Oxygen isotope compositions of phosphate from Middle Miocene–Early Pliocene marine vertebrates of Peru

https://doi.org/10.1016/j.palaeo.2008.04.001Get rights and content

Abstract

Phosphatic remains of marine vertebrates recovered from five fossil sites of the Pisco Formation ranging from the latest Middle/earliest Late Miocene (Ca 11–13 Ma) to the Early Pliocene (Ca 3.5 Ma) have been analysed for their oxygen isotope compositions (δ18Op). Coexisting seals, dolphins, whales, penguins and sharks from each locality have distinct δ18Op values reflecting ecology and physiology differences, ranging from 18.2‰ to 21.4‰ for marine mammals, from 19.5‰ to 21.5‰ for marine birds and from 20.9‰ to 23.1‰ for sharks. Systematic offsets observed between dolphin teeth and bones as well as between dolphin and whale bones indicate that the fractionation equation established by using data from extant cetaceans may not be directly applicable to Miocene cetaceans in order to estimate water δ18Ow values. Assuming that polar ice-caps were not totally developed during this time interval, marine palaeotemperatures ranging from 13.0 ± 1.3 °C to 17.2 ± 1.3 °C were estimated. Comparison of our results with those obtained in other World's areas suggests that the oxygen isotope ratios of Pisco vertebrates reflect the influence of both global and local events, such as the setting of the Atacama Desert, the cold Humboldt Current or the global phases of ice-cap growth and decay.

Introduction

Quantitative reconstitutions of temperatures and oxygen isotope compositions of past seawater (δ18Osw) mainly rely on the oxygen isotope analysis of skeletal phosphate (δ18Op) and carbonate (δ18Oc) secreted by marine organisms. As the oxygen isotope fractionation between these biominerals and water is temperature-dependent, the δ18Op or δ18Oc value of aquatic ectothermic organisms (such as marine invertebrates and most fish) both reflect environmental temperature and water composition whereas endothermic organisms (marine mammals) provide δ18Ow estimates. Temporal variations in both temperature and δ18Ow value of sea or fresh waters can be tracked by analysing the phosphatic tissues of both coexisting marine mammals and fish or reptiles and fish (Barrick et al., 1993, Lécuyer et al., 1996, Barrick et al., 1999) using the fractionation equations established between phosphate and water for cetaceans (δ18Op = 0.773 δ18Ow + 17.8; Yoshida and Miyazaki, 1991), fish (T °C = 113.3  4.38 (δ18Op  δ18Ow); Kolodny et al., 1983), turtles (δ18Ow = 1.01 δ18Op  22.3; Barrick et al., 1999) or crocodilians (δ18Ow = 0.82 δ18Op  19.13; Amiot et al., 2007). It is noteworthy that using cetaceans is not possible for periods older than the Eocene when these marine mammals appeared (Fordyce, 1994).

The Miocene and Pliocene were periods of great changes in the Earth's global climatic regime marked by the development of polar ice-caps and by the progressive global cooling that followed the Middle Miocene Climatic Optimum (see Zachos et al., 2001 for a review). Oxygen isotope compositions of phosphatic remains from coexisting cetaceans and fish have been used as proxies of thermal changes and ice volume fluctuations (Barrick et al., 1992, Barrick et al., 1993). Surprising conclusions were drawn from their study of fossil samples recovered from Miocene deposits of the Chesapeake Bay (North America, Atlantic coast). Assuming that the studied samples were not diagenetically altered, Barrick et al. (1992) applied the oxygen isotope fractionation equation established for modern cetaceans (Yoshida and Miyazaki, 1991) to Miocene porpoises and whales. They obtained unrealistically high δ18Ow values ranging from + 2 to + 4.7‰. They also observed a positive correlation between estimated marine temperatures and δ18Ow values, meaning that warmer marine conditions prevailed during polar glaciations and conversely cooler temperatures during ice-cap melting. Moreover, the inferred variations in δ18Ow values suggest larger volumes of polar ice involved during the Miocene growth and decay stage than previously suspected (Haq et al., 1987, Miller et al., 1991). These results raised several questions concerning the meaning of the oxygen isotope compositions recorded in Miocene cetacean bones from Chesapeake Bay. Do they reflect global palaeoenvironmental conditions or local ones (Barrick et al., 1992, Barrick et al., 1993)? To which extent is the equation established by Yoshida and Miyazaki (1991) applicable to any fossil cetacean given the excessively high δ18Ow values estimated from the δ18Op values of their Miocene counterparts?

The fossil sites of the Pisco Formation (Peru, Pacific coast) have yielded rich and well-preserved marine vertebrate fauna of Miocene and Pliocene ages that include cetaceans (porpoises, whales), pinnipeds (seals), birds (among others, penguins), turtles, crocodiles and sharks. Fossil remains of these animals having various ecologies and physiologies have been analysed for their oxygen isotope contents in order to estimate the variations in marine temperatures and ice volume changes as well as to search for answers to questions raised by the results provided by Barrick et al. (1993)'s study.

Section snippets

Geological settings

The Pisco Formation consists of marine Neogene deposits located along the southern coast of Peru, and is known for its abundant marine vertebrate fauna. The Pisco Formation extends about 300 km from the city of Pisco south to Yauca (Fig. 1) and is about 640 m thick (Brand et al., 2003). Its geology and palaeoecology in the Sacaco area were studied by Muizon and DeVries (1985). The age of the sediments is constrained by vertebrate and molluscan biostratigraphies and radioisotopic dating. Samples

Sample collection

Phosphatic remains of marine vertebrates (whales, dolphins, seals, penguins, marine sloths, turtles, crocodilians and sharks) recovered from five fossil sites of the Pisco Formation were collected, cleaned and analysed for their oxygen isotope compositions. Except for small teeth for which bulk analyses were performed as well as two aquatic sloth teeth covered with durodentine, enamel was preferentially selected and sampled with a microdrill. Cortical regions of various bones which are the most

Analytical techniques

Measurements of oxygen isotope ratios of apatite consist of isolating phosphate ions using acid dissolution and anion-exchange resin, according to a protocol derived from the original method published by Crowson et al. (1991) and slightly modified by Lécuyer et al. (1993). Silver phosphate is quantitatively precipitated in a thermostatic bath set at a temperature of 70 °C. After filtration, washing with double deionised water, and drying at 50 °C, 8 mg of Ag3PO4 are mixed with 0.5 mg of pure

Results

Oxygen isotope measurements of tooth and bone phosphate are reported in Table 1. The whole δ18O dataset ranges from 17.5‰ to 23.1‰. For each of the five localities sampled, mean δ18Op values of teeth and bones for each taxonomic group are plotted against their relative age in Fig. 2. Significant isotopic differences are observed at any given locality between the various groups of vertebrates, with ranges from 18.2‰ to 21.4‰ for marine mammals, 19.5‰ to 21.5‰ for marine birds and 20.9‰ to 23.1‰

Pristine oxygen isotope conservation of fossil apatite

Secondary precipitation of apatite and isotopic exchange during microbially-mediated reactions may scramble the primary isotopic signal (Blake et al., 1997, Zazzo et al., 2004). However, apatite crystals that make up tooth enamel are large and densely packed, and isotopic exchange under inorganic conditions has little effect on the oxygen isotope composition of phosphates even at geological time scales (Kolodny et al., 1983, Lécuyer et al., 1999). Although no method is available to demonstrate

Conclusion

The study of the oxygen isotope compositions of marine vertebrates from Miocene and Pliocene localities of the Pisco formation leads to the following conclusions:

  • Coexisting marine mammals, reptiles, birds and fish have different δ18Op values as a result of different ecologies and physiologies. Fish and reptiles tend to have higher δ18Op values than mammals and birds due to their lower body temperature, and birds have higher δ18Op values than marine mammals due to their higher metabolic rates.

Acknowledgements

The authors would like to thank E. Frey from the Staatliches Museum für Naturkunde Karlsruhe, Germany (SMNK) for providing samples of the Miocene dolphin Tursiops oligodon. Thomas Tütken and an anonymous reviewer provided critical reviews that greatly helped us to improve the manuscript. U. G. was supported by a fellowship in the Feodor-Lynen-Program by the Alexander von Humboldt Foundation.

References (56)

  • LécuyerC. et al.

    Thermal excursions in the ocean at Cretaceous–Tertiary boundary (northern Morocco): δ18O record of phosphatic fish debris

    Palaeogeography, Palaeoclimatology, Palaeoecology

    (1993)
  • LécuyerC. et al.

    Deciphering “temperature” and “salinity” from biogenic phosphates: the δ18O of coexisting fishes and mammals of the Middle Miocene sea of western France

    Palaeogeography, Palaeoclimatology, Palaeoecology

    (1996)
  • LécuyerC. et al.

    Oxygen isotope exchange between dissolved phosphate and water at temperatures < 135 °C: inorganic versus biological fractionations

    Geochimica et Cosmochimica Acta

    (1999)
  • LécuyerC.

    Stable isotope composition and rare earth element content of vertebrate remains from the Late Cretaceous of northern Spain (Lano); did the environmental record survive?

    Palaeogeography, Palaeoclimatology, Palaeoecology

    (2003)
  • LonginelliA. et al.

    Revised phosphate-water isotopic temperature scale

    Earth and Planetary Science Letters

    (1973)
  • VennemannT.W. et al.

    Isotopic composition of recent shark teeth as a proxy for environmental conditions

    Geochimica et Cosmochimica Acta

    (2001)
  • WilliamsT.M. et al.

    A killer appetite: metabolic consequences of carnivory in marine mammals

    Comparative Biochemistry and Physiology; Part A

    (2001)
  • ZazzoA. et al.

    Experimentally-controlled carbon and oxygen isotope exchange between bioapatites and water under inorganic and microbially-mediated conditions

    Geochimica et Cosmochimica Acta

    (2004)
  • BarrickR.E. et al.

    Paleotemperatures versus sea-level — oxygen isotope signal from fish bone phosphate of the Miocene Calvert Cliffs, Maryland

    Paleoceanography

    (1993)
  • BarrickR.E. et al.

    Cetacean bone oxygen isotopes as proxies for Miocene ocean composition and glaciation

    Palaios

    (1992)
  • BarrickR.E. et al.

    Oxygen isotopes from turtle bone: application for terrestrial paleoclimates?

    Palaios

    (1999)
  • BillupsK. et al.

    Paleotemperatures and ice volume of the past 27 Myr revisited with paired Mg/Ca and 18O/16O measurements on benthic foraminifera

    Paleoceanography

    (2002)
  • BrandL.R. et al.

    Stratigraphy of the Miocene/Pliocene Pisco Formation in the Pisco Basin, Peru

  • BrochuC.A. et al.

    A gavialoid crocodylian from the Lower Miocene of Venezuela

  • BushM. et al.

    A bleeding technique for nonpalpable vessels in anesthetized two-toed sloths (Choloepus didactylus)-plus hematological data

    Journal of Zoo Animal Medicine

    (1979)
  • CappettaH.
  • CrowsonR.A. et al.

    A method for preparation of phosphate samples for oxygen isotope analysis

    Analytical Chemistry

    (1991)
  • FordyceR.E.

    The evolutionary history of whales and dolphins

    Annual Review of Earth and Planetary Sciences

    (1994)
  • Cited by (35)

    • Strontium Isotope Stratigraphy and the thermophilic fossil fauna from the middle Miocene of the East Pisco Basin (Peru)

      2020, Journal of South American Earth Sciences
      Citation Excerpt :

      It should be noted that the deposition of the P0 strata likely took place during the last phases of the Middle Miocene Climatic Optimum (= MMCO, the last major warming interval of the Cenozoic, occurred between 17 and 14 Ma; Loughney et al., 2019, and references therein) or the onset of the Middle Miocene Climatic Transition (= MMCT, the subsequent interval of gradual change towards cooler climatic conditions), i.e., during a period of globally high temperatures relative to the modern. As reported above, the warm, tropical paleoenvironment here reconstructed for the P0 sequence sets it apart from the remainder of the Pisco Formation, which is thought to reflect a cooler setting (e.g., Dunbar et al., 1990; DeVries and Frassinetti, 2003; Amiot et al., 2008; Di Celma et al., 2017). In particular, oxygen-isotope analyses on phosphatic remains of marine vertebrates from upper Miocene horizons of the Pisco Formation exposed in the Ica Desert (i.e., the CLB vertebrate level of Muizon and DeVries, 1985; referred to the late Miocene according to Di Celma et al., 2017) and the Sacaco area (i.e., the ELJ, AGL, SAS and SAO vertebrate levels of Muizon and DeVries, 1985; late Miocene in age according to Ehret et al., 2012) have revealed marine paleotemperatures that, on the whole, match those observed today off the coasts of Peru (Amiot et al., 2008).

    • Refining the temperature dependence of the oxygen and clumped isotopic compositions of structurally bound carbonate in apatite

      2019, Geochimica et Cosmochimica Acta
      Citation Excerpt :

      The low δ18OH2O value of C. megalodon sample (−1.3 ± 0.9‰) is consistent with oceanic δ18O values in a world with only small ice shields (Miller et al., 2005; their Fig. 1; Lear et al., 2000; their Fig. 1E). It is apparent that the reconstructed δ18OH2O and Δ47-based temperature fit to the respective models from partly cosmopolitan marine vertebrates of Amiot et al. (2008) during the late Miocene. Strontium isotopes analysis of C. megalodon tooth enameloid yields an age of 5.75 ± 0.06 Ma (87Sr/86Sr = 0.709006 ± 0.000013; McArthur et al., 2001).

    • Massive middle Miocene gypsic paleosols in the Atacama Desert and the formation of the Central Andean rain-shadow

      2019, Earth and Planetary Science Letters
      Citation Excerpt :

      We agree that a cold Humboldt Current and strong temperature inversion along the Pacific Coast is essential for hyperaridity in the Atacama Desert, but argue that the extreme hyperaridity over tectonic time scales (106 yr) as indicated by the massive gypsic paleosols during the late middle Miocene could only be maintained with a strong Central Andes rain-shadow in place. Marine records from the southeastern Pacific suggest that cold water and upwelling along the west coast of South America has been a persistent feature of ocean circulation during the late Cenozoic (Amiot et al., 2008; Ibaraki, 1997). There is also good evidence for significant cooling of the eastern equatorial Pacific between ∼14 to 12 Ma following the Miocene Climatic Optimum (16.8–14.7 Ma; Holbourn et al., 2014).

    • More than ten million years of hyper-aridity recorded in the Atacama Gravels

      2018, Geochimica et Cosmochimica Acta
      Citation Excerpt :

      The rise could have created an enhanced rain shadow effect that blocked moisture entry from the continental interior in the east to the Atacama (Garzione et al., 2014). During the same time, the Humboldt cold current off the west coast of South America was also strengthened due to the cold polar current linked to an expansion of Antarctic Ice Sheet (Amiot et al., 2008). This reinforced cold water upwelling further prevented precipitation along the coastal region of South America (Garreaud et al., 2010).

    View all citing articles on Scopus
    1

    Now at the Institute of Vertebrate Paleontology and Paleoanthropology, Chinese Academy of Sciences, #142 XiZhiMenWai DaJie, Beijing 100044, China.

    2

    Now at the Natural History Museum Vienna, Geological–Paleontological Department, Burgring 7, A-1010 Vienna, Austria.

    3

    Also at Institut Universitaire de France.

    View full text