checking for cross-correlation with the ggcorr function in the Ggally package v2.1.2 (Schloerke 2021):

• Population size: log-transformed estimated number of mature individuals in the population.

• Area of Occupancy: log-transformed area of occupancy in km².

• Habitat type: 3-level factor being terrestrial, oceanic or coastal.

• Generation length.

• IUCN Red List Status: 4-level factor being Critically Endangered (CR), Endangered (E), Vulnerable (V) or Near Threatened (NT).

• Presence of recovery plan or conservation advice document: 3-level factor, recovery plan present, only conservation advice present or neither document present.

• Taxon group: 5-level factor being shorebirds, seabirds, passerines, parrots or other.

• Site accessibility score: a 10-point scale summarising the legality and proportion of a taxon’s AOO that can be accessed.

• Taxon weight: mean taxon body mass in grams.

• Genetic distinctiveness: number of close relatives each taxon has up until Order, following Daniels et al. (1991).

Package MuMIn V1.47.1 (Bartoń 2022) was used to identify automatically the most parsimonious model (i.e., subset of the global model) for each average feasibility measure based on lowest Akaike Information Criterion values, corrected for small sample size (AICc).

Bar graphs were then used to illustrate the current impact of all threats facing Australia’s birds domestically. The bars for each taxon were labelled with the most prominent overarching threat class they currently face, and were colour-coded according to the maximum scaled feasibility of addressing that threat as defined in Table 1. Where taxa had > 1 threat considered equally prominent, the authors reached agreement on which threat was most logical to present in figures, with the remainder of the threats presented in Table S7.

To compare patterns in feasibility of addressing threats across the overarching threat classes and across the five feasibility measures, the proportion of all combinations of threats and taxa were calculated for each threat class and feasibility score separately for taxa while in Australia and while they are outside Australia.

Finally, the scaled feasibility for each threat (F_xwt) was summarised for all threats and threat classes to determine the most tractable threat currently having the greatest impact on the persistence of each taxon. For this summary, bar graphs were produced to show the scaled feasibility of addressing the most tractable threat facing each taxon. Bars were colour coded by the feasibility measure (or measures where > 1 measure were assigned an equal lowest feasibility score) that was currently most limiting each threat.

Results

Feasibility data were collected for 217 threatened or Near Threatened bird taxa facing 202 threats (Table S2) either inside Australia (182 threats affecting 195 taxa; Table S8) or outside (47 threats affecting 54 taxa; Table S9).

A higher percentage of all threats to all taxa was assessed as having low or very low financial (43%), technical (32%) or temporal (29%) feasibility than political (9%) or social feasibility (7%; Figure 1). The feasibility of managing the most difficult threat to a taxon was low or very low for over half the threats identified (52%).

Figure 1. Percentage of threats to Australian threatened birds in each feasibility class for the five types of feasibility considered and for the minimum feasibility across all classes.

Correlations between the overall feasibility of addressing these threats and the five feasibility components were positive for 14 of the 15 pairwise comparisons but varied in strength (Figure S1). Only mean social and technical feasibility estimates showed a weakly negative correlation (Figure S1).

The overall feasibility of addressing domestic threats to Australian birds was lower for shorebirds, passerines and parrots than for seabirds and other taxa, and this also held when assessing technical and financial feasibility (Figure 2a, Table 3). Overall feasibility (as well as technical, social and financial feasibility) was lower in species lacking a recovery plan or conservation advice document (Figure 2b, Table 3). None of the feasibility metrics for domestic threats varied with the conservation status of taxa (Figure 2c). Overall feasibility appeared higher for oceanic taxa than for terrestrial and coastal taxa (Figure 2d), but the only statistically significant difference relating to broad habitat type was for political feasibility, which was lower for coastal taxa than for oceanic and terrestrial taxa (Table 3). Overall feasibility, as well as technical and financial feasibility, was correlated positively with generation length and negatively with population size (Figure 2e,f, Table 3). Overall feasibility as well as social, political and financial feasibility were also positively associated with site accessibility but negatively associated with the genetic distinctiveness of taxa (Table 3). The temporal feasibility of addressing threats is only weakly and positively associated with generation length.

Figure 2. Summary of the overall feasibility of addressing identified threats to Australian birds by (a) taxon group; (b) presence/absence of species recovery plan or conservation advice document; (c) IUCN red list status; (d) habitat type; (e) population size; (f) generation length; and (g) area of occupancy. See methods for calculation of the overall feasibility score.

Table 3. Modelled relationships between demographic, life-history and conservation traits of Australia’s threatened birds and measures of the feasibility of addressing the threats they face within Australia. Predictions are derived from the top model for each feasibility measure based on lowest AICc scores. Only significant effects or factors with significant factor levels (p < .05) are shown.

Feasibility measure	Covariate	Estimate	SE	Z	P
Minimum scaled feasibility (WFS)	Generation length	−0.042	0.017	−2.412	0.017
	Taxon group: passerines	0.347	0.151	2.297	0.023
	Site accessibility	0.051	0.020	2.505	0.013
Social	Conservation document: neither*	−0.069	0.033	−2.075	0.039
	Detectability	−0.028	0.010	−2.759	0.006
	Generation length	0.017	0.004	4.106	<.001
	Site accessibility	−0.016	0.006	−2.618	0.001
	Taxon group: passerines	0.145	0.043	3.381	<.001
	Taxon group: shorebirds	−0.221	0.078	−2.825	0.005
Political	Generation length	0.009	0.004	2.394	0.018
	Habitat: oceanic*	0.381	0.077	4.947	<.001
	Habitat: terrestrial*	0.313	0.069	4.555	<.001
	Site accessibility	−0.021	0.006	−3.710	<.001
Technical	Population size	−0.181	0.030	−6.001	<.001
	Conservation document: neither*	−0.489	0.164	−2.988	0.003
	Generation length	0.060	0.023	2.650	0.009
	Taxon group: parrots	−0.780	0.253	−3.078	0.002
	Taxon group: passerines	−0.645	0.192	−3.356	0.001
	Taxon group: shorebirds	−1.080	0.343	−3.141	0.002
Financial	Population size	−0.124	0.022	−5.218	<.001
	Conservation document: neither*	−0.353	0.125	−2.826	0.005
	Generation length	0.063	0.016	3.951	<.001
	Site accessibility	0.055	0.019	2.894	0.004
	Taxon group: parrots	−0.648	0.179	−3.614	<.001
	Taxon group: passerines	−0.545	0.144	−3.783	<.001
	Taxon group: seabirds	0.432	0.191	2.268	0.024
	Taxon group: shorebirds	−1.203	0.243	−4.939	<.001
	Genetic distinctiveness	−0.014	0.005	−2.643	0.009
Temporal	Generation length	0.018	0.006	2.961	0.003

*Habitat factor effects are relative to predictions for coastal taxa. Taxon group factor effects are relative to taxon group other. Conservation document factor effects are relative to predictions for taxa with a recovery plan.

Current impact of threats within Australia and the feasibility of addressing the most prominent threats

Within Australia, the estimated current impact of the threats facing the 84 taxa considered Endangered or Critically Endangered varies substantially (Figure S2 and S3). For the 33 taxa for which the current impact estimates exceed a score of 200 (very high relative impact), including Black-eared Miner Manorina melanotis and Regent Honeyeater Anthochaera phrygia, the overarching threats are predominantly associated with climate change, habitat destruction and altered fire regimes, for which the scaled management feasibility is predominantly low to medium. Seventeen of the Critically Endangered or Endangered taxa, mostly seabirds such as the White-necked Petrel Pterodroma cervicalis cervicalis, have a high or very high feasibility score of addressing the most prominent threat they face, which are typically invasive species or over-exploitation (Figures S2 and S3). The scaled feasibility of addressing the most prominent threat is considered very low for 27 Critically Endangered or Endangered taxa including Eastern Rockhopper Penguin Eudyptes chrysocome filholi and Eastern Pink Cockatoo Cacatua leadbeateri leadbeateri, for which the predominant threat is climate change.

For the 106 taxa considered Near Threatened or Vulnerable, the patterns in threat impact and feasibility within Australia are similar to those for Endangered and Critically Endangered taxa (Figures S4 and S5). The most frequently occurring prominent threats are climate change and altered fire regimes, including for the 14 taxa for which current impact estimates exceed 200, including Malleefowl Leipoa ocellata and South-eastern Glossy Black-Cockatoo Calyptorhynchus lathami lathami, for which the scaled feasibility of addressing these threats is generally low or very low (Figures S4 and S5). The estimated scaled feasibility of addressing the most prominent threat is considered high or very high for only 18 of 106 taxa, and these threats are again mainly linked to invasive species (e.g. Tasman White-bellied Storm-Petrel Fregetta grallaria grallaria) or, within Australian waters, over-exploitation (e.g. Campbell Albatross Thalassarche impavida).

Patterns in the feasibility of addressing all threats across threat classes

Within Australia, across all threat classes, there are social or political impediments to reducing the impacts of habitat destruction and degradation (habitat loss) for a small number of taxa with the major constraints being time and money (Figure 3a). In contrast, for the small number of taxa threatened by over-exploitation, the solutions are relatively simple across all feasibility classes. For altered fire regimes, technical and financial constraints are major impediments to action. Control of invasive species, which has the largest number of threat taxon combinations, is largely socially feasible but, particularly on the mainland where eradication is rarely possible, there are major technical, financial and temporal challenges. Pollution, which affects few taxa, and climate change, which could be classed with pollution, which affects many taxa, have similar feasibility profiles: few would oppose reducing their impacts but the technical, financial and temporal problems of doing so are almost insuperable for most of the affected taxa. Finally, solutions for many of the small number of other threats, mostly genetic constraints, are gradually being developed.

Figure 3. Major impediments to addressing every threat to every threatened Australian bird taxon/population and the feasibility of doing so (a) within Australia; and (b) outside Australia by overarching threat class and feasibility component. Bars for each threat class show, first, the percentage of the major impediments to reducing threats to taxa/populations within threat class and secondly the percentage of threats, for each component of feasibility, in each feasibility class. Numbers on the y-axis labels signify the number of threats identified within each overarching threat class. Note altered fire regimes are not shown in (b) as no threats were identified outside Australian territory in this threat class for Australian threatened birds while overseas.

The overall feasibility of addressing the most tractable threat facing taxa within Australian territory is generally considered high where the most tractable threats are related to habitat destruction and degradation, such as for the Christmas Island Goshawk Accipiter fasciatus natalis or over-exploitation such as Grey-headed Albatross Thalassarche chrysostoma (Figures S6–S9). In contrast, the scaled feasibility of addressing the most tractable threats is low for taxa where the most tractable threat is linked to climate change, including the Noisy Scrub-bird Atrichornis clamosus and Eungella Brown Gerygone Gerygone mouki amalia or altered fire regimes including the Mallee Emu-wren Stipiturus mallee and Flinders Ranges Short-tailed Grasswren Amytornis merrotsyi merrotsyi (Figures S6–S9). Few of the most tractable threats facing taxa domestically were limited by a single feasibility component (Figures S6–S9).

Predictors of threat feasibility scores across taxa

The overall feasibility of addressing threats to birds outside of Australian territory is lower for shorebirds than for seabirds (Figure 2a, Table 4). The feasibility of addressing overseas threats is no higher for taxa with an existing conservation document than for taxa currently lacking one (Figure 2b). Feasibility also does not differ between IUCN conservation status classes (Figure 2c). Threats are more feasible to address for marine taxa than for coastal taxa (Figure 2d) but population size is not linked to the mean feasibility of addressing threats outside of Australia (Figure 2e). Generation length is associated positively with most components of threat feasibility (Figure 2f, Table 4), while there are negative associations between financial and temporal feasibility and both detectability and genetic distinctiveness (Table 4).

Table 4. Modelled relationships between demographic and life-history traits of Australia’s threatened birds and measures of feasibility of addressing the threats they face outside of Australian territory. Predictions are derived from the top model for each feasibility measure based on lowest AICc scores. Only significant effects or factors with significant factor levels (p < .05) are shown.

Feasibility measure	Covariate	Estimate	SE	Z	P
Overall	Generation length	0.125	0.021	6.098	<.001
	Detectability score	−0.168	0.060	−2.792	0.007
	Genetic distinctiveness	−0.027	0.009	−2.797	0.007
Social	Generation length	0.072	0.013	5.521	<.001
	Habitat: oceanic*	0.975	0.219	4.457	<.001
	Genetic distinctiveness	−0.024	0.008	−3.120	0.003
Political	Generation length	0.102	0.009	11.06	<.001
Technical	Generation length	0.091	0.027	3.392	0.001
	Area of Occupancy	0.099	0.044	2.253	0.028
	Site accessibility	0.362	0.154	2.354	0.022
	Genetic distinctiveness	−0.030	0.012	−2.415	0.019
Financial	Area of Occupancy	0.110	0.036	2.989	0.004
	Detectability score	−0.136	0.060	−2.190	0.033
	Generation length	0.088	0.020	4.284	<.001
	Genetic distinctiveness	−0.035	0.010	−3.456	0.001
Temporal	Generation length	0.122	0.023	5.310	<.001
	Detectability score	−0.216	0.075	−2.844	0.006
	Habitat: oceanic	−0.878	0.391	−2.244	0.029
	Area of Occupancy	0.118	0.049	2.405	0.020
	Genetic distinctiveness	−0.028	0.011	−2.024	0.048

*Habitat factor effects are relative to predictions for coastal taxa.

Current impact of threats outside of Australia and the feasibility of addressing the most prominent threats

The most frequent prominent threats facing Australian birds outside of Australia are linked to invasive species, over-exploitation and climate change, both for Endangered and Critically Endangered taxa (Figure S10) and for Near Threatened and Vulnerable taxa (Figure S11). The current impacts of threats are considered greatest for migratory shorebirds including Bar-tailed Godwit Limosa lapponica and Greenshank Tringa nebularia. The feasibility of addressing the most prominent overseas threats is considered medium or high for 24 taxa where the current scaled impact of threats exceeds 100. However, for all other taxa where the feasibility of addressing the most prominent overseas threat is considered medium or high (n = 27), the current impact of those threats is thought to be relatively low (<100, Figures S10 and S11).

Patterns in the feasibility of addressing all threats across threat classes

In contrast to threats within Australia, empirical evidence for policy and practice within the habitats of many Australian taxa shared with other countries suggests there are major social and policy constraints on slowing the rate of habitat destruction and degradation, and addressing climate change (Figure 3b). Threats associated with habitat destruction and degradation and species over-exploitation are most limited from social and political perspectives with technical impediments to stopping habitat loss or species over-exploitation being relatively few, and the actions likely to be relatively quick and inexpensive, once policy changes. Addressing almost all overseas threats from pollution currently has low estimated feasibility from technical, financial and temporal perspectives, whilst addressing threats linked to climate change is most limited financially and technically (Figure 3b).

The overall feasibility of addressing the most tractable threat facing taxa outside of Australian territory is generally considered medium to high where the most tractable threats are related to species over-exploitation (e.g. Eastern Amsterdam Albatross Diomedea amsterdamensis) and some invasive species (e.g. Matsudaira’s Storm-Petrel Oceanodroma matsudairae, Figures S12 and S13). Such threats are mostly limited by political, technical and financial feasibility. As for domestic taxa, the overall feasibility off addressing the most tractable threats outside Australia is low for taxa where the most tractable threat is linked to climate change, including multiple penguin species. It is also low for many migratory shorebirds where invasive species threats are linked to cordgrass invasion of mudflats (Figures S12 and S13). None of the most tractable threats facing Endangered or Critically Endangered taxa, and very few of the most tractable threats facing Near Threatened or Vulnerable taxa overseas, were limited by a single feasibility component (Figures S12 and S13).

Discussion

With the goal of assisting prioritisation of future research, funding and advocacy, we used a database developed for the updated Action Plan for Australian Birds (Garnett & Baker 2021; Garnett et al. 2024) to estimate the feasibility of addressing threats currently facing threatened Australian birds. We also identified broad trends in conservation feasibility and the perspective from which the feasibility of abating threats is currently most limited, but there was also substantial variation in feasibility scores. Broadly, birds are most threatened within Australia by invasive species, climate change, altered fire regimes and habitat loss or degradation. The feasibility of addressing these threats within Australia is most limited by technical and financial considerations, reiterating the urgent need for sufficient funding to implement recovery actions (Wintle et al. 2019), as well as additional research to help inform those actions where optimal conservation strategies are currently unclear (Nichols and Williams 2006). Outside Australia, birds are additionally impacted by over-exploitation and pollution and the feasibility of addressing such threats is often most limited from social and political perspectives. This highlights the important role of advocacy in complementing funding and research to conserve Australia’s migratory species (Yong et al. 2018). We discuss the implications of our findings, the strengths and weaknesses of our approach and future improvements to it.

Although we acknowledge that our measure of feasibility at a taxon level (the mean of the minimum feasibility of addressing a threat across five feasibility measures weighted by the threat impact) is a general measure that unavoidably overlooks much species-specific detail, we considered it useful to explore general patterns in the feasibility of addressing threats to Australian birds. While our analysis did not control for phylogeny (Prudent et al. 2016), this reflects a publicly held view that all taxa are of equal social value regardless of taxonomic uniqueness (Garnett et al. 2018). As such, we show that greater focus is needed to increase the average feasibility of conserving shorebirds and passerines relative to seabirds, parrots and other taxa. We also found that, within Australia, the mean feasibility of mitigating threats to taxa with a recovery plan or conservation advice document in place was higher than for taxa lacking one but the same was not true outside Australian territory. This could simply be because conservation documents focus on the most imperilled taxa which are also the most challenging to recover (Walsh et al. 2013). However, it could also hint that formal conservation documents do not by themselves increase the feasibility of addressing the threats faced by migratory species (Bottrill et al. 2011). Even if they effectively identify threats and measures to address them, this does not necessarily increase the feasibility of successfully implementing those actions, especially for pervasive, large-scale threats such as climate change and invasive species. The mean feasibility of addressing threats for seabirds is higher than those facing terrestrial or coastal taxa. The identified threats facing seabirds are predominantly associated with fisheries bycatch (overharvesting) and island invasive species which are, in theory, easier to address than threats in other threat classes.

Feasibility metric classes

Scores for social and political feasibility were largely high within Australia with differences between the two often representing a lag between social acceptability and policy change. For example, while polls indicated that the social feasibility of stronger climate action was high over the last decade (Crowley 2021), political feasibility was low until a change in Australian Government in 2022 (MacNeil and Edwards 2023). Differences can also occur when local or individual action contrasts with national policy. For example, while studies of social acceptability of conservation actions within Australia show that killing invasive species is acceptable if it reduces extinction risk (Zander et al. 2021), the ongoing toll from domestic cats (Legge et al. 2020) is evidence of social resistance to cat containment. Impediments can also arise when local beliefs contrast with those of the broader population. For example, shooting of Baudin’s Black-Cockatoo Zanda baudini by orchardists in Western Australia is illegal but has been a threat for decades (Chapman 2007), partly because advocacy has focused on the species’ sympatric congener Carnaby’s Black-Cockatoo Z. latirostris (Ainsworth et al. 2016). Where politics is the major barrier to extinction risk reduction, change could occur rapidly. For example, the threat of habitat loss to the five endemic taxa on King Island could instantly be reduced by reimposition of a moratorium on land clearing (e.g., Webb et al. 2016).

Technical and financial constraints are less closely aligned than social and political feasibility: there are many problems, like climate change, for which innovative ideas about adaptation are only just starting to emerge – money for research is needed but not yet funds for rolling out solutions. The reverse is also true – there are actions such as invasive species and weed control where management that is effective at a local scale needs funds to be scaled up to have an impact at the population level. Often, however, the feasibility scores interact. For example, fires of the scale and intensity of those that occurred in south-eastern Australia in 2019–20 cannot currently be stopped (Clarke et al. 2022) and the only strategies to reduce the risk to threated species from catastophic events – such as emergency rescues – are expensive and difficult (Selwood et al. 2022). Funding will therefore be less of a constraint on risk reduction if technical impediments to fire control or rescue are reduced.

For some taxa, time alone is the major constraint. The effects of habitat loss and degradation may take centuries to reverse if a taxon occurs widely across the landscape. Tree hollows of the type used by Gang-gang Cockatoos Callocephalon fimbriatum and Superb Parrots Polytelis swainsonii for nesting are only found in very old trees (Gibbons et al. 2000) and linking fragments to allow recovery of species like Brown Treecreeper Climacteris picumnus (Doerr et al. 2011) will also take many decades of population level improvement. Long-term actions face the added risk of interruption if, for example, fires or timber harvesting kill trees that would otherwise have developed hollows (McLean et al. 2015).

Patterns in the feasibility of threat abatement

Although the broad patterns in the feasibility of threat abatement identified above are useful for informing high-level policy decisions (Ross and Dovers 2008), they shroud taxon-specific variation in the number and type of threats Australian birds currently face, their level of impact, as well as the feasibility of addressing them. As shown for frogs (Gillespie et al. 2020), our approach is perhaps best suited to identify current barriers to Australian bird conservation at the taxon level, although the challenge is arguably greater for birds (n = 273 threatened taxa) than it is for frogs (n = 45 threatened taxa). Assessing feasibility at a taxon level also shows that the feasibility of threat management may vary among taxa for the same threat. For example, translocations of the Eastern Bristlebird Dasyornis brachypterus have already reduced extinction risk from an increased severity, scale and frequency of fire and drought associated with climate change (Baker et al., 2012), so further risk reduction is likely to be relatively quick, cheap and effective. For most taxa threatened by climate change, even with strong social and political support, there remain major technical, financial and temporal barriers to amelioration of the same threats.

That said, many of the recovery actions frequently identified as most feasible (scaled for impact), such as invasive species’ management and habitat preservation and restoration (Figures S6–S9, S12–S13), will yield ecosystem-level benefits that are not limited to birds (Tonietto et al. 2018). Consideration of the potential ecosystem-level benefits of threat abatement is but one way our current approach to assessing feasibility could be improved. Our goal here is to provide a first step towards a comprehensive, standardised way of assessing systematically the current impediments to addressing threats. As for all such processes, our results are only as good as the information available to be assessed. Unfortunately, for many taxa (not just birds), current monitoring data are inadequate for accurately and comprehensively assessing contemporary threats (Scheele et al. 2019). Some threats may therefore be overlooked, or the impact of known threats could be over- or under-estimated. For example, few would have foreseen the impact of parasitic flies on Forty-spotted Pardalotes Pardalotus quadragintus were it not for targeted monitoring effort (Edworthy et al. 2019), allowing the implementation of a simple, highly-feasible solution to this threat (Alves et al. 2021). Other novel threats are likely to arise, whilst the impact of known threats may increase or diminish over time. Improved monitoring of Australia’s threatened birds is therefore fundamental to allow iterative refinement of the accuracy of threat assessments as well as assessment of the feasibility of addressing them (Verdon et al. 2024).

Whilst it will take years to improve the quality of available monitoring data for many of Australia’s threatened birds, there are other, shorter-term options to improve our feasibility assessment approach. We do not consider uncertainty, yet uncertainty surrounding the current impact of threats and the feasibility of addressing them is likely to be high for many taxa and could have major impacts on conservation decision-making (Regan et al. 2005). For example, is it better to invest in highly feasible actions with moderate benefits or low feasibility actions that, if achieved, would provide more substantial benefits? Our current approach scales feasibility measures by their current estimated impact, but incorporating uncertainty into the assessment process may yield alternative priorities.

Conclusion

We provide here the first systematic analyses of the feasibility of reducing threats to Australian birds based on five interacting metrics measured on a five-point scale. The results suggest some broad trends that can assist advocacy and planning for threatened bird conservation. For taxa that are confined to Australia, the social and political environment is largely sympathetic to conservation but there are technical and financial constraints on the management of many taxa. Such constraints are greatest for those taxa living on the Australian mainland where threat management is required at a landscape level. On islands, which have historically faced the greatest rates of extinction, feasibility scores were higher on the basis of proven success in island threat management at many sites. For Australian bird taxa that also rely on habitats outside Australia, greater investment in diplomacy is likely to be the most beneficial approach given that the Australian government can rarely support direct action.

The approach is novel and can be refined, particularly to manage uncertainty around the impact of threats as this could lead to errors in prioritisation of conservation actions. Methodological improvements could also be made in understanding and incorporating interactions among different types of feasibility, and between the threats themselves. Nevertheless, given that feasibility is already built into many prioritisation processes, having a systematic set of measures should help improve the effectiveness of conservation targeting.

Acknowledgments

The authors wish to thank the many contributors to the Action Plans for Australian Birds whose insights Form the basis for this research. Without them, feasibility scores for many taxa would be far lower.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

The overall threats database relevant to all papers in the Action Plan for Australian Birds 2020 special issue is available by contacting the lead editor of the plan (STG). Data and R-code used for analysis in this manuscript are available in the online Supplementary Material.

Supplementary material

Supplemental data for this article can be accessed at https://doi.org/10.1080/01584197.2023.2295355.

Correction Statement

This article has been corrected with minor changes. These changes do not impact the academic content of the article.

Additional information

Funding

This work was supported by the Australian Bird Environment Fund, BirdLife Australia, Charles Darwin University, Biosis Pty Ltd, New South Wales Local Land Services via the National Landcare Program 2 and Auchmeddan.