Relation of Phanerozoic stable isotope excursions to climate, bacterial metabolism, and major extinctions

Abstract

Conspicuous global stable carbon isotope excursions that are recorded in marine sedimentary rocks of Phanerozoic age and were associated with major extinctions have generally paralleled global stable oxygen isotope excursions. All of these phenomena are therefore likely to share a common origin through global climate change. Exceptional patterns for carbon isotope excursions resulted from massive carbon burial during warm intervals of widespread marine anoxic conditions. The many carbon isotope excursions that parallel those for oxygen isotopes can to a large degree be accounted for by the Q10 pattern of respiration for bacteria: As temperature changed along continental margins, where ∼90% of marine carbon burial occurs today, rates of remineralization of isotopically light carbon must have changed exponentially. This would have reduced organic carbon burial during global warming and increased it during global cooling. Also contributing to the δ¹³C excursions have been release and uptake of methane by clathrates, the positive correlation between temperature and degree of fractionation of carbon isotopes by phytoplankton at temperatures below ∼15°, and increased phytoplankton productivity during “icehouse” conditions. The Q10 pattern for bacteria and climate-related changes in clathrate volume represent positive feedbacks for climate change.

Many possible explanations have been advanced for one or more of the abrupt excursions in the δ¹³C of marine carbonates and organic matter that have been documented in the geologic record (see Fig. 1). Most of these hypotheses have focused on factors that change the global rate of burial of organic carbon, which is isotopically light: changes in upwelling and primary productivity; fluctuations of sea level; changes in ocean dynamics, including ones affecting the extent of anoxic conditions; changes in carbonate weathering rates; release of methane from sediments; changes in nutrient input from the land to the oceans; volcanic degassing; and release of isotopically light CO₂ from the deep sea (for literature, see SI Text, Section I). Given the strong positive correlation between δ¹³C and δ¹⁸O excursions in shallow marine sediments (Fig. 1), however, parsimony suggests that one or more unifying explanations should be sought to explain all of these phenomena.

Fig. 1.

Stable isotope excursions that have been documented in shallow marine strata and are associated with mass extinctions. Eighteen intervals (A–R) contain a total of 26 such δ¹³C excursions. Corresponding to these, and trending in the same direction, are 19 published δ¹⁸O excursions, which are displayed in the plots to the right of those depicting δ¹³C. Temporal positions of excursions are indicated by encircled letters on the left. Blue indicates association with global cooling and red with global warming; black indicates the absence of published evidence of associated climate change. Horizontal scales represent magnitudes of δ¹³C and δ¹⁸O excursions in parts per thousand. Light δ¹³C in N is for organic carbon rather than carbonates, and heavy δ¹⁸O in H is for conodonts rather than bulk or skeletal carbonate. Ordinates represent stratigraphic positions of samples and are neither precisely linear with respect to time nor scaled the same for all graphs. (See SI Text, Section II for sources of figures and for citations of literature on isotope excursions, climate change, and major extinctions.)

Missing from previous explanations for geologically brief δ¹³C excursions has been a consideration of the role of microbial processes, even though microbes have an enormous global biomass and play a prominent role in the exogenic carbon cycle. Particulate organic matter in ecosystems has three possible fates: It can be consumed by primary consumers, decomposed by bacteria (or fungi, in the case of lignocellulose), or buried. The relative importance of these fates varies from place to place and time to time. For example, carbon burial on land increased markedly during the late Paleozoic, when coal swamps became widespread (1). In the oceanic realm beyond continental rises, organic carbon burial plays a minor role in the global carbon cycle. Owing to consumption by primary consumers and microbes, less than 1% of particulate organic matter in the upper oceanic realm ends up being buried on the deep seafloor (2). In contrast, continental shelves, slopes, and rises account for ∼90% of global carbon burial in the ocean (3). In these regions, elevated nutrient levels produce high rates of primary productivity; rivers and marshes supply large amounts of organic detritus; and high rates of sedimentation result in the burial and preservation of ∼75% of the organic carbon reaching the seafloor, about half of this amount being buried in deltaic depositional systems (4).

Temperature has a profound effect on the role of decomposers, as indicated by the effective fungal destruction of plant debris in tropical rainforests, which results in very thin, humus-poor soils. The Arrhenius pattern for chemical reactions, an exponential increase in rate with temperature, yields a Q10 value: the fractional increase in the rate of a reaction per 10 °C of temperature elevation (Fig. 2). Bacteria and fungi adhere approximately to the Q10 pattern because of the effect of temperature on enzymes utilized in their metabolic processes. For marine and freshwater sediments above very low temperatures, the average Q10 for respiration is ∼2.2 (Fig. 2) (5). This relationship has been incorporated into models of the carbon cycle on long time scales (6) but has not previously been invoked to explain the δ¹³C excursions of Fig. 1, which occurred on short time scales (generally < 10⁶ y) and are distinct from long-term secular isotopic trends (2).

Fig. 2.

Relationship to temperature of respiration rate and specific growth rate (fractional increase in cell mass per unit time) for freshwater and marine benthic bacteria (data from ref. 5). An increase in mean temperature from 15–20 ºC results in an increase in respiration rate of ∼40% (stippled area).

It is true that growth rates of phytoplankton also increase with temperature, but the net primary productivity of marine phytoplankton nonetheless decreases with global warming, which results in decreased turbulent mixing and upwelling of ocean waters. With warming, the reduced supply of nutrients exerts a much stronger negative influence on the productivity of phytoplankton than the positive effect of the Q10 relationship (7).

In this paper, I investigate the causes of abrupt δ¹³C excursions that occurred in association with major extinctions. I attribute these excursions to four factors: first, and perhaps most importantly, changes in global respiration rates of remineralizing marine bacteria; second, release or uptake of methane by clathrates; third, the positive correlation between carbon isotope fractionation by marine phytoplankton and temperature at relatively low temperatures; and fourth, changes in the productivity of phytoplankton and hence in the burial rate of organic carbon, with changes in global temperature that alter wind stress and rates of nutrient supply.

Results

A compilation of conspicuous isotopic excursions associated with major extinctions reveals a close correspondence between δ¹³C and δ¹⁸O excursions throughout Phanerozoic time. Fig. 1 displays 18 plots of δ¹³C for carbonates and organic matter in shallow marine stratigraphic sections. These exhibit a total of 26 δ¹³C excursions, each associated with a major extinction. Of these 26, 19 are known to have been accompanied by a δ¹⁸O excursion, and in every case the paired excursions have both been either positive or negative. It is common knowledge that δ¹⁸O excursions reflect global temperature changes and sometimes expansion or contraction of glaciers, which preferentially accumulate ¹⁶O. Among the global extinctions associated with the 26 δ¹³C excursions depicted in Fig. 1 are four of the five most severe Phanerozoic mass extinctions: those of the Late Ordovician, Late Devonian, Late Permian, and Late Triassic. Of the 17 positive δ¹³C excursions of Fig. 1, 16 have been identified as being associated with climate change and, in every case, the change entailed global cooling. In complementary fashion, all six of the nine negative excursions for which there is evidence of simultaneous climatic change were associated with global warming. The implication is that climate changes have been connected to carbon and oxygen excursions and, as indicated by a variety of other evidence (8), have also played an important role in major extinctions. Excluded from Fig. 1 are exceptional positive δ¹³C excursions during three Mesozoic intervals of global warming when massive amounts of organic carbon were buried under widespread anoxic conditions in the ocean. I will discuss these episodes in the Discussion section.

It is inevitable that changes in mean global temperature will alter the δ¹³C of seawater via the Q10 relationship for microbes. Although the total rate of net primary productivity and total rate of respiration in the ocean are similar (2), an increase in the rate of bacterial respiration would reduce but not eliminate carbon burial along continental margins, where ∼75% of organic carbon reaching the seafloor is buried today. The flux of remineralized carbon from modern continental shelf/margin sediments is 3.72 × 10¹⁴ g/y (6). The amount of dissolved inorganic carbon in the modern ocean is larger by a factor of almost exactly 10⁵(3.74 × 10¹⁹ g) (9). A temperature increase of 5° would elevate the flux of remineralized carbon by ∼40% (Fig. 2). For marine organic carbon remineralized in shelf/slope sediments, δ¹³C is ∼-23‰, which is slightly more negative than that for marine organic matter because of an admixture of isotopically light carbon derived from terrestrial organic detritus (10). It would be difficult to estimate the time that would be required for a 40% increase in the rate of remineralization along continental margins to shift δ¹³C for dissolved inorganic carbon in the ocean from its present value of ∼-2‰ to -6‰ (a 4‰ excursion, which is a typical magnitude for the excursions illustrated in Fig. 1). The complicating factor is that, although less isotopically light carbon would be buried, a negative feedback would be triggered: Dissolved inorganic carbon would become progressively isotopically lighter, and so would the carbon of phytoplankton and, therefore, of buried carbon. Also, to the degree that warming or cooling of the climate has been protracted, δ¹³C excursions will also have been protracted. In fact, this pattern is evident in Fig. 1, where gradual shifts of δ¹⁸O are paralleled by gradual shifts of δ¹³C.

The bacterial mechanism I am proposing is also consistent with the observation that many carbon isotope excursions, such as those of the Late Ludlow (Fig. 1L) (11) and Late Permian (Fig. 1E) (12), began at about the time of the associated major extinction, when a global climate shift occurred, and peaked slightly later. The high-resolution record for the Eocene/Oligocene transition shows the associated δ¹³C excursion to have begun slightly (< 10,000 y) after the associated δ¹⁸O excursion (13). These patterns are understandable, given the exponential response of respiration rate to temperature change (Fig. 2).

Changes in global rates of bacterial respiration at times of CO₂-induced climate change have represented positive feedbacks because changes in bacterial respiration rates as a function of temperature are accompanied by little change in the fraction of acquired carbon that is respired (14). Every rapid global shift of δ¹³C for shallow marine sediments has been reversed at some point, however, either by one or more negative feedbacks, such as a change in terrestrial weathering rates, or by a reversal of whatever factor or factors produced the initial climate change. In fact, many intervals of glacial expansion and contraction have been quite brief. For example, the Late Ordovician Hirnantian interval and its δ¹³C and δ¹⁸O excursions were confined to just 0.5–1 Myr (15).

Three additional factors have presumably also contributed to the δ¹³C excursions associated with climate change. The co-occurrence of positive global δ¹³C and δ¹⁸O excursions is compatible with the possibility that sequestration or release of methane hydrates in marine and terrestrial settings has been partly responsible for the δ¹³C excursions (16). However, quantitative evaluations have concluded that release of methane hydrates from the seafloor cannot fully account for the Paleocene/Eocene (17) or terminal Permian (18) negative δ¹³C excursions; one or more positive feedbacks are required. The response of bacterial respiration to climatic warming would have been one such feedback.

Another, secondary factor contributing to the δ¹³C excursions of Fig. 1 would have been the positive correlation between temperature and the δ¹³C of phytoplankton and particulate organic carbon for temperatures below ∼15 °C (Fig. 3) (19). During “greenhouse” intervals of the past, seawater temperatures were above 15 °C throughout the photic zone, except at relatively high latitudes. The degree to which this condition would have elevated the global average δ¹³C of buried organic carbon above its level during icehouse times is difficult to quantify, but temperature changes equatorward of high-latitude regions would have had no effect.

Fig. 3.

Positive correlation between temperature and δ¹³C of marine particulate organic matter, including phytoplankton cells, for temperatures below ∼15 °C (after ref. 19).

A third, secondary factor contributing to the δ¹³C excursions would have been the strong turbulent mixing and upwelling that characterize icehouse oceans, which elevate the supply of nutrients and therefore enhance phytoplankton productivity (7) and burial of isotopically light carbon.

Discussion

Not all carbon isotope excursions of Fig. 1 have configurations closely resembling those of the associated oxygen isotope excursions, presumably because of second-order regional and global influences. Relative magnitudes of excursions are difficult to interpret. It is puzzling, for example, that the terminal Eocene excursions were small despite having been associated with substantial global cooling.

During three Mesozoic intervals of global warming when vertical mixing of the ocean weakened and anoxia became widespread, black muds were deposited over broad areas. At each of these times, massive burial of isotopically light carbon overrode the reduction of carbon burial via increased bacterial remineralization in areas that were better oxygenated. The result was a positive excursion for δ¹³C. At the end of the Cenomanian, for example, a positive global δ¹³C excursion coincided with a negative global δ¹⁸O excursion (20). Thus, a pulse of global warming quickly reduced vertical mixing of the ocean to provide widespread anoxia. A different pattern, however, is observed throughout the world for oceanic changes in the mid-Toarcian and late early Aptian (21). At the latter time, a sudden negative δ¹³C excursion of ∼2‰ coincided with a negative δ¹⁸O excursion of ∼3‰, in keeping with the standard pattern for global warming (Fig. 1). Only after the widespread burial of organic-rich sediment began was there a strong positive shift of δ¹³C (Fig. 4). Similarly, a 4–7‰ negative shift in mid-Toarcian time was followed by a positive shift after deposition of dark mud began; Hesselbo et al. (21) interpreted the negative excursion as a result of massive dissociation of gas hydrate, but elevated bacterial remineralization via climate warming may largely account for it.

Fig. 4.

Excursions of δ¹³C and δ¹⁸O near the end of the early Aptian recorded in strata of northwestern Sicily and displaying a global pattern: Both isotopic ratios shifted in a negative direction when global warming began (lower arrow), but the δ¹³C trend was reversed when anoxic conditions produced widespread burial of organic carbon (upper arrow) (after ref. 31). OAE, oceanic anoxic event.

A sudden negative δ¹³C excursion following the terminal Cretaceous mass extinction, which, perhaps uniquely, was caused by an extraterrestrial impact, was unrelated to factors discussed in the present paper, having resulted from a brief annihilation of marine phytoplankton populations that resulted in a drastic reduction of organic carbon burial (22).

During the Pleistocene, global δ¹³C did not follow the temperature-related pattern of earlier intervals discussed herein: It declined slightly during glacial maxima and rose slightly during glacial minima; the mean value for the ocean was 0.32‰ more negative than today during the most recent glacial maximum (23). One contributing factor must have been the transfer to the ocean of a large amount of isotopically light terrestrial organic carbon during the shrinkage of forests (24), which contracted to about 35% of their present global area during the last glacial maximum (25). It is also possible that increased burial of organic material along continental margins during eustatic sea level rises reduced [PO₄] in the ocean, and hence phytoplankton productivity and burial of isotopically light carbon (26). Another likely factor relates to the dramatic change of global carbon dynamics during the Miocene caused by the ecological expansion of marine diatoms. Diatoms assimilate carbon for which δ¹³C is typically ∼6‰ heavier than that of other phytoplankton (27), and they are responsible for ∼40% of the net primary productivity in the modern ocean (28). Their expanded presence has caused a conspicuous increase in the average δ¹³C of carbon buried in the ocean since the early Miocene and, reciprocally, a decrease in the δ¹³C of seawater and hence in marine calcium carbonate (29). Because of steep latitudinal temperature gradients during glacial maxima, winds were strengthened and, as a result, so were turbulent mixing of the ocean and upwelling along continental margins. Consequently, primary productivity of diatoms increased dramatically (30), largely because a special vacuole in which they store nutrients permits them to respond more effectively than other kinds of phytoplankton to increased nutrient supply (28). An increase in the rate at which diatoms sent relatively heavy carbon to the seafloor during glacial maxima must have significantly reduced the δ¹³C of seawater.

Of the four consequences of global climate change identified here, which in some combination can account for the δ¹³C excursions depicted in Fig. 1, three must be of concern with regard to future global warming: increased respiration by bacteria along continental margins, release of methane by clathrates, and reduced assimilation of CO₂ by phytoplankton. All of these processes represent significant feedbacks that will enhance global warming that results from human-induced increases in the partial pressure of atmospheric CO₂.

Methods

The δ¹³C excursions depicted in Fig. 1 are the set of excursions associated with major extinctions that were encountered in a literature survey. Where multiple plots were encountered depicting a particular excursion, one was selected for Fig. 1. The remainder are cited in SI Text, Section II.