Outpacing the Math: Why Range-Shifting Species Are in Greater Danger Than We Think

Introduction to Climate-Driven Biodiversity Loss and Extinction Forecasting of Native Species

The acceleration of anthropogenic climate change poses an unprecedented and multifaceted threat to global biodiversity, fundamentally altering the ecological, thermal, and geographical parameters that dictate species survival. Historically, conservation biology has focused heavily on mitigating direct, highly localized threats, such as habitat destruction, overexploitation, agricultural expansion, and the introduction of invasive alien species. However, as global mean temperatures continue to rise, shifting climatic envelopes are forcing flora and fauna into a profound adapt-or-perish paradigm. Species must either adapt rapidly in situ to novel climatic stressors or physically migrate across increasingly fragmented landscapes to track suitable environmental conditions¹. The immense biological complexity of these spatial and demographic responses has introduced profound challenges for scientists tasked with predicting future biodiversity loss. While it is unequivocally accepted that a substantial fraction of the planet's species faces heightened extinction risk over the next century, the precise magnitude of this risk—and the specific ecological mechanisms driving it—remain a subject of intense academic scrutiny and methodological evolution.

Early macroscopic efforts to quantify climate-related extinction risks heavily utilized the species-area relationship, a foundational ecological principle positing that larger geographic areas harbor more species. In the early 2000s, pioneering studies applied this power-law relationship to climate scenarios, assuming that as climate change reduced the total area of suitable habitat, a proportional number of species would be mathematically committed to extinction². Some of these early global estimates suggested that between fifteen and thirty-seven percent of terrestrial species could face extinction by the year 2050³. While these foundational models were instrumental in highlighting the sheer scale of the climate crisis to policymakers, they were subsequently criticized for their biological oversimplifications. Specifically, species-area models treat species as static entities uniformly distributed across abstract geographical spaces, effectively ignoring the complex mechanics of individual dispersal, life-history traits, evolutionary adaptation, and landscape fragmentation².

Subsequent research sought to refine these projections by identifying specific climatic drivers and biological traits that predispose species to high risk. For instance, detailed analyses of historical and modern survey data have revealed that absolute increases in maximum annual temperatures—specifically the hottest daily highs during summer months—are far more strongly correlated with localized population extinctions than changes in mean annual temperatures⁶. While some populations possess the phenotypic plasticity or evolutionary capacity to tolerate warmer conditions through in-situ thermal niche shifts, these shifts are remarkably slow. Observations indicate that positive thermal niche shifts occur at a mean rate of approximately 0.02 degrees Celsius per year, a pace that is estimated to be three to four times slower than the projected velocity of anthropogenic climate change⁶. Consequently, even when factoring in both dispersal capabilities and potential niche shifts, contemporary forecasts still project that sixteen to thirty percent of analyzed terrestrial plant and animal species could disappear within the next fifty years under moderate to severe warming scenarios⁶.

Despite these advances in macroecology and thermal biology, translating broad ecological forecasts into actionable, species-specific conservation assessments remains fraught with severe methodological limitations. The International Union for Conservation of Nature Red List of Threatened Species stands as the definitive global standard for evaluating extinction risk¹. To accommodate the realities of climate change, the International Union for Conservation of Nature has continuously expanded its guidelines to incorporate predictive modeling, ostensibly allowing species to be categorized based on forecasted habitat loss and future population declines¹. However, recent systematic evaluations of these assessment guidelines have uncovered critical blind spots in the underlying mathematical models. Most notably, research published in 2026 demonstrates that the standard correlative models used for these assessments consistently and dangerously underestimate the extinction risks for species that are forced to shift their geographical ranges to survive¹. This systemic underestimation stems from flawed, overly simplistic assumptions regarding the mathematical relationship between habitat loss and population size, raising urgent questions about the efficacy of current conservation early-warning systems¹.

The Framework of Extinction Risk Assessment

The Red List Criteria and Climate Change

Since its inception in 1964, the International Union for Conservation of Nature Red List has become the planet's barometer of biodiversity, utilizing a rigorous scientific process to document the global status of over 160,000 species¹. The framework relies on five quantitative criteria, labeled A through E, to classify species into categories of relative extinction risk, ranging from Least Concern to Critically Endangered, and ultimately Extinct⁹. These criteria are deeply rooted in the theoretical trajectories of single-population declines and rely heavily on observable demographic variables such as population size, geographic range extent, and biological generation length¹⁰.

When assessing risks driven specifically by future, slow-acting threats like climate change, conservation assessors primarily lean on Criterion A3 and Criterion E. Criterion A3 is designed to measure projected future population reductions over a specific, biologically relevant timeframe—typically calculated as three generations or ten years, whichever is longer, up to a maximum of one hundred years¹⁴. Because empirical census data on future population sizes are inherently impossible to obtain, the Red List guidelines permit the use of predicted habitat loss as a direct, one-to-one proxy for population decline⁹.

Criterion E, on the other hand, utilizes complex, quantitative mathematical models, such as population viability analyses, to estimate the absolute probability of a species going extinct within a given timeframe¹⁶. A species may qualify as Vulnerable under Criterion E if its modeled probability of extinction reaches at least ten percent within one hundred years¹⁸.

The application of these standard criteria to climate change has sparked considerable academic debate, largely centered around temporal scales and biological assumptions. A major point of contention is the calculation of the assessment time horizon. For many short-lived species, empirical data on exact generation lengths are unavailable. Assessors frequently default to a naive ten-year window to represent the three-generation horizon¹⁴. Recent studies analyzing forest-dependent amphibian species have demonstrated that relying on this simplified ten-year proxy systematically distorts and underestimates extinction risk compared to using scientifically predicted, trait-based generation lengths¹⁴. When predicted generation lengths were utilized, the percentage of amphibian species identified as declining due to future climate-driven habitat change increased substantially, from one percent up to twenty percent¹⁴.

Beyond temporal scaling, the assumption that habitat loss translates perfectly and linearly to population reduction under Criterion A3 has become the primary focal point of modern ecological critique¹. The Red List guidelines acknowledge this linearity as a source of uncertainty, yet the assumption has persisted due to a historical lack of scalable alternatives for multi-species assessments⁹.

Red List Criterion	Primary Focus of Measurement	Application to Climate Change Forecasts	Key Methodological Limitations
Criterion A3	Projected future population reduction	Uses modeled future habitat loss as a direct 1:1 proxy for population decline.	Assumes a linear relationship between habitat area and population; highly sensitive to temporal horizons.
Criterion B	Geographic range size and fragmentation	Identifies restricted spatial distributions combined with ongoing environmental threats.	Can overestimate risk for naturally rare but stable localized endemics if future threats are only suspected.
Criterion C & D	Small population sizes and declines	Focuses on absolute abundance thresholds and extreme rarity.	Highly difficult to project future absolute abundances without complex demographic data.
Criterion E	Quantitative probability of extinction	Uses population viability analysis to forecast extinction probability (e.g., 10% in 100 years).	Requires immense amounts of species-specific empirical data; often produces highly conservative estimates.

Predictive Ecological Modeling Paradigms

To fulfill the rigorous data requirements of the Red List criteria under future climate scenarios, conservation scientists deploy two primary paradigms of spatial modeling: Species Distribution Models and Spatially Explicit Population Models.

Species Distribution Models are correlative tools that construct a mathematical representation of a species' current environmental niche. They achieve this by statistically linking known geographical occurrence data with prevailing abiotic variables, such as precipitation gradients, maximum temperatures, and topographic complexity⁹. Algorithms utilized in these models range from traditional Generalized Additive Models to sophisticated machine-learning techniques like Random Forests, Maximum Entropy, and boosted regression trees²⁰. To improve predictive reliability, researchers frequently use ensemble modeling frameworks, such as the biomod2 package in the R programming environment, which aggregates multiple modeling techniques to produce consensus forecasts²¹. These niche models are subsequently projected onto forecasted future climate layers to estimate where suitable habitat will persist, newly emerge, or disappear entirely²¹.

While computationally efficient and widely accessible for thousands of species, Species Distribution Models are inherently static and correlative. They operate on the foundational assumption that species are currently in perfect equilibrium with their environments⁹. Crucially, they omit vital biological mechanisms such as individual dispersal constraints, interspecies competition, and complex demographic rates (like age-specific survival and fecundity)⁹. In the context of Red List assessments, the forecasted reduction in suitable habitat area generated by these models is often plugged directly into Criterion A3 as a proxy for population decline, under the heavily scrutinized linear assumption⁹.

Conversely, Spatially Explicit Population Models are process-based, mechanistic frameworks primarily used to satisfy the rigorous requirements of Criterion E¹. These models integrate structural demography—such as life stages, varying reproductive success, and carrying capacity—alongside spatial dynamics, including habitat fragmentation, inter-patch connectivity, and active dispersal behavior²⁴.

Modern iterations of these mechanistic models, such as the RangeShiftR platform, represent the cutting edge of computational ecology. Built using object-oriented C++ for computational speed and interfaced through R, RangeShiftR simulates individual-based spatial eco-evolutionary dynamics²⁵. It allows researchers to model how individual organisms move through temporally changing landscapes, incorporating evolving genetic traits, complex mating systems, and density-dependent mortality²⁵. While they offer a profoundly more biologically realistic portrayal of population viability than correlative niche maps, Spatially Explicit Population Models are notoriously data-hungry. They require specific empirical estimates for vital rates, dispersal kernels, and environmental stochasticity—data that simply do not exist for the vast majority of the world's threatened biodiversity¹⁶.

Simulating Climate-Driven Extinction Dynamics

Given the extended, multi-decadal timelines of anthropogenic climate change and the logistical impossibility of observing future extinctions empirically before they happen, researchers have increasingly turned to virtual ecology to rigorously test the robustness of international conservation guidelines. Advanced simulation platforms allow ecologists to generate complete, omniscient datasets of virtual species, free from the observation errors and sampling biases that plague empirical field data.

In a landmark 2026 study published in the journal Nature Ecology & Evolution, a team of researchers from the University of Potsdam systematically evaluated the International Union for Conservation of Nature guidelines by simulating the fates of virtual species facing a severe, ninety-year climate change scenario¹. To ensure their findings were ecologically robust and not merely artifacts of a single, specific geographical configuration, the researchers utilized a spatially explicit, individual-based modeling framework across three stochastically replicated, artificially generated landscapes⁹.

The experimental design was meticulous: all simulated species were programmed in such a way that they would eventually face total extinction due to progressive habitat loss driven by the shifting climate⁹. This deterministic endpoint allowed the researchers to work backward. By iteratively evaluating the species' simulated populations every single year of the ninety-year period, they could pinpoint the exact moment the models would theoretically trigger an International Union for Conservation of Nature conservation warning, and compare that moment to the true time of the species' functional extinction⁹.

The Influence of Life-History Traits on Range Dynamics

To accurately capture the vast diversity of the natural world, the researchers simulated sixteen distinct virtual species, endowing them with contrasting life-history and dispersal traits. These traits were categorized across four primary axes: niche position, niche breadth (wide versus narrow environmental tolerances), intrinsic growth rate (fast versus slow reproduction), and dispersal distance (long versus short)⁹. All virtual species were simulated with a standardized generation time of one generation per year to maintain demographic parity⁹.

The position of a species' climatic niche relative to the available landscape fundamentally dictated its spatial response to the changing climate. Half of the virtual species were assigned marginal, cold-adapted niches. These species were restricted to the extreme "northern" edges of the simulated landscapes. As the climate warmed, these species experienced pure range-contracting dynamics; their suitable habitats simply shrank against the boundary of the landscape with nowhere left to go⁹.

The other half of the simulated species were assigned central niche positions, meaning their initial habitats were located in the interior of the landscape. As temperatures shifted, these species exhibited range-shifting dynamics. To survive, they had to physically migrate across the landscape to track their moving climatic envelopes toward the cooler "northern" boundaries⁹.

The interactions between these life-history traits and the shifting environment produced highly variable extinction speeds. As anticipated by classical ecological theory, species hampered by slow reproductive growth rates, narrow environmental tolerances, and short dispersal distances perished significantly earlier than highly vagile, fast-growing generalist species⁹. However, while dispersal and growth traits influenced the timing of extinction, the most profound insight generated by the simulations related to the mathematical shape of the populations' declines, which was governed almost entirely by whether a species was contracting or shifting its range.

Deconstructing the Fallacy of the Linear Assumption

Prior to the onset of the simulated climate change, the researchers sampled occurrence data from the virtual species to train an ensemble of correlative Species Distribution Models. These models demonstrated exceptionally high predictive accuracy, achieving an Area Under the Receiver Operating Characteristic Curve score of 0.95 and a True Skill Statistic of 0.78⁹. Such high scores indicate that the models were perfectly capable of identifying suitable habitat based on environmental variables⁹.

The researchers then projected these models forward through the ninety-year climate scenario to predict future habitat loss. To evaluate the validity of the Red List Criterion A3 guidelines, they utilized a Bayesian generalized mixed-model analysis of covariance—specifically employing ordered beta regression techniques—to statistically compare the "true" simulated population sizes against the habitat loss predicted by the Species Distribution Models⁹.

The standard application of Species Distribution Models to Red List Criterion A3 hinges on a strict linear assumption: if a species loses fifty percent of its climatically suitable habitat area, it is assumed to have lost exactly fifty percent of its population abundance¹. The 2026 statistical analyses proved this assumption to be empirically unsupported, highly misleading, and dangerously dependent on the underlying range dynamics of the species in question¹.

Range-Contracting Species: The Convex Reality

For species with marginal niches that experienced purely range-contracting dynamics, the mathematical relationship between habitat loss and population size was distinctly convex⁹. In these scenarios, the total population size remained relatively stable and highly resilient during the initial and intermediate phases of habitat degradation²⁸. Because these species were not forced to migrate through novel or unsuitable terrain, they could persist in situ, effectively packing into higher densities within the remnants of their shrinking historical ranges.

It was only after habitat loss crossed a severe, critical threshold—often exceeding fifty percent of the original area across all replicate landscapes—that the populations experienced a steep, sudden demographic collapse⁹. Because the observable loss of geographic habitat preceded the final population crash, the Species Distribution Models successfully served as reliable leading indicators. The correlative models accurately predicted the disappearance of the habitat well before the population plummeted, thereby providing the simulated conservationists with adequate and timely warnings to intervene¹.

Range-Shifting Species: The Concave Reality

The dynamics for range-shifting species revealed a starkly different and deeply concerning pattern. For these species, the relationship between predicted habitat loss and actual population size was distinctly concave⁹. When these species attempted to track their shifting climatic envelopes across the landscape, they encountered intense spatial and demographic friction. Even minor losses of original optimal habitat triggered severe, immediate, and disproportionate population drops⁹.

This concave decline is driven by several interacting biological realities that static, correlative models inherently ignore. First, populations at the trailing edge of a shifting range often collapse rapidly due to acute thermal stress, drought, or changing biotic interactions long before populations at the leading edge can successfully disperse and colonize new territories⁶. Second, colonization time-lags and intrinsic dispersal limitations mean that while a Species Distribution Model might accurately identify a newly warmed region as "climatically suitable," the species simply cannot physically traverse the landscape fast enough to establish viable populations there⁹.

Furthermore, migrating populations rarely traverse pristine, contiguous environments. They often face fragmented landscapes, geographic obstacles, suboptimal microhabitats, and novel predator-prey dynamics that severely depress reproductive success and elevate mortality rates during transit²³.

Consequently, the actual demographic abundance of a range-shifting species collapses much faster than the projected loss of its overall suitable habitat area⁹. By the time a correlative Species Distribution Model registers enough geographic habitat loss to trigger a threatened categorization under the linear assumption of Criterion A3, the species' population has already plummeted. The species may already be trapped in an irreversible extinction vortex, rendering the conservation warning moot⁹.

Range Dynamic Type	Biological Niche Position	Habitat-Population Relationship	Primary Ecological Mechanism	Efficacy of Model Warning Time
Range-Contracting	Marginal / Boundary	Convex	In-situ persistence at high densities until a critical spatial threshold is breached.	Adequate (Geographic habitat loss clearly precedes the demographic population crash)
Range-Shifting	Central / Interior	Concave	Dispersal lags, landscape friction, and high mortality during migration.	Severely Delayed (Demographic population crashes occur long before habitat loss registers)

The Crisis of Warning Times and Conservation Implications

The ultimate purpose of a formal conservation assessment is not merely to document the tragic mathematics of an extinction after the fact. Rather, the goal is to provide sufficient "warning time"—defined as the temporal window between the formal identification of a species as threatened and its inevitable functional extinction in the absence of human intervention⁹. Effective warning times must be long enough to allow for the drafting of complex environmental policy, the mobilization of international funding, and the physical implementation of conservation strategies. Such strategies might include large-scale habitat restoration, the establishment of protected wildlife corridors, or highly controversial interventions like managed relocation and assisted migration³¹.

The reliance on Species Distribution Models for range-shifting species systematically truncates this vital warning time, lulling policymakers and environmental managers into a false sense of security⁹. The standard models suggest a slow, linear, proportional decline, whereas the biological reality is a rapid, early demographic collapse¹. This fundamental mismatch means that by the time a range-shifting species is officially flagged as Vulnerable or Endangered on the Red List, the remaining isolated populations may be too small, geographically fragmented, or genetically impoverished to save through traditional means.

If simple correlative models fail so drastically for range-shifting species, one might assume that the complex, mechanistic Spatially Explicit Population Models used under Criterion E would provide the ultimate solution. However, the simulation analyses revealed a different set of flaws. Probabilistic extinction estimates derived from Spatially Explicit Population Models were found to be overly conservative across the board¹. Because these advanced models require exceedingly high thresholds of statistical certainty to declare a specific probability of absolute extinction within a set timeframe, they systematically issued delayed warnings across all species types¹. This delay was particularly pronounced for highly threatened species experiencing rapid environmental degradation. Therefore, while biologically accurate, Criterion E is currently rendered largely ineffective as a practical, early-warning system for the rapid pace of modern climate change⁹.

These modeling biases do not exist in a vacuum; they interact dangerously with known empirical extinction risks. Research integrating life-history traits with spatial data has shown that while the background risk of extinction for certain taxonomic groups (like endemic reptiles and amphibians) might be less than one percent without climate change, shifting climates raise that probability to nearly twenty-eight percent by the year 2100³³. The failure of our primary assessment models to accurately capture the rapid decline of range-shifters means that this massive wave of impending extinctions is currently slipping under the radar of global conservation monitoring.

The implications for global biodiversity governance are profound. The International Union for Conservation of Nature Red List is not merely an academic ledger; it is the foundational architecture upon which international treaties, national environmental protection acts, and billions of dollars in corporate and governmental conservation funding are based¹¹. If the extinction risks for range-shifting species are consistently undercounted, current global conservation portfolios are fundamentally misaligned with biological reality²³. Conservation planners may be unknowingly prioritizing resources for range-contracting species—whose plights are accurately captured by correlative models—while inadvertently abandoning highly vulnerable range-shifting species whose populations are secretly collapsing behind the facade of seemingly abundant, yet unreachable, suitable habitat.

Furthermore, this methodological flaw risks actively undermining the achievement of international biodiversity targets, such as the Kunming-Montreal Global Biodiversity Framework³⁶. Achieving these ambitious targets requires highly accurate environmental baselines and responsive, causality-informed monitoring systems. The failure to detect rapid demographic collapses in migrating species could lead to a cascading loss of functional ecological diversity and critical ecosystem services long before the threat is even formally recognized in policy documents³⁷.

Strategic Recommendations for Modernizing Assessment Guidelines

To close the dangerous gap between predictive ecological modeling and on-the-ground biological reality, the scientific community and conservation practitioners must adopt a more integrated, mechanistic approach to risk assessment. The exhaustive findings from recent virtual simulations and historical macroecological studies provide a clear, evidence-based mandate for updating the standard guidelines used to evaluate climate-related threats¹.

Abandoning the Linear Assumption

The most immediate and critical priority is the formal, institutional recognition that habitat loss does not equate to a linear decline in population size under climate change scenarios. Assessment guidelines must be revised to explicitly acknowledge the concave population-habitat relationship inherent to range-shifting species¹. When assessors utilize correlative Species Distribution Models under Criterion A3, they must be instructed to apply rigorous correction factors or heavier risk weightings for species that are forced to track shifting climates. These adjustments must mathematically compensate for the anticipated, severe demographic crashes caused by dispersal lags and spatial friction⁹.

Integrating Dispersal and Macroecological Constraints

While pure Spatially Explicit Population Models may remain too computationally heavy and data-intensive for widespread, rapid application across tens of thousands of species, hybrid modeling approaches offer a highly viable middle ground. Conservation scientists must move beyond basic, unconstrained correlative niche mapping and incorporate, at minimum, basic dispersal constraints into Species Distribution Models²⁹. By applying spatial buffer zones that represent a species' maximum realistic annual migration capacity, modelers can filter out geographically inaccessible "suitable habitat," providing a much more accurate, conservative estimate of true habitat availability⁹.

Furthermore, integrating readily available macroecological life-history traits—such as adult body mass, clutch or litter size, and average generation length—into predictive frameworks can help identify which species are most highly susceptible to the demographic costs of range shifting¹². Developing criterion-specific extinction risk models that weigh these biological covariates differently depending on the specific threat pathway will greatly enhance the sensitivity and accuracy of the Red List¹². For instance, a species with a massive geographic range but an exceptionally slow intrinsic growth rate might not trigger traditional restricted-range criteria, but a criterion-specific model could correctly flag it as highly vulnerable to the friction of climate migration³⁴.

Redefining Early-Warning Metrics

Because traditional assessment metrics relying on raw habitat area percentages or highly certain probabilistic extinction thresholds provide belated warnings, the conservation community must identify and validate novel early-warning indicators¹. These could include monitoring real-time changes in functional diversity and trait composition within ecological communities, which have been shown to decline and homogenize even when raw taxonomic species counts appear temporarily stable³⁸. Additionally, tracking shifts in absolute population density specifically at the trailing edges of a species' range may provide a much more sensitive, immediate leading indicator of climate stress than measuring overall range extent⁶.

Prioritizing Landscape Connectivity

From a practical, boots-on-the-ground management perspective, the extreme vulnerability of range-shifting species underscores the absolute critical importance of landscape connectivity. Because these species experience their most severe demographic losses during the physical act of migration, conservation efforts must prioritize the creation of permeable, highly connected landscapes²³. Expanding protected area networks to encompass not just current habitats, but projected climate refugia and the transitional corridors between them, is essential. Establishing robust, ecologically viable wildlife corridors will directly reduce the spatial friction that drives the concave population collapse, thereby giving range-shifting species a fighting chance to track their climatic niches successfully into the next century²³.

Works cited

1. When species are forced to move: Prediction models underestimate climate-related extinction risk | EurekAlert!, https://www.eurekalert.org/news-releases/1135648

2. Climate change, species–area curves and the extinction crisis - Royal Society Publishing, https://royalsocietypublishing.org/rstb/article/361/1465/163/20679/Climate-change-species-area-curves-and-the

3. (PDF) Extinction risk from climate change - ResearchGate, https://www.researchgate.net/publication/8928870_Extinction_risk_from_climate_change

4. Extinction risks from climate change: macroecological and historical insights - PMC, https://pmc.ncbi.nlm.nih.gov/articles/PMC2924687/

5. Accelerating extinction risk from climate change - UBC Zoology, https://www.zoology.ubc.ca/edg/pdfs/Urban2015.pdf

6. Climate Crisis Could Cause a Third of Plant and Animal Species to Disappear Within 50 Years: Study | Common Dreams, https://www.commondreams.org/news/2020/02/14/climate-crisis-could-cause-third-plant-and-animal-species-disappear-within-50-years

7. Recent responses to climate change reveal the drivers of species extinction and survival, https://lindseyresearch.com/wp-content/uploads/2021/12/NHTSA-2021-0053-1556-90-Roman-Palacios-2020_Recent-responses-to-climate-change.pdf

8. How many species will Earth lose to climate change? - ResearchGate, https://www.researchgate.net/publication/377152993_How_many_species_will_Earth_lose_to_climate_change

9. Red List criteria underestimate climate-related extinction risk of range-shifting species - bioRxiv, https://www.biorxiv.org/content/10.1101/2025.08.20.671260v2.full.pdf

10. Models used for Red List assessments underestimate climate-related extinction risk of range-shifting species - ResearchGate, https://www.researchgate.net/publication/408733027_Models_used_for_Red_List_assessments_underestimate_climate-related_extinction_risk_of_range-shifting_species

11. IUCN RED LIST, https://iucn.org/sites/default/files/2025-10/red-list-60-years-of-success-report-03.10.25.pdf

12. Modelling the probability of meeting IUCN Red List criteria to support reassessments, https://www.researchgate.net/publication/371499467_Modelling_the_probability_of_meeting_IUCN_Red_List_criteria_to_support_reassessments

13. GUIDELINES FOR APPROPRIATE USES OF RED LIST DATA - IUCN Portals, https://portals.iucn.org/library/sites/library/files/documents/Rep-2009-007.pdf

14. The risk of underestimating generation length for extinction risk assessments | Cambridge Prisms, https://www.cambridge.org/core/journals/cambridge-prisms-extinction/article/risk-of-underestimating-generation-length-for-extinction-risk-assessments/65C1E90126782C65AAE01F8180A85360

15. Improvements to the Red List Index | PLOS One - Research journals, https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0000140

16. Population Viability Analysis in the Classification of Threatened Species: Problems and Potentials | Request PDF - ResearchGate, https://www.researchgate.net/publication/259798922_Population_Viability_Analysis_in_the_Classification_of_Threatened_Species_Problems_and_Potentials

17. Use and Misuse of the IUCN Red List Criteria in Projecting Climate Change Impacts on Biodiversity - Polar bear science, https://polarbearscience.com/wp-content/uploads/2014/11/akc3a7akaya-et-al-use-and-misude-iucn-redlistclimatechange_2006-author-copy.pdf

18. (PDF) A general relationship between carrying capacity and extinction risk - ResearchGate, https://www.researchgate.net/publication/392121717_A_general_relationship_between_carrying_capacity_and_extinction_risk

19. (PDF) Standards for distribution models in biodiversity assessments - ResearchGate, https://www.researchgate.net/publication/330442703_Standards_for_distribution_models_in_biodiversity_assessments

20. Which species distribution models are more (or less) likely to project broad-scale, climate-induced shifts in species ranges? - Western Sydney University, https://researchers.westernsydney.edu.au/en/publications/which-species-distribution-models-are-more-or-less-likely-to-proj/

21. Predictive performance of presence‐only species distribution models: a benchmark study with reproducible code - ResearchGate, https://www.researchgate.net/publication/355148233_Predictive_performance_of_presence-only_species_distribution_models_a_benchmark_study_with_reproducible_code

22. Habitat Suitability and Distribution Models: With Applications in R - ResearchGate, https://www.researchgate.net/publication/325677986_Habitat_suitability_and_distribution_models_With_applications_in_R

23. Climate models may undercount extinctions as species shift ranges, study warns - Medium, https://medium.com/@ramilmdelrosario/climate-models-may-undercount-extinctions-as-species-shift-ranges-study-warns-ce975f4eae8c

24. (PDF) Population viability analyses with demographically structured models - ResearchGate, https://www.researchgate.net/publication/234014125_Population_viability_analyses_with_demographically_structured_models

25. RangeShifter 2.0: an extended and enhanced platform for modelling spatial eco‐evolutionary dynamics and species' responses to environmental changes - ResearchGate, https://www.researchgate.net/publication/354209731_RangeShifter_20_an_extended_and_enhanced_platform_for_modelling_spatial_eco-evolutionary_dynamics_and_species'_responses_to_environmental_changes

26. RangeShiftR: an R package for individual‐based simulation of spatial eco‐evolutionary dynamics and species' responses to environmental changes - ResearchGate, https://www.researchgate.net/publication/354209568_RangeShiftR_an_R_package_for_individual-based_simulation_of_spatial_eco-evolutionary_dynamics_and_species'_responses_to_environmental_changes

27. Distribution of virtual species and SDM-predicted habitat suitability... | Download Scientific Diagram - ResearchGate, https://www.researchgate.net/figure/Distribution-of-virtual-species-and-SDM-predicted-habitat-suitability-for-two-example_fig1_394922562

28. Population size -habitat loss relationship for the different range... - ResearchGate, https://www.researchgate.net/figure/Population-size-habitat-loss-relationship-for-the-different-range-dynamics-Relative_fig2_394922562

29. Range and climate niche shifts in European and North American breeding birds, https://royalsocietypublishing.org/rstb/article/379/1902/20230013/42857/Range-and-climate-niche-shifts-in-European-and

30. Quantifying the extinction vortex | Request PDF - ResearchGate, https://www.researchgate.net/publication/6831239_Quantifying_the_extinction_vortex

31. Expert opinion on extinction risk and climate change adaptation for biodiversity - UC Press Journals, https://online.ucpress.edu/elementa/article/doi/10.12952/journal.elementa.000057/112719/Expert-opinion-on-extinction-risk-and-climate

32. Detecting Extinction Risk from Climate Change by IUCN Red List Criteria - ResearchGate, https://www.researchgate.net/publication/260151562_Detecting_Extinction_Risk_from_Climate_Change_by_IUCN_Red_List_Criteria

33. Climate Change Causes High, but Predictable, Extinction Risks - SBU News, https://news.stonybrook.edu/newsroom/press-release/general/02252014_climatechangepredictableextinction/

34. Life history and spatial traits predict extinction risk due to climate change - ResearchGate, https://www.researchgate.net/publication/260496420_Life_history_and_spatial_traits_predict_extinction_risk_due_to_climate_change

35. Bridging the research-implementation gap in IUCN Red List assessments - ResearchGate, https://www.researchgate.net/publication/357934266_Bridging_the_research-implementation_gap_in_IUCN_Red_List_assessments

36. Use and misuse of the IUCN Red List Criteria in projecting climate change impacts on biodiversity | Request PDF - ResearchGate, https://www.researchgate.net/publication/236679159_Use_and_misuse_of_the_IUCN_Red_List_Criteria_in_projecting_climate_change_impacts_on_biodiversity

37. Integrative Biodiversity Science Informing Transformative Multiscale Governance through Diversifying Values of Nature in Scenarios and Models - ResearchGate, https://www.researchgate.net/publication/406316820_Integrative_Biodiversity_Science_Informing_Transformative_Multiscale_Governance_through_Diversifying_Values_of_Nature_in_Scenarios_and_Models

38. Susanne A. Fritz's research while affiliated with University of Wuerzburg and other places, https://www.researchgate.net/scientific-contributions/Susanne-A-Fritz-2162916053

39. Predicting extinctions with species distribution models - ResearchGate, https://www.researchgate.net/publication/376311116_Predicting_extinctions_with_species_distribution_models

40. Modelling the probability of meeting IUCN Red List criteria to support reassessments - IIASA PURE, https://pure.iiasa.ac.at/id/eprint/19407/1/Global%20Change%20Biology%20-%202024%20-%20Henry%20-%20Modelling%20the%20probability%20of%20meeting%20IUCN%20Red%20List%20criteria%20to%20support%20reassessments.pdf

41. Raya Keuth - Ecology / Macroecology, https://www.uni-potsdam.de/en/ibb-macroecology/team/raya-keuth