Extinction selectivity among marine fishes during multistressor global change in the end-Permian and end-Triassic crises

Ancient mass extinction events such as the end-Permian and endTriassic crises provide analogues for multistressor global change of ocean warming, pH reduction, and deoxygenation. Organism physiology is hypothesized to be a key trait influencing vulnerability to these stressors, but it is not certain how physiology predicts survival over evolutionary time scales and when organisms are faced with opposing or synergistic stressors. Fishes (bony fishes and chondrichthyan fishes) are active organisms with high aerobic scope for thermal tolerance and well-developed acid-base regulation, traits that should confer resilience to global change. To test this, we compiled a database of fossil marine fish occurrences to quantify extinction rates during background and mass extinctions from the Permian through Early Jurassic, using maximum likelihood estimation to compare extinction trajectories with marine invertebrates. Our results show that fewer chondrichthyan fishes underwent extinction than marine invertebrates during the end-Permian crisis. End-Triassic chondrichthyan extinction rates also were not elevated above background levels. In contrast, bony fishes underwent an end-Triassic extinction comparable to that of marine invertebrates. The differing responses of these two groups imply that a more active physiology can be advantageous during global change, although not uniformly. Permian–Triassic chondrichthyan fishes may have had broader environmental tolerances, facilitating survival. Alternatively, the larger offspring size of chondrichthyan fishes may provide greater energy reserves to offset the demands of warming and acidification. Although more active organisms have adult adaptations for thermal tolerance and pH regulation, some may nevertheless be susceptible to global change during early life stages. INTRODUCTION Rapid emission of carbon dioxide (CO2) to the atmosphere triggers a chain of perturbations leading to ocean warming, pH reduction, and deoxygenation. These stressors threaten many marine organisms, with consequences for growth, reproduction, calcification, and ultimately survival (Doney et al., 2012; Kroeker et al., 2013). Although the negative consequences of multistressor global change span the tree of life, some organisms are more vulnerable than others. Organism physiology is likely to play a critical role in resisting or adapting to stresses from warming, reduced pH, and hypoxia (Melzner et al., 2009; Pörtner, 2010; Deutsch et al., 2015); however, the degree to which physiology can predict survival at the ecosystem scale and over evolutionary time scales is not as well understood (Queirós et al., 2015). Physiological differences between fishes and invertebrates predict that marine fishes should be more resistant to many global change stresses. While invertebrates may undergo acidosis (decreased pH of body fluids) under elevated pCO2, fishes are typically better able to compensate via active ion exchange and buffering (Claiborne et al., 2002). Although active acid-base compensation incurs energetic costs and may force trade-offs, fishes also tend to have higher internal pCO2 that can maintain diffusive CO2 excretion under reduced ocean pH (Melzner et al., 2009). Fishes may similarly be less vulnerable to thermal stresses if more active organisms have greater thermal tolerance (Peck et al., 2009). The need for short bursts of elevated performance should lead to greater aerobic scope (the excess energy available beyond metabolic maintenance for growth, reproduction, and locomotion), which may confer greater thermal tolerance (Pörtner, 2010). In contrast, fishes are probably more vulnerable than invertebrates, on average, to hypoxia (Vaquer-Sunyer and Duarte, 2008). The synergistic effects of temperature, pH, and dissolved oxygen further complicate predictions of extinction susceptibility during global change (McBryan et al., 2013; Vaquer-Sunyer and Duarte, 2011; Pörtner et al., 2005), so the effects of multistressor global change on active organisms such as fishes remain poorly understood. The fossil record provides opportunities to assess the response of marine fishes to multistressor global change over evolutionary time scales. In particular, the end-Permian and end-Triassic mass extinctions were triggered by rapid CO2 release from flood basalt eruptions, leading to ocean warming and reductions in pH and dissolved oxygen (Payne and Clapham, 2012; Greene et al., 2012; Richoz et al., 2012). Ancient CO2-driven extinctions exhibited characteristic selectivity against less active invertebrate clades (Knoll et al., 2007; Clapham and Payne, 2011; Kiessling and Simpson, 2011; Clapham, 2017). However, there has been little detailed work on extinction selectivity among active marine vertebrate groups during ancient global change events (Friedman and Sallan, 2012). Previous studies of the end-Permian extinction found little loss in richness among bony fishes (Schaeffer, 1973), although they may have undergone elevated turnover following the extinction (Romano et al., 2016). In contrast, multiple Paleozoic chondrichthyan lineages (sharks, rays, and chimaeras) passed through the Permian-Triassic boundary (Mutter and Neuman, 2008; Mutter et al., 2007; Guinot et al., 2013). Chondrichthyan fishes may have had somewhat elevated extinction at the end-Triassic crisis (Guinot and Cavin, 2015; Guinot et al., 2012), but its effect on other fishes remains unclear (Friedman and Sallan, 2012). As a result, it is not certain whether chondrichthyan or bony fishes were more resistant than marine invertebrates, as predicted by physiological differences among the groups, during the end-Permian and end-Triassic mass extinctions. To test the hypothesis that active fishes are less vulnerable to global change stressors, we compare the extinction rates of marine bony fishes and chondrichthyan fishes to those of marine invertebrates from the Permian to the Early Jurassic. These results will enable us to reconstruct the fate of marine fishes during multistressor global change and will test the importance of physiological adaptations on survival.