Australia: The Land Where Time Began

A biography of the Australian continent 

Atmospheric Carbon Dioxide - A 300-Million-Year record from Plant Cuticles

According to the author1 geochemical proxies are used to reconstruct past atmospheric CO2 levels in order to better understand the link between atmospheric CO2 and climate over geological time (Berner, 1997; Pearson & Palmer, 2000; Ekart et al., 1999; Pagani, Freeman & Arthur, 1999). A broad picture of the variation of atmospheric CO2 throughout the Phanerozoic, that covers the last 544 My, has been provided by these records, though there are still some inconsistencies and gaps requiring attention. In this article Retallack presents evidence, based on stomatal abundance in fossil leaves from 4 plant genera closely related to the Ginkgo of the present, for a continuous record spanning 300 My. He reconstructed atmospheric CO2 concentrations based on the known relationship between the number of stomata on a leaf and CO2 concentrations in the growing season. The record he constructed indicated that over the last 300 My there were 2 episodes in which there were intervals when the atmospheric CO2 concentrations were low, less than 1,000 ppmv, and both of these intervals coincided with known ice ages occurring in the Neogene  8-1 Ma and Early Permian, 290-275 Ma. The levels of atmospheric CO2 were high, 1,000-2,000 ppmv, 250-65 Ma, most of the Mesozoic, though in this period there were intervals during which there were excursions in the atmospheric CO2 levels above 2,000 ppmv. The results of the author's study are consistent with some past CO2 record reconstructions (Berner, 1997; Pearson & Palmer, 2000), and palaeotemperature records (Veizer, Godderis & François, 2000), though they suggest the reconstructions based on carbon isotope proxies (Ekart et al., 1999; Pagani, Freeman & Arthur, 1999) may be compromised by isotopically light methane that was released as episodic outbursts (MacLeod & Wing, Krull et al., 2000). According to the author1 water vapour, CO2 and methane were involved in greenhouse climate warming over the last 300 My.

Over the past 200 years of industrial fossil fuel consumption as atmospheric levels of CO2 have risen, plants have responded by decreasing the density of stomates on their leaves (Pagani, Freeman & Arthur, 1999; Beerling, McElwain & Osborne, 1998). In this study Retallack has used the relationship, which is well-known, between concentration of CO2 and stomatal density as a palaeobarometer of atmospheric CO2 during the growth of fossil plant leaves. Stomatal density has, however, been related to differences in insolation, water stress and the position of stomata within leaves, which also affect the size of the cell, and thus stomatal density. The use of stomatal index (percentage stomata over stomata plus epidermal cells), instead of stomatal density, stomata per unit area (Berner, 1997), in order to minimise the effects of these competing variables: the habitats of fossil leaves that have been studied, such as mostly humid, lowland, fluvial and swamp environments that are suitable for the preservation of plant cuticles (Meyen, 1987), have been suggested to further limit such competing variables. These plants grew in substrates that were peaty, nutrient-poor and siliceous, underwent periods of flood disturbance and were in open vegetation early in the ecological succession following disturbance, as is indicated by sedimentological and taphonomic studies of several of the studied fossil sites (Meyen, 1987; Retallack, 1999). The differences between trees of different sex in Gingko: trees that produce pollen have branches reaching upwards, and are therefore lit evenly, whereas horizontal branches tend to shade branches that are lower down in trees that produce ovules. Counts of stomatal index obtained from SEM images of leaves from a male and female tree that were recovered from sites in Eugene, Oregon, show identical stomatal index (mean SI was 8.6 ± 0.4; Fig. 1).

The different SI among species growing in the same location is the most serious obstacle to the use of SI as a CO2 palaeobarometer. Application of this technique have, therefore, focused on particular species such as Quercus patraea, which is known in the fossil record as far back as the Miocene, 10 Ma (Kürschner et al., 1996). In this paper the Ginkgo genus has primarily been used, which has a fossil record of leaf cuticles that are swell preserved back to at least the Triassic period 229 Ma (Anderson & Anderson, 1989). The Pteridosperm genera Lepidopteris (Permian to Triassic), Tatarina (Permian) were chosen for several reasons. Where they occur in the same places they have similar SI (7.6 ± 1.4) and Lepidopteris africana SI (6.5 ± 1.7) at Little Switzerland (Late Triassic, South Africa (Anderson & Anderson, 1989)), and Tatarina conspicua (SI 5.5 ± 1.1) and Lepidopteris sp. (SI 6.3 ± 0.9 at Baizovka (Late Permian, Russia (Gomankov & Meyen, 1986). Furthermore, there is a relationship between ginkgoes and Peltasperm seed ferns, as has been revealed by analysis of their fossilised reproductive structures (Mayen, 1987). All of the 4 genera have stomata that are morphologically similar. The cuticles of their cells are darker (because they are thicker) than are those of other epidermal cells. Subsidiary cells that are arranged in an even ring each have a hollow papilla that is oriented in such a way that they occlude the stomatal pit. This kind of stomate is also found in cycads such as Dioon, and conifers such as Torreya, Agathis and Thuja (Florin, 1931). A degree of functional equivalence is given by such partial papillar occlusion to the stomatal apparatus of Ginkgo, Lepidopteris, Tatarina and Rachiphyllum that is absent in other fossil genera such as Dicroidium, Pachypteris or Baiera (Anderson & Anderson, 1989).

Illustrations of fossil cuticle that had been published previously were used in this study, except for the 2 species from the Triassic (250-245 Ma) from near Sydney Australia. A current geological time scale provided the numerical ages of the fossil cuticles (Gradstein et al., 1995), but published accounts of the cuticles provided relative age. It was demonstrated by the compilation of Retallack that SI in Permian (275-296 Ma) Rachiphyllum and Lepidopteris was comparable with Neogene (0-8 Ma) Ginkgo. Such high SI (>9) is a reflection of low levels of atmospheric CO2 that was predicted for the Permian by models of sedimentary mass balance (Berner, 1997). From the role of CO2 as a greenhouse gas cool climates would be predicted (Beerling et al., 1998) and the Neogene and the Permian are both known from evidence of tillites, stratified pavements, and debris that had been ice-rafted to be times when there were large polar ice caps (Retallack, 1999). It is indicated by these 2 endpoints that plants in the Permian were similar to modern ones in their relationship between stomatal index and atmospheric CO2, and there has been no long-term drift in this proxy measure of atmospheric CO2.

It has been found that fossil plant CI can be calibrated to levels of atmospheric CO2 from greenhouse experiments that used Ginkgo biloba (Beerling, McElwain & Osborne, 1998). Ginkgo biloba herbarium specimens that ranged in age back to 1888 were also counted for stomatal index and levels of CO2 for each taken from global estimates for their year of collection (Kürschner et al., 1996). It is indicated by herbarium and greenhouse results, together with additional greenhouse studies that are in progress (Royer, Berner & Hickey, 2000), that the response of SI to concentrations of CO2 is markedly nonlinear. Greenhouse (Meyen, 1987) and herbarium data were fitted with a quadratic equation for the concentration of CO2: [CO2] I = 155.17 (SI)2 – 2897.8 (SI) + 13773, where r2 = 0.96 (adjusted to sample size of 8). Retallack used this transport function in order to estimate past partial pressures of CO2 from individual measurements of SI, and smoothed by a 5-point moving average to give a curve comparable to past estimates of atmospheric CO2.

According to Retallack the comparison of the new results with estimates from the mass balance model is good, given the coarse resolution (10 Myr intervals) of the model in comparison with the present study (mean sample gap = 3.6 ± 5.4 Myr). Also, high CO2 levels shown for the cuticular estimates during the early Tertiary comparable to those by high-resolution (mean sample gap 1.8 ± 2.7 Myr) boron isotope studies of marine foraminifera, though if correction for secular changes of riverine boron input into the ocean was carried out, these estimates could be reduced substantially (Lemarchand et al., 2000). Evidence from petrographic and chemical studies of palaeosols (Retallack, 1999; Retallack, Bestland & Fremd, 2000), the migration of thermophilic organisms to high latitudes (Thomas, Zachos & Bralower, 2000; Retallack, 1999), and oxygen isotope analysis of marine skeletal remains from different palaeolatitudes (Veizer, Godderis & Francois, 2000; Thomas, Zachos & Brawler, 2000), support the increased warmth and weathering that occurred during times when high CO2 has been demonstrated in this study. There are also indications from palaeosols and palynology that times of high CO2 were times when the palaeoclimate was humid and, conversely, that times of cool intervals were dry. E.g., the abundance of fossil pollen that is believed to be evidence of aridity in many parts of Asia (Vakrameev, 1991), there are rises at all 4 times of low CO2 and cooling during the Jurassic and Cretaceous. From the Eocene to the Oligocene the low CO2 and cooling, which is indicated by the cuticular record, there is evidence for synchronous palaeoclimatic drying in the palaeosol record of central and western North America (Retallack, Bestland & Fremd, 2000).

Many transient excursions to very high CO (>2,000 ppmv) are shown by the cuticular time series. It is indicated by high resolution studies of the SI minimum in fossil Ginkgo and cycad leaves across the Triassic-Jurassic (200 Ma) in Greenland and Sweden that a transient spike in CO2 was coincident with excursion to carbon that was isotopically lighter (δ13Corg) of the same leaves, and mass extinction (That claimed Lepidopteris among others) (McElwain, Beerling & Woodward, 1999).  Following mass extinction of the earliest Triassic (250 Ma), and faunal overturn of the early Jurassic (190 Ma), Early Cretaceous (117 Ma), and Late Palaeocene (55 Ma), other CO2 and carbon isotopic transients, have been related to catastrophic outbursts of methane that was isotopically light from marine permafrost and reservoirs of marine methane hydrate (Thomas, McElwain & Bralower, 2000; Krull, Retallack, Campbell & Lyon, 2000; Jahren et al., 2001). At the Cretaceous-Tertiary boundary there was yet another spike of CO that coincided with an excursion to carbon isotopes values that were lighter, on land as well as in the sea, asteroid impact in Yucatan, which claimed dinosaurs and ammonites (Arens & Jahren, 2000). In the cuticular time series transient CO2 maxima represent strong perturbations in the carbon cycle.

Between cuticular estimates of atmospheric CO and estimates that are based on the carbon isotopic composition of palaeosol carbonates, which indicate CO2 minima (which includes even concentrations that are negative) at 250 Ma (earliest Triassic, 190 Ma (Early Jurassic), 117 Ma (Early Cretaceous) and 55 Ma (Latest Palaeocene (Ekart, et al., 1999). According to Retallack the timing of these events may be a clue to the timing of these discrepancies, as carbon isotopic studies indicate that these were times of catastrophic release of methane that was isotopically light from permafrost and marine gas-hydrate reservoirs (Thomas, Zachos & Bralower, 2000, Krall et al., 2000; Jahren & Arens, 2001). Methane is oxidised in 2-7 years to carbon dioxide once it is in the atmosphere, which retains the unusual isotopic signature of gas-hydrate reservoirs that are isolated from global surficial systems (Khalil (ed.), 2000). The atmospheric addition of methane isotopes that are isotopically light may also explain why the palaeosol isotopic palaeobarometer (Ekart et al., 1999) and high-resolution carbon isotopic studies of deep oceanic organic matter (Pagani, freeman & Arthur, 1999) failed to detect the high CO2 levels and the warm palaeoclimate of the Middle Miocene, which is evident from the few Miocene results that were presented in this paper, as well as from studies of foraminifera (McGowran & Li, 1997), plants (Utescher, Mossbrugger & Ashraf, 2000), palaeosols (Schwartz, 1997) and the oxygen isotopic composition of marine shells (Veizer, Godderis & Francois, 2000; McGowran & Li, 1997). The observation that carbon isotopic palaeobarometers fail to show high CO2 at times of warm palaeoclimate has led to the suggestion that climatic warming and increasing atmospheric CO2 concentrations were decoupled (Veizer, Godderis & Francois, 2000; Cowling, 1999). If carbon isotopic CO2 palaeobarometers are compromised by episodic methane bursts, as suggested by this study, then the accepted role of CO2, methane and water vapour remains intact for at least the past 300 Myr (Veizer, Godderis & Francois, 2000). 


Retallack, G. J. "A 300-Million-Year Record of Atmospheric Carbon Dioxide from Fossil Plant Cuticles." Nature 411 (// 2001): 287-90.


Author: M. H. Monroe
Last updated 06/11/2020

Journey Back Through Time
Experience Australia
Aboriginal Australia
National Parks
Photo Galleries
Site Map
                                                                                           Author: M.H.Monroe  Email:     Sources & Further reading