Hostname: page-component-7c8c6479df-27gpq Total loading time: 0 Render date: 2024-03-29T11:54:14.350Z Has data issue: false hasContentIssue false

Selectivity of mass extinctions: Patterns, processes, and future directions

Published online by Cambridge University Press:  09 May 2023

Jonathan L. Payne*
Affiliation:
Department of Earth and Planetary Sciences, Stanford University, Stanford, CA, USA
Jood A. Al Aswad
Affiliation:
Department of Earth and Planetary Sciences, Stanford University, Stanford, CA, USA
Curtis Deutsch
Affiliation:
Department of Geosciences, Princeton University, Princeton, NJ, USA
Pedro M. Monarrez
Affiliation:
Department of Earth and Planetary Sciences, Stanford University, Stanford, CA, USA
Justin L. Penn
Affiliation:
Department of Geosciences, Princeton University, Princeton, NJ, USA
Pulkit Singh
Affiliation:
Department of Earth and Planetary Sciences, Stanford University, Stanford, CA, USA
*
Corresponding author: Jonathan L. Payne; Email: jlpayne@stanford.edu
Rights & Permissions [Opens in a new window]

Abstract

A central question in the study of mass extinction is whether these events simply intensify background extinction processes and patterns versus change the driving mechanisms and associated patterns of selectivity. Over the past two decades, aided by the development of new fossil occurrence databases, selectivity patterns associated with mass extinction have become increasingly well quantified and their differences from background patterns established. In general, differences in geographic range matter less during mass extinction than during background intervals, while differences in respiratory and circulatory anatomy that may correlate with tolerance to rapid change in oxygen availability, temperature, and pH show greater evidence of selectivity during mass extinction. The recent expansion of physiological experiments on living representatives of diverse clades and the development of simple, quantitative theories linking temperature and oxygen availability to the extent of viable habitat in the oceans have enabled the use of Earth system models to link geochemical proxy constraints on environmental change with quantitative predictions of the amount and biogeography of habitat loss. Early indications are that the interaction between physiological traits and environmental change can explain substantial proportions of observed extinction selectivity for at least some mass extinction events. A remaining challenge is quantifying the effects of primary extinction resulting from the limits of physiological tolerance versus secondary extinction resulting from the loss of taxa on which a given species depended ecologically. The calibration of physiology-based models to past extinction events will enhance their value in prediction and mitigation efforts related to the current biodiversity crisis.

Type
Review
Creative Commons
Creative Common License - CCCreative Common License - BY
This is an Open Access article, distributed under the terms of the Creative Commons Attribution licence (http://creativecommons.org/licenses/by/4.0), which permits unrestricted re-use, distribution and reproduction, provided the original article is properly cited.
Copyright
© The Author(s), 2023. Published by Cambridge University Press

Impact statement

Mass extinction events represent the greatest catastrophes in the history of animal life and only five major extinction events have occurred across the past 550 million years. Geological evidence can reveal the physical and chemical processes that caused environmental change, but differences in morphological, ecological, and physiological traits between extinction victims and survivors provide our best record of actual kill mechanisms. In recent years, this field has advanced both through the compilation of experimental data on organismal traits, enabling new insights into extinction patterns, and through the development of mechanistic models for biological response to environmental change, enabling incorporation of physiological tolerance into climate models to predict extinction patterns. Ultimately, mass extinction events are a critical source of data to calibrate the magnitude and rate of biological response to climate change over timescales longer than those of experiments and field studies. In this way, integration of information from the fossil record is becoming essential to the task of predicting and mitigating taxonomic losses due to current environmental change.

Introduction

Earth is currently undergoing a biodiversity crisis on a scale unprecedented in the history of the human species (Barnosky et al., Reference Barnosky, Matzke, Tomiya, Wogan, Swartz, Quental, Marshall, McGuire, Lindsey, Maguire, Mersey and Ferrer2011; Dirzo et al., Reference Dirzo, Young, Galetti, Ceballos, Isaac and Collen2014; McCauley et al., Reference McCauley, Pinsky, Palumbi, Estes, Joyce and Warner2015), but crises of similar or greater magnitude have occurred at least five times across the 600-million-year history of animal life (Figure 1A) (Raup and Sepkoski, Reference Raup and Sepkoski1982; Barnosky et al., Reference Barnosky, Matzke, Tomiya, Wogan, Swartz, Quental, Marshall, McGuire, Lindsey, Maguire, Mersey and Ferrer2011). All major mass extinction events are associated with evidence of rapid environmental change. In some cases, such as the end-Permian (252 million years ago [Mya]) and end-Triassic (201 Mya) mass extinctions, there is evidence for rapid and pronounced climate warming (Kiessling and Simpson, Reference Kiessling and Simpson2011; Payne and Clapham, Reference Payne and Clapham2012; Blackburn et al., Reference Blackburn, Olsen, Bowring, McLean, Kent, Puffer, McHone, Rasbury and 'Et-Touhami2013; Burgess et al., Reference Burgess, Bowring and Shen2014; Bond and Sun, Reference Bond, Sun, Ernst, Dickson and Bekker2021). By contrast, the Late Ordovician (443 Mya) and Late Devonian (372 Mya) extinctions occurred in association with climate cooling (Joachimski and Buggisch, Reference Joachimski and Buggisch2002; Finnegan et al., Reference Finnegan, Bergmann, Eiler, Jones, Fike, Eisenman, Hughes, Tripati and Fischer2011). The end-Cretaceous extinction (66 Mya) was associated with an asteroid impact event whose aftermath resembled the consequences of a hypothetical global thermonuclear war (Pollack et al., Reference Pollack, Toon, Ackerman, McKay and Turco1983; Turco et al., Reference Turco, Toon, Ackerman, Pollack and Sagan1983). Due to the magnitude and global scale of the current “Sixth” extinction, these events from Earth’s past provide historical reference points for predicting the long-term magnitude, ecological impact, and recovery timescale from the current crisis or other, potential, human-mediated catastrophes.

Figure 1. Extinction patterns in the fossil record. (A) Graph of marine animal diversity across the past 600 million years, illustrating the diversity declines associated with the five major mass extinction events (modified from Raup and Sepkoski, Reference Raup and Sepkoski1982). (B) Extinction selectivity with respect to geographic range, illustrating the preferential survival of broadly distributed genera during background intervals and the greatly reduced selectivity during mass extinction events (modified from Payne and Finnegan, Reference Payne and Finnegan2007). (C) Principal components analysis of logistic regression coefficients of ecological traits and body size selectivity of the Big Five mass extinction events and the modern oceans, demonstrating the unique selectivity of the modern extinction threat (modified from Payne et al., Reference Payne, Bush, Heim, Knope and McCauley2016b). (D) Extinction selectivity during the end-Permian mass extinction, illustrating the preferential extinction of heavily calcified marine animal classes with less complex respiratory and circulatory systems (modified from Knoll et al., Reference Knoll, Bambach, Payne, Pruss and Fischer2007; Knoll and Fischer, Reference Knoll, Fischer, Gattuso and Hansson2011). (E) Extinction selectivity with respect to body size for major classes of marine animals, illustrating the general bias of background extinction against smaller-bodied genera versus the variable direction of selectivity for classes that exhibit distinct patterns during mass extinction (modified from Monarrez et al., Reference Monarrez, Heim and Payne2021).

While mass extinctions have been identified in the fossil record based largely on the magnitude of diversity loss across many higher taxa (Newell, Reference Newell1963, Reference Newell1967; Raup and Sepkoski, Reference Raup and Sepkoski1982), causal inference has relied more on geological and geochemical evidence of potential triggers (Alvarez et al., Reference Alvarez, Alvarez, Asaro and Michel1980; Svensen et al., Reference Svensen, Planke, Polozov, Schmidbauer, Corfu, Podladchikov and Jamtveit2009; Finnegan et al., Reference Finnegan, Bergmann, Eiler, Jones, Fike, Eisenman, Hughes, Tripati and Fischer2011) and patterns of extinction selectivity interpreted to reflect proximal kill mechanisms (Jablonski, Reference Jablonski1986; Sheehan and Hansen, Reference Sheehan and Hansen1986; Valentine and Jablonski, Reference Valentine and Jablonski1986; Jablonski and Raup, Reference Jablonski and Raup1995; Knoll et al., Reference Knoll, Bambach, Canfield and Grotzinger1996, Reference Knoll, Bambach, Payne, Pruss and Fischer2007; Smith and Jeffery, Reference Smith and Jeffery1998; Finnegan et al., Reference Finnegan, Heim, Peters and Fischer2012; Penn et al., Reference Penn, Deutsch, Payne and Sperling2018). Selectivity patterns have been assessed with respect to a wide range of traits (Figure 1B–E), including geographic range (Jablonski, Reference Jablonski1986; Kiessling and Aberhan, Reference Kiessling and Aberhan2007; Payne and Finnegan, Reference Payne and Finnegan2007; Dunhill and Wills, Reference Dunhill and Wills2015), body size (Jablonski and Raup, Reference Jablonski and Raup1995; Friedman, Reference Friedman2009; Longrich et al., Reference Longrich, Bhullar and Gauthier2012; Allen et al., Reference Allen, Stubbs, Benton and Puttick2019; Payne and Heim, Reference Payne and Heim2020; Monarrez et al., Reference Monarrez, Heim and Payne2021), abundance (Lockwood, Reference Lockwood2003; Payne et al., Reference Payne, Truebe, Nützel, Chang and Nu2011), larval ecology (Valentine and Jablonski, Reference Valentine and Jablonski1986), diet (Wilson, Reference Wilson2013), functional ecology (Bambach et al., Reference Bambach, Knoll and Sepkoski2002; Payne et al., Reference Payne, Bush, Heim, Knope and McCauley2016b; Hughes et al., Reference Hughes, Berv, Chester, Sargis and Field2021), environmental breadth (Jablonski and Raup, Reference Jablonski and Raup1995), respiratory and circulatory anatomy (Knoll et al., Reference Knoll, Bambach, Canfield and Grotzinger1996, Reference Knoll, Bambach, Payne, Pruss and Fischer2007; Clapham, Reference Clapham2017), and shell mineralogy (Clapham and Payne, Reference Clapham and Payne2011; Kiessling and Simpson, Reference Kiessling and Simpson2011).

Extinction selectivity provides our most direct evidence of proximal kill mechanisms (Raup, Reference Raup1986), but to date, most testing of observed extinction patterns against hypothesized kill mechanisms has been semi-quantitative, focused on establishing consistency between predicted and observed directions of selectivity under various hypothesized kill mechanisms. Recently, advances in paleontological databases, geochemical proxies, physiological experiments, and Earth system and ecosystem models have enabled the comparison of observed and predicted extinction patterns within quantitative, self-consistent frameworks (Figure 2) (Penn et al., Reference Penn, Deutsch, Payne and Sperling2018). Although quantitative model-data comparison between observed and predicted extinction patterns is still in its early days, the door for direct comparison of past and future biotic response to climate change is now open, increasing the value of the fossil record in the mitigation of the current biotic crisis.

Figure 2. Workflow illustrating the use of geological and geochemical data to constrain Earth system models (ESMs), physiological experiments to constrain parameters used to populate models with species of different ecophysiotypes, and fossil occurrence data to conduct model-data comparison. Ecosystem structure remains to be incorporated into such models and can be used to predict extinction cascades. Calibration of models against selectivity patterns in ancient extinction events will improve their use in forecasting biotic response to current and future environmental change. Panels on right showing CO2 emissions curves and future biodiversity projections are from Penn and Deutsch (Reference Penn and Deutsch2022).

Pattern

Analyses of selectivity for individual mass extinction events date back many decades (Jablonski, Reference Jablonski2005). Studies synthesizing and comparing selectivity patterns across all major mass extinctions (and intervening background intervals) have emerged more recently, alongside publicly available databases of fossil occurrences and other traits (Alroy, Reference Alroy and MacPhee1999; Payne and Finnegan, Reference Payne and Finnegan2007; Peters, Reference Peters2008; Kiessling and Simpson, Reference Kiessling and Simpson2011; Payne et al., Reference Payne, Bush, Heim, Knope and McCauley2016b; Smith et al., Reference Smith, Elliott Smith, Lyons and Payne2018; Payne and Heim, Reference Payne and Heim2020; Monarrez et al., Reference Monarrez, Heim and Payne2021).

Geographic range is one of the traits most commonly hypothesized to correlate with extinction risk due to its influence on the extent to which populations of a given taxon may avoid a regional disturbance or have broad enough physiological tolerance limits or ecological capacities to survive a global one. Analyses of fossil data have confirmed that widely distributed taxa survive preferentially during background intervals (Figure 1C) (Jablonski, Reference Jablonski1986, Reference Jablonski2005; Payne and Finnegan, Reference Payne and Finnegan2007). Broader geographic range is also significantly associated with survival during at least some major mass extinction events (Jablonski and Raup, Reference Jablonski and Raup1995; Finnegan et al., Reference Finnegan, CMØ and DAT2016), but the strength of this association (i.e., the change in odds or probability of extinction per unit change in geographic range) is greatly reduced relative to background intervals (Figure 1C) (Kiessling and Aberhan, Reference Kiessling and Aberhan2007; Payne and Finnegan, Reference Payne and Finnegan2007). Due to the consistency of the association and the expectation of selectivity on total geographic range under most extinction scenarios, these patterns have rarely yielded direct insight into kill mechanisms. By contrast, the biogeography of extinction can be more informative. For example, end-Cretaceous echinoid extinction was significantly more severe in areas proximal to the Chicxulub impact site (Smith and Jeffery, Reference Smith and Jeffery1998), and differences in extinction intensity across latitude often correspond with expectations due to climate change (Finnegan et al., Reference Finnegan, Heim, Peters and Fischer2012; Penn et al., Reference Penn, Deutsch, Payne and Sperling2018; Reddin et al., Reference Reddin, Kocsis and Kiessling2019, Reference Reddin, Kocsis, Aberhan and Kiessling2021). Quantifying the expected magnitude of spatial gradients in extinction intensity and differences in such gradients across higher taxa (or functional groupings) is the key to linking these findings with hypothesized kill mechanisms, and one that is already being partially realized (Penn et al., Reference Penn, Deutsch, Payne and Sperling2018).

The extinctions of large mammals during the Pleistocene (0.0117 Ma) and of large, non-avian dinosaurs during the Maastrichtian (66 Ma) have long prompted speculation that large-bodied animals are at systematically higher risk of extinction during times of environmental change (Raup, Reference Raup1986; Wallace, Reference Wallace1889; Brown, Reference Brown1995). Analyses of the fossil record reveal a more heterogeneous relationship, and one that may differ across taxa and habitats. For example, smaller body size is generally associated with greater extinction risk during background times for many classes of marine animals (Figure 1D) (Payne and Heim, Reference Payne and Heim2020; Monarrez et al., Reference Monarrez, Heim and Payne2021). By contrast, body size was not generally associated with extinction probability for terrestrial mammals until the Pleistocene (Alroy, Reference Alroy and MacPhee1999; Smith et al., Reference Smith, Elliott Smith, Lyons and Payne2018). End-Cretaceous extinctions preferentially eliminated larger-bodied fish, lizards, and snakes (Friedman, Reference Friedman2009; Longrich et al., Reference Longrich, Bhullar and Gauthier2012) but were unbiased in bivalves and gastropods (Jablonski and Raup, Reference Jablonski and Raup1995). End-Permian extinctions preferentially affected larger foraminifera and brachiopods (Schaal et al., Reference Schaal, Clapham, Rego, Wang and Payne2016). Many taxon-size combinations have yet to be examined systematically. In marine animals, size selectivity changes between background and mass extinction in many classes but the direction and magnitude of the size bias during mass extinction differs among classes (Figure 1D) (Payne and Heim, Reference Payne and Heim2020; Monarrez et al., Reference Monarrez, Heim and Payne2021). The differences in responses among classes remain to be explained. Because body size correlates with many ecological and physiological traits (Peters, Reference Peters1983), size bias on its own is insufficient to diagnose proximal kill mechanisms but may be useful in conjunction with other traits or in testing against predictions of specific kill mechanisms (Deutsch et al., Reference Deutsch, Penn, Verberk, Inomura, Endress and Payne2022).

Some mass extinction events exhibit selectivity patterns that can be mapped onto respiratory and circulatory anatomy, potentially reflecting underlying differences in susceptibility to metabolic stress from hypercapnia, anoxia, climate warming, or their interactive effects. For example, the end-Permian mass extinction preferentially affected heavily calcified marine animal genera with limited respiratory and circulatory systems (Figure 1B), suggesting a role for hypercapnia and/or direct and indirect fitness effects of acidification on shell dissolution (Calosi et al., Reference Calosi, Melatunan, Turner, Artioli, Davidson, Byrne, Viant, Widdicombe and Rundle2017) in driving the extinction (Knoll et al., Reference Knoll, Bambach, Canfield and Grotzinger1996). At the same time, the lack of sophisticated oxygen-supply mechanisms would also make these taxa more sensitive to temperature-dependent hypoxia (Deutsch et al., Reference Deutsch, Penn and Seibel2020; Endress et al., Reference Endress, Boag, Burford, Penn, Sperling and Deutsch2022) and metabolic differences among groups likely influence taxonomic selectivity patterns from changes in CO2, temperature, and O2. Similar patterns as seen in the end-Permian apply to other extinction events, including the end-Triassic mass extinction (Clapham, Reference Clapham2017; Kiessling and Simpson, Reference Kiessling and Simpson2011), consistent with shared kill mechanisms. By contrast, the end-Cretaceous mass extinction exhibits the opposite pattern, with taxa thought to be more sensitivity to ocean acidification surviving preferentially (Kiessling and Simpson, Reference Kiessling and Simpson2011), potentially reflecting differences in extinction patterns triggered primarily by volcanism versus impact events. The extent to which these patterns stand out from background extinction remains incompletely studied. A study controlling for differences between benthic versus planktonic and nektonic taxa indicates that many background intervals show the same selectivity, often of similar magnitude (Payne et al., Reference Payne, Bush, Chang, Heim, Knope and Pruss2016a). As discussed below, results of physiological experiments on living relatives of species in the fossil record are enabling quantitative prediction of biological response to past environmental changes inferred from geological and geochemical proxies. This is currently an area of rapid progress.

Simultaneous analysis of extinction selectivity across multiple traits and time intervals enables quantitative comparison of selectivity patterns between background and mass extinction as well as among mass extinction events (Figure 1E). Such analyses generally confirm that mass extinction events differ in selectivity from background patterns (Figure 1C, E) (Payne and Finnegan, Reference Payne and Finnegan2007; Kiessling and Simpson, Reference Kiessling and Simpson2011; Finnegan et al., Reference Finnegan, Heim, Peters and Fischer2012; Payne et al., Reference Payne, Bush, Heim, Knope and McCauley2016b; Monarrez et al., Reference Monarrez, Heim and Payne2021) and that the pronounced size bias of the modern extinction makes it an outlier relative to major mass extinctions as well as to recent background intervals (Figure 1D) (Payne et al., Reference Payne, Bush, Heim, Knope and McCauley2016b; Smith et al., Reference Smith, Elliott Smith, Lyons and Payne2018).

Overall, selectivity patterns accord with geological and geochemical data, indicating that mass extinction events are typically associated with large and rapid environmental perturbations rather than intensification of background extinction processes (Alvarez et al., Reference Alvarez, Alvarez, Asaro and Michel1980; Hallam and Wignall, Reference Hallam and Wignall1997; Finnegan et al., Reference Finnegan, Bergmann, Eiler, Jones, Fike, Eisenman, Hughes, Tripati and Fischer2011). Testing hypothesized kill mechanisms requires simultaneous consideration of selectivity across multiple variables because physiological and ecological traits are often linked in complex ways. For example, body size is related to the supply and demand of oxygen (Deutsch et al., Reference Deutsch, Ferrel, Seibel, Pörtner and Huey2015, Reference Deutsch, Penn, Verberk, Inomura, Endress and Payne2022) and food (Gearty et al., Reference Gearty, McClain and Payne2018) as well as to trophic level (Romanuk et al., Reference Romanuk, Hayward and Hutchings2011).

Process

Introduction

Understanding the causes of extinction selectivity in the fossil record requires additional information about the patterns of environmental change, the sensitivity of species to those changes, and disruptions in ecological networks. The interpretation of extinction selectivity thus relies on geochemical reconstructions of climate, understanding of the ecological and physiological traits of living taxa and, increasingly, on models that incorporate all these aspects of ecological and Earth system dynamics into an internally consistent, quantitative framework (Figure 2).

Patterns of extinction selectivity can arise simply from the fact that environmental changes can be highly variable in strength or even direction across space. Extinction selectivity could also arise from taxonomic or geographic differences in physiological sensitivity to environmental change, even if climate trends were globally uniform. In general, these factors are likely to be connected, as the tolerance limits of taxa to environmental conditions will shape the pre-extinction geographic distribution, which may confer greater or lesser sensitivity to environmental change in certain regions. Contemporary studies have advanced a mechanistic approach to investigating the causes of selectivity in mass extinctions by integrating many of these elements, from geochemical proxies of climate change, the modern diversity of ecophysiological traits, and the climate dynamics of Earth system models. In ocean studies, emphasis has been on integrating climate and physiological constraints (Penn et al., Reference Penn, Deutsch, Payne and Sperling2018; Stockey et al., Reference Stockey, Pohl, Ridgwell, Finnegan and Sperling2021). Terrestrial studies, by contrast, have tended to focus on ecological (food web) mechanisms largely missing from marine analyses (Roopnarine, Reference Roopnarine2006; Roopnarine and Angielczyk, Reference Roopnarine and Angielczyk2015). These dichotomous approaches have made significant advances in their respective domains, paving the way for more unified marine and terrestrial studies.

Example: Metabolic Index

One promising avenue for examining physiological kill mechanisms for ancient extinction events is the Metabolic Index, which was initially developed to test whether the biogeographic distributions of species are physiologically limited by O2 supply and demand in the modern ocean (Deutsch et al., Reference Deutsch, Ferrel, Seibel, Pörtner and Huey2015). This ecophysiological model quantifies habitat viability for a species, in terms of its ability to carry out aerobic respiration, by taking a ratio of environmental oxygen supply to biological oxygen demand as a function of temperature and taxon-specific metabolic and O2 supply traits (Eq. (1)). The metabolic energy demands of water-breathing marine animals increase with water temperature and body size (Gillooly et al., Reference Gillooly, Brown, West, Savage and Charnov2001), raising corresponding biological O2 requirements. Temperature and body size also impact the rates of organismal O2 supply through diffusion, ventilation, and internal circulation (Deutsch et al., Reference Deutsch, Penn, Verberk, Inomura, Endress and Payne2022; Endress et al., Reference Endress, Boag, Burford, Penn, Sperling and Deutsch2022), while warmer water holds less ambient O2. The ratio of temperature and body size (B)-dependent rates of potential O2 supply and organismal metabolic demand, termed the Metabolic Index (ɸ), quantifies the metabolic viability of a habitat for a given species:

(1) $$ \Phi ={A}_{\mathrm{o}}{B}^{\varepsilon }{p\mathrm{O}}_2\mathit{\exp}\left\{\frac{E_{\mathrm{o}}}{k_{\mathrm{B}}}\left[\frac{1}{T}-\frac{1}{T_{\mathrm{ref}}}\right]\right\}, $$

where A 0 (atm−1) is the ratio of O2 supply to resting demand rate coefficients, or hypoxia tolerance at a reference temperature and body size (B), with allometric scaling exponent $ \varepsilon $ and Arrhenius temperature sensitivity, E 0 (eV), and pO2 and T are the oxygen partial pressure and temperature of ambient water, respectively (Figure 3) (Deutsch et al., Reference Deutsch, Ferrel, Seibel, Pörtner and Huey2015, Reference Deutsch, Penn and Seibel2020). These physiological traits and their distributions across taxa can be estimated from critical oxygen thresholds in respirometry experiments conducted for diverse marine biota over the past half century (Rogers et al., Reference Rogers, Urbina, Reardon, McKenzie and Wilson2016; Chu and Gale, Reference JWF and KSP2017). Critical oxygen thresholds define the Metabolic Index to be 1 (i.e., ɸ = 1), allowing the traits to be estimated for organisms in a resting state under laboratory conditions. In the environment, O2 requirements are elevated by more strenuous activities important for population persistence, such as growth, reproduction, feeding, defense, or motion. These additional energy demands require the O2 supply to be raised by a factor, ɸcrit, corresponding to sustained metabolic scope (Peterson et al., Reference Peterson, Nagy and Diamond1990). Stable aerobic habitat barriers thus arise in ocean regions where the Metabolic Index falls below ɸcrit, while the geographic positions of these barriers depend on the species’ traits (Deutsch et al., Reference Deutsch, Penn and Seibel2020). The habitability of any given parcel of water can therefore be determined from the temperature and oxygen partial pressure given the species values of A 0, E 0, and ɸcrit. Earth system models can be populated with hypothetical species by drawing combinations of values from the trait distributions (Figure 3). The promise of this framework for paleontological application is that trait distributions can be used to predict the patterns of biodiversity, providing a means for testing the model against the fossil record. Indeed, the observed tropical dip in marine species richness observed for diverse animal groups in the modern ocean (Chaudhary et al., Reference Chaudhary, Richardson, Schoeman and Costello2021) can be explained by aerobic habitat limitation implied by modern species Metabolic Index traits (Penn and Deutsch, Reference Penn and Deutsch2022). Environmental temperature and oxygen concentration can be quantified using geochemical proxies for ancient events to calibrate Earth system models and body size can be measured from fossil specimens. In principle, ecological interactions can be further incorporated to model, allowing extinction cascades to be accounted for alongside direct, climate-driven habitat loss (Figure 4).

Figure 3. Graphs illustrating the key species traits of the Metabolic Index (ɸ) along with how ɸ relates to temperature and oxygen partial pressure. (A–C) Frequency distributions of the Metabolic Index parameters for marine animals. (D, E) Graphs of variation in ɸ as a function of temperature and oxygen for species with negative (D) and positive (E) temperature sensitivities (E o) of hypoxia tolerance (A o), which is the inverse of the critical oxygen threshold (red circle) at a reference temperature (T ref), as derived from respirometry experiments. For species in a resting state, the aerobic habitat limit occurs when ɸ = 1, but in the environment a species’ activity level or sustained metabolic scope (SMS) elevates the habitat limit to ɸcrit. For species with negative E o, aerobic habitat availability increases with temperature, whereas for those with positive E o (i.e., most species; panel B), aerobic habitat declines with warming. Changes in PO2 has the potential to lower aerobic habitat availability, and thus the amount of warming a species can withstand, as exemplified for two scenarios of with different fractions of present atmospheric levels of O2 (PAL; yellow dots and arrows). A change in CO2 also has the potential to alter hypoxia tolerance, but the magnitude and direction of this effect is unknown across marine biota and is illustrated here from experimental data for a single species under ∆ pH = +0.5 (Rosa et al., Reference Rosa, Trübenbach, Repolho, Pimentel, Faleiro, Boavida-Portugal, Baptista, Lopes, Dionísio, Leal, Calado and Pörtner2013). Arrows in A–C denote species traits in D and E.

Figure 4. Hypothetical progression of a mass extinction highlighting sources of trait-based and geographic selectivity and potential ecological amplification. (A) An initial distribution of species (or “ecophysiotypes”) defined by traits under selection by large-scale environmental conditions will likely result in systematic correlations between traits and geographic range. The range metric here can be considered overall range size (area and volume), or centroid (e.g., low-latitude versus high-latitude, shallow versus deep). (B) The initial biota are subjected to climate perturbation that poses a direct stress through a reduction in fitness whose magnitude depends on species traits and on local climate trends. The resulting change in available habitat (ΔH; contours) presents an ecophysiological extinction risk that is geographically selective because it is trait selective (but may also be caused by climate patterns themselves). In this hypothetical case, habitat loss (ΔH < 0) selects against species with high values of two traits (habitat “Losers”) and may even benefit species with low values of those traits (habitat “Gainers”; ΔH > 0). (C) Physiological extinction poses further ecological risks (or advantages) depending on the mutualistic or adversarial interactions with ecophysiotypes (nodes in graph) that are under trait-selective risk. Ecological risk is complex and for any particular species will depend on the physiological risk faced by the other species with which it interacts, which may be positive (green lines) or negative (brown lines), and strong (thick lines) or weak (thin lines). The results of these associations, which may be multiple and indirect, could alter extinction risk by either preserving ecological fitness (“+” symbol) or reducing it (“−” symbol). Changes in extinction risk are likely to be most pronounced for those in the neutral zone whose antagonists go extinct or who are buoyed by prey/mutualists that are under positive selection. (D) Post-extinction ecosystem, equal to the initial one (A) minus the ecotypes that have gone extinct from either primary (B) or secondary (C) effects.

During periods of climate warming, rising water temperatures can drive the metabolic O2 demand above a supply declining from ocean deoxygenation, leading to the loss of available aerobic habitat, and eventually species extinctions at local and global scales (Penn et al., Reference Penn, Deutsch, Payne and Sperling2018; Reddin et al., Reference Reddin, Nätscher, Kocsis, Pörtner and Kiessling2020). At regional scales, such as in the California Current System, aerobic habitat changes have been linked to multi-decadal fluctuations of anchovy populations, including near-extirpation of larvae from portions of their range (Howard et al., Reference Howard, Penn, Frenzel, Seibel, Bianchi, Renault, Kessouri, Sutula, McWilliams and Deutsch2020). At global scales, aerobic habitat loss under the climate change simulated for the end-Permian mass extinction predicted a geographic selectivity of extinction consistent with the fossil record (Figure 5A): Extinction risk was greater for species inhabiting higher latitudes. This geographic selectivity arises because species previously occupying the tropics would already have been adapted to warm, low-O2 conditions that became more widespread, whereas polar habitat niches disappeared more completely (Penn et al., Reference Penn, Deutsch, Payne and Sperling2018). In contrast to the geographic selectivity predicted for warming, periods of global cooling, such as during the Late Ordovician, are expected to generate extinctions focused on the low latitudes (Saupe et al., Reference Saupe, Qiao, Donnadieu, Farnsworth, Kennedy-Asser, Ladant, Lunt, Pohl, Valdes and Finnegan2020), consistent with the patterns observed for that mass extinction (Finnegan et al., Reference Finnegan, Heim, Peters and Fischer2012) and may also occur through aerobic habitat loss if accompanied by deoxygenation (Finnegan et al., Reference Finnegan, CMØ and DAT2016) or due to declining hypoxia tolerance in cold water in species with thermal optima (Boag et al., Reference Boag, Stockey, Elder, Hull and Sperling2018; Endress et al., Reference Endress, Boag, Burford, Penn, Sperling and Deutsch2022). Aerobic habitat loss is also predicted to select against large-bodied species, with a strong variability within size classes that depends on a species’ temperature sensitivity (Deutsch et al., Reference Deutsch, Penn, Verberk, Inomura, Endress and Payne2022). Extinctions driven by aerobic habitat loss may also explain the amplified background extinction rates observed for the early Phanerozoic, because of dramatically lower atmospheric O2 levels and thus species living closer to their ecophysiological limits (Stockey et al., Reference Stockey, Pohl, Ridgwell, Finnegan and Sperling2021). Trait adaption to different past climate states (Bennett et al., Reference Bennett, Sunday, Calosi, Villalobos, Martínez, Molina-Venegas, Araújo, Algar, Clusella-Trullas, Hawkins and Keith2021) has the potential to buffer or amplify predicted extinction risks. The role of differences in ecophysiological traits across taxonomic groups in explaining observed patterns of extinction selectivity across higher taxa (Knoll et al., Reference Knoll, Bambach, Canfield and Grotzinger1996, Reference Knoll, Bambach, Payne, Pruss and Fischer2007) remains an open area of research.

Figure 5. Geographic patterns of extinction and ocean changes in Earth system model simulations of the end-Permian climate transition (left column) and under anthropogenic greenhouse gas forcing (C) to 2300 C.E. (middle column). Line plot comparisons of end-Permian and potential future environmental changes versus latitude are shown in panels on the right (F, I, L, O). Model extinctions (A, B) are driven by ocean warming (D, E) and O2 loss (G, H), as quantified through the Metabolic Index, and in (A) reproduce the latitudinal pattern from the fossil record of the end-Permian (red points). These primary extinctions have the potential to be amplified by other environmental stressors like changes in net primary productivity (NPP) (J, K) or pH (M, N) or through secondary extinctions via the food web. Shaded region in (A) shows uncertainty in end-Permian extinction magnitudes across a range of potential extinction threshold parameters. Solid line in (B) shows future extinction risk averaged across Earth system models, using an extinction threshold calibrated from the end-Permian (same as the solid line in A) (see Penn et al., Reference Penn and Deutsch2022 for calibration details), while the shading in (B) shows the inter-model range. Future changes are projected under a high greenhouse gas emissions scenario, leading to a net radiative forcing of 8.5 W m−2 in 2100 C.E. (C) and are relative to the pre-industrial era (1850–1900). Model fields are averaged over the upper 500 m, and for the future projections, they are averaged across Earth system models (n = 5). Model details are provided in Penn et al. (Reference Penn, Deutsch, Payne and Sperling2018, Reference Penn and Deutsch2022). Panels A–C are modified from Penn et al. (Reference Penn, Deutsch, Payne and Sperling2018, Reference Penn and Deutsch2022), respectively.

Primary extinctions driven by the loss of aerobic habitat have the potential to be amplified by secondary extinctions arising from food web effects (Figure 4) or co-occurring environmental stressors that exacerbate direct aerobic habitat loss (Figure 5J–O). Aerobically tolerant species could still be lost if they are ecologically tied to vulnerable ones, for example, through the food web (Figure 4) or other critical interactions. Ocean acidification (Figure 5M–O) has the potential to further deplete aerobic habitat through direct CO2 effects on critical oxygen thresholds, but the magnitude and direction of this effect is uncertain and variable across limited available experimental studies (Figure 3E) (Rosa et al., Reference Rosa, Trübenbach, Repolho, Pimentel, Faleiro, Boavida-Portugal, Baptista, Lopes, Dionísio, Leal, Calado and Pörtner2013; Lefevre et al., Reference Lefevre, Watson, Munday and Nilsson2015). On its own, the magnitude of primary extinction from climate warming and associated physiological stresses depends on the amount of habitat loss beyond which a species can no longer sustain a viable population (i.e., the extinction threshold) (Urban, Reference Urban2015; Penn et al., Reference Penn, Deutsch, Payne and Sperling2018; Penn and Deutsch, Reference Penn and Deutsch2022), even if population decline takes a long time to occur. Extinction thresholds may vary across species, but the average value at the global ecosystem level has been estimated from comparison of end-Permian model simulations to the fossil record, and assuming a similar loss of habitat that drove extinctions in the past would apply in the modern ocean (Penn and Deutsch, Reference Penn and Deutsch2022). Calibration of this parameter from the fossil record has recently been used to project future extinction risk from climate changes resembling those of the end-Permian, which are arising today due to accelerating anthropogenic greenhouse gas emissions (Figure 5).

Example: Food webs

Terrestrial paleo-community dynamics are usually modeled according to trophic ecology and body size to investigate the role of food-web topology in the propagation of disruptions caused by environmental change. Models of extinction cascades suggest that responses can be complex, resulting from both bottom-up and top-down effects (Kaneryd et al., Reference Kaneryd, Borrvall, Berg, Curtsdotter, Eklöf, Hauzy, Jonsson, Münger, Setzer, Säterberg and Ebenman2012), with debate about whether simple or complex communities are more susceptible to such cascades and whether trophic versus other ecological interactions are most important (Eklöf and Ebenman, Reference Eklöf and Ebenman2006; Donohue et al., Reference Donohue, Petchey, Kéfi, Génin, Jackson, Yang and O’Connor2017). Explicit consideration of extinction cascades during mass extinctions has generally focused on the consequences of collapse in primary production (Tappan, Reference Tappan1968; Vermeij, Reference Vermeij2004). Bottom-up models predict extinction of smaller-bodied species in both the marine and terrestrial realms, due to the correlation of body size with trophic level, and exacerbated paleo-community instability post-extinction, which are consistent with investigations conducted on patterns of selectivity in relation to body size (Dunne et al., Reference Dunne, Williams and Martinez2002; Roopnarine, Reference Roopnarine2006; Roopnarine et al., Reference Roopnarine, Angielczyk, Wang and Hertog2007; Dunne and Williams, Reference Dunne and Williams2009; de Visser et al., Reference de Visser, Freymann and Olff2011; Lotze et al., Reference Lotze, Coll and Dunne2011). Interestingly, the end-Cretaceous mass extinction, for which we have the strongest evidence for collapse of primary production, is associated with preferential extinction of larger-bodied species in some clades (Friedman, Reference Friedman2009; Longrich et al., Reference Longrich, Bhullar and Gauthier2012) but not with the preferential extinction of smaller-bodied species, suggesting that physiology or other ecological factors (including top-down extinction cascades) were important in determining survivorship.

Two challenges remain in the modeling of extinction via networks of ecological interactions. First, evidence that “primary” extinctions may often occur via environmental change that exceeds the physiological tolerance limits of species at many positions in the food web creates a need for further investigation of how food webs respond to such losses. Are extinction cascades more, or less, extensive when driven by primary extinctions occurring simultaneously at multiple trophic levels? Second, there is the challenge of integrating physiological and ecological models such that the full response of the marine or terrestrial ecosystem could be predicted in an integrated manner from the modeling of climate change to the loss of species that cannot physiologically tolerate the modified world, to the loss of species that depended on ecological interactions with species lost via primary extinctions (Figure 4). Differences in timescale and level of biological organization at which physiological and ecological processes dominate add to this challenge.

Application to the sixth extinction

Mass extinction events provide our best source of information regarding the response of the biosphere to planetary-scale environmental disruption and the timescales and mechanisms of subsequent recovery. This information may be particularly important for the oceans, where observing biological response to environmental change is challenging and where the fossil record is particularly complete and diverse. Since the industrial revolution, the oceans have experienced substantial changes in ocean biogeochemistry, mainly because of rapid injection of CO2 into the atmosphere from anthropogenic sources. Under the accelerating future anthropogenic emissions scenario consistent with historical trends (Figure 5C), the oceans are expected to warm by 4–5°C and pH is expected to decrease, on average, by 0.44 pH units by the end of the 21st century, with changes increasing even further over the next few centuries (Figure 5E, N) (Kwiatkowski et al., Reference Kwiatkowski, Torres, Bopp, Aumont, Chamberlain, Christian, Dunne, Gehlen, Ilyina, John, Lenton, Li, Lovenduski, Orr, Palmieri, Santana-Falcón, Schwinger, Séférian, Stock, Tagliabue, Takano, Tjiputra, Toyama, Tsujino, Watanabe, Yamamoto, Yool and Ziehn2020). High temperatures are also expected to reduce the ocean’s oxygen content while also altering nutrient cycles (Sweetman et al., Reference Sweetman, Thurber, Smith, Levin, Mora, Wei, Gooday, Jones, Rex, Yasuhara, Ingels, Ruhl, Frieder, Danovaro, Würzberg, Baco, Grupe, Pasulka, Meyer, Dunlop, Henry and Roberts2017). Unabated anthropogenic emissions could drive the oceans toward widespread oxygen deficiency over the rest of the 21st century and beyond (Figure 5H) (Breitburg et al., Reference Breitburg, Levin, Oschlies, Grégoire, Chavez, Conley, Garçon, Gilbert, Gutiérrez, Isensee, Jacinto, Limburg, Montes, Naqvi, Pitcher, Rabalais, Roman, Rose, Seibel, Telszewski, Yasuhara and Zhang2018).

Such changes would have drastic consequences for marine ecosystems as evident from declining fish stocks, expansion of marine dead zones, and reduced primary productivity across different parts of the globe (Figure 5K) (Blanchard et al., Reference Blanchard, Jennings, Holmes, Harle, Merino, Allen, Holt, Dulvy and Barange2012). Efforts are already underway to project changes in species’ ranges and abundances in response to climate change on land and in the oceans (Thuiller, Reference Thuiller2004; Cheung et al., Reference Cheung, Lam, Sarmiento, Kearney, Watson and Pauly2009; Chen et al., Reference Chen, Hill, Ohlemüller, Roy and Thomas2011; Pinsky et al., Reference Pinsky, Selden and Kitchel2020). Extrapolating results from experiments and field observations over days or years to timescales of centuries, millennia, and beyond is challenging because different processes may dominate the biospheric response on different timescales, although there is emerging evidence that responses to some stresses are concordant across timescales (Reddin et al., Reference Reddin, Nätscher, Kocsis, Pörtner and Kiessling2020). Furthermore, the primary phase of extinction, dominated by physiology, may give way over time to a secondary phase of extinction, dominated by the effects of changing ecological interactions. Connecting the physiological and ecological processes driving extinction remains a research frontier.

Studies from the fossil record show that the ecophysiological constraints on marine taxa due to global warming and ocean deoxygenation will exert a key role in determining their risk to extinction under current and future emissions scenarios. The fossil record can even be used to calibrate the Earth system models used to predict future extinctions and changes in geographic range, just as paleoclimate records are used to calibrate models providing climate projections (Zhu et al., Reference Zhu, Otto-Bliesner, Brady, Gettelman, Bacmeister, Neale, Poulsen, Shaw, McGraw and Kay2022). Under a high emissions scenario (Figure 5C), the marine biological richness could be reduced to 65% of its current state due to global warming and oxygen loss from oceans by 2,300 (Penn and Deutsch, Reference Penn and Deutsch2022). The combined climate-ecophysiological models indicate that the local loss of species is expected to be the highest in tropical to temperate regions where taxa are expected to undergo a significant loss of aerobic habitat at their warm/low-O2 range boundaries. In contrast, in terms of global habitat loss and extinction risk, the equatorial taxa are expected to fare better overall in low oxygen and warmer oceans compared to polar species due to their higher tolerance limits to warm climates and opportunities to expand their available habitats as the poles become more like the present-day tropics. This scenario has precedent in the fossil record with the end-Permian mass extinction where a similar latitudinal extinction pattern unfolded (Figure 5A, B) (Penn et al., Reference Penn, Deutsch, Payne and Sperling2018; Reddin et al., Reference Reddin, Kocsis and Kiessling2019). Further work to integrate the effects of changes in pH, pCO2, salinity, and other key environmental variables into physiological performance models has the potential to make these models more general and accurate in reconstructing the causes of past extinction and predicting the consequences of future global change.

The ecological functions disrupted by global warming and marine defaunation are also bound to have cascading effects which could lead to extinction of vulnerable taxa. Modeling such effects is challenging due to the complexity of the interactions involved. The fossil record is our only source of data on the effects of major environmental disturbance at global scale. Fortunately, calibration of environmental change to physiologically expected extinction is becoming possible due to parallel advances in geochemistry, Earth system modeling, and physiological experimentation. The next decade will require integration of food webs and other types of ecosystem models to extract the full value of the lessons from Earth’s past in forecasting and guiding its future.

Open peer review

To view the open peer review materials for this article, please visit http://doi.org/10.1017/ext.2023.10.

Data availability statement

No data were collected or analyzed as part of this review paper.

Acknowledgments

The authors thank G. Wilson Mantilla, P. Calosi, C. Rasmussen, and an anonymous reviewer for their constructive feedback on the manuscript.

Author contribution

All authors contributed to the conceptualization, original draft preparation, review, and editing of the manuscript.

Financial support

This work was supported by the National Science Foundation (J.L.P., grant number EAR-2121392; C.D., grant number EAR-2121466).

Competing interest

The authors declare no competing interests exist.

References

Allen, BJ, Stubbs, TL, Benton, MJ and Puttick, MN (2019) Archosauromorph extinction selectivity during the Triassic–Jurassic mass extinction. Palaeontology 62(2), 211224.CrossRefGoogle Scholar
Alroy, J (1999) Putting North America’s end-Pleistocene megafaunal extinction in context. In MacPhee, RDE (ed.), Extinctions in near Time: Causes, Contexts, and Consequences. Boston: Springer, pp. 105143.CrossRefGoogle Scholar
Alvarez, LW, Alvarez, W, Asaro, F and Michel, HV (1980) Extraterrestrial cause for the Cretaceous-tertiary extinction. Science 208, 10951108.CrossRefGoogle ScholarPubMed
Bambach, RK, Knoll, AH and Sepkoski, JJ (2002) Anatomical and ecological constraints on Phanerozoic animal diversity in the marine realm. Proceedings of the National Academy of Sciences 99(10), 68546859.CrossRefGoogle ScholarPubMed
Barnosky, AD, Matzke, N, Tomiya, S, Wogan, GOU, Swartz, B, Quental, TB, Marshall, C, McGuire, JL, Lindsey, EL, Maguire, KC, Mersey, B, Ferrer, EA (2011) Has the Earth’s sixth mass extinction already arrived? Nature 471, 5157.CrossRefGoogle ScholarPubMed
Bennett, JM, Sunday, J, Calosi, P, Villalobos, F, Martínez, B, Molina-Venegas, R, Araújo, MB, Algar, AC, Clusella-Trullas, S, Hawkins, BA and Keith, SA (2021) The evolution of critical thermal limits of life on Earth. Nature Communications, 12(1), 1198.CrossRefGoogle ScholarPubMed
Blackburn, TJ, Olsen, PE, Bowring, SA, McLean, NM, Kent, DV, Puffer, J, McHone, G, Rasbury, ET, 'Et-Touhami, M (2013) Zircon U-Pb geochronology links the end-Triassic extinction with the Central Atlantic Magmatic Province. Science 340(6135), 941945.CrossRefGoogle ScholarPubMed
Blanchard, JL, Jennings, S, Holmes, R, Harle, J, Merino, G, Allen, JI, Holt, J, Dulvy, NK, Barange, M (2012) Potential consequences of climate change for primary production and fish production in large marine ecosystems. Philosophical Transactions of the Royal Society B: Biological Sciences 367(1605), 29792989.CrossRefGoogle ScholarPubMed
Boag, TH, Stockey, RG, Elder, LE, Hull, PM and Sperling, EA (2018) Oxygen, temperature and the deep-marine stenothermal cradle of Ediacaran evolution. Proceedings of the Royal Society B: Biological Sciences 285, 20181724.CrossRefGoogle ScholarPubMed
Bond, DPG and Sun, Y (2021) Global warming and mass extinctions associated with large Igneous Province volcanism. In Ernst, RE, Dickson, AJ and Bekker, A (eds.), Large Igneous Provinces. Hoboken: John Wiley and Sons, Inc., pp. 83102.CrossRefGoogle Scholar
Breitburg, D, Levin, LA, Oschlies, A, Grégoire, M, Chavez, FP, Conley, DJ, Garçon, V, Gilbert, D, Gutiérrez, D, Isensee, K, Jacinto, GS, Limburg, KE, Montes, I, Naqvi, SWA, Pitcher, GC, Rabalais, NN, Roman, MR, Rose, KA, Seibel, BA, Telszewski, M, Yasuhara, M, Zhang, J (2018) Declining oxygen in the global ocean and coastal waters. Science 359(6371), eaam7240.CrossRefGoogle ScholarPubMed
Brown, JH (1995) Macroecology. Chicago: University of Chicago Press.Google Scholar
Burgess, SD, Bowring, S and Shen, S (2014) High-precision timeline for Earth’s most severe extinction. Proceedings of the National Academy of Sciences 111(9), 33163321.CrossRefGoogle ScholarPubMed
Calosi, P, Melatunan, S, Turner, LM, Artioli, Y, Davidson, RL, Byrne, JJ, Viant, MR, Widdicombe, S, Rundle, SD (2017) Regional adaptation defines sensitivity to future ocean acidification. Nature Communications 8(1), 13994.CrossRefGoogle ScholarPubMed
Chaudhary, C, Richardson, AJ, Schoeman, DS and Costello, MJ (2021) Global warming is causing a more pronounced dip in marine species richness around the equator. Proceedings of the National Academy of Sciences 118(15), e2015094118.CrossRefGoogle Scholar
Chen, I-C, Hill, JK, Ohlemüller, R, Roy, DB and Thomas, CD (2011) Rapid range shifts of species associated with high levels of climate warming. Science 333(6045), 10241026.CrossRefGoogle ScholarPubMed
Cheung, WWL, Lam, VWY, Sarmiento, JL, Kearney, K, Watson, R and Pauly, D (2009) Projecting global marine biodiversity impacts under climate change scenarios. Fish and Fisheries 10(3), 235251.CrossRefGoogle Scholar
JWF, Chu and KSP, Gale (2017) Ecophysiological limits to aerobic metabolism in hypoxia determine epibenthic distributions and energy sequestration in the Northeast Pacific Ocean. Limnology and Oceanography 62(1), 5974.Google Scholar
Clapham, ME (2017) Organism activity levels predict marine invertebrate survival during ancient global change extinctions. Global Change Biology 23(4), 14771485.CrossRefGoogle ScholarPubMed
Clapham, ME and Payne, JL (2011) Acidification, anoxia, and extinction: A multiple logistic regression analysis of extinction selectivity during the Middle and Late Permian. Geology 39(11), 10591062.CrossRefGoogle Scholar
de Visser, SN, Freymann, BP and Olff, H (2011) The Serengeti food web: Empirical quantification and analysis of topological changes under increasing human impact. Journal of Animal Ecology 80(2), 484494.CrossRefGoogle ScholarPubMed
Deutsch, C, Ferrel, A, Seibel, B, Pörtner, HO and Huey, RB (2015) Climate change tightens a metabolic constraint on marine habitats. Science 348, 11321135.CrossRefGoogle ScholarPubMed
Deutsch, C, Penn, JL and Seibel, B (2020) Metabolic trait diversity shapes marine biogeography. Nature 585(7826), 557562.CrossRefGoogle ScholarPubMed
Deutsch, C, Penn, JL, Verberk, WCEP, Inomura, K, Endress, M-G and Payne, JL (2022) Impact of warming on aquatic body sizes explained by metabolic scaling from microbes to macrofauna. Proceedings of the National Academy of Sciences 119(28), e2201345119.CrossRefGoogle ScholarPubMed
Dirzo, R, Young, HS, Galetti, M, Ceballos, G, Isaac, NJB and Collen, B (2014) Defaunation in the Anthropocene. Science 345, 401406.CrossRefGoogle ScholarPubMed
Donohue, I, Petchey, OL, Kéfi, S, Génin, A, Jackson, AL, Yang, Q, O’Connor, NE (2017) Loss of predator species, not intermediate consumers, triggers rapid and dramatic extinction cascades. Global Change Biology 23(8), 29622972.CrossRefGoogle Scholar
Dunhill, AM and Wills, MA (2015) Geographic range did not confer resilience to extinction in terrestrial vertebrates at the end-Triassic crisis. Nature Communications 6(May), 18.CrossRefGoogle Scholar
Dunne, JA and Williams, RJ (2009) Cascading extinctions and community collapse in model food webs. Philosophical Transactions of the Royal Society B: Biological Sciences 364(1524), 17111723.CrossRefGoogle ScholarPubMed
Dunne, JA, Williams, RJ and Martinez, ND (2002) Food-web structure and network theory: The role of connectance and size. Proceedings of the National Academy of Sciences 99(20), 1291712922.CrossRefGoogle ScholarPubMed
Eklöf, A and Ebenman, BO (2006) Species loss and secondary extinctions in simple and complex model communities. Journal of Animal Ecology 75, 239246.CrossRefGoogle ScholarPubMed
Endress, M-GA, Boag, TH, Burford, BP, Penn, JL, Sperling, EA and Deutsch, CA (2022) Physiological causes and biogeographic consequences of thermal optima in the hypoxia tolerance of marine ectotherms. BioRxiv, 2022.02.03.478967.CrossRefGoogle Scholar
Finnegan, S, Bergmann, K, Eiler, JM, Jones, DS, Fike, DA, Eisenman, I, Hughes, NC, Tripati, AK, Fischer, WW (2011) The magnitude and duration of Late Ordovician–Early Silurian glaciation. Science 331(6019), 903906.CrossRefGoogle ScholarPubMed
Finnegan, S, Heim, NA, Peters, SE and Fischer, WW (2012) Climate change and the selective signature of the Late Ordovician mass extinction. Proceedings of the National Academy of Sciences of the United States of America 109(18), 68296834.CrossRefGoogle ScholarPubMed
Finnegan, S, CMØ, Rasmussen and DAT, Harper (2016) Biogeographic and bathymetric determinants of brachiopod extinction and survival during the Late Ordovician mass extinction. Proceedings of the Royal Society B: Biological Sciences 283(1829), 20160007.CrossRefGoogle ScholarPubMed
Friedman, M (2009) Ecomorphological selectivity among marine teleost fishes during the end-Cretaceous extinction. Proceedings of the National Academy of Sciences of the United States of America 106(13), 52185223.CrossRefGoogle ScholarPubMed
Gearty, W, McClain, CR and Payne, JL (2018) Energetic tradeoffs control the size distribution of aquatic mammals. Proceedings of the National Academy of Sciences of the United States of America 115(16), 41944199.CrossRefGoogle ScholarPubMed
Gillooly, JF, Brown, JH, West, GB, Savage, VM and Charnov, EL (2001) Effects of size and temperature on metabolic rate. Science 293(5538), 22482251.CrossRefGoogle ScholarPubMed
Hallam, A and Wignall, PB (1997) Mass Extinctions and Their Aftermath. Oxford: Oxford University Press.Google Scholar
Howard, EM, Penn, JL, Frenzel, H, Seibel, BA, Bianchi, D, Renault, L, Kessouri, F, Sutula, MA, McWilliams, JC, Deutsch, C (2020) Climate-driven aerobic habitat loss in the California current system. Science Advances 6(20), eaay3188.CrossRefGoogle ScholarPubMed
Hughes, JJ, Berv, JS, Chester, SGB, Sargis, EJ and Field, DJ (2021) Ecological selectivity and the evolution of mammalian substrate preference across the K–Pg boundary. Ecology and Evolution 11(21), 1454014554.CrossRefGoogle ScholarPubMed
Jablonski, D (1986) Background and mass extinctions: The alternation of macroevolutionary regimes. Science 231(4734), 129133.CrossRefGoogle ScholarPubMed
Jablonski, D (2005) Mass extinctions and macroevolution. Paleobiology 31(sp5), 192210.CrossRefGoogle Scholar
Jablonski, D and Raup, DM (1995) Selectivity of end-Cretaceous marine bivalve extinctions. Science 268(5209), 389391.CrossRefGoogle ScholarPubMed
Joachimski, MM and Buggisch, W (2002) Conodont apatite δ18O signatures indicate climatic cooling as a trigger of the Late Devonian mass extinction. Geology 30(8), 711714.2.0.CO;2>CrossRefGoogle Scholar
Kaneryd, L, Borrvall, C, Berg, S, Curtsdotter, A, Eklöf, A, Hauzy, C, Jonsson, T, Münger, P, Setzer, M, Säterberg, T, Ebenman, B (2012) Species‐rich ecosystems are vulnerable to cascading extinctions in an increasingly variable world. Ecology and Evolution 2(4), 858874.CrossRefGoogle Scholar
Kiessling, W and Aberhan, M (2007) Geographical distribution and extinction risk: Lessons from Triassic–Jurassic marine benthic organisms. Journal of Biogeography 34(9), 14731489.CrossRefGoogle Scholar
Kiessling, W and Simpson, C (2011) On the potential for ocean acidification to be a general cause of ancient reef crises. Global Change Biology 17(1), 5667.CrossRefGoogle Scholar
Knoll, AH, Bambach, RK, Canfield, DE and Grotzinger, JP (1996) Comparative earth history and Late Permian mass extinction. Science 273(5274), 452457.CrossRefGoogle ScholarPubMed
Knoll, AH, Bambach, RK, Payne, JL, Pruss, S and Fischer, WW (2007) Paleophysiology and end-Permian mass extinction. Earth and Planetary Science Letters 256(3–4), 295313.CrossRefGoogle Scholar
Knoll, AH and Fischer, WW (2011) Skeletons and ocean chemistry: the long view. In Ocean Acidification, ed. Gattuso, J-P and Hansson, L, 6782. Oxford: Oxford University Press.Google Scholar
Kwiatkowski, L, Torres, O, Bopp, L, Aumont, O, Chamberlain, M, Christian, JR, Dunne, JP, Gehlen, M, Ilyina, T, John, JG, Lenton, A, Li, H, Lovenduski, NS, Orr, JC, Palmieri, J, Santana-Falcón, Y, Schwinger, J, Séférian, R, Stock, CA, Tagliabue, A, Takano, Y, Tjiputra, J, Toyama, K, Tsujino, H, Watanabe, M, Yamamoto, A, Yool, A, Ziehn, T (2020) Twenty-first century ocean warming, acidification, deoxygenation, and upper-ocean nutrient and primary production decline from CMIP6 model projections. Biogeosciences 17(13), 34393470.CrossRefGoogle Scholar
Lefevre, S, Watson, S-A, Munday, PL and Nilsson, GE (2015) Will jumping snails prevail? Influence of near-future CO2, temperature and hypoxia on respiratory performance in the tropical conch Gibberulus gibberulus gibbosus. Journal of Experimental Biology 218(19), 29913001.CrossRefGoogle ScholarPubMed
Lockwood, R (2003) Abundance not linked to survival across the end-Cretaceous mass extinction: Patterns in north American bivalves. Proceedings of the National Academy of Sciences of the United States of America 100(5), 24782482.CrossRefGoogle Scholar
Longrich, NR, Bhullar, B-AS and Gauthier, JA (2012) Mass extinction of lizards and snakes at the Cretaceous-Paleogene boundary. Proceedings of the National Academy of Sciences of the United States of America 109(52), 2139621401.CrossRefGoogle ScholarPubMed
Lotze, HK, Coll, M and Dunne, JA (2011) Historical changes in marine resources, food-web structure and ecosystem functioning in the Adriatic Sea, Mediterranean. Ecosystems 14(2), 198222.CrossRefGoogle Scholar
McCauley, DJ, Pinsky, ML, Palumbi, SR, Estes, JA, Joyce, FH and Warner, RR (2015) Marine defaunation: Animal loss in the global ocean. Science 347(6219), 247.CrossRefGoogle ScholarPubMed
Monarrez, PM, Heim, NA and Payne, JL (2021) Mass extinctions alter extinction and origination dynamics with respect to body size. Proceedings of the Royal Society B: Biological Sciences 288(1960), 1681.Google ScholarPubMed
Newell, ND (1963) Crises in the history of life. Scientific American 208(2), 7692.CrossRefGoogle Scholar
Newell, ND (1967) Revolutions in the history of life. Geological Society of America Special Papers 89, 6392.CrossRefGoogle Scholar
Payne, JL, Bush, AM, Chang, ET, Heim, NA, Knope, ML and Pruss, SB (2016a) Extinction intensity, selectivity and their combined macroevolutionary influence in the fossil record. Biology Letters 12(10), 20160202.CrossRefGoogle ScholarPubMed
Payne, JL, Bush, AM, Heim, NA, Knope, ML and McCauley, DJ (2016b) Ecological selectivity of the emerging mass extinction in the oceans. Science 353, 12841286.CrossRefGoogle ScholarPubMed
Payne, JL and Clapham, ME (2012) End-Permian mass extinction in the oceans: An ancient analog for the twenty-first century? Annual Review of Earth and Planetary Sciences 40(1), 89111.CrossRefGoogle Scholar
Payne, JL and Finnegan, S (2007) The effect of geographic range on extinction risk during background and mass extinction. Proceedings of the National Academy of Sciences of the United States of America 104(25), 1050610511.CrossRefGoogle ScholarPubMed
Payne, JL and Heim, NA (2020) Body size, sampling completeness, and extinction risk in the marine fossil record. Paleobiology 46(1), 2340.CrossRefGoogle Scholar
Payne, JL, Truebe, S, Nützel, A, Chang, ET and Nu, A (2011) Local and global abundance associated with extinction risk in Late Paleozoic and early Mesozoic gastropods. Society 37(4), 616632.Google Scholar
Penn, JL and Deutsch, C (2022) Avoiding ocean mass extinction from climate warming. Science 376(6592), 524526.CrossRefGoogle ScholarPubMed
Penn, JL, Deutsch, C, Payne, JL and Sperling, EA (2018) Temperature-dependent hypoxia explains biogeography and severity of end-Permian marine mass extinction. Science 362, eaat1327.CrossRefGoogle ScholarPubMed
Peters, RH (1983) The Ecological Implications of Body Size. Cambridge: Cambridge University Press.CrossRefGoogle Scholar
Peters, SE (2008) Environmental determinants of extinction selectivity in the fossil record. Nature 454(7204), 626629.CrossRefGoogle ScholarPubMed
Peterson, CC, Nagy, KA and Diamond, J (1990) Sustained metabolic scope. Proceedings of the National Academy of Sciences of the United States of America 87, 23242328.CrossRefGoogle ScholarPubMed
Pinsky, ML, Selden, RL and Kitchel, ZJ (2020) Climate-driven shifts in marine species ranges: Scaling from organisms to communities. Annual Review of Marine Science 12, 153179.CrossRefGoogle ScholarPubMed
Pollack, JB, Toon, OB, Ackerman, TP, McKay, CP and Turco, RP (1983) Environmental effects of an impact-generated dust cloud: Implications for the Cretaceous-tertiary extinctions. Science 219(4582), 287289.CrossRefGoogle ScholarPubMed
Raup, DM (1986) Biological extinction in earth history. Science 231(4745), 15281533.CrossRefGoogle ScholarPubMed
Raup, DM and Sepkoski, JJ (1982) Mass extinctions in the marine fossil record. Science 215, 15011503.CrossRefGoogle ScholarPubMed
Reddin, CJ, Kocsis, ÁT, Aberhan, M and Kiessling, W (2021) Victims of ancient hyperthermal events herald the fates of marine clades and traits under global warming. Global Change Biology 27(4), 868878.CrossRefGoogle ScholarPubMed
Reddin, CJ, Kocsis, ÁT and Kiessling, W (2019) Climate change and the latitudinal selectivity of ancient marine extinctions. Paleobiology 45(1), 7084.CrossRefGoogle Scholar
Reddin, CJ, Nätscher, PS, Kocsis, ÁT, Pörtner, H-O and Kiessling, W (2020) Marine clade sensitivities to climate change conform across timescales. Nature Climate Change 10(3), 249253.CrossRefGoogle Scholar
Rogers, NJ, Urbina, MA, Reardon, EE, McKenzie, DJ and Wilson, RW (2016) A new analysis of hypoxia tolerance in fishes using a database of critical oxygen level (Pcrit). Conservation Physiology 27(4), cow012.CrossRefGoogle Scholar
Romanuk, TN, Hayward, A and Hutchings, JA (2011) Trophic level scales positively with body size in fishes. Global Ecology and Biogeography 20(2), 231240.CrossRefGoogle Scholar
Roopnarine, PD (2006) Extinction cascades and catastrophe in ancient food webs. Paleobiology 32(1), 119.CrossRefGoogle Scholar
Roopnarine, PD and Angielczyk, KD (2015) Community stability and selective extinction during the Permian-Triassic mass extinction. Science 350(6256), 9093.CrossRefGoogle ScholarPubMed
Roopnarine, PD, Angielczyk, KD, Wang, SC and Hertog, R (2007) Trophic network models explain instability of early Triassic terrestrial communities. Proceedings of the Royal Society B: Biological Sciences 274(1622), 20772086.CrossRefGoogle ScholarPubMed
Rosa, R, Trübenbach, K, Repolho, T, Pimentel, M, Faleiro, F, Boavida-Portugal, J, Baptista, M, Lopes, VM, Dionísio, G, Leal, MC, Calado, R, Pörtner, HO (2013) Lower hypoxia thresholds of cuttlefish early life stages living in a warm acidified ocean. Proceedings of the Royal Society B: Biological Sciences 280(1768), 20131695.CrossRefGoogle Scholar
Saupe, EE, Qiao, H, Donnadieu, Y, Farnsworth, A, Kennedy-Asser, AT, Ladant, JB, Lunt, DJ, Pohl, A, Valdes, P, Finnegan, S (2020) Extinction intensity during Ordovician and Cenozoic glaciations explained by cooling and palaeogeography. Nature Geoscience 13(1), 6570.CrossRefGoogle Scholar
Schaal, EK, Clapham, ME, Rego, BL, Wang, SC and Payne, JL (2016) Comparative size evolution of marine clades from the Late Permian through middle Triassic. Paleobiology 42(1), 127142.CrossRefGoogle Scholar
Sheehan, PM and Hansen, TA (1986) Detritus feeding as a buffer to extinction at the end of the Cretaceous. Geology 14(10), 868870.2.0.CO;2>CrossRefGoogle Scholar
Smith, AB and Jeffery, CH (1998) Selectivity of extinction among sea urchins at the end of the Cretaceous period. Nature 392(6671), 6971.CrossRefGoogle Scholar
Smith, FA, Elliott Smith, RE, Lyons, SK and Payne, JL (2018) Body size downgrading of mammals over the late quaternary. Science 360(6386), 310313.CrossRefGoogle ScholarPubMed
Stockey, RG, Pohl, A, Ridgwell, A, Finnegan, S and Sperling, EA (2021) Decreasing Phanerozoic extinction intensity as a consequence of earth surface oxygenation and metazoan ecophysiology. Proceedings of the National Academy of Sciences 118(41), e2101900118.CrossRefGoogle ScholarPubMed
Svensen, H, Planke, S, Polozov, AG, Schmidbauer, N, Corfu, F, Podladchikov, YY, Jamtveit, B (2009) Siberian gas venting and the end-Permian environmental crisis. Earth and Planetary Science Letters 277(3–4), 490500.CrossRefGoogle Scholar
Sweetman, AK, Thurber, AR, Smith, CR, Levin, LA, Mora, C, Wei, CL, Gooday, AJ, Jones, DOB, Rex, M, Yasuhara, M, Ingels, J, Ruhl, HA, Frieder, CA, Danovaro, R, Würzberg, L, Baco, A, Grupe, BM, Pasulka, A, Meyer, KS, Dunlop, KM, Henry, LA, Roberts, JM (2017) Major impacts of climate change on deep-sea benthic ecosystems. Elementa: Science of the Anthropocene 5, 4.Google Scholar
Tappan, H (1968) Primary production, isotopes, extinctions and the atmosphere. Palaeogeography, Palaeoclimatology, Palaeoecology 4(3), 187210.CrossRefGoogle Scholar
Thuiller, W (2004) Patterns and uncertainties of species’ range shifts under climate change. Global Change Biology 10(12), 20202027.CrossRefGoogle Scholar
Turco, RP, Toon, OB, Ackerman, TP, Pollack, JB and Sagan, C (1983) Nuclear winter: Global consequences of multiple nuclear explosions. Science 222(4630), 12831292.CrossRefGoogle Scholar
Urban, MC (2015) Accelerating extinction risk from climate change. Science 348(6234), 571573.CrossRefGoogle ScholarPubMed
Valentine, JW and Jablonski, D (1986) Mass extinctions: Sensitivity of marine larval types. Proceedings of the National Academy of Sciences of the United States of America 83(September), 69126914.CrossRefGoogle ScholarPubMed
Vermeij, GJ (2004) Ecological avalanches and the two kinds of extinction. Evolutionary Ecology Research 6(3), 315337.Google Scholar
Wallace, AR (1889) Darwinism: An Exposition of the Theory of Natural Selection, with Some of its Applications. London: Macmillan and Co.CrossRefGoogle Scholar
Wilson, GP (2013) Mammals across the K/Pg boundary in northeastern Montana, USA: Dental morphology and body-size patterns reveal extinction selectivity and immigrant-fueled ecospace filling. Paleobiology 39(3), 429469.CrossRefGoogle Scholar
Zhu, J, Otto-Bliesner, BL, Brady, EC, Gettelman, A, Bacmeister, JT, Neale, RB, Poulsen, CJ, Shaw, JK, McGraw, ZS, Kay, JE (2022) LGM paleoclimate constraints inform cloud parameterizations and equilibrium climate sensitivity in CESM2. Journal of Advances in Modeling Earth Systems 14(4), e2021MS002776.CrossRefGoogle Scholar
Figure 0

Figure 1. Extinction patterns in the fossil record. (A) Graph of marine animal diversity across the past 600 million years, illustrating the diversity declines associated with the five major mass extinction events (modified from Raup and Sepkoski, 1982). (B) Extinction selectivity with respect to geographic range, illustrating the preferential survival of broadly distributed genera during background intervals and the greatly reduced selectivity during mass extinction events (modified from Payne and Finnegan, 2007). (C) Principal components analysis of logistic regression coefficients of ecological traits and body size selectivity of the Big Five mass extinction events and the modern oceans, demonstrating the unique selectivity of the modern extinction threat (modified from Payne et al., 2016b). (D) Extinction selectivity during the end-Permian mass extinction, illustrating the preferential extinction of heavily calcified marine animal classes with less complex respiratory and circulatory systems (modified from Knoll et al., 2007; Knoll and Fischer, 2011). (E) Extinction selectivity with respect to body size for major classes of marine animals, illustrating the general bias of background extinction against smaller-bodied genera versus the variable direction of selectivity for classes that exhibit distinct patterns during mass extinction (modified from Monarrez et al., 2021).

Figure 1

Figure 2. Workflow illustrating the use of geological and geochemical data to constrain Earth system models (ESMs), physiological experiments to constrain parameters used to populate models with species of different ecophysiotypes, and fossil occurrence data to conduct model-data comparison. Ecosystem structure remains to be incorporated into such models and can be used to predict extinction cascades. Calibration of models against selectivity patterns in ancient extinction events will improve their use in forecasting biotic response to current and future environmental change. Panels on right showing CO2 emissions curves and future biodiversity projections are from Penn and Deutsch (2022).

Figure 2

Figure 3. Graphs illustrating the key species traits of the Metabolic Index (ɸ) along with how ɸ relates to temperature and oxygen partial pressure. (A–C) Frequency distributions of the Metabolic Index parameters for marine animals. (D, E) Graphs of variation in ɸ as a function of temperature and oxygen for species with negative (D) and positive (E) temperature sensitivities (Eo) of hypoxia tolerance (Ao), which is the inverse of the critical oxygen threshold (red circle) at a reference temperature (Tref), as derived from respirometry experiments. For species in a resting state, the aerobic habitat limit occurs when ɸ = 1, but in the environment a species’ activity level or sustained metabolic scope (SMS) elevates the habitat limit to ɸcrit. For species with negative Eo, aerobic habitat availability increases with temperature, whereas for those with positive Eo (i.e., most species; panel B), aerobic habitat declines with warming. Changes in PO2 has the potential to lower aerobic habitat availability, and thus the amount of warming a species can withstand, as exemplified for two scenarios of with different fractions of present atmospheric levels of O2 (PAL; yellow dots and arrows). A change in CO2 also has the potential to alter hypoxia tolerance, but the magnitude and direction of this effect is unknown across marine biota and is illustrated here from experimental data for a single species under ∆ pH = +0.5 (Rosa et al., 2013). Arrows in A–C denote species traits in D and E.

Figure 3

Figure 4. Hypothetical progression of a mass extinction highlighting sources of trait-based and geographic selectivity and potential ecological amplification. (A) An initial distribution of species (or “ecophysiotypes”) defined by traits under selection by large-scale environmental conditions will likely result in systematic correlations between traits and geographic range. The range metric here can be considered overall range size (area and volume), or centroid (e.g., low-latitude versus high-latitude, shallow versus deep). (B) The initial biota are subjected to climate perturbation that poses a direct stress through a reduction in fitness whose magnitude depends on species traits and on local climate trends. The resulting change in available habitat (ΔH; contours) presents an ecophysiological extinction risk that is geographically selective because it is trait selective (but may also be caused by climate patterns themselves). In this hypothetical case, habitat loss (ΔH < 0) selects against species with high values of two traits (habitat “Losers”) and may even benefit species with low values of those traits (habitat “Gainers”; ΔH > 0). (C) Physiological extinction poses further ecological risks (or advantages) depending on the mutualistic or adversarial interactions with ecophysiotypes (nodes in graph) that are under trait-selective risk. Ecological risk is complex and for any particular species will depend on the physiological risk faced by the other species with which it interacts, which may be positive (green lines) or negative (brown lines), and strong (thick lines) or weak (thin lines). The results of these associations, which may be multiple and indirect, could alter extinction risk by either preserving ecological fitness (“+” symbol) or reducing it (“−” symbol). Changes in extinction risk are likely to be most pronounced for those in the neutral zone whose antagonists go extinct or who are buoyed by prey/mutualists that are under positive selection. (D) Post-extinction ecosystem, equal to the initial one (A) minus the ecotypes that have gone extinct from either primary (B) or secondary (C) effects.

Figure 4

Figure 5. Geographic patterns of extinction and ocean changes in Earth system model simulations of the end-Permian climate transition (left column) and under anthropogenic greenhouse gas forcing (C) to 2300 C.E. (middle column). Line plot comparisons of end-Permian and potential future environmental changes versus latitude are shown in panels on the right (F, I, L, O). Model extinctions (A, B) are driven by ocean warming (D, E) and O2 loss (G, H), as quantified through the Metabolic Index, and in (A) reproduce the latitudinal pattern from the fossil record of the end-Permian (red points). These primary extinctions have the potential to be amplified by other environmental stressors like changes in net primary productivity (NPP) (J, K) or pH (M, N) or through secondary extinctions via the food web. Shaded region in (A) shows uncertainty in end-Permian extinction magnitudes across a range of potential extinction threshold parameters. Solid line in (B) shows future extinction risk averaged across Earth system models, using an extinction threshold calibrated from the end-Permian (same as the solid line in A) (see Penn et al., 2022 for calibration details), while the shading in (B) shows the inter-model range. Future changes are projected under a high greenhouse gas emissions scenario, leading to a net radiative forcing of 8.5 W m−2 in 2100 C.E. (C) and are relative to the pre-industrial era (1850–1900). Model fields are averaged over the upper 500 m, and for the future projections, they are averaged across Earth system models (n = 5). Model details are provided in Penn et al. (2018, 2022). Panels A–C are modified from Penn et al. (2018, 2022), respectively.

Author comment: Selectivity of mass extinctions: Patterns, processes, and future directions — R0/PR1

Comments

September 15, 2022

Dear Profs. Alroy and Brook,

Please find attached a manuscript entitled Selectivity of mass extinction: patterns, processes, and future directions, submitted for publication in Cambridge Prisms: Extinction. This manuscript is submitted in response to your invitation and is intended as a review of the topic with a focus on progress over the past two decades. To produce the review, I have collaborated with several advisees in my research group as well as two scientists at the forefront of applying Earth system models to the study of biological response to climate change in the modern and ancient worlds. We have addressed, in particular, the use of Earth system models and data from physiological experiments as quantitative bridges between geochemical constraints on environmental change and paleontological data on extinction selectivity during mass extinctions. We have briefly addressed the use of the fossil record for calibrating model forecasts of future extinction in response to climate change. We have attempted to write the manuscript in a way that will be accessible to the broad readership of this new journal. We hope that you will find it acceptable for publication and appreciate the opportunity to contribute.

Best,

Jonathan Payne and co-authors

Review: Selectivity of mass extinctions: Patterns, processes, and future directions — R0/PR2

Conflict of interest statement

not applicable

Comments

Comments to Author: Review of the article by Payne et al. entitled “Selectivity of mass extinctions: patterns, processes, and

future directions”

The work by Jonathan L. Payne, Jood A. Al Aswad, Curtis Deutsch, Pedro M. Monarrez, Justin L. Penn and Pulkit Singh is really interesting y, and will make a significant contribution to the research field of extinctions, but also to the fields of global change ecology and biology. However, the article requires some work before to be ready for publication. Particularly, provide a more balanced discussion of the existing evidence and limitations of the approach proposed, verify semantics in few cases, and consider to integrate existing references and concept I highlight below. I believe you can relatively easy deal with the points I raise, and whilst the comments may lead to minor or major revision. I selected major only to enable you have more time if you require so.

Major Comments

Line 149-151. In my opinion, here, you are only skimming the surface. You correctly mention the simple Bauplans of heavy calcified marine invertebrates can in part explain the extinction patterns described in the fossil record for the end-Permian. The point is valid and interesting. I could argue, however, that Bauplan “simplicity” can be an advantage, enabling an organism to be less impacted directly by (at least some) environmental drivers, and indirectly possess the advantage to have lower metabolic-energetic requirements. In this sense, organisms with more complex, high-performance cardio-circulatory and respiration, systems may possess greater homeostatic abilities (see for example Melzner et al. 2009 for ocean acidification), but they will not be able to sustain high homeostatic activities for extended periods of time. Complexity requires more energy for maintenance and repair. Modern ecosystem collapse, such as in the North Sea following a switch of a keystone copepod species (from more to less nutritional), the loss in high level predators was documented. In addition, you seem to (I say this respectfully and with the best of intention to help you further improving your excellent work) miss the importance of the effects of other drivers, which can act along different pathways. For example, here specifically, you are not discussing of the direct (i.e. corrosion of carbonated structure) and indirect (e.g. increase in energetic cost for compensatory calcification to maintain a positive net mineralisation in light of passive dissolution) effects of ocean acidification conditions on marine heavy calcifiers. Please ref. to the literature on this topic. It is vast, but it needs to (even if synthetically) be acknowledged, as it is super relevant. I can refer to my own work to help, but the choice of ref. to use sits entirely with you: please see Calosi et al. 2019 Ann. Rev. Mar. Sc. for the discussion of putative mechanisms-pathways of action (see summary Figure 4), and to Calosi et al. 2017 Nat. Comms. and Findlay et al. 2011 Mar. Biol. Res. where differences in passive corrosion-net mineralisation in different population of the same species and different species are reported. but as I suggest, the literature on this topic is extensive.

Line 193. I feel that in the paper, “ecological-ecosystemic” aspects of mass extinction should be discuss even more thoroughly. The authors do so, but more thinking and work should go into this to push the work and the MS to where it can, and deserved to be. Ensure to bring this as forward as possible in the MS. The method you propose is great to paint a picture on the direct (physiological) impacts of temperature and oxygen (not all drivers characterising current global change for example) but cannot be used to depict the indirect (ecological) impacts. You recognise the issue, but this can be stated even more clearly. You need to discuss in greater depth how best the tool you propose can be used, to what advantage, and where it cannot be used or used recognising the presence of limitations. In line 280 and following you could integrate the work by Reddin et al. 2020 NCC (which you cite) to the discussion, related to early extinctions being caused primarily by direct physiological effects, and later extinctions being caused by indirect ecological effects. All considered, I suggest to enhanced the discussion, and I fully recognise the authors’ effort in the intellectual integration of “physiological” and “ecological” extinctions …. and figure 4 is excellent by the way!!

Line 110, 203-204, 280 and elsewhere. Please the first time you mention “tolerance” and “resilience” ensure to provide a definition in parenthesis: 110 and 204, respectively. I have the impression that in most cases you use it “correctly” but in other not (ex. line 282. Tolerance is the ability of a biological system to resist change following or under a perturbation, and resilience is the ability of a biological system to come back to the original stable state it was in after a perturbation (i.e. resilience is the inverse of the time needed to get back to the starting status).

Line 266. Pre-adaptation does not exist. It is an erroneous concept. An organism cannot be adapted to something it has not experienced before. And if it is adapted to be able to face a certain challenge, it is because it has already experienced it in his phylogenetic history, or it has been selected to be tolerant to another stressor which grants it “protection” to the stressor we are investigating. In either of these cases, we talk of “exaptation” (i.e. existing adaptation), please see Gould’s seminal work on this topic.

For your interest, we have recently published a work on the diversity and evolution of thermal limits, testing for the presence of paleo signals on these traits, and we found one. It was not the strongest of signals of the evolutionary drivers we investigated, but it was non the less significant. In short, the era of emergence of a taxonomic group has a significant influence of defining its thermal limits. See Bennet et al. 2021 Nat. Comms. “The evolution of critical thermal limits of life on Earth”.

Line 287-289. Consider to integrate here and elsewhere some discussion of the importance of extreme and intense, but temporary, climatic-environmental events. These are primary drivers for local and regional extinction, helping shaping physiological diversity (see for example (1) for the CEH Bozinovic et al. 2011 Ann. Rev. Ecol. Evol. Syst. and references within, and (2) for extreme drivers as evolutionary driver for thermal limits Bennet et al. 2021 Nat. Comms.) participating to global extinctions. I think for example to Marine Heat Waves, intense salinity changes, intense upwelling events, etc.

Minor Comments

Line 19. Specify “differences” in what to make the statement explicit and clearer to readers.

Line 20. << .. has advanced .. >>

Line 26. Add after << … in this way, .. >> something along the line of << ..the integration of information emerging from .. >>

Line 26. Change << … the ask ... >> to something more descriptive << .. to help us in the arduous task .. >>

Line 31, 46, 158, 306 and elsewhere. Give Latinisms in Italics << .. versus .. >>.

Line 128-145. Here consideration to the fact that not the same mechanisms often apply to terrestrial and aquatic (marine-freshwater) organisms should be given. Oxygen for example is 20 times lower in water than in air. The same point must be considered for body plans and body plans’ complexity, which differ greatly between terrestrial and aquatic organisms. This applying not only across phyla, but also within phyla, see for example body plan and physiological differences between terrestrial (reptiles, birds, mammals) and aquatic chordates (fish, and secondarily marine mammals), and also between terrestrial (insects and arachnids) and aquatic (crustaceans).

Line 153. “bake”?

Line 157. Unclear what << .. poorly buffered taxa … >> means. Do you mean << … organisms with poor homeostatic abilities .. >>?

Line 190-191. This is obvious! Conceptually tautological.

Line 242. What about intra and interspecific competition?

Line 320-329. I see the challenge here to integrate data that have a different “grain”, considering that physiological and ecological processes occur at different time scales, and are studied at different time scales and level of biological complexity.

Line 341. Incorrect terminology. We cannot say that seawater become more acidic, in fact it is less alkaline, but even so it is always best to refer to the phenomenon of “ocean acidification” or to talk about “a reduction in pH”, but to avoid to say “more acidic seawater”. It would be semantically incorrect. Also note that a reduction in pH it is but one of the symptoms of ocean acidification, as the increase in seawater pCO2 and the reduction in carbonate ions and omega values are also important element to mention. Finally remember that 0.44 an estimate for the global ocean, locally and regionally conditions vary greatly.

Line 351-355. I am not convinced of this point. Based on physiological abilities, rare species should be sensitive (more sensitive) in the early stage of extinction based on the fact that they possess narrower physiological windows of tolerance: based on extant organisms’ physiology of course, as we cannot define empirically that of extinct species. To provide an example, I have shown, and I am not the only one, that there is a thermal limit physiology of rare versus common species (Calosi et al. 2008 J. Biogeog., Calosi et al. 2008 Biol. Letts.), and that the breadth and central position of the latitudinal range of extent of modern species, as well as their southern and northern most geographical limits, are predicted by the breadth of their thermal window and their CTmax and CTmin, respectively. I bring this as an example among many.

However, I do not know whether in secondary phases of extinction, what were before rare versus common species, are more or less favoured.

All consider the term “long-term” confuses me, as I think here it is more relevant to state in which “phases” of the extinction we are: the early or primary stage when direct physiological effects are more relevant, or the later or secondary phases when indirect ecological effects matter more? Consider to change the discussion in this direction. Apologise if the “binary” view of an extinction I used here does not make justice to a far more-complex phenomenon.

Line 363-379. I agree fully with the statement, but what about other major drivers? Ocean acidification, changes in salinity, etc?

More in general your discussion only focuses on temperature and oxygen, and whilst I recognise the primary impacts of these major drivers, they are not the only one, and they do not occur in isolation: see for example works by Côté, Piggot, Carrier-Belleau and others on multi-stressors occurrence and non-linear effects on aquatic organisms.

Review: Selectivity of mass extinctions: Patterns, processes, and future directions — R0/PR3

Conflict of interest statement

I have no competing interests.

Comments

Comments to Author: Review of the manuscript ”Selectivity of mass extinctions: patterns, processes, and future directions” submitted to Cambridge Prisms: Extinction by Payne et al.

This manuscript seeks to address mainly intrinsic processes governing the selectivity of mass extinction events. The manuscript is nicely written and in clear language. There are many good examples highlighting the different mass extinctions and I think therefore that this manuscript will serve as an excellent instalment within the field of mass extinctions as it reviews past research highlighting many relevant references in the field in a nice way while at the same way calls out specific directions within this research field that may of particular relevance for the current Anthropocene biodiversity crisis.

While I do not have any ‘objections’ to anything written I do have a few comments that the authors might find relevant. Firstly, it stumbled upon the statement in the abstract (l. 34 and again l. 108–126) that ‘geographic range matters less during mass extinction’. While this may be true, I just note that in the case of the Late Ordovician crisis, we do actually see a lot of selectivity with respect to geographic ranges (see, for instance, Finnegan et al., 2016, Proc. Roy. B, http://dx.doi.org/10.1098/rspb.2016.0007). This reference would also be relevant for lines 268–272 (aerobic habitat loss).

l. 132. Regarding body size, it could perhaps be of relevance to throw in a few references on the ‘lilliput-effect’ during mass extinctions. From the top of my head, I recall Huang Bing had a paper on the end-Ordovician and Richard Twitchett one on the end-Permian, but there are quite a few more.

Also, perhaps of relevance to the discussion in l. 222–388 (latitudinal extinction patterns), I think the authors might find a new preprint by Ontiveros et al: ‘Cooling Oceans Did Trigger Ordovician Biodiversifcation’ (it should be on Research Square) interesting as it models the latitudinal biodiversity gradient during the Middle Ordovician cooling, showing how cooling climate affected biodiversity accumulation positively (i.e. the opposite scenario than what is seen during extinctions where warming reduces biodiversity levels).

Minor pedantic edits:

l. 20: add ‘d’ after ‘advance’ ->’advanced’

l. 153: change ‘b’ to ‘m’ so that it reads ‘make’ instead of ‘bake’ :-)

To sum up, I enjoyed reading this very comprehensive and interesting manuscript and I, therefore, encourage publication almost as is.

-Christian Rasmussen.

Review: Selectivity of mass extinctions: Patterns, processes, and future directions — R0/PR4

Conflict of interest statement

No competing interests.

Comments

Comments to Author: This is a concise and well-written review discussing patterns of extinction selectivity in the fossil record. A particular emphasis is placed on mass extinctions and how fossil data--when combined with Earth system models--may help to provide constraints on likely patterns of extinction stemming from anthropogenic global change. The paper provides an admirable overview of a diverse field and condenses the topic into a manageable size perfect for a general audience, student seminars, and so on. My only suggestions to the authors concern a few minor edits (issues of consistency, typos, parts that the authors themselves have flagged as needing an update) indicated below. I look forward to seeing the final version of this contribution.

Line 20: advance -> advanced

Line 31: changing -> change

Line 56: be sure to include social media summary

Line 153: bake -> make

Line 192: “marine” is indicated here, but is the approach potentially more general?

Line 212-213: condense this line. could read “a way for more unified marine and terrestrial studies” or similar.

Line 232: replace “-” with “:” given that it sets off a list of items?

Line 241 (and elsewhere): check that “2” is in subscript for O2

Line 399: replace “), (” with “;”

Lines 686, 689: subscript “2” for O2

Line 701: write out “low latitude vs high latitude”

Figure 1, panel B: Chondrichtyes -> Chondrichthyes

Figure 1 (and elsewhere): use shared conventions for axis labels; currently a mix of first word capitalized, all words capitalized, no words capitalized

Recommendation: Selectivity of mass extinctions: Patterns, processes, and future directions — R0/PR5

Comments

Comments to Author: Dear Dr. Payne and coauthors,

Thank you for your submission on selectivity of mass extinctions. It was an enjoyable read and will make a fine contribution with some minor revision. Of the three reviewers, two had very minor comments, mostly typological that should be easy to address. Reviewer #1 had more substantial comments reflecting minor disagreement on some points, clarification on others, and additional references that should be considered for citation to better reflect the literature. I encourage you to carefully consider their comments. With this last point about literature citation, I would agree that in some sections (the K/Pg terrestrial section, an area that I study) the coverage of the literature could be increased. I don’t mean to suggest that you need to cite my work but there are a number of relevant papers on extinction selectivity across the K/Pg in terrestrial biota (Field et al. 2018 Curr Biol; Hughes et al. 2021 Ecol & Evol; Wilson 2013 Paleobiology) that you should consider to bolster your survey of the literature.

We look forward to your revisions so that we can move this contribution to accepted.

Sincerely,

Greg Wilson Mantilla

Decision: Selectivity of mass extinctions: Patterns, processes, and future directions — R0/PR6

Comments

No accompanying comment.

Decision: Selectivity of mass extinctions: Patterns, processes, and future directions — R0/PR7

Comments

No accompanying comment.

Author comment: Selectivity of mass extinctions: Patterns, processes, and future directions — R1/PR8

Comments

March 2, 2023

Dear Dr. Brook and Dr. Alroy,

We appreciate the opportunity to submit a revised version of our manuscript. We have addressed each of the reviewer and editor comments and hope that you will find the revised manuscript acceptable for publication. We append a point-by-point response below for your reference. We have accepted and adopted most of the suggestions. Some comments from reviewer 1 ask for substantially more detail or depth on certain topics. We have added content where possible but found that the edits required to fully accommodate many of these comments are beyond the scope of what we can accomplish within the length limits available.

Sincerely yours,

Jonathan Payne and co-authors

Handling Editor’s Comments to Author:

Handling Editor: Wilson Mantilla, Gregory

Comments to the Author:

Dear Dr. Payne and coauthors,

Thank you for your submission on selectivity of mass extinctions. It was an enjoyable read and will make a fine contribution with some minor revision. Of the three reviewers, two had very minor comments, mostly typological that should be easy to address. Reviewer #1 had more substantial comments reflecting minor disagreement on some points, clarification on others, and additional references that should be considered for citation to better reflect the literature. I encourage you to carefully consider their comments. With this last point about literature citation, I would agree that in some sections (the K/Pg terrestrial section, an area that I study) the coverage of the literature could be increased. I don’t mean to suggest that you need to cite my work but there are a number of relevant papers on extinction selectivity across the K/Pg in terrestrial biota (Field et al. 2018 Curr Biol; Hughes et al. 2021 Ecol & Evol; Wilson 2013 Paleobiology) that you should consider to bolster your survey of the literature.

Response: We have checked these papers. We note that the Hughes paper was already cited in the manuscript (line 85 of initial submission). We have added a reference to Wilson 2013, which we had missed, because it adds another trait (diet) that was not part of the list but is of obvious relevance and interest. As noted elsewhere, length limitations on this manuscript prevent fully comprehensive review of prior work or exhaustive citation lists on any given topic addressed here.

We look forward to your revisions so that we can move this contribution to accepted.

Sincerely,

Greg Wilson Mantilla

Reviewer(s)' Comments to Author:

Reviewer: 1

Comments to the Author

Review of the article by Payne et al. entitled “Selectivity of mass extinctions: patterns, processes, and

future directions”

The work by Jonathan L. Payne, Jood A. Al Aswad, Curtis Deutsch, Pedro M. Monarrez, Justin L. Penn and Pulkit Singh is really interesting y, and will make a significant contribution to the research field of extinctions, but also to the fields of global change ecology and biology. However, the article requires some work before to be ready for publication. Particularly, provide a more balanced discussion of the existing evidence and limitations of the approach proposed, verify semantics in few cases, and consider to integrate existing references and concept I highlight below. I believe you can relatively easy deal with the points I raise, and whilst the comments may lead to minor or major revision. I selected major only to enable you have more time if you require so.

Major Comments

Line 149-151. In my opinion, here, you are only skimming the surface. You correctly mention the simple Bauplans of heavy calcified marine invertebrates can in part explain the extinction patterns described in the fossil record for the end-Permian. The point is valid and interesting. I could argue, however, that Bauplan “simplicity” can be an advantage, enabling an organism to be less impacted directly by (at least some) environmental drivers, and indirectly possess the advantage to have lower metabolic-energetic requirements. In this sense, organisms with more complex, high-performance cardio-circulatory and respiration, systems may possess greater homeostatic abilities (see for example Melzner et al. 2009 for ocean acidification), but they will not be able to sustain high homeostatic activities for extended periods of time. Complexity requires more energy for maintenance and repair. Modern ecosystem collapse, such as in the North Sea following a switch of a keystone copepod species (from more to less nutritional), the loss in high level predators was documented. In addition, you seem to (I say this respectfully and with the best of intention to help you further improving your excellent work) miss the importance of the effects of other drivers, which can act along different pathways. For example, here specifically, you are not discussing of the direct (i.e. corrosion of carbonated structure) and indirect (e.g. increase in energetic cost for compensatory calcification to maintain a positive net mineralisation in light of passive dissolution) effects of ocean acidification conditions on marine heavy calcifiers. Please ref. to the literature on this topic. It is vast, but it needs to (even if synthetically) be acknowledged, as it is super relevant. I can refer to my own work to help, but the choice of ref. to use sits entirely with you: please see Calosi et al. 2019 Ann. Rev. Mar. Sc. for the discussion of putative mechanisms-pathways of action (see summary Figure 4), and to Calosi et al. 2017 Nat. Comms. and Findlay et al. 2011 Mar. Biol. Res. where differences in passive corrosion-net mineralisation in different population of the same species and different species are reported. but as I suggest, the literature on this topic is extensive.

Response: We agree with the reviewer that we are only scratching the surface here, and have only limited ability to do more given length limitations. Regarding direct and indirect acidification effects on heavy calcifier fitness, we have included a reference to Calosi et al., 2017. To the reviewer’s point about being able to argue the opposite – we do exactly that later in the paragraph, noting that the end-Cretaceous mass extinction shows an opposing direction of selectivity. We agree that body plan complexity can be an advantage or disadvantage depending on the circumstances and have noted that metabolic differences among higher taxa should be taken into account in future studies of taxonomic extinction selectivity. From our point of view, the key value of this review is showing that different events play out in different ways.

Line 193. I feel that in the paper, “ecological-ecosystemic” aspects of mass extinction should be discuss even more thoroughly. The authors do so, but more thinking and work should go into this to push the work and the MS to where it can, and deserved to be. Ensure to bring this as forward as possible in the MS. The method you propose is great to paint a picture on the direct (physiological) impacts of temperature and oxygen (not all drivers characterising current global change for example) but cannot be used to depict the indirect (ecological) impacts. You recognise the issue, but this can be stated even more clearly. You need to discuss in greater depth how best the tool you propose can be used, to what advantage, and where it cannot be used or used recognising the presence of limitations. In line 280 and following you could integrate the work by Reddin et al. 2020 NCC (which you cite) to the discussion, related to early extinctions being caused primarily by direct physiological effects, and later extinctions being caused by indirect ecological effects. All considered, I suggest to enhanced the discussion, and I fully recognise the authors’ effort in the intellectual integration of “physiological” and “ecological” extinctions …. and figure 4 is excellent by the way!!

Response: As noted in our response to the reviewer’s prior comment, we are constrained in terms of how much depth we can use to address these points by the length limit on the manuscript. We appreciate the desire for more depth but believe that this is generally beyond the scope available to us. We have already made the point that the reviewer is emphasizing here in the preceding paragraph. We added another citation to Reddin’s work earlier in that paragraph, where it was warranted, but have chosen to otherwise leave the text unchanged. We agree with and understand the reviewer’s desire to see more depth in the discussion of these factors but find ourselves constrained by the length limit of the manuscript.

Line 110, 203-204, 280 and elsewhere. Please the first time you mention “tolerance” and “resilience” ensure to provide a definition in parenthesis: 110 and 204, respectively. I have the impression that in most cases you use it “correctly” but in other not (ex. line 282. Tolerance is the ability of a biological system to resist change following or under a perturbation, and resilience is the ability of a biological system to come back to the original stable state it was in after a perturbation (i.e. resilience is the inverse of the time needed to get back to the starting status).

Response: We have provided greater specificity to our use of these terms where relevant in the text.

Line 266. Pre-adaptation does not exist. It is an erroneous concept. An organism cannot be adapted to something it has not experienced before. And if it is adapted to be able to face a certain challenge, it is because it has already experienced it in his phylogenetic history, or it has been selected to be tolerant to another stressor which grants it “protection” to the stressor we are investigating. In either of these cases, we talk of “exaptation” (i.e. existing adaptation), please see Gould’s seminal work on this topic.

Response: We agree with the reviewer and regret the poor phrasing here. We were not arguing for pre-adaptation in this sense but rather for the fact that some species were already adapted to the conditions that become more widespread after the extinction. We have edited to the text to read “…species previously occupying the tropics would already have been adapted to warm, low-O2 conditions that became more widespread”.

For your interest, we have recently published a work on the diversity and evolution of thermal limits, testing for the presence of paleo signals on these traits, and we found one. It was not the strongest of signals of the evolutionary drivers we investigated, but it was non the less significant. In short, the era of emergence of a taxonomic group has a significant influence of defining its thermal limits. See Bennet et al. 2021 Nat. Comms. “The evolution of critical thermal limits of life on Earth”.

Response: This is an interesting finding. We have added a citation to this study within the paragraph on climate warming within Section 3.2.

Line 287-289. Consider to integrate here and elsewhere some discussion of the importance of extreme and intense, but temporary, climatic-environmental events. These are primary drivers for local and regional extinction, helping shaping physiological diversity (see for example (1) for the CEH Bozinovic et al. 2011 Ann. Rev. Ecol. Evol. Syst. and references within, and (2) for extreme drivers as evolutionary driver for thermal limits Bennet et al. 2021 Nat. Comms.) participating to global extinctions. I think for example to Marine Heat Waves, intense salinity changes, intense upwelling events, etc.

Response: We added an example of a regional extirpation connected to fluctuating aerobic habitat at a relevant location in the text. Given that the focus of this review is on patterns of global extinction selectivity and their drivers, and due to space constraints, we refrain from discussing short-term, regionalized extreme events in further detail, which have a strong influence on patterns of local diversity and extirpation.

Minor Comments

Line 19. Specify “differences” in what to make the statement explicit and clearer to readers.

Response: Excellent point. We have rephrased as: differences in morphological, ecological, and physiological traits

Line 20. << .. has advanced .. >>

Done

Line 26. Add after << … in this way, .. >> something along the line of << ..the integration of information emerging from .. >>

Done

Line 26. Change << … the ask ... >> to something more descriptive << .. to help us in the arduous task .. >>

Done

Line 31, 46, 158, 306 and elsewhere. Give Latinisms in Italics << .. versus .. >>.

Done

Line 128-145. Here consideration to the fact that not the same mechanisms often apply to terrestrial and aquatic (marine-freshwater) organisms should be given. Oxygen for example is 20 times lower in water than in air. The same point must be considered for body plans and body plans’ complexity, which differ greatly between terrestrial and aquatic organisms. This applying not only across phyla, but also within phyla, see for example body plan and physiological differences between terrestrial (reptiles, birds, mammals) and aquatic chordates (fish, and secondarily marine mammals), and also between terrestrial (insects and arachnids) and aquatic (crustaceans).

Response: This is a great point that was not made clearly enough in the manuscript. We have rephrased the second sentence of this paragraph to draw this distinction out explicitly. Due to length limits, we have not been able to incorporate a more extensive discussion of this issue.

Line 153. “bake”?

Corrected to “make”

Line 157. Unclear what << .. poorly buffered taxa … >> means. Do you mean << … organisms with poor homeostatic abilities .. >>?

We rephrased to “taxa thought to be more sensitivity to ocean acidification”

Line 190-191. This is obvious! Conceptually tautological.

Fair enough. Sentence has been removed.

Line 242. What about intra and interspecific competition?

Feeding, growth, motion, etc, are all activities that may be part of this intra and interspecific competition. Therefore we have not edited the text here.

Line 320-329. I see the challenge here to integrate data that have a different “grain”, considering that physiological and ecological processes occur at different time scales, and are studied at different time scales and level of biological complexity.

Good point. We have added a sentence paraphrasing this point from the reviewer.

Line 341. Incorrect terminology. We cannot say that seawater become more acidic, in fact it is less alkaline, but even so it is always best to refer to the phenomenon of “ocean acidification” or to talk about “a reduction in pH”, but to avoid to say “more acidic seawater”. It would be semantically incorrect. Also note that a reduction in pH it is but one of the symptoms of ocean acidification, as the increase in seawater pCO2 and the reduction in carbonate ions and omega values are also important element to mention. Finally remember that 0.44 an estimate for the global ocean, locally and regionally conditions vary greatly.

We have rephrased to say “pH is expected to decrease, on average, by 0.44 pH units by the end of the 21st century” to make the phrasing accurate with respect to acidity/alkalinity and spatial complexity.

Line 351-355. I am not convinced of this point. Based on physiological abilities, rare species should be sensitive (more sensitive) in the early stage of extinction based on the fact that they possess narrower physiological windows of tolerance: based on extant organisms’ physiology of course, as we cannot define empirically that of extinct species. To provide an example, I have shown, and I am not the only one, that there is a thermal limit physiology of rare versus common species (Calosi et al. 2008 J. Biogeog., Calosi et al. 2008 Biol. Letts.), and that the breadth and central position of the latitudinal range of extent of modern species, as well as their southern and northern most geographical limits, are predicted by the breadth of their thermal window and their CTmax and CTmin, respectively. I bring this as an example among many.

However, I do not know whether in secondary phases of extinction, what were before rare versus common species, are more or less favoured.

All consider the term “long-term” confuses me, as I think here it is more relevant to state in which “phases” of the extinction we are: the early or primary stage when direct physiological effects are more relevant, or the later or secondary phases when indirect ecological effects matter more? Consider to change the discussion in this direction. Apologise if the “binary” view of an extinction I used here does not make justice to a far more-complex phenomenon.

Response: We appreciate the reviewer’s comment here and have edited to remove the sentence regarding the IUCN Red List. Instead, we have added a final sentence to the paragraph highlighting the challenge noted so nicely by the reviewer, about connecting the primary and secondary (physiological and ecological) phases of extinction within predictive models.

Line 363-379. I agree fully with the statement, but what about other major drivers? Ocean acidification, changes in salinity, etc?

Response: Good point. We have added a sentence to make this more explicit for the readers: Further work to integrate the effects of changes in pH, pCO2, salinity, and other key environmental variables into physiological performance models has the potential to make these models more general and accurate in reconstructing the causes of past extinction and predicting the consequences of future global change.

More in general your discussion only focuses on temperature and oxygen, and whilst I recognise the primary impacts of these major drivers, they are not the only one, and they do not occur in isolation: see for example works by Côté, Piggot, Carrier-Belleau and others on multi-stressors occurrence and non-linear effects on aquatic organisms.

Response: Although we have focused on oxygen and temperature in the description of the Metabolic Index, the revised text acknowledges the importance of other stressors for more complete understanding of physiological stresses leading to extinction. Due to limits of space, we have not been able to address each of these in detail.

Reviewer: 2

Comments to the Author

Review of the manuscript ”Selectivity of mass extinctions: patterns, processes, and future directions” submitted to Cambridge Prisms: Extinction by Payne et al.

This manuscript seeks to address mainly intrinsic processes governing the selectivity of mass extinction events. The manuscript is nicely written and in clear language. There are many good examples highlighting the different mass extinctions and I think therefore that this manuscript will serve as an excellent instalment within the field of mass extinctions as it reviews past research highlighting many relevant references in the field in a nice way while at the same way calls out specific directions within this research field that may of particular relevance for the current Anthropocene biodiversity crisis.

While I do not have any ‘objections’ to anything written I do have a few comments that the authors might find relevant. Firstly, it stumbled upon the statement in the abstract (l. 34 and again l. 108–126) that ‘geographic range matters less during mass extinction’. While this may be true, I just note that in the case of the Late Ordovician crisis, we do actually see a lot of selectivity with respect to geographic ranges (see, for instance, Finnegan et al., 2016, Proc. Roy. B, http://dx.doi.org/10.1098/rspb.2016.0007). This reference would also be relevant for lines 268–272 (aerobic habitat loss).

Response: We have added the citations suggested here. We believe our characterization of selectivity as present during both mass extinction and background intervals but generally more pronounced during background intervals remains an accurate description of comparative results.

l. 132. Regarding body size, it could perhaps be of relevance to throw in a few references on the ‘lilliput-effect’ during mass extinctions. From the top of my head, I recall Huang Bing had a paper on the end-Ordovician and Richard Twitchett one on the end-Permian, but there are quite a few more.

Response: The Lilliput Effect is an interesting observation, describing size reduction within (and sometimes among) species across extinction events. However, it is often documented by showing changes in size distributions before and after extinction events. Such changes result from the combined effects of extinction and origination. Consequently, many studies documenting the Lilliput Effect do not provide an explicit analysis of extinction selectivity and so are difficult to cite as evidence in this context. The Huang et al. (2010) study on the Ordovician and the Twitchett (2007) study on the end-Permian both fall into this category.

Also, perhaps of relevance to the discussion in l. 222–388 (latitudinal extinction patterns), I think the authors might find a new preprint by Ontiveros et al: ‘Cooling Oceans Did Trigger Ordovician Biodiversifcation’ (it should be on Research Square) interesting as it models the latitudinal biodiversity gradient during the Middle Ordovician cooling, showing how cooling climate affected biodiversity accumulation positively (i.e. the opposite scenario than what is seen during extinctions where warming reduces biodiversity levels).

Response: We thank the reviewer for highlighting this new study. As it is not yet peer reviewed and the focus of this review is on extinction selectivity, as opposed to broader evolutionary processes, we have chosen not to cite it in this review.

Minor pedantic edits:

l. 20: add ‘d’ after ‘advance’ ->’advanced’

Done

l. 153: change ‘b’ to ‘m’ so that it reads ‘make’ instead of ‘bake’ :-)

Done

To sum up, I enjoyed reading this very comprehensive and interesting manuscript and I, therefore, encourage publication almost as is.

-Christian Rasmussen.

Reviewer: 3

Comments to the Author

This is a concise and well-written review discussing patterns of extinction selectivity in the fossil record. A particular emphasis is placed on mass extinctions and how fossil data--when combined with Earth system models--may help to provide constraints on likely patterns of extinction stemming from anthropogenic global change. The paper provides an admirable overview of a diverse field and condenses the topic into a manageable size perfect for a general audience, student seminars, and so on. My only suggestions to the authors concern a few minor edits (issues of consistency, typos, parts that the authors themselves have flagged as needing an update) indicated below. I look forward to seeing the final version of this contribution.

Line 20: advance -> advanced

Done

Line 31: changing -> change

Done

Line 56: be sure to include social media summary

Done

Line 153: bake -> make

Done

Line 192: “marine” is indicated here, but is the approach potentially more general?

Response: It is more general. We have revised by removing the specification about marine species.

Line 212-213: condense this line. could read “a way for more unified marine and terrestrial studies” or similar.

Response: We agree with the reviewer and have adopted their suggested phrasing.

Line 232: replace “-” with “:” given that it sets off a list of items?

Done

Line 241 (and elsewhere): check that “2” is in subscript for O2

We have rechecked and fixed these cases.

Line 399: replace “), (” with “;”

Done

Lines 686, 689: subscript “2” for O2

Done

Line 701: write out “low latitude vs high latitude”

Done

Figure 1, panel B: Chondrichtyes -> Chondrichthyes

Done

Figure 1 (and elsewhere): use shared conventions for axis labels; currently a mix of first word capitalized, all words capitalized, no words capitalized

Done

Review: Selectivity of mass extinctions: Patterns, processes, and future directions — R1/PR9

Conflict of interest statement

n/a

Comments

Comments to Author: Thank you for addressing previous comments. I caught only a few issues here, all of them very minor and easily sorted:

Throughout: there are variable abbreviations with the same meaning (Ma, Mya).

Line 141: fish -> fishes.

Line 165: sensitivity -> sensitive.

Lines 352-3: is it redundant to indicate oceans experienced change to “ocean biogeochemistry”?

Recommendation: Selectivity of mass extinctions: Patterns, processes, and future directions — R1/PR10

Comments

Comments to Author: Dear Authors,

Thank you for the careful revisions made to your manuscript “Selectivity of mass extinctions: patterns, processes, and future directions.” We are happy at this time to accept this excellent contribution to the journal. The reviewer made a few very minor suggested corrections that we would like you to address before publication. Again, we greatly appreciate the attention to your revision process and the excellent resulting product.

Sincerely,

Greg Wilson Mantilla

Decision: Selectivity of mass extinctions: Patterns, processes, and future directions — R1/PR11

Comments

No accompanying comment.

Decision: Selectivity of mass extinctions: Patterns, processes, and future directions — R1/PR12

Comments

No accompanying comment.