Aspects of Ocean History

MA03CH01-Prefatory-Berger ARI 17 November 2010 6:37

Geologist at Sea: Aspects of Ocean History

Wolfgang H. Berger

Scripps Institution of Oceanography, University of California, La Jolla, California 92093-0244; email: [email protected]

Annu. Rev. Mar. Sci. 2011. 3:1–34 Keywords First published online as a Review in Advance on ocean history, deep-sea sediments, Quaternary, Cenozoic, whale evolution, October 26, 2010

by Dr John Klinck on 07/15/11. For personal use only. climate history The Annual Review of Marine Science is online at marine.annualreviews.org Abstract This article’s doi: Ocean history is largely read from deep-sea sediments, using microscopic 10.1146/annurev-marine-120709-142831 Annu. Rev. Marine. Sci. 2011.3:1-34. Downloaded from www.annualreviews.org fossils, notably foraminifers. Ice age fluctuations in the ocean’s sediments Copyright c 2011 by Annual Reviews. provided for a new geologic understanding of climate change. The discovery All rights reserved of rapid decay of ice masses at the end of glacial periods was especially 1941-1405/11/0115-0001$20.00 important, yielding rates of sea level rise reaching values of 1 to 2 m per century for millennia. Thanks to deep-ocean drilling, the overall planetary cooling trend in the Cenozoic was recognized as occurring in three large steps. The first step is at the Eocene–Oligocene boundary and is marked by a great change in sedimentation patterns; the second is in the middle Miocene, associated with a major pulse in the buildup of Antarctic ice masses and the intensification of upwelling regimes; and the third is within the late Pliocene and led into the northern ice ages. Evolution in the sea is linked to these various steps.

1 MA03CH01-Prefatory-Berger ARI 17 November 2010 6:37

INTRODUCTION In recent decades, geologists have learned to see geologic history in entirely new ways. Whatever was taught in freshman geology in the 1950s is now considered wrong in large parts. Those of us who were students in the 1960s realized that a revolution was under way, as we watched distinguished professors struggle with new ideas. Of the new ideas that became the new dogma as a result of drilling, we must mention orbitally driven climate cycles, the great cooling steps in the Cenozoic, the oxygen-stressed seas in the Cretaceous, and especially, plate tectonics and its ramifications. My own work, which informs the main areas of focus for this essay, has been largely concerned with two of these items, that is, certain aspects of the reconstruction of the ice ages (time scale measured in terms of 10,000-year steps) and of Cenozoic cooling steps (million-year time scale), with much attention to associated changes in the operation of the marine carbon cycle. The intended audience for this essay is not my colleagues in paleoceanography—many of them know more than I do about the topics discussed. But I do hope that my colleagues in related fields might find things of interest in this review, and perhaps it might serve as a useful introduction to some of the problems arising in paleoceanography, for graduate students and for college teachers concerned with the subject. Much background on the various thoughts here discussed may be found in my recent book Ocean, from which I have drawn some of the material here presented (Berger 2009).

ON THE RISE OF HISTORICAL STUDIES IN OCEAN SCIENCE Marine geology and marine biology have common origins. The iconic founding hero of this connection was Charles Darwin (1809–1882), who started his career as a geologist on the H.M.S. Beagle. Having Darwin on board made its circumnavigation (1831–1836) famous. To many scientists, Darwin’s book On the Origin of Species, which is commonly linked to observations on the Galapagos Islands during said voyage, is the most important biologic work ever written and certainly the most important scientiﬁc treatise published in the nineteenth century. Some, on the other hand, have found much to criticize, armed with new knowledge gained in the twentieth century.

by Dr John Klinck on 07/15/11. For personal use only. Whatever the merits of such discussions, the fact is clear that Darwin was a geologist first, and that he saw geology and biology and environmental change as complementary aspects of the same thing, that is, natural history. He followed Charles Lyell (1797–1875) in emphasizing the dominant role of observable processes active over incredibly long time spans. Annu. Rev. Marine. Sci. 2011.3:1-34. Downloaded from www.annualreviews.org In spite of Darwin’s iconic stature, his various insights must not be treated as dogma. For example, regarding the origin of atolls, it was the realization that sea level fluctuated over a considerable depth owing to the buildup and decay of major ice sheets that provided a new focus—away from Darwin’s sinking-volcano concept. With sea level many times dropping well below the present sea surface, erosion of exposed reef surface during glacial periods became an important topic (Daly 1934, Purdy & Winterer 2006, Winterer 2009). Likewise, the question about potential rates of upward growth of reefs became urgent, considering that sea level rose at rates of more than 1 m per century for millennia whenever the great ice sheets wasted. John Murray (1841–1914) was another illustrious pioneer with regard to the combination of different fields of knowledge in the pursuit of ocean history. He was among the first to realize that the calcareous sediments on the seafloor are largely produced in surface waters, by shelled plankton. He was present when the Challenger Expedition established that there is no abyssal azoic environment, as was earlier surmised by Edward Forbes (1815–1854), the marine biologist who

2Berger MA03CH01-Prefatory-Berger ARI 17 November 2010 6:37

established that bottom-living organisms prefer their own depth zones. Yet another task brought negative results: Murray and colleagues did not find the “living fossils” that some of the expedition organizers may have hoped for, figuring that the deep sea would be unchanging, hence likely to harbor ancient life forms. Instead, the life forms found in the dredges of the Challenger looked thoroughly modern. We are no longer surprised: The deep ocean has changed markedly in several great steps of cooling of the planet, for the last 40 million years or so. Thus, like everything else on the planet, the organisms of the deep sea are geologically young, adapted to an ever-changing environment. The purely geologic legacy of Murray’s is equally remarkable: The sediments he studied became the means for the detailed reconstruction of ocean history, for the entire Cenozoic and the preceding Cretaceous. ( Jurassic sediments have largely disappeared into the trenches, unless they rest on continental crust.) The nature of these sediments, first clarified by Murray, define the scope of what the ocean can remember. Much has been learned since the time of these pioneers, from an intensive study of long cores, and from drilling into ocean sediments. Analysis of cores from the Swedish Albatross expedition led the way (e.g., Arrhenius 1952, Phleger et al. 1953, Emiliani 1955, Parker 1958, Olausson 1965). In later years, much more material became available from coring expeditions at several oceanographic institutions, especially at Lamont Geological Observatory. The enormous Lamont collection formed the basis for important studies on the nature of glacial periods (e.g., CLIMAP Project Members 1976, Hays et al. 1976, and later work). From 1968, material from deep-ocean drilling became available. Some of the results are summarized in textbooks of the 1980s and 1990s (Kennett 1982, Seibold & Berger 1996), and in various specialty symposia (e.g., Warme et al. 1981, Kennett & Warnke 1992, Summerhayes et al. 1992, Cullen & Clark 1994, Wefer et al. 1996, and references therein). Valuable summaries and highlights are in various workshop reports concerned with the planning of drilling operations (e.g., Mountain & Katz 1991, Baker & McNutt 1996, Kappel & Farrell 1997, Becker et al. 2002). In addition to a voluminous specialty literature and a long series of reports of the expeditions (e.g., Initial Reports of the Deep Sea Drilling Project; Proceedings of the Ocean Drilling Program), there are a few publications aimed at education and outreach (e.g., Hsu¨ 1992). Of special interest to ocean historians are treatises on biostratigraphy and paleoceanography (e.g., Bolli et al. 1985, Gersonde & Hodell 2002, McGowran 2005). Also, much relevant infor-

by Dr John Klinck on 07/15/11. For personal use only. mation may be found in the encyclopedias edited by Steele et al. (2001) and by Gornitz (2009), with numerous entries reﬂecting recent insights into the workings of the sea and its history. The report by Arrhenius (1952) on cores retrieved during the Swedish Deep Sea Expedition established a new paradigm for the study of ocean history: The environment of growth of shelled Annu. Rev. Marine. Sci. 2011.3:1-34. Downloaded from www.annualreviews.org plankton organisms can be reconstructed, and thus its changes can be used to probe the biological response of the ocean to climatic change, in a heuristic approach that simulates experiment— otherwise impossible on time scales beyond the human life span. This paradigm has since been much applied, both with respect to the great cycles of climate change in the Quaternary (e.g., Hays et al. 1976, Imbrie et al. 1984, Berger & Herguera 1992) and with regard to the great cooling steps of the Cenozoic (e.g., Savin et al. 1975, Shackleton & Kennett 1975, Miller et al. 1987, Berger & Wefer 1996, and references therein). For the geologic history beyond the Quaternary, cores raised in steel tubes from regular research vessels have long lost their status as a preferred source of sediment for study. Instead, it was deep-ocean drilling, pioneered by the oil industry, that opened a path to the abyssal memories of the sea (Berger 1979). The Deep-Sea Drilling Project started in the 1960s. It was managed by Scripps Institution of Oceanography. The ﬁrst expedition set out in 1968, in the Gulf of Mexico, with the director of Lamont Geological Observatory, the marine geophysicist Maurice Ewing

www.annualreviews.org • Geologist at Sea: Aspects of Ocean History 3 MA03CH01-Prefatory-Berger ARI 17 November 2010 6:37

Hydrophone

Drill pipe

Max. depth 6,300 m (21,000 ft) Television camera Acoustic Reentry cone beacon

Sediment layers by Dr John Klinck on 07/15/11. For personal use only. Basalt Annu. Rev. Marine. Sci. 2011.3:1-34. Downloaded from www.annualreviews.org Figure 1 Deep-ocean drilling by the drilling vessel JOIDES Resolution, using multiple thrusters to maintain position, controlled by an acoustic beacon on the seaﬂoor and hydrophones in the hull listening to beacon output. The typical sediment thickness is between 100 m and one or several kilometers. Basalt is reached after drilling through the sediment cover on the deep-sea ﬂoor (but not usually in the margins). Illustration by Christie Newman, based on the Ocean Drilling Program.

(1906–1974), as chief scientist. The vessel was the GLOMAR Challenger, a ﬂoating drilling platform in the shape of a ship. The project metamorphosed into the Ocean Drilling Program (ODP) in the 1980s, with management moving to Texas A&M, in College Station, and with drilling done from a bigger and more stable platform (Figure 1). The project ran for many decades and has been called one of the most successful international scientiﬁc operations ever. Its successes, presumably, result mainly from the diligence of dedicated scientists and technicians but also to a

4Berger MA03CH01-Prefatory-Berger ARI 17 November 2010 6:37

number of clever technical and organizational arrangements based on the working knowledge of competent managers and engineers and on some painful lessons learned from earlier attempts at deep drilling into the sea floor (summarized in A Hole in the Bottom of the Sea, by Willard Bascom [1916–2000], published in 1961). Hundreds of kilometers of core material were recovered in numerous expeditions of the GLOMAR Challenger and its successor, the JOIDES Resolution, and stored in a number of reposito- ries. (Each two-month leg of the expedition brought in hundreds of meters to several kilometers, depending on circumstances.) After piston coring was adapted to drilling—an outstanding engineering feat allowing cores to be taken through the drill bit and raised within the drill string— the quality of sediment cores became excellent, matching the best results of traditional coring operations. The workday of scientists on an ODP leg consisted of a shift of 12 hours (and often more) every day at sea. The same was true for their support teams (the ship’s crew, the drilling crew, and the stewards and kitchen staff ). Approximately one-half of the scientists would be from the United States, the others from affiliated countries: the United Kingdom, France, Germany, Japan, Sweden, Norway, Denmark, the Netherlands, Switzerland, Italy, Spain, and several others. Strong bonds of long-term scientific collaboration were formed during these two months, and lifelong friendships emerged among former strangers. In the memory of the participants, such an expedition looms large forever, as a milestone in their careers and their lives. I was fortunate to benefit from participation in both the Deep Sea Drilling Project (Leg 14) and in the Ocean Drilling Program (Legs 130 and 175). Typically, each leg had a specific task, focusing on one or two of four general topics: (1) the regional paleogeography, that is, the horizontal and vertical motions of the Earth’s crust that keep changing the position and the topographic setting of the seafloor, including such complications as triple plate junctions, jumping ridge crests, and clogged subduction zones; (2) the nature of the basaltic basement below the sediment and the messages it bears regarding processes within the Earth’s mantle and at the seafloor spreading centers and regarding the history of volcanism; (3) the processes of exchange between seafloor and the ocean, including alteration of volcanic rocks and sediments, expulsion of solutions from sediments, and production of gas in the thick piles of deposits at the continental margins; and (4) the record of change of climate and of evolution in the ocean, including changes in water temperature at the surface and at depth, circulation and

by Dr John Klinck on 07/15/11. For personal use only. productivity, and extinction and diversification. All these aspects must be studied for a thorough understanding of the history of the ocean and of the life within it. The motions must be reconstructed to get the past geography right (and the associated circulation patterns); the intensity of volcanism bears on the amount of carbon dioxide in Annu. Rev. Marine. Sci. 2011.3:1-34. Downloaded from www.annualreviews.org the atmosphere and hence helps determine the overall climate state of the planet; and the recycling of phosphorus and other nutrients is important for the changing productivity of the ocean. Thus, ocean historians (more commonly labeled paleoceanographers) tend to have the broadest range of interests among the many specialists in a scientific shipboard party. In the 1960s and 1970s, many paleoceanographers started their career as experts in biostratigraphy or sedimentology. The time resolution of deep-sea sediments, even when coring results are excellent, is rarely better than one millennium and commonly worse, owing to the mixing of sediments by burrowing organisms. There is one way to escape this conundrum: It is the use of finely layered deposits found in anaerobic conditions. The absence of oxygen prevents the stirring of the sediments as it makes the site of deposition off-limits to large organisms. Another way to avoid the rather severe abyssal resolution limit is to study the growth bands in corals and the rings in trees influenced by precipitation sent by the sea. Naturally, such high-resolution studies are largely confined to the last several tens of thousands of years and are concentrated on recent history and the late Holocene.

www.annualreviews.org • Geologist at Sea: Aspects of Ocean History 5 MA03CH01-Prefatory-Berger ARI 17 November 2010 6:37

In studies at Scripps, the varved sediments of the Santa Barbara Basin (which is anaerobic at the bottom) became a favorite target for high-resolution studies from the 1980s (e.g., Baumgartner et al. 1989, Schimmelmann & Lange 1996, Berger et al. 2004, Field et al. 2006). These deposits had previously become extremely important in the reconstruction of fish population abundance and hence in gathering insights into the limitations of fishery management. John Dove Isaacs (1913–1980) initiated these studies, working with his erstwhile student, Andrew Soutar (Soutar & Isaacs 1969). Isaacs was a major figure in ocean sciences, a former fisherman with a background in engineering and a strong interest in marine life and the factors controlling its environment. For many years he was a key figure in the multidecadal California Cooperative Fisheries Investigation (CalCOFI) program, and he deeply appreciated the importance of long time-series, including series from sediments. Indeed, the varved sediments in the Santa Barbara Basin turned out to be a veritable treasure chest for many unusual insights on the nature of environmental variations in a coastal ocean setting.

ON THE RECONSTRUCTION OF OCEAN HISTORY FROM BIOGENIC SEDIMENTS The physicist Richard Feynman (1918–1988) is reported as having said “The sole test of any idea is experimental” (Simmons 1996). If that were true, and “experiment” were deﬁned as something devised by humans, all of the historical sciences, including geology and cosmology, would be excluded from the world of ideas. However, history is not subject to experiment—things either happened or they did not. If something happened, it is possible, and no experiment or calculation can prove otherwise. History is the experiment. The power of this truth has been widely recognized in a new kind of modeling of climate change, that is, a testing of programs for bottom-up simulation (based on physics and theory) with top-down simulation (observed patterns in nature). (The assumption here is that hypothesis is at the bottom and reality is at the top of the hierarchy of knowledge, with kudos to the intelligent concept that there is no reality without hypothesis.) Attempts at modeling the last glacial maximum, with the Climate Long-Range Investigation Mapping and Prediction (CLIMAP) data set (CLIMAP Project Members 1976, 1981) as target, are widely noted examples of this new approach (Gates 1976, Manabe & Broccoli 1985, Crowley & North 1991, Broccoli &

by Dr John Klinck on 07/15/11. For personal use only. Marciniak 1996, Shin et al. 2003). The systematic time-series measurement of ocean features related to climate and productivity does not normally extend into periods before the early nineteenth century. Thus, all of our instrumentally documented historical experience, however valuable, suffers from myopia. Also, Annu. Rev. Marine. Sci. 2011.3:1-34. Downloaded from www.annualreviews.org there are few such series to begin with, because monitoring is not commonly appreciated as an important scientific activity. There is therefore a built-in difficulty for hypotheses based on recent experience (commonly concatenated with physics) in dealing with truly unusual events. The same problem arises when applying the doctrine of uniformitarianism, assuming that the present is the key to the past. This idea dominated textbook thinking ever since Lyell, till the final decades of the twentieth century, when the limits to its usefulness emerged from detailed stratigraphic studies. The past turned out to have had major events of a type unknown within the “present” or even within the entire Holocene. Detailed stratigraphic studies on continuous sequences in ice cores, in lake deposits, and in pelagic sediments revealed that abrupt change and catastrophe are real and not just an artifice of a gap-ridden record. Outstanding pioneer efforts include the studies of Dansgaard et al. (1971), regarding the end of the last ice age, and those of Alvarez et al. (1980), regarding the end of the Mesozoic. Within the Cenozoic, the dramatic changes at the end of the Eocene are the

6Berger MA03CH01-Prefatory-Berger ARI 17 November 2010 6:37

subject of many detailed studies (e.g., Kennett et al. 1974, Shackleton & Kennett 1975, Kennett & Shackleton 1976, Pomerol & Premoli-Silva 1986, Prothero & Berggren 1992, DeConto & Pollard 2003, Prothero et al. 2003, Coxall et al. 2005, Zachos & Kump 2005, DeConto et al. 2007). Since 1980, an increased focus on impacts from space is evident (see, e.g., Koeberl & Montanari 2009). The topic of abrupt climate change from threshold dynamics and internal feedback mechanisms attracted increased attention in the 1980s, encouraged by the Belgian astronomer and climatologist Andre´ Berger, as well as by the physicists and climate historians studying ice cores in Grenoble, France, and Berne, Switzerland. But mass extinctions soon took center stage, as the profound change in pelagic fauna and flora at the end of the Cretaceous emerged from the study of pelagic sediments and the mechanism of impact (Alvarez et al. 1980) as the agent of that change gained credence. In all reconstruction of climate and ocean history (or any other type of history), it is difficult to significantly reduce a residual level of uncertainty by increased study of clues. The reason is that at some point additional study almost automatically identifies additional sources of uncertainty rather than just reducing uncertainty for known sources. In a cross-linked system of feedback mechanisms, any additional source of uncertainty immediately affects all others, potentially changing the total in surprising ways. The pervasiveness of uncertainty in complex systems is basically the reason why there are so few so-called one-armed exemplars among environmental scientists (that is, scientists who will forego the phrase “on the other hand”). In what follows, I summarize aspects of the study of pelagic fossils, chiefly planktonic foraminifers, to exemplify the information contained in deep-sea sediments, as well as problems arising in finding the clues for reconstruction. The intent here is to lay the groundwork for developing arguments about history in later sections of this review and warn the reader that all history is tentative, even when this fact is not constantly emphasized. The principles of pattern-matching of biogeographic distributions and physical parameters of the environment—chiefly temperature and nutrient supply—are quite well established (see Fischer & Wefer 1999 for a recent review). The basics are the content of textbooks (e.g., Seibold & Berger 1996). Of course, biostratigraphers have long used abundance patterns of fossils for reconstruction of environments. Everyone knew, since Murray, which fossils are of planktonic origin and which are benthic and that the ratio found in sediment samples says something about the distance

by Dr John Klinck on 07/15/11. For personal use only. from shore (and hence about depth, which increases with distance, and productivity, which decreases with distance). Likewise, the tropical forms in the plankton were known and readily distinguished from the cold-water forms. John Murray of the Challenger Expedition could tell, within a few degrees Celsius, what the temperature of the water might have been in which a given as- Annu. Rev. Marine. Sci. 2011.3:1-34. Downloaded from www.annualreviews.org semblage was formed. Wolfgang Schott of the Meteor Expedition (1925–1927) and later David Ericson at Lamont used the presence and absence of the tropical species Globorotalia menardii to determine glacial and interglacial sections in tropical Atlantic cores (Schott 1935, Ericson & Wollin 1956). Also, when studying deep-sea samples for foraminifers, one cannot miss the fact that some of the benthic foraminifers indicate high overlying production (e.g., Bolivina, Bulimina, and Uvigerina). What sets the more recent studies apart from the earlier ones is an emphasis on quantitative reconstruction, based on quantitative counts of microfossils. Both the German geologist Wolfgang Schott and the American micropaleontologist Frances Parker summarized their fossil inventories in tables showing the percentages of the species noted, in the samples they studied, starting in the 1930s. A mathematical treatment of such counts was for later developments, aided by computing facilities, as is well exempliﬁed by the work of Imbrie & Kipp (1971) and in the subsequent CLIMAP studies and similar work since.

www.annualreviews.org • Geologist at Sea: Aspects of Ocean History 7 MA03CH01-Prefatory-Berger ARI 17 November 2010 6:37

1.5

2.0 Excellent preservation

2.5

3.0 Good preservation

Depth (km) 3.5 Lysocline Poor preservation

4.0

Residual

4.5

Calcite compensation depth

5.0 0 20 40 60 80 100 Calcite dissolved (%) Figure 2 Modification of calcareous fossil assemblages with increased dissolution of carbonate (here, planktonic foraminifers and increased depth in the deep sea). Note the definitions of the dissolution levels called lysocline and calcite compensation depth (CCD), as based on preservation. Sediment becomes darker with depth, as the carbonate content decreases. From Berger & Wefer (2009). by Dr John Klinck on 07/15/11. For personal use only. General problems in calculating aspects of past environments arise in connection with the fact that Pacific and Atlantic plankton distributions are not necessarily congruent, suggesting subtle shifts in the environmental preferences of taxa that are not reflected in morphology (and hence will Annu. Rev. Marine. Sci. 2011.3:1-34. Downloaded from www.annualreviews.org not be reflected in a taxonomic count). The great variability of morphological character in individual plankton species (Parker 1962) underlines yet another source of uncertainty: misidentification. In addition, dissolution of carbonate on the seafloor changes assemblages, thereby removing the more susceptible species (Phleger et al. 1953, Berger 1968, McIntyre & McIntyre 1971, Parker & Berger 1971), so that the link between abundance tables of fossils and the environment of growth becomes distorted for samples taken below a certain depth level, which varies in time and place (Figure 2). The single most important item in the toolbox for the reconstruction of past environments in the Quaternary and in the Cenozoic is the application of oxygen isotopes. It was the Italian American paleontologist and physicist Cesare Emiliani (1922–1995) who pioneered the method in its application to deep-sea sediments. The data produced by Dansgaard & Tauber (1969), on the oxygen isotope composition of Greenland ice, showed that the fluctuations documented by Emiliani (1955) predominantly track ice mass and sea level rather than temperature (as urged by

8Berger MA03CH01-Prefatory-Berger ARI 17 November 2010 6:37

Emiliani). Whatever the exact ratio of these two factors (and there are more than two), the ratio of 18Oto16O turns out to be indispensable for correct dating, and for correct assessment of the situation with respect to climatic conditions. For many purposes, this ratio has become the master variable in all of ocean history, and the details of its origin are commonly ignored. It is generally accepted that it is strongly related to ice mass and also to temperature, an assumption that has proved constructive. In any case, assumptions about the ratio of the importance of ice mass versus temperature have to be verified for each specific time period under consideration. The calcium carbonate that makes up the shells of the signal carriers also has carbon in addition to oxygen. It turned out that the carbon isotopes have useful information about the environment of growth of the organisms whose shells are being studied (Broecker 1973). Specifically, the δ13Cof the carbonate of certain pelagic foraminifers reflects the composition of the surface water layer in which the shells were grown. A high content of 13C indicates surface waters stripped of nutrients, whereas a low content indicates waters that are relatively rich in nutrients (and hence in dissolved carbon dioxide). The gradient in carbon isotopes between surface layer and thermocline (captured in the appropriate species) reflects the intensity of biological pumping, which is the process moving carbon out from the productive layer into deeper waters, and which decreases carbon dioxide in the surface layer and increases 13C relative to 12C. To the extent that biological pumping is involved in controlling the carbon dioxide of the atmosphere, the 13C record holds clues to changes in the atmospheric content of carbon dioxide. However, the carbon isotope ratio found in shells is the result of a considerable number of processes, some having to do with preferences of the various shell-building organisms. Thus, the interpretation of changing carbon isotope ratios is commonly quite difficult, and it cannot be assumed that what is proposed in any given report is in fact correct (Wefer & Berger 1991). Not all reconstruction, naturally, is based on the study of individual fossils. Important clues come from the overall chemical nature of the sediments. For example, the carbonate content of deep-sea sediments contains information about the state of deep-water chemistry at the time of deposition (as water high in carbon dioxide tends to dissolve carbonate unless there is sufficient compensation from high alkalinity). Likewise, the content of opal (made of the skeletons of diatoms and radiolarians, mostly) reflects the intensity of pelagic diatom production of a region, along with the saturation of deep waters and waters in contact with the sediment, with respect to silica.

by Dr John Klinck on 07/15/11. For personal use only. The type of important clues to be garnered from information of the distribution of pelagic sediments may be appreciated when considering that the deep North Atlantic seafloor has much carbonate and little silica, whereas the reverse is true for the North Pacific. This asymmetry is generated by the fact that there is much deepwater production in the North Atlantic but not in Annu. Rev. Marine. Sci. 2011.3:1-34. Downloaded from www.annualreviews.org the North Pacific. In turn, the North Atlantic deepwater production is linked to heat transport from the tropical Atlantic into high northern latitudes. Thus, the asymmetries of deep-sea deposits between ocean basins (that is, the differences in the elevation of the CCD and the differences in the abundance of opal) have implications for the patterns of global heat transport. Roughly, the position of the CCD depends on the apparent oxygen utilization, that is, the amount of oxygen in deepwater that has been used up in making carbon dioxide. Because of this relationship, fluctuations of the CCD through time are useful in the reconstruction of the marine carbon cycle and provide (some) clues to changes in the sharing of carbon dioxide between ocean and atmosphere. Quite generally, the deposition of biogenic sediments is crucial for the definition of the machinery of global biogeochemical processes. Overall, calcareous nannofossils are the dominant component of calcareous ooze, especially in the large regions of low productivity below the central gyres. Thus, the typical biogenic de- posit on the planet is coccolith ooze. The study of the minute coccoliths (platelets shed from

www.annualreviews.org • Geologist at Sea: Aspects of Ocean History 9 MA03CH01-Prefatory-Berger ARI 17 November 2010 6:37

coccolithophorids) requires the use of powerful microscopes. The invention and spread of scan- ning electron microscopes greatly aided in producing inventories and detailed descriptions of these fossils (e.g., Winter & Siesser 1994). The pioneers (M.N. Bramlette, M. Black, D. Bukry, J.A. Crux, G. Deflandre, A. Farinacci, S. Gartner, B.U. Haq, W.W. Hay, E. Kamptner, H. Lohmann, H. Manivit, E. Martini, A. McIntyre, C. Muller,¨ D. Noel,¨ H. Okada, K. Perch-Nielsen, P. Reinhardt, P.H. Roth, H. Stradner, H.R. Thierstein, and J.W. Verbeek) routinely used petro- logic microscopes, and these instruments are still the preferred tools for shipboard analysis, for dating of sediments during drilling expeditions. The presence of easily recognized nannofossils known as discoasters indicates an age greater than Quaternary, for example. Discoasters died out at the end of the Pliocene. The process of seafloor spreading slowly moves calcareous sediments to greater depth, on a million-year time scale. Thus, as a general rule, deep-sea sediments overlying basaltic basement (generated at depths near 2.5 km at the spreading center) should be rich in carbonate, whereas sediments overlying these older deposits should have less carbonate, having been deposited at greater depth. Reality is more complicated because of large fluctuations of the CCD through geologic time (Berger & Winterer 1974).

ON THE LAST GLACIAL MAXIMUM AND ITS END The discovery of the ice ages began with the invention of the “Great Ice Age” by Louis Agassiz (1807–1873) in the first half of the nineteenth century (cf. Emiliani 1992, p. 569). His ideas were shaped by the interpretation of the frozen remains of large mammals found in Siberia, first noted by Georges Cuvier (1769–1832), the leading vertebrate paleontologist (and zoologist) of his time. It seems quite clear that a concept of a former ruling of northern environments by snow and ice is thousands of years old—and that its origin was apparently tightly linked to the remains from mammoths and other Quaternary megafauna (Berger 2007b). Similarly, the rise of the Great Ice Age concept among naturalists and geologists of the nineteenth century was linked largely to paleontology, making the mammoth an iconic symbol of the Quaternary. Yes, the explanation of erratics (large rocks transported over long distances) surely played some role as well (see Imbrie & Imbrie 1979, Hsu¨ 1992). But I think, for Agassiz, the erratics confirmed what he surmised from megafauna fossils, after having worked with Cuvier.

by Dr John Klinck on 07/15/11. For personal use only. It is perhaps surprising that the modern discovery of the ice age took so long, in that so many of the grand geographic features of the planet—including the fjords of Norway, the jagged peaks of the Alps and the Sierra Nevada, and even the Great Barrier Reef—are legacies of former glaciation. In the case of fjords and reefs, it is relevant that approximately 900,000 years ago small ice age Annu. Rev. Marine. Sci. 2011.3:1-34. Downloaded from www.annualreviews.org fluctuations gave way to big ones (Pisias & Moore 1981). From that time, glacial-period ice buildup was large enough to lower sea level by more than 100 m. Great ice tongues flowing out from the enormous ice caps of Scandinavia eroded the fjord valleys, which then filled with seawater upon the retreat of those mighty glaciers. And the repeated rise and fall of sea level produced the coral rubble that built the Great Barrier Reef, abetted by a rerouting of warm equatorial water to the south due to the clogging of the Indonesian escape to the west (Berger & Wefer 2003). Fjords and jagged mountains and reefs are impressive features on our planet, and worthy witnesses of the past ice ages. However, the point must be made that a great multitude of other geographic features can only be understood as legacies of the ice ages. Physical geography focused on present processes is woefully inadequate in explaining the enormous ancient dune fields near Arcachon along the shores of the Bay of Biscay or the daunting walls of Yosemite Valley in the Sierras of California. Only ice age history can explain these features.

10 Berger MA03CH01-Prefatory-Berger ARI 17 November 2010 6:37

The Great Ice Age was a big improvement over the then current ideas on landscape modification, which emphasized the Great Flood of the Bible and Lyell’s “glacial drift,” a process whereby erratic blocks from Scandinavia were delivered by icebergs to a submerged northern Europe. In a sense, Agassiz turned out to be right, whereas Lyell’s proposition proved wrong. Much of North America and all of Scandinavia, including the Baltic Sea and the North Sea, were indeed deeply buried under a huge ice cover that attained a thickness of several kilometers in places. Much of the northern North Atlantic was covered with ice, both sea ice and icebergs calved from the surrounding glaciers. Frozen ground was extensive in middle Europe and deep into the Russian plains. The Alps were covered with ice, sending glacier tongues well into Bavaria, France, and Italy. One can certainly appreciate Agassiz’s nightmarish vision of a permanent Siberian winter and his doubts that a rich megafauna could be supported in that climate. Nevertheless, strictly speaking, the Great Ice Age as envisaged by Agassiz is a figment of the imagination. It never happened the way he thought it did, as a sudden change from a tropical climate to a frozen world. Also, the onset of the ice age had nothing whatsoever to do with the extinction of the mammoth and the woolly rhinoceros. On the contrary, the giant mammals were creatures of the period of ice ages, not its victims. They died at the end of the last ice age (as correctly related by ancient Scandinavian myths). The one great idea that stood the test of time was Agassiz’s insistence that there had been a lot of ice around, not too long ago. It set the stage for climate reconstruction in Earth history. As far as climate history, there were at least two major tasks: (1) to determine how extensive the ice was at its maximum extent, and (2) to determine how ocean circulation and ocean productivity were affected by the changes in climatic conditions associated with the ice fields. Quaternary geologists attacked the first question by mapping moraines and other evidence of former glaciation (e.g., Daly 1934, Denton & Hughes 1981). Their efforts demonstrated that ice had indeed been widespread, in places moving to latitudes near 40◦N. Clues as to the thickness of ice caps emerged from the amount of depression of continental crust under the weight of ice, with the present-day uplift a consequence of the unloading (Peltier 1994). Paleoceanographers attacked the question about changes in the geography of the sea by recon- structing (1) the biogeography of plankton, (2) the preservation patterns of calcareous deposits on the deep-sea floor, and (3) the changes in the isotope patterns of both planktonic and benthic fossils (Berger & Wefer 1996). The first set of clues yields proxies for the calculation of temperature

by Dr John Klinck on 07/15/11. For personal use only. distributions in surface waters, and for identifying regions of upwelling and the outlines of the great subtropical gyres (Imbrie & Kipp 1971; CLIMAP Project Members 1976, 1981; Pflaumann et al. 1996). The second set of clues tracks changes in deep circulation, specifically changes in the lower depth boundary of North Atlantic Deep Water. The third set, isotopic clues, in essence Annu. Rev. Marine. Sci. 2011.3:1-34. Downloaded from www.annualreviews.org provides the means for reconstruction of the physical environment, oxygen isotopes being linked to both temperature and ice mass, and carbon isotopes to upwelling and the apparent oxygen utilization in deep waters. Other chemical proxies are also useful in this respect (see Wefer et al. 1996, Fischer & Wefer 1999). And ice-rafted debris on the seafloor tracks the motion of icebergs. Two major topics of inquiry emerged in these and many other studies (e.g., Flohn 1985, Broecker & Denton 1989, Crowley & North 1991, Boyle 1995, Berger & Wefer 1996, Hewitt et al. 2003): (1) How much telescoping of climate zones toward the equator did the northern hemisphere experience, and how did this influence wind fields, heat transport, and the position of the Intertropical Convergence Zone? (2) Just how did the ice age deep-sea circulation differ from the present one? Mainly, the northern climate zones move toward the equator, along with the Intertropical Convergence Zone and the heat subsidy that the northern hemisphere receives from the southern one is greatly reduced, especially in the Atlantic. Also, the production of North Atlantic Deep

www.annualreviews.org • Geologist at Sea: Aspects of Ocean History 11 MA03CH01-Prefatory-Berger ARI 17 November 2010 6:37

Water is reduced during glacial time, with a corresponding reduction in the Pacific–Atlantic asymmetry of deepwater chemistry. Of other topics, the one addressing the question of why there are sporadic iceberg releases during the glacial maximum (Heinrich 1988) is of special interest. Perhaps not surprisingly, the likelihood of such release seems to be linked to the amount of polar ice available (Schulz et al. 1999); that is, they are typical for maximum glaciation. The releases document the buildup of instability within portions of the ice sheets (MacAyeal 1993), a signal that is important for understanding terminations. Also, this same instability presumably drove the erratic climate patterns that may have helped to reduce Neanderthal populations to the point of no return (Finlayson & Carrion´ 2007). Another important topic is the question concerning the mechanisms of reducing atmospheric carbon dioxide to the level seen in ice cores (Petit et al. 1999). As Broecker (1982) pointed out, such changes have to be mediated by the ocean’s reservoir of carbon, both because of its size and its ability to react rapidly. The problem is unsolved, and not for want of trying. Of the various efforts of constraining the nature of the glacial world, one stands out. It is the systematic mapping done in the late 1960s and early 1970s by a large group of paleoceanographers led by Andrew McIntyre, James Hays, and Neil Opdyke of Lamont Geological Observatory, by Theodore C. Moore of Oregon State University, and by John Imbrie of Brown University. Dubbed the CLIMAP project (an acronym memorable for containing allusions to climate and mapping), the well-coordinated interinstitutional collaboration produced a semiquantitative map for surface- water temperatures of the ice age ocean (CLIMAP Project Members 1976, 1981). As mentioned, the reconstructed temperature patterns were based on statistical correlations linking present-day biogeography of plankton organisms to present-day conditions of growth. Results were somewhat surprising in suggesting relatively modest change for tropical latitudes, changes whose amplitudes had to be revised upward later. Presumably, a number of factors combined to make biogeography change less than would seem appropriate from the change of physical conditions. The map nevertheless reflects the profound difference in the responsiveness of tropical and polar latitudes to large-scale climate change, with high latitudes showing the greater amplitudes. Presumably, this difference is largely a matter of albedo feedback, which involves snow in high latitudes (positive feedback) and clouds in low latitudes (neutral or negative feedback). The question of feedback distribution is exceedingly difficult and is being studied in general circulation models.

by Dr John Klinck on 07/15/11. For personal use only. The mapping of the Last Glacial Maximum (LGM) furthermore suggested increased upwelling along the equator, increased coastal upwelling, and strengthened eastern boundary currents, and only very modest change within the central gyres. The CLIMAP Last Glacial Maximum map was commonly referred to as the 18-k map, assigning Annu. Rev. Marine. Sci. 2011.3:1-34. Downloaded from www.annualreviews.org an age of 18,000 years to the Last Glacial Maximum. Early estimates of the age of the glacial period were based on counting varves in periglacial lake sediments. In the 1950s, radiocarbon dating became available and was widely applied to calcareous deep-sea sediments (Suess 1956). From such dating, it soon emerged that the transition from glacial conditions to postglacial ones was remarkably fast, and remarkably close to our own time. The associated change in sea level would have been noticeable in one person’s lifetime. It was of the order of 1 m per century, for thousands of years (Figure 3), using the principle that the average shelf edge around the world ocean is close to the glacial sea level position (Shepard 1963). Difﬁculties in radiocarbon dating arise from uncertainties in the ambient production of the isotope (which is linked to solar activity), contamination, sediment mixing by organisms, a relatively short half-life (near 6,000 years), and uncertainties in the half-life. At an age of 30,000 years (ﬁve half-lives), the concentration of radiocarbon is down to 1/32 of the original value and is quite vulnerable to contamination. Beyond that age, therefore, radiocarbon dating becomes unreliable.

12 Berger MA03CH01-Prefatory-Berger ARI 17 November 2010 6:37

–1.5

Holocene –1.0

1 m century–1

–0.5

Last Glacial Maximum 0.0 Oxygen isotope index (permil) isotope index Oxygen 0.5 01020304050 Age of sediment (kyr before present) Figure 3 Oxygen isotope stratigraphy of the uppermost part of Atlantic Core P6304-9, based on analysis of Globigerinoides sacculifer, as published by Cesare Emiliani (inset) in 1966 (Emiliani 1966). Sedimentation rate is near 2.7 cm per 1,000 years, taken as invariable for this graph. Average rate of sea level rise (∼1mper century) is based on taking the deglaciation period as 10,000 years, and the total rise as ca. 100 m, and postulating a linear relationship between change in δ18O and change in sea level. The Last Glacial Maximum is centered near 25,000 years before present. Inset image Rosenstiel School of Marine and Atmospheric Science, University of Miami.

In a given core, the extrapolation of a sedimentation rate obtained from the uppermost sediments can still provide estimates for ages of older deposits, but such estimates are no better than the assumption of steady sedimentation rate. (Modern methods are based on more reliable radiometric methods and on porting the time scale of magnetic reversals from terrestrial basaltic sequences into marine sequences.) The end of the ice age came much as expected from Milankovitch theory (within a thousand years or so), but in essence without particular warning. The ice started melting in earnest approx-

by Dr John Klinck on 07/15/11. For personal use only. imately 16,000 years ago and sea level rose rapidly, within some 3,000 years, by approximately 60 m (Figure 4). It then stood still for a dozen centuries, while a severe cold spell held the North Atlantic realm in its grip. Then, again without warning, extremely rapid warming set in and the remainder of the vulnerable ice melted within another 3,000 years, to raise the sea level another Annu. Rev. Marine. Sci. 2011.3:1-34. Downloaded from www.annualreviews.org 60 m. The reasons for the steepness of the two steps, why there are steps at all, or why there are two steps rather than three or four, and the reasons for the duration of the cold spell are obscure. From this circumstance, one might conclude that we do not understand the climate machine very well, especially where large ice masses are involved. This means, for clarity, that we have not much of a grip on how the ice masses will react to global warming. We shall return to that topic, after discussing ice age cycles and their repeated periods of rapid decay of ice masses (the “terminations” of Broecker & van Donk 1970). Although understanding is limited, there are some clues as to what happened, as follows. The ﬁrst of the two great meltwater pulses can be ascribed to Milankovitch forcing, as it is centered near 14,000 years ago, when July insolation in high northern latitudes was at its maximum. The implication is that ice sheets (or rather some ice sheets) were ready to go, and they did so when the summer sunlight was of the right intensity. The ice to go ﬁrst, presumably, was that which was

www.annualreviews.org • Geologist at Sea: Aspects of Ocean History 13 MA03CH01-Prefatory-Berger ARI 17 November 2010 6:37

0 1960s

2 60 1990s

Sea level rise (m) Sea level 80

100

120

0 5 1015 20 Thousands of years before present Figure 4 Sea level rise following the last glacial maximum: reconstructions in the 1960s (red ), based on submerged nearshore deposits (Shepard & Curray 1967; note large error bars) and around 1990 (blue), based on detailed dating of shallow-water coral (Fairbanks 1989, Bard et al. 1990; note the two-step nature of the rise). From Berger (2009). Numbers 1 and 2 denote the two grand steps of deglaciation, when the rate of sea level greatly exceeded the overall average.

farthest from the poles and most vulnerable to being invaded by the sea, that is, the ice that was grounded below sea level (Berger & Jansen 1995). This ice was subject to positive feedback from a rise of sea level: As sea level rose, it lifted more of the ice off the ground. A runaway situation ensued. When this rather vulnerable ice was used up, the process stopped. During a pause, while the

by Dr John Klinck on 07/15/11. For personal use only. remaining glacial ice warmed in its interior, the climate went back to the state it had left during the ﬁrst melting step. This return to glacial conditions is known as the Younger Dryas cold spell. There is nothing mysterious about the cold spell. After all, large ice sheets still dominated the scene with their whiteness and high elevation, so its arrival is not a surprise and no esoteric explanations Annu. Rev. Marine. Sci. 2011.3:1-34. Downloaded from www.annualreviews.org are needed. The central question is about the warming: What made the ice decay in such a short time? As suggested, a likely answer is that ice that is grounded below sea level is especially vulnerable to collapse, as emphasized by the geologist T. J. Hughes (Hughes 1987; also see Alley et al. 2005). In addition, as noted by Olausson (1965) and many since, the addition of large amounts of meltwater to the northern North Atlantic interferes with the formation of deepwater and thus has many ramiﬁcations for the global thermohaline circulation and associated heat transport (Broecker & Denton 1989).

ON ICE AGE CYCLES AND THE QUESTION OF ORBITAL FORCING It soon emerged, from the mapping of moraines and other ice age deposits, that the Great Ice Age of Agassiz was more properly described as a long period of repeated glaciations separated by

14 Berger MA03CH01-Prefatory-Berger ARI 17 November 2010 6:37

warm intervals. The task became to identify these different glaciations and establish whether they were cyclic phenomena and whether they were global or regional events. On learning that there had been multiple ice ages, James Croll (1821–1890), a brilliant Scottish Earth scientist, rather quickly concluded that the ice ages were cyclic and that the cycles were driven by changes in insolation on the northern hemisphere, linked to orbital forcing (i.e., the eccentricity of the planet’s orbit) (Croll 1875). Colleagues on the continent (renowned geologists such as Albrecht Penck, Eduard Bruckner,¨ and Gustaf Steinmann) advised caution: Would it not be better to establish cyclicity first, before attempting to explain it, and is it not true that the known facts do not fit the theory? Croll’s fellow geologist and Scotsman James Geikie, on the other hand, confirmed the reality of multiple glaciations (in 1874) and supported Croll’s approach. Steinmann’s main criticism concerned the fact that ice ages in the northern and southern hemispheres seemed to have been contemporaneous, so could not derive their forcing from changes in insolation that are opposite in sign on the two hemispheres. With such powerful arguments, Croll’s speculations were soon made irrelevant in the pursuit of Quaternary history. Today, however, most geologists studying the Quaternary agree that orbital forcing is the cause of cyclical ice ages. In hindsight, we can see that Croll argued well, but he was misled in at least one fundamental way—he took the present to be typical. Thus, because at present the northern hemisphere is almost ice-free, he defined the problem in terms of making the northern hemisphere cold enough to bear large ice caps. However, the problem is not how to make ice in a warm world, the problem is how to get rid of the ice in an unusually cold world. The Earth is already programmed, as it were, to have its high northern latitudes covered by ice—snow falls every winter—the real question is how to get rid of the snow and ice to make and keep an interglacial period on a cold planet. The effect of this shift in emphasis about what it is that needs explaining is profound. Instead of focusing on the amount of sunshine in winter, which Croll thought important, we now follow Milankovitch (1930) in focusing attention on whether northern summers are warm enough to melt the snow and ice that the winters readily provide. Concentrating the argument on summer melting power turns out to give the correct answer. By adjusting a few seemingly minor details in Croll’s scheme—focusing solely on ice growth and decay in the northern hemisphere and identifying summer insolation in high latitudes as the crucial guide to climatic change—the Serbian astronomer and engineer Milutin Milankovitch (1879–

by Dr John Klinck on 07/15/11. For personal use only. 1958) succeeded where Croll had failed. Also, where Croll had attempted to master oceanography and climatology on his own, Milankovitch took the reasonable course to ask the experts. One of the greatest experts around was Wladimir Koppen¨ (1846–1940). He encouraged Milankovitch in his quest for astronomical forcing of ice age climates and supported his approach. It was in a Annu. Rev. Marine. Sci. 2011.3:1-34. Downloaded from www.annualreviews.org book by Koppen¨ (and his son-in-law Alfred Wegener) that Milankovitch’s graph of the summer radiation for 65◦N, for the last 600,000 years, was ﬁrst published, along with the claim that this radiation curve is relevant to the timing of glaciations (Koppen¨ & Wegener 1924). Milankovitch’s concept is quite simple. First of all, the important action is in high northern latitudes, where relatively modest changes in insolation can be translated into large changes in land-based snow cover, and in snow retention. Second, what is important in insolation is the apparent size of the sun in the summer months, and how high it is over the horizon at noon in high summer. If the sun is high (which happens during times of high obliquity of the Earth’s axis) and if the sun’s disk is large (which happens when the Earth is close to the sun), the insolation is maximized and the sun will melt snow and ice from the last winter. If not, snow and ice will build up and make polar ice sheets, reﬂecting more and more of the incoming sunlight in the process. Thus, positive albedo feedback is essential in translating changing sunlight into ice buildup and decay.

www.annualreviews.org • Geologist at Sea: Aspects of Ocean History 15 MA03CH01-Prefatory-Berger ARI 17 November 2010 6:37

Milankovitch’s reconstruction of summer insolation, in essence, describes seasonal contrast, because the overall annual insolation does not change much. So, if there is more sunlight in summer, there is less in winter. A link between ice buildup and decay on the one hand and seasons on the other makes the ice ages cyclic, naturally, as the astronomical forcing of seasonal contrast is strictly cyclic. Milankovitch’s proposition was a bold move. No one in fact knew whether the ice ages were cyclic or not, or whether they were spaced more or less according to the intervals prescribed by Milankovitch’s mechanisms. In the absence of methods for dating the glacial deposits, all age assignments were rather crude guesswork. Using the Milankovitch time scale for assigning dates to glacial deposits led into a circular argument. Thus, without independent dates, the hypothesis stood on weak feet indeed. Some say it still does (e.g., Muller & Macdonald 2000). Whereas Milankovitch theory is now the ruling paradigm [thanks to research summarized in a milestone compendium edited by A. Berger et al. (1984)], there are still some puzzling problems that await explanation. Right now, for example, our planet is closest to the Sun in January and farthest in July, along its elliptical path around the central star. According to Milankovitch (and his climate advisor Koppen),¨ this is a good situation, in principle, to make glaciers. If this is correct, why have we not been building ice sheets for the last 2,000 years? The reason is that the sun-distance effect is just one of the important factors; the long-term state of the climate system plays a role as well. Importantly, the land surfaces that hold the winter snow have not sufﬁciently rebounded from the depressed position they acquired when covered by ice. They need to rise to be able to keep that snow through the following summer. Canada and Scandinavia are still rising. When elevations are high enough, snow can stay and the positive albedo feedback can then begin to work its magic (unless frustrated by increased levels of greenhouse gas, of course). Thus, Milankovitch theory is an incomplete scheme for ice age oscillations when applied to ice mass all by itself. However, we should be able to apply the concept to the change in ice mass, that is, to the buildup and decay of polar ice. After all, that is how the theory was designed in the ﬁrst place, as a driver of change, not as a driver of the integration of that change. To test any aspect of Milankovitch theory, we need two ingredients: a reliable reconstruction of ice mass for the last 600,000 years (i.e., the period Milankovitch was attempting to explain) and a reliable reconstruction of the solar insolation curve. Both types of reconstructions are available, and we can proceed provided that we call on some simplifying assumptions.

by Dr John Klinck on 07/15/11. For personal use only. The ﬁrst assumption is that the oxygen isotope record of both the benthic foraminifer Cibicides spp. and the planktonic species Globigerinoides sacculifer and other surface-water forms are linearly related to ice mass. This assumption allows us to present all the appropriate stratigraphies in terms of mean and standard deviation and see them as ice-mass variation, for easy comparison. Annu. Rev. Marine. Sci. 2011.3:1-34. Downloaded from www.annualreviews.org The second assumption is that it is best to avoid oxygen isotope data that have been “tuned” to Milankovitch theory, that is, which owe their dates to taking the theory to be correct. If derived under this assumption, obviously, the data can no longer be used to test the theory. To avoid circularity, I shall refer the available isotope data (Zachos et al. 2001, Lisiecki & Raymo 2005) to a time scale based on linear interpolation between the last deglaciation (set to 12,000 years) (Figure 4) and the Brunhes–Matuyama boundary in a record from ODP Site 806 (Berger et al. 1993). The Brunhes–Matuyama boundary denotes a major magnetic reversal and is dated on land using basalt sequences and radiometric techniques (Baksi et al. 1992). We may take the age to be near 790,000 years (in rough agreement with a number of investigations, including Shackleton et al. 1990). The strategy for deriving the “untuned” time scale is as follows: (1) Assign a date to each sample point in Site 806, for the last 900,000 years (the time since the Mid-Pleistocene climate shift, or Mid-Pleistocene Revolution). (2) Identify marker points for control age points (6 to 5 transition,

16 Berger MA03CH01-Prefatory-Berger ARI 17 November 2010 6:37

12 to 11 transition, 16 to 15 transition). (3) Port these control points to the stacked records of Zachos et al. (2001) and Lisiecki & Raymo (2005) and reassign corresponding ages to all sample points by linear interpolation. After this sequence of redating is completed, we can construct the series to be tested against Milankovitch predictions. The chief difficulty in assigning dates to the sample points in Site 806 is the fact that Isotope Stage 19 is missing: It is in the core gap between cores 2H and 3H. The gap has to be adjusted to provide space for the missing isotope signal. This is done by assuming that Stage 19 has the same width as Stage 21 (which is reasonable, based on the record of Site 805 of ODP Leg 130, described in the same report as that for 806, and from considering other records, such as the one by Bassinot et al. 1994). To make the gap wide enough, one must assume expansion of core material upon retrieval from great depth. Specifically, the required expansion for Core 2H is 14% (significantly greater than assumed by Berger et al. 1993). For Core 1H, it is 3% (to ensure that Stage 9, which falls into the core gap, is equal in width to Stage 11). At this stage of the analysis, by comparing unrelated records from two different regions of the ocean (ODP Site 806 and the record of Bassinot et al. 1994), one can already see that the assumption of ice-mass control is well supported—the records are largely congruent after the adjustments described. The last 650 ky in both records are marked by prominent 100-ky cycles. The label Milankovitch Chron has been applied to this time span. The nature of the cycles is as yet unresolved (Muller & MacDonald 2000). We next identify the age points for redating the combined stacked records of Zachos et al. (2001) and Lisiecki & Raymo (2005), based on the reconstruction of oxygen isotope variations for Site 806, as follows: transition from Marine Isotope Stage 2 to MIS 1: 12,000 years (set); transition from Marine Isotope Stage 6 to MIS 5: 123,100 years; transition from Marine Isotope Stage 12 to MIS 11: 415,900 years; transition from Marine Isotope Stage 16 to MIS 15: 616,700 years; Brunhes–Matuyama Boundary: 790,000 years (set). The resulting record is (numerically) differentiated for comparison with Milankovitch forcing, taken as the insolation in June, at 60◦N (data in Berger & Loutre 1991) (Figure 5). One notes (in Figure 5a) that there is near-coincidence of unusual periods of high insolation and periods of rapid melting (shown as peaks in the differentiated record). The typical adjust- ment of ages, to make these two types of periods congruent, is ±6 ±3 kyr. Adjustments in age

by Dr John Klinck on 07/15/11. For personal use only. estimates for the sediment data are positive after 650,000 years ago and negative before that time back to 800,000 years ago. One implication of this systematic variation may be that during the Milankovitch Chron, the summer radiation was maximally effective after June, and before that chron, maximum effectiveness occurred earlier in the year. However, the discrepancies presum- Annu. Rev. Marine. Sci. 2011.3:1-34. Downloaded from www.annualreviews.org ably largely reflect difficulties in the dating procedure, which rests on assuming no change in sedimentation rate for Site 806 for the last million years. The average interval between tie-points is rather large and uneven: 83,000 ± 64,000 years. After making the adjustments using the tie-points (adding between +6and−6 years to the dates of the record for best fit), the record is now “tuned” to Milankovitch forcing (Figure 5b;see also Supplemental Table 1; follow the Supplemental Material link from the Annual Reviews Supplemental Material home page at http://www.annualreviews.org) and is no longer suitable for testing orbital theory. However, with these relatively minor adjustments (the uncertainty of any dating is on the order of 3,000 years in the sediments, to begin with), we obtain a remarkable record of correlations of isotope differences with Milankovitch insolation over large portions of the series (Figure 5c). Correlations fall to low values within the time span corresponding to the isotopic stages labeled 11 to 14, apparently a time (approximately 500,000 years ago) when the climate system underwent strong reorganization and did not “listen” faithfully to orbital forcing.

www.annualreviews.org • Geologist at Sea: Aspects of Ocean History 17 MA03CH01-Prefatory-Berger ARI 17 November 2010 6:37

2.0 a

1.5

1.0

0.5 Ox iso Orb F

0.0

2.0 b

1.5

1.0

Index (arb. units) (arb. Index 0.5 units) (arb. Index

0.0

1.0 c 0.8

0.6

0.4

0.2 (50-kyr window) 0.0 by Dr John Klinck on 07/15/11. For personal use only. Correlation coefficient Correlation

–0.2 0 200 400 600 800 1,000 Age (kyr) Annu. Rev. Marine. Sci. 2011.3:1-34. Downloaded from www.annualreviews.org Figure 5 The deep-sea oxygen isotope record of the last million years compared with Milankovitch forcing—a comparison. (a) Combined record of Zachos et al. (2001) and Lisiecki & Raymo (2005), but time scale for the record from Site 806, ODP Leg 130 (see text). Milankovitch forcing (orb F) from Berger & Loutre (1991). (b)Sameas(a) but time scale adjusted (typically by 6 kyr) to enhance matches between forcing and melting peaks. (c) Correlation of isotope record and orbital forcing, for time scale in (b) (50-kyr window). Gray shaded areas indicate periods of high correlation. Abbreviations: DT, system deaf to orbital forcing; Ox iso, oxygen isotope record, differentiated.

If Milankovitch forcing were indeed unimportant over the entire time span of the last million years (as has been implied by some students of the matter), we should expect similar evidence for orbital deafness elsewhere in the record. That this is not the case suggests that Milankovitch forcing is effective over much of the last million years, although not equally so everywhere.

18 Berger MA03CH01-Prefatory-Berger ARI 17 November 2010 6:37

Two spells of maximum orbital deafness are seen in the actual record (Figure 5c): One is 50 kyr long and centered on 485 kyr ago, the other is 20 kyr long and centered on 830 kyr ago. There is some indication that the times of maximum attention to orbital forcing include terminations, that is, the periods of fast melting (this may in part be a consequence of the tuning process). The same is true for fast buildup of ice. Thus, it seems, when the climate system “listens” to orbital forcing, ice masses can change more quickly than otherwise (and presumably attain larger extremes, as well). It appears, from the correlation series (Figure 5c), that the level of response of the system to orbital forcing changed after isotopic Stage 12, toward improved correlations. Thus, some factors became active within the system that increased its ability to respond to Milankovitch stimulation. Possible candidates are positive feedback from albedo or greenhouse effects or both. Larger and higher maximum ice ﬁelds might have increased relevant snowﬁeld sizes or the extent of sea ice. Or the stronger melting events affected deep circulation. Or the buildup of the Great Barrier Reef since isotopic Stage 11, or an increased expansion of permafrost, affected the carbon cycle in fundamental ways. No one knows. What we can be reasonably sure of is that modeling of the last 400,000 years has to deal with additional feedback mechanisms, compared with the time before that. The implication for understanding climate change is that the times before 450,000 years ago were different, and therefore their lessons are not as applicable to present conditions as are those of the last 450 kyr. Much of the discussion about the nature of ice age variations and their relationship to orbital forcing has revolved around the properties of the series as they appear in periodograms (Fourier-type spectra), based on analysis in terms of sinusoidal cycles (Hays et al. 1976, Imbrie et al. 1984, Muller & MacDonald 2000). However, inasmuch as Milankovitch theory is about the growth and decay of ice masses, a focus on the integration of the ice-mass changes (as commonly seen in the literature) seems misplaced when discussing the merits of the theory. To appreciate the similarities and differences between presumed forcing and response, we can compare the two spectra of the differentiated versions of the combined stacked record (Zachos et al. 2001, Lisiecki & Raymo 2005) (one set to the 806 time scale and one tuned to Milankovitch forcing) with the forcing itself (here, June insolation at 60◦N; Berger & Loutre 1991) (Figure 6).

20 23.8 by Dr John Klinck on 07/15/11. For personal use only. Orbital forcing 22.3 U 15 Stacked record Stacked record T 41.0 18.9 Annu. Rev. Marine. Sci. 2011.3:1-34. Downloaded from www.annualreviews.org 10

5 96.0 Amplitude (arb. units) (arb. Amplitude

0 –2.4 –2.1 –1.8 –1.5 –1.2 –0.9 log F

Figure 6 Spectra for orbital forcing for the last million years (data from Berger & Loutre 1991) and for the differentials of the combined stacked records of Zachos et al. (2001) and Lisiecki & Raymo (2005). “U ”isfor the time scale based on ODP Site 806 (Leg 130, which is untuned), and “T ” is for the scale tuned to Milankovitch forcing as in Figure 5. The two spectra were generated by Fourier scan of autocorrelation series (in steps of 1/1.01).

www.annualreviews.org • Geologist at Sea: Aspects of Ocean History 19 MA03CH01-Prefatory-Berger ARI 17 November 2010 6:37

The x-axis “F ” (for frequency) is in terms of 1/(length of period in thousands of years) and is given as a logarithm (that is, to obtain the period, simply change the sign). From a visual comparison of the spectra (Figure 6, left to right), we can deduce the following: (a) There is almost no forcing at 100 kyr (at log F =−2); therefore, the ice age cycles near that length are not a direct response to orbital forcing. Presumably, such long cycles reflect internal properties within the system that favor oscillations near 100 ky (possibly capturing multiples of shorter forcings, e.g., 4 × 24) or else listen to some other type of forcing (Muller & MacDonald 2000). (b) The response to obliquity-related forcing (near 41 kyr) is relatively strong, and it shows more strength with tuning (even though tuning in this case did not explicitly use obliquity cycles). The correspondence between forcing and response at this period strongly supports Milankovitch theory. (c) The precession-related periods (19 to 24 kyr) are strong in the forcing but not so well expressed in the growth and decay of ice masses. Tuning produces a split in the response that is reminiscent of the split in the forcing. The record seems to discriminate against periods shorter than 20 kyr (log F < −1.3), either because the climate system has a preference for long periods, or because the recording mechanisms are not suitable for short-period cycles, or both. The climate system integrates external forcing and internal oscillation. Thus, the commonly seen spectra of ice age fluctuations in various proxy records reflect systemic integration (and associated filtering of periods) as much as they do external forcing. The main internal oscillation is well represented by the shape of isotopic Stage 11 (Figure 7), which reflects the course of events during a time when external forcing was weak. The prime use of the stacked and tuned records is for stratigraphy, that is, for the dating of oxygen isotope records obtained from deep-sea sediments, and also for other types of Quaternary records that show variations similar to the standard. The records compiled by Zachos et al. (2001) and by Lisiecki & Raymo (2005) readily serve that purpose, at least for the last million years. The so-called SPECMAP standard (Imbrie et al. 1984), largely based on Atlantic data, remains valid

1.8 157911 1.6 13 15 17 19 21 1.4 1.2 1.0 by Dr John Klinck on 07/15/11. For personal use only. 0.8 0.6

Index (arb. units) (arb. Index 0.4 Annu. Rev. Marine. Sci. 2011.3:1-34. Downloaded from www.annualreviews.org 0.2 Modern DT Premodern 0.0 0 100 200 300 400 500 600 700 800 900 1,000 Age (kyr)

Zachos et al. 2001 Lisiecki & Raymo 2005 Combined and retuned, as in Figure 5b

Figure 7 Stacked record of oxygen isotopes (standardized to a mean of unity and a standard deviation of 0.25) for the last million years (the transform happens to be close to the original δ18O records). Modern, time span since isotopic Stage 12, with high correlations between response and orbital forcing (see also Figure 5c); premodern, time span before isotopic stage 13, with medium correlations. Abbreviation: DT, system deaf to orbital forcing (i.e., deaf to tuning).

20 Berger MA03CH01-Prefatory-Berger ARI 17 November 2010 6:37

back to approximately 650,000 years. It is incorrect for times preceding the Milankovitch Chron. A standard for the last million years, based on combining the two recent compilations, and retuned (Figure 5b), is given in Figure 7 (see also Supplemental Table 1). Supplemental Material With a standard oxygen isotope series in hand, for the last million years, we can now find clues to a very simple question: Just how fast can sea level rise in any one century? We find the answer by interpreting the standard oxygen isotope series as a record of sea level changes, setting the sea-level change for the last deglaciation equal to 120 m (Figure 4). When plotting the rates of sea-level change against sea level position, we find a maximum of rate of change intermediate sea level positions, as might be expected (Berger 2008). At those times, at intermediate positions, we are within the time span of deglaciation, when large ice masses are collapsing, at overall rates of up to 2 m per century. What is of interest in the context of the present situation, when the planet is warming, is how fast the present sea level can rise in response to ice- mass collapse. After all, much of the more vulnerable ice is already gone, although there remain some ice masses (on Greenland and in the western Antarctic) that are deemed quite vulnerable still. The typical fast decay rate, for past millennia of “modern” times (Figure 7) for sea level positions close to the present one, is near 1 m per century. The averaging of rates of deglaciation on a millennium scale yields a minimum risk for this item, rather than a typical risk, with regard to the present situation.

PASSAGE INTO THE MODERN WORLD: THE GREAT COOLING The ice age variations inform our modern world; they have dominated the climate narrative for the last 3 million years, at the end of a long period of cooling. The Earth was warm 50 mya, and it since became cold at the poles and acquired enormous ice fields in high latitudes, first in the south, then in the north. The question of how we got there has great relevance for understanding the evolution of all existing organisms that are linked to geography and oceanography. As mentioned, if we ask why the Challenger Expedition (1873–1876) did not find “living fossils” at the bottom of the sea, we must consider that the abyssal environment has changed fundamentally within the last 40 million years. This has led to the extinction of ancient forms and encouraged invasion from shallower waters, especially from high latitudes. Within this general framework of thought, it is clear what the discussion of Ceno-

by Dr John Klinck on 07/15/11. For personal use only. zoic cooling has to focus on the following questions: (1) Just what sets the modern cold ocean apart from one that is warm? It is the presence of a strong thermocline and a great abundance of oxygen throughout the water column. (2) What is the single most important element in the enormous collection of special properties that is relevant for the evo- Annu. Rev. Marine. Sci. 2011.3:1-34. Downloaded from www.annualreviews.org lution of the more conspicuous organisms? It is the shortness of the food chain in upwelling regions. (3) What were the steps in achieving the modern situation? They are (a) loss of shelf seas and opening of Drake Passage, (b) buildup of enormous ice masses in Antarctica, and (c) buildup of ice masses around the Arctic, ice that was vulnerable to small changes in seasonal insolation patterns. (4) Are there examples for major events in the evolution of selected marine organisms? Yes; they have to do with the acquisition of baleen in whales and the diversification of seals and sea birds. To find clues to these various questions, we must consider a few crucial facts (in addition to many others, not here mentioned). The ice buildup in high latitudes depends on a solid land base, and once it starts, it provides for positive albedo feedback. Naturally, therefore, ice buildup in the south, with a continent centered on the pole, preceded that in the north, where the pole is in the middle of an ocean basin. Large temperature differences between warm and cold regions drive strong zonal winds and hence currents. The cold ocean has a well-defined thermocline. A

www.annualreviews.org • Geologist at Sea: Aspects of Ocean History 21 MA03CH01-Prefatory-Berger ARI 17 November 2010 6:37

Bottom waters Cold Warm

Ice sheets Some ice Ice free Plio- 0 Pleistocene Northern ice sheets build LPNH Strong regional upwelling 10 High plankton diversity Southern ice sheets grow Miocene MMAA 20 Modern patterns Thermocline of circulation begin to evolve Low plankton Oligocene 30 diversity Major changes in circulation and sedimentation 40 LEFS

Cenozoic Eocene Isotope values 50 Overall cooling of benthic High plankton foraminifers diversity

Age (millions of years before present) before (millions of years Age Hot spot

60 Low plankton Paleocene diversity

Extinction Cretaceous 70 3.0 2.0 1.0 0.0 –1.0 –2.0 δ18O of foraminifers from the Atlantic deep-sea floor (‰; PDB)

Figure 8 Overview of the ocean history in the Cenozoic, showing the general cooling (isotope values of deep-sea benthic foraminifers) of the sea and the associated ice buildup, first in Antarctica (MMAA) and then in the far northern lands (LPNH). In the late Eocene, there was a fundamental change in the operation of the climate system and the ocean, reflected in a major shift of deep-sea sediments in many regions from quickly alternating facies types (limestones, chert, and mudstone) to long series of pure carbonate (LEFS). Base graph by Miller et al. (1987), with additions. Abbreviations: LEFS, late Eocene facies shift; LPNH, late Pliocene northern hemisphere; MMAA, mid-Miocene Antarctic event; PDB, Pee Dee Belemnite (the fossil used to establish the original oxygen isotope standard, in Chicago). by Dr John Klinck on 07/15/11. For personal use only. shallow thermocline provides for biologically effective upwelling wherever there is mixing and divergence. The presence of dissolved silica within the mix of nutrients is crucial. The evolution of warm-blooded marine vertebrates (the ones that invaded the sea from land early in the Tertiary) Annu. Rev. Marine. Sci. 2011.3:1-34. Downloaded from www.annualreviews.org is closely linked to the history of upwelling and silica availability, as I shall argue in what follows. The overriding fact in the historical narrative of the cooling planet in the Cenozoic is the observation that cooling occurs in a few great steps (Figure 8). These steps introduce instability into the system as it moves from one dominant condition into another. The number of large steps is three: The first is the change from the Eocene to the Oligocene, from a planet with shelf seas and warm deep water to one without such seas and with cool deep water (Figure 8, LEFS, or late Eocene facies shift). This shift also involved an overall change from tropical circumglobal circulation (in the Tethys) toward a circumpolar circulation in the Southern Ocean (Haq 1981). The second step involves the buildup of Antarctic ice (Figure 8, MMAA, or mid-Miocene Antarctic event), as well as further cooling of deep waters. It is preceded by the development of a thermocline, which is a response to the production of cold intermediate waters and is a somewhat gradual process rather than a step. The appearance of this feature forces increased diversity in the

22 Berger MA03CH01-Prefatory-Berger ARI 17 November 2010 6:37

plankton as it provides for vertical zonation and for a host of different lifestyles involving vertical migration. Finally, the third step (Figure 8, LPNH, or late Pliocene northern hemisphere, ice buildup) concerns the cooling of the northern hemisphere to a point where ice can reach the northern continents, which are separated from the North Pole by the Arctic Sea and thus do not readily cover themselves with ice. In fact, upon slight increases in heat received from the summer sun, these regions readily part with their ice, giving rise to the Quaternary ice age cycles discussed above. The existence of steps naturally begs the question of what factor or factors are responsible for turning an overall cooling trend into cooling steps. The most obvious proposition is one that calls on positive feedback associated with snow and ice. The last great step of cooling is generally taken as having happened approximately 3 mya, within the late Pliocene. It is since then that we have a continuous record of ice-rafted debris in the northern North Atlantic (Berggren 1972). Naturally, ice-rafted debris is a more reliable indicator of the beginning of glaciation than is a change in isotope ratios, although considerable efforts have been expended to refine the beginning of northern glaciations on the basis of isotope records. Some have chosen to link the initiation of the northern ice ages to the emergence of the Panama Isthmus (see discussion in Berger & Wefer 1996). This idea is highly questionable as the closing of the Panama gateway presumably increased the heat supply to the northern North Atlantic, owing to interdiction of the flow of warm water from the Caribbean to the tropical Pacific and its rerouting into the Gulf Stream. The argument that increased heat led to increased snowfall is unconvincing (even though many modern glaciers in western Norway do grow with more precipitation). More satisfactory (at least to this geologist) is an assumption that if ice is to begin building in northern lands, the climate must cool below the freezing point and stay there. In this line of argument, the buildup of ice in the far north is simply another consequence of the general cooling of the planet that governs the history of the entire Cenozoic. If this reasoning has merit, the closing of the Panama Straits hindered rather than helped the buildup of ice in northern lands. Concerning the start of northern glaciations, we know from drilling off Greenland that glaciers in Greenland had already grown within the late Miocene, some 6 or 7 mya ( Jansen & Sjøholm 1991, Larsen et al. 1994). The corresponding fall of sea level apparently affected the access of

by Dr John Klinck on 07/15/11. For personal use only. ocean water to the Tethys-derived remnant then present in the region of today’s Mediterranean. As a consequence, the ancient Mediterranean Sea dried out near the end of the Miocene (Ryan & Hsu¨ 1973, Hsu¨ et al. 1973, Hsu¨ 1992). What needs explaining is not so much the initiation of ice buildup in the late Pliocene (it is part and parcel of the general cooling and related to mountain Annu. Rev. Marine. Sci. 2011.3:1-34. Downloaded from www.annualreviews.org building) but the fact that ice buildup was interrupted in the early Pliocene, that is, that the trend was reversed for several million years. Presumably, the answer is the transport of excess heat into the northern North Atlantic, caused by the closing of the Panama Straits. Heat, in this scheme, prevents the growth of ice rather than promoting it. Closing the Panama gateway may not have been the only factor keeping the far northern lands too warm for ice buildup for much of the Pliocene. The Antarctic buildup, in the middle Miocene, may have had a similar effect. With ice on Antarctica, and none in the far north, there is a strong asymmetry in the global zonal wind systems, such that heat is transported across the equator from the southern seas to the northern ones (Flohn 1985). Much of the heat is available to fuel evaporation, so that high salinity anomalies ensue in the subtropical North Atlantic. In turn, upon cooling, such high-salinity water provides a source for deepwater production. Serious deepwater production in the North Atlantic turned on late in the middle Miocene, presumably both in response to ice buildup in the Antarctic and to cooling in the far north.

www.annualreviews.org • Geologist at Sea: Aspects of Ocean History 23 MA03CH01-Prefatory-Berger ARI 17 November 2010 6:37

Two fundamental issues are involved in the heat transport from south to north of the equator (i.e., northern heat piracy). One is the fact that landmasses are concentrated in the northern hemisphere, which results in the pulling in of heat with summer monsoon winds, some of which originate south of the equator. The other is the fact that the stronger zonal winds in the southern hemisphere (with the icy shores of Antarctica fixing a low end point of the temperature gradient) tend to push the Intertropical Convergence Zone into latitudes north of the equator, with both sensible and latent heat moving across from south to north. Northern heat piracy, then, has been pervasive since the middle Miocene—and it has been especially strong in the Atlantic since approximately 12 mya. The timing is readily deduced from the great silica shift at about that time (Keller & Barron 1983), a change in sedimentation patterns that indicates the turning on (or up) of the production of North Atlantic Deep Water (which flushes silica out of the deep Atlantic). The rise of northern heat piracy along with the ice buildup in the Antarctic in the middle Miocene [which was associated with a major disturbance in the carbon cycle as well, according to Vincent & Berger (1985)] delayed northern glaciations through the introduction of global north–south asymmetries in wind patterns, which favored northern heat piracy. By the same token, the asymmetry helped stabilize Antarctic ice sheets, creating an important environment for evolution—for the great whales and also for penguins (which are restricted to productive waters in the southern hemisphere). Before we turn to the intriguing story of the evolution of whales and other warm-blooded marine vertebrates, we need to take a closer look at the nature of the late Eocene facies shift (Figure 8, LEFS), a geologic event also referred to as Auversian Facies Shift (Berger & Wefer 1996, McGowran 2005). In essence, it describes a major and lasting change in deep-sea sedimentation patterns, marking the beginning of widespread deep-ocean carbonate deposition. An important aspect of the shift is the apparent drop in the carbonate compensation depth (CCD), which emerges when considering the general sinking of the seafloor to the carbonate record (Berger & Winterer 1974, van Andel et al. 1977) (Figure 9). It is commonly considered a shift of carbonate deposition from (disappearing) shelves to the deep-sea floor, as somewhat crudely exemplified in Figure 9 by contrasting Eocene nummulite limestone from an ancient Tethys shelf sea in northeastern Spain with the most common components of Cenozoic deep-sea carbonate fossils, that is, nannofossils. The Auversian Facies Shift is not simply an expression of the great drop of the CCD at the end of the Eocene. It reflects, in addition, a reorganization of sediment patterns involving opaline

by Dr John Klinck on 07/15/11. For personal use only. fossils, that is, radiolarians and diatoms. In the Eocene, we find siliceous sediments (chert) over large parts of the seafloor (e.g., Calvert 1971, Muttoni & Kent 2007); in the Oligocene, this is not the case. Something changed in the way silica (derived from the weathering of continental rocks and from the reaction of seawater with hot basalt) was removed from the system. Annu. Rev. Marine. Sci. 2011.3:1-34. Downloaded from www.annualreviews.org In the present ocean, opaline sediments are greatly concentrated below upwelling regions and especially in the greatest vertical mixing region of all, the Antarctic Circumpolar Ocean (Figure 10). There are indications that this great sink for silica arose during the late Eocene and early Oligocene. At that time, the Drake Passage broke open as South America separated from Antarctica and thus arose the great mixing ring that is the Circumpolar Current, driven by the howling winds around the land of ice fields. Silica, being incorporated into the skeletons of planktonic organisms (diatoms and radiolarians) that sink readily in seawater, moves to the seafloor and from there into deep waters, released by dissolving frustules. The deep mixing around the Antarctic brings dissolved silicate back up into the sunlit zone, stimulating diatom production there. Thus, much of the ocean’s recycling of silica takes place in the Antarctic Ring Current. Also, the Ring acts as a trap, being fed a steady supply of silica by exchange with the deep waters of the ocean, and especially (since 12 mya) by the advection of deep water from the Atlantic. By upward mixing, silicate becomes available for intermediate

24 Berger MA03CH01-Prefatory-Berger ARI 17 November 2010 6:37

3 AFS

South Atlantic

4 Depth (km) Nonequatorial Pacific

Equatorial Pacific Pacific (5° off equator)

0 10 20 30 40 50 Age (mya) b

Figure 9 by Dr John Klinck on 07/15/11. For personal use only. (a) Reconstructions of calcite compensation depth (CCD) ﬂuctuations for various oceanic regions, and position of the Auversian Facies Shift (AFS; Berger & Wefer 1996) with respect to the shift of sedimentation of carbonate from shelves to the deep sea. CCD graph from Seibold & Berger (1996), redrawn, based on data in Berger & Roth (1975) (dashed lines) and van Andel et al. (1977) (solid lines). Note agreement of all

Annu. Rev. Marine. Sci. 2011.3:1-34. Downloaded from www.annualreviews.org reconstructions in regard to the change at the AFS (b) Nannofossils from Bukry & Bramlette (1969); nummulite limestone photo by W.H.B. (Girona Cathedral, Spain). Arrow symbolizes the change in carbonate deposition during the AFS, in moving the site of sedimentation from shelf seas to the deep sea. Scale: nannofossils, 10 μm; nummulites, 1–10 mm.

waters generated south of 50◦S and thus ﬁnds its way into the southern upwelling regions off Peru and Chile, and off Namibia, and also into the equatorial upwelling zone of the eastern Paciﬁc, all regions that are fed by thermocline waters. Since the great CCD event, the increased removal of silica by precipitation within upwelling zones created an ocean that was successively more depleted in silica for the last 40 million years. The result was a thinning of radiolarian skeletons, as the raw material for making the skeletons became limiting (Moore 1969, Lazarus et al. 2009). In that sense, there is competition between diatoms and radiolarians for silicate, or rather between upwelling regions and the rest of the sea.

www.annualreviews.org • Geologist at Sea: Aspects of Ocean History 25 MA03CH01-Prefatory-Berger ARI 17 November 2010 6:37

Few or no siliceous fossils Very abundant Siliceous fossils common Main constituent of sediment Figure 10 Global distribution of siliceous fossils within marine sediments. High abundance is restricted to regions of upwelling, and the highest values are found below the Circumpolar Current around Antarctica. Adapted from Berger & Herguera (1992).

The thinning of radiolarian skeletons, especially within the early Oligocene (Lazarus et al. 2009), has clues to changes in the nature of productivity, which in turn affected the evolution not just of some planktonic protists that happen to make fossils but of everything else in the sea. The main point is this: The segregation of silicate-precipitating regions into a few privileged regions (those where upwelling occurred) created patches of “green ocean” but impoverished the rest of the sea. There were at least two important corollaries of creating green patches in a desert sea. First of all, the increase in the range of productivity stimulated speciation by providing opportunities for both starvation specialists and luxury-feeding opportunists. Second, the silica-rich, green-ocean patches shortened the food chain. As pointed out by Ryther (1969), a short food chain results from starting the base with relatively large diatoms, which serve as food for organisms large enough to be eaten by ﬁsh. In short, silica provides for food-rich waters in a few selected places and thus provides opportunities for energy-inefﬁcient (warm-blooded) consumers, such as cormorants and seals.

by Dr John Klinck on 07/15/11. For personal use only. The highly productive green places are concentrated along the shores of the world (especially at the shores bathed by eastern boundary currents). Thus, there was an opportunity, during the time of cooling, for terrestrial animals to invade the sea to take advantage of the increasing food supply. We see this happening today with the sea otters hunting in the kelp forests and with the Annu. Rev. Marine. Sci. 2011.3:1-34. Downloaded from www.annualreviews.org marine iguanas grazing the algae in the (geologically young) Galapagos Islands. For whales— whose four-legged ancestors invaded the sea sometime during the early Eocene—the Oligocene local greenings occasioned the evolution of baleen some 15 million years after the invasion, filters for straining food-laden waters (Berger 2007a). For toothed whales, which hunt below the mixed layer of the sea and go for prey involved in vertical migration, the rise of the thermocline in the early Miocene was a crucial event. Hunting in the dark, using echo sounding, their skulls were greatly modified for sending and receiving acoustic signals. The timing of the change of their skulls, therefore, is linked to the rise of the thermocline and the greening of the sea. For seals, greatly increased rates of upwelling in the middle and late Miocene proved vitally important, with conditions in the Antarctic playing a special role. It may be noted that the Antarctic, shrimp-eating seal (Lobodon carcinophagus), named for its filtering teeth, is considered the most abundant of all seals. It lives off the shortest possible food chain: diatom → krill → seal. Blue whales and their relatives, incredibly abundant at the time of James Cook (before his reports

26 Berger MA03CH01-Prefatory-Berger ARI 17 November 2010 6:37

helped start the great slaughter in Antarctic seas), took advantage of this same ultrashort chain, and the same is true for the most abundant penguins and even for the leopard seal, a large predator that has filtering teeth in addition to impressive canines. There is another important aspect to the patchiness of food-rich regions. The pattern is patchy not just geographically but also seasonally, and to some degree it is patchy on the scale of mul- timillennial climate variations (and everything shorter). The strong seasonality that comes with winter mixing followed by spring or summer bloom encouraged a habit of annual long-distance migrations in the more mobile vertebrates, from fish to whales, and in a host of seabirds. Whether migration targets are separated into feeding and breeding areas depends on a number of factors, not necessarily related to the physical environment but to social habits, to skill of predators, and to the ability of juveniles to cope with the stress of migration. In gray whales and humpbacks and many others, feeding and breeding are separated. But for the penguins of Antarctica, the summer feeding areas are precisely next to those where eggs are laid and the young are raised. The possible role of a rising thermocline in modulating the environment has been mentioned a number of times. The reason why a thermocline should develop and rise is straightforward: With cooling comes increased production of deep, cold water, which fills the deep-sea basins and moves the warm–cold boundary upward. The principle of this argument is that used by Munk (1966), balancing the upward motion of cold water with downward mixing of warm water. The drastic reduction of shelves at the end of the Eocene moved the destruction of tidal energy into the deep sea, stimulating mixing at depth, which favors the addition of cold bottom water from sources in high latitudes. Once the thermocline reached the biologically active upper waters, productivity was greatly enhanced according to the principle that photosynthesizing organisms have to stay in sunlit waters to grow (Sverdrup 1953). This condition is not achieved, evidently, as long as the thermocline is well below the sunlit zone. The low plankton diversity during the Oligocene may owe much to a highly variable environment, with a thermocline close enough to the bottom of the sunlit layer for occasional blooms after great storms but not close enough to provide predictable conditions for adaptive evolution for a great number of plankton species. A reliably shallow thermocline, and upwelling, have largely defined the plankton ecology of the modern ocean for the last 20 million years and increasingly since. Such profound changes in the productivity systems of the sea must have greatly influenced the

by Dr John Klinck on 07/15/11. For personal use only. evolution of warm-blooded marine animals (Lipps & Mitchell 1976).

ON THE END OF THE MESOZOIC

Annu. Rev. Marine. Sci. 2011.3:1-34. Downloaded from www.annualreviews.org Of all the various new ideas that shook the foundations of traditional geology in the past half- century, none are more important than the new appreciation of the importance of mass extinction in evolution. It came with the recognition of the overriding importance of an impact of a celestial body with Earth at the end of the Cretaceous, that is, at the Cretaceous-Tertiary (K-T) boundary (see Berggren & van Couvering 1984 for early discussions). The extinction event was ﬁrst clearly documented in marine biogenic deposits, in particular, in calcareous shallow-water and pelagic deposits exposed on land but also in the deep sea (e.g., Bramlette 1965, Thierstein & Okada 1979). In deep-sea sediments, also, there was evidence for marked changes in oxygen and carbon isotopes in surface waters at the time (Thierstein & Berger 1978, Boersma et al. 1979), suggesting collapse of the marine production system (Hsu¨ & McKenzie 1985). The direct evidence for impact was found in a famous pelagic carbonate section near Gubbio in Italy (Alvarez et al. 1980). Here, the exact horizon of extinction had been established by paleontologists from Switzerland and Italy (Luterbacher & Premoli-Silva 1964), based on the study of foraminifers. The evidence for

www.annualreviews.org • Geologist at Sea: Aspects of Ocean History 27 MA03CH01-Prefatory-Berger ARI 17 November 2010 6:37

impact consisted of unusually great concentrations of a rare metal, iridium, at the level of the K-T boundary. The metal is thought to have arrived with extraterrestrial bombardment. Numerous studies of the K-T boundary have resulted in an unusually detailed reconstruction of the crucial time interval. Nevertheless, it is probably fair to say that the actual mechanisms of extinction have remained a matter of discussion rather than of consensus. Clearly, when a large body from space hits the planet at a speed of more than 11 km sec−1, there are untoward consequences. In a sense, the devastation is not the surprising thing. What is at least equally worthy of study is the resilience of the ecosystem to such destruction: What we would like to know is how the system responded to stress and how it recovered. The early response to stress involves strange plankton blooms (Thierstein & Okada 1979). Recovery from the event in terms of restoration of biodiversity is counted in millions of years. One would expect the K-T boundary to be best preserved on the deep-sea floor rather than in shelf seas. In the deep sea, there should be places where sediments accumulate in steady fashion, undisturbed by tides and storms. However, the K-T disturbance presumably set off huge tsunamis (through submarine landslides); that is, it produced waves large enough to erode seafloor at a depth of several kilometers. Thus, undisturbed sections, if they exist, are likely rare. In addition, burrowing organisms will normally turn a sharp boundary into a fuzzy one, producing a thin zone that has both old and young fossils side by side, even though the organisms did not live together. This is not to claim that complete sections do not exist. But assuming one were to find one, how would one prove it is in fact complete? Despite all problems of precise documentation, there is now no doubt that the change at the K-T boundary was sudden and profound and that it was essentially unannounced by any preceding stressful events (e.g., a drop in sea level, or volcanic activity). Of course, one cannot exclude the possibility that conditions had changed in the latest part of the Cretaceous in a manner that made an impact more effective in its destructive potential. Impact happens. The questions are does it happen often enough to drive evolution, and could it even be cyclic? Speculations about the sinister activities of a hypothetical death star (Raup 1985) and about sporadic rains of comets and other missives from space attracted some following, and discussions arose that are reminiscent of the early stages of geology when catastrophism reigned supreme. Darwin’s mentor Charles Lyell changed the level of discussion when he ruled catastrophism

by Dr John Klinck on 07/15/11. For personal use only. inadmissible. But external and fundamentally unpredictable events cannot be dismissed any longer as a force in Earth’s history. They do give occasion to the study of the behavior of Earth’s systems in response to severe disturbance, a theme that is relevant for the study of history throughout the Cenozoic, as well (Thomas 2007). Unfortunately, the relevancy of such studies to our own time is Annu. Rev. Marine. Sci. 2011.3:1-34. Downloaded from www.annualreviews.org rapidly becoming evident, with measurable acceleration in sea level rise within the past century, and measurable increases in ocean temperature.

DISCLOSURE STATEMENT The author is not aware of any afﬁliations, memberships, funding, or ﬁnancial holdings that might be perceived as affecting the objectivity of this review.

ACKNOWLEDGMENTS Like any other scientist writing a review, I owe thanks to a multitude of colleagues and students. I do not list them here, fearing that in naming some I would neglect others, with equal claims

28 Berger MA03CH01-Prefatory-Berger ARI 17 November 2010 6:37

on my gratitude. My debts to my teachers, collaborators, shipmates, and other colleagues are documented to some degree in the citations but with signiﬁcant gaps. Apologies.

LITERATURE CITED Alley R, Clark PU, Huybrechts P, Joughin I. 2005. Ice sheets and sea-level changes. Science 310:456–60 Alvarez LW, Alvarez W, Asaro F, Michel HV. 1980. Extraterrestrial cause for the Cretaceous-Tertiary extinction. Science 208:1095–108 Arrhenius GOS. 1952. Sediment cores from the east Pacific. Rep. Swed. Deep-Sea Exped. 1947–1948 5(1):1–227 Baker P, McNutt M. 1996. The Future of Marine Geology and Geophysics. Report of a Workshop, Ashland Hills, Oregon. National Science Foundation, Washington, DC. 259 pp. and append. Baksi AK, Hsu V, McWilliams MO, Farrar E. 1992. 40Ar/39Ar dating of the Brunhes-Matuyama geomagnetic field reversal. Science 256:356–57 Bard E, Hamelin B, Fairbanks RG, Zindler A. 1990. Calibration of the 14C timescale over the past 30,000 years using mass spectrometric U-Th ages from Barbados corals. Nature 345:405–10 Bassinot FC, Labeyrie LD, Vincent E, Quidelleur X, Shackleton NJ, Lancelot Y. 1994. The astronomical theory of climate and the age of the Brunhes-Matuyama magnetic reversal. Earth Planet. Sci. Lett. 126:91– 108 Baumgartner TR, Michaelsen JM, Thompson LG, Shen GT, Soutar A, Casey RE. 1989. The recording of interannual climatic change by high-resolution natural systems: tree-rings, coral bands, glacial ice layers, and marine varves. In Aspects of Climate Variability in the Pacific and the Western Americas,ed. DH Peterson, vol. 55, pp. 1–14. Am. Geophys. Union Monogr. Geophys. Monogr. AGU, Washington, DC Becker K, et al., eds. 2002. Achievements and opportunities of scientific ocean drilling. The legacy of the Ocean Drilling Program. JOIDES J. 28(1):1–97 Berger A, Imbrie J, Hays J, Kukla G, Saltzman B, eds. 1984. Milankovitch and Climate: Understanding the Response to Astronomical Forcing. Dordrecht, Neth.: Reidel. 2 vols., 895 pp. Berger A, Loutre MF. 1991. Insolation values for the climate of the last 10 million of years. Quat. Sci. Rev. 10:297–317 Berger WH. 1968. Planktonic foraminifera: selective solution and paleoclimatic interpretation. Deep-Sea Res. 15:31–43 Berger WH. 1979. Impact of deep-sea drilling on paleoceanography. In Deep Drilling Results in the Atlantic Ocean: Continental Margins and Paleoenvironment, ed. M Talwani, W Hay, WBF Ryan. Am. Geophys. Union Maurice Ewing Ser. 3:297–314 by Dr John Klinck on 07/15/11. For personal use only. Berger WH. 2007a. Cenozoic cooling, Antarctic nutrient pump, and the evolution of whales. In Paleoceanog- raphy and Paleoclimatology of the Southern Ocean: A Synthesis of Three Decades of Scientific Ocean Drilling,ed. D Warnke, GM Filippelli, J-A Flores. Deep-Sea Res. II 54:2399–421 Berger WH. 2007b. On the discovery of the Ice Age: science and myth. In Myth and Geology,ed. Annu. Rev. Marine. Sci. 2011.3:1-34. Downloaded from www.annualreviews.org L Piccardi, WB Masse, vol. 273, pp. 271–78. Geol. Soc. Lond., Spec. Publ., London Berger WH. 2008. Sea level in the Late Quaternary: patterns of variation and implications. Int. J. Earth Sci. (Geol. Rundsch.) 97:1143–50 Berger WH. 2009. Ocean—Reflections on a Century of Exploration. Berkeley: University of California Press. 519 pp. Berger WH, Bickert T, Schmidt H, Wefer G. 1993. Quaternary oxygen isotope record of pelagic foraminifers: Site 806, Ontong Java Plateau. Proc. Ocean Drilling Program, Sci. Results 130:381–95 Berger WH, Herguera JC. 1992. Reading the sedimentary record of the ocean’s productivity. In Primary Productivity and Biogeochemical Cycles in the Sea, ed. PG Falkowski, AD Woodhead, pp. 455–86. New York: Plenum Berger WH, Jansen E. 1995. Younger Dryas episode: ice collapse and super-fjord heat pump. In The Younger Dryas, ed. SR Troelstra, JE van Hinte, GM Ganssen, pp. 61–105. Amsterdam: North-Holland Berger WH, Roth PH. 1975. Oceanic micropaleontology: progress and prospects. Rev. Geophys. Space Phys. 13:561–85, 624–35

www.annualreviews.org • Geologist at Sea: Aspects of Ocean History 29 MA03CH01-Prefatory-Berger ARI 17 November 2010 6:37

Berger WH, Schimmelmann A, Lange CB. 2004. Tidal cycles in the sediments of Santa Barbara Basin. Geology 32:329–32 Berger WH, Wefer G. 1996. Expeditions into the past: paleoceanographic studies in the South Atlantic. In The South Atlantic: Present and Past Circulation, ed. G Wefer et al., pp. 363–410. Berlin: Springer Berger WH, Wefer G. 2003. On the dynamics of the ice ages: Stage-11 Paradox, mid-Brunhes climate shift, and 100-ky cycle. In Earth’s Climate and Orbital Eccentricity: The Marine Isotope Stage 11 Question,ed. AW Droxler, RZ Poore, LH Burckle, vol. 137, pp. 41–59. Am. Geophys. Union Monogr. AGU, Wash- ington, DC Berger WH, Wefer G. 2009. Marine biogenic sediments. In Encyclopedia of Paleoclimatology and Ancient Envi- ronments, ed. V Gornitz, pp. 525–33. Dordrecht, Neth.: Springer Berger WH, Winterer EL. 1974. Plate stratigraphy and the ﬂuctuating carbonate line. In Pelagic Sediments on Land and Under the Sea,ed.KJHsu,¨ H Jenkyns, vol. 1, pp. 11–48. Spec. Publ. Intl. Assoc. Sediment Water Sci., Nedlands, Aust. Berggren WA. 1972. Late Pliocene–Pleistocene glaciation. In Initial Reports DSDP Vol. 12,ed. A Laughton, WA Berggren, et al., pp. 953–63. Washington DC: U.S. Gov. Print. Off Berggren WA, van Couvering JA, Eds. 1984. Catastrophes and Earth History. Princeton, NJ: Princeton Univ. Press. 464 pp. Boersma A, Shackleton N, Hall M, Given Q. 1979. Carbon and oxygen isotope records at DSDP Site 384 (North Atlantic) and some Paleocene paleotemperatures and carbon isotope variations in the Atlantic Ocean. In Initial Reports DSDP, ed. BE Tucholke, PR Vogt, et al., vol. 43, pp. 695–717. Washington, DC: U.S. Gov. Print. Off. Bolli HM, Saunders JB, Perch-Nielsen K, eds. 1985. Plankton Stratigraphy. Cambridge, UK: Cambridge Univ. Press. 1032 pp. Boyle E. 1995. Last-glacial maximum North Atlantic Deep Water: on, off or somewhere in-between. Philos. Trans. R. Soc. Lond. B 348:243–53 Bramlette MN. 1965. Massive extinctions of biota at the end of Mesozoic time. Science 148:1696–99 Broccoli AJ, Marciniak EP. 1996. Comparing simulated glacial climate and paleodata: a reexamination. Pale- oceanography 11:3–14 Broecker WS. 1973. Factors controlling CO2 content in the oceans and atmosphere. In Carbon and the Biosphere, ed. GM Woodwell, EV Pecan, vol. 30, pp. 32–50. At. Energ. Comm., Symp. Ser., Washington, DC Broecker WS. 1982. Ocean chemistry during glacial time. Geochim. Cosmochim. Acta 46:1689–705 Broecker WS, Denton GH. 1989. The role of ocean-atmosphere reorganizations in glacial cycles. Geochim. Cosmochim. Acta 53:2465–501 Broecker WS, Van Donk J. 1970. Insolation changes, ice volumes, and the O-18 record in deep-sea cores. Rev. Geophys. Space Phys. 8:169–98 by Dr John Klinck on 07/15/11. For personal use only. Bukry D, Bramlette MN. 1969. Coccolith age determinations—Leg 1, Deep Sea Drilling Project. Initial Rep. Deep Sea Drill. Proj. 1:369–87 Calvert SE. 1971. Composition and origin of North Atlantic deep-sea cherts. Contr. Min. Petrol. 33:273–88 CLIMAP Project Members. 1976. The surface of the ice-age Earth. Science 191:1131–37 Annu. Rev. Marine. Sci. 2011.3:1-34. Downloaded from www.annualreviews.org CLIMAP Project Members. 1981. Seasonal reconstructions of the Earth’s surface at the last glacial maximum. Map Chart Ser., MC36. Geol. Soc. Am., Boulder, Colo. Coxall HK, Wilson PA, Palike¨ H, Lear C, Backman J. 2005. Rapid stepwise onset of Antarctic glaciation and deeper calcite compensation in the Paciﬁc Ocean. Nature 433:53–57 Croll J. 1875. Climate and Time in Their Geological Relations—A Theory of Secular Changes of the Earth’s Climate. New York: Appleton. 577 pp. Crowley TJ, North GR. 1991. Paleoclimatology. New York: Oxford Univ. Press. 339 pp. Cullen V, Clark L, eds. 1994. 25 years of ocean drilling. Oceanus 36:136 Daly RA. 1934. The Changing World of the Ice Age. New Haven, CT: Yale Univ. Press. 271 pp. Dansgaard W, Johnsen SJ, Clausen HB, Langway CC. 1971. Climatic record revealed by the camp century ice core. In The Late Cenozoic Glacial Ages, ed. KK Turekian, pp. 37–56. New Haven, CT: Yale Univ. Press Dansgaard W, Tauber H. 1969. Glacier oxygen-18 content and Pleistocene ocean temperatures. Science 166:499–502

30 Berger MA03CH01-Prefatory-Berger ARI 17 November 2010 6:37

DeConto R, Pollard D. 2003. Rapid Cenozoic glaciation of Antarctica induced by declining atmospheric CO2. Nature 421:245–49 DeConto R, Pollard D, Harwood D. 2007. Sea ice feedback and Cenozoic evolution of Antarctic climate and ice. Paleoceanography 22:1–18 Denton GH, Hughes TJ. 1981. The Last Great Ice Sheets. New York: Wiley. 484 pp. Emiliani C. 1955. Pleistocene temperatures. J. Geol. 63:538–78 Emiliani C. 1966. Paleotemperature analysis of Caribbean cores P6304-8 and P6304-9 and a generalized temperature curve for the past 425,000 years. J. Geol. 74:109–26 Emiliani C. 1992. Planet Earth—Cosmology, Geology, and the Evolution of Life and Environment. Cambridge, UK: Cambridge Univ. Press. 719 pp. Ericson DB, Wollin G. 1956. Correlation of six cores from the equatorial Atlantic and the Caribbean. Deep-Sea Res. 3:104–25 Fairbanks RG. 1989. A 17,000-year glacio-eustatic sea level record: influence of glacial melting rates on the Younger Dryas event and deep-ocean circulation. Nature 342:637–42 Field DB, Baumgartner TR, Charles CD, Ferreira-Bartrina V, Ohman MD. 2006. Planktonic foraminifera of the California Current reflect 20th-century warming. Science 311:63–66 FinlaysonC,Carrion´ JS. 2007. Rapid ecological trunover and its impact on Neanderthal and other human populations. Trends Ecol. Evol. 22:213–22 Fischer G, Wefer G, Eds. 1999. Use of Proxies in Paleoceanography. Berlin: Springer. 735 pp. Flohn H. 1985. Das Problem der Klimaänderungen in Vergangenheit und Zukunft. Darmstadt, Ger.: Wissenschaftliche Buchgesellschaft. 228 pp. Gates WL. 1976. Modeling the ice-age climate. Science 191:1138–44 Gersonde R, Hodell DA, Eds. 2002. Southern Ocean Eocene-Pleistocene stratigraphies—insight from ocean drilling. Mar. Micropaleontol. 45:191–400 Gornitz V, Ed. 2009. Encyclopedia of Paleoclimatology and Ancient Environments. Dordrecht, Neth.: Springer. 1047 pp. Haq BU. 1981. Paleogene paleoceanography: early Cenozoic oceans revisited. Oceanol. Acta 4(Suppl.):71–82 Hays JD, Imbrie J, Shackleton NJ. 1976. Variations in the Earth’s orbit: pacemaker of the ice ages. Science 194:1121–32 Heinrich H. 1988. Origin and consequences of cyclic ice rafting in the Northeast Atlantic Ocean during the past 130,000 years. Quat. Res. 29:142–52 Hewitt CD, Stouffer RJ, Broccoli AJ, Mitchell JFB, Valdes PJ. 2003. The effect of ocean dynamics in a coupled GCM simulation of the Last Glacial Maximum. Clim. Dyn. 20:203–18 Hsu¨ KJ. 1992. Challenger at Sea—A Ship That Revolutionized Earth Science. Princeton, NJ: Princeton Univ. by Dr John Klinck on 07/15/11. For personal use only. Press. 417 pp. Hsu¨ KJ, Cita MB, Ryan WBF. 1973. The origin of the Mediterranean evaporites. Initial. Rep. DSDP 13:1203– 31 Hsu¨ KJ, McKenzie J. 1985. A “Strangelove” ocean in the earliest Tertiary. In The Carbon Cycle and Atmospheric Annu. Rev. Marine. Sci. 2011.3:1-34. Downloaded from www.annualreviews.org CO2: Natural Variations Archean to Present, ed. WS Broecker, ET Sundquist. Am. Geophys. Union Geophys. Monogr. Ser. 32:487–92 Hughes T. 1987. Ice dynamics and deglaciation models when ice sheets collapsed. In North America and Adjacent Oceans During the Last Deglaciation, ed. WF Ruddiman, HE Wright, vol. K-3, pp. 183–220. Geol. Soc. Am., Boulder, Colo. Imbrie J, Hays JD, Martinson DG, McIntyre A, Mix AC, et al. 1984. The orbital theory of Pleistocene climate: support from a revised chronology of the marine δ18O record. In Milankovitch and Climate, ed. AL Berger, J Imbrie, JD Hays, J Kukla, J Saltzman, Part 1, pp. 269–305. Hingham, Mass.: Reidel Imbrie J, Imbrie KP. 1979. Ice ages: solving the mystery. Short Hills, NJ: Enslow. 224 pp. Imbrie J, Kipp NG. 1971. A new micropaleontological method for quantitative paleoclimatology: application to a late Pleistocene Caribbean core. In The Late Cenozoic Glacial Ages, ed. KK Turekian, pp. 71–181. New Haven, CT: Yale Univ. Press Jansen E, Sjøholm J. 1991. Reconstruction of glaciation over the past 6 million years from ice-borne deposits in the Norwegian Sea. Nature 349:600–604

www.annualreviews.org • Geologist at Sea: Aspects of Ocean History 31 MA03CH01-Prefatory-Berger ARI 17 November 2010 6:37

Kappel ES, Farrell JW, Eds. 1997. ODP’s Greatest Hits. Jt. Oceanogr. Inst., Washington, DC. 29 pp. Keller G, Barron JA. 1983. Paleoceanographic implications of Miocene deep-sea hiatuses. Geol. Soc. Am. Bull. 94:590–613 Kennett JP. 1982. Marine Geology. Englewood Cliffs, NJ: Prentice-Hall. 813 pp. Kennett JP, Houtz RE, Andrews PB, Edwards AR, Gostin VA, et al. 1974. Development of the Circum- Antarctic Current. Science 186:144–47 Kennett JP, Shackleton NJ. 1976. Oxygen isotopic evidence for the development of the psychrosphere 38 Myr ago. Nature 260:513–15 Kennett JP, Warnke DA, eds. 1992 . The Antarctic Paleoenvironment: A Perspective on Global Change. Antarc. Res. Ser., Part I, vol. 56. Am. Geophys. Union, Washington, DC. 384 pp. Koeberl C, Montanari A, eds. 2009. The Late Eocene Earth—Hothouse, Icehouse, and Impacts, vol. 452. Geol. Soc. Am., Spec. Pap. Boulder, Colo. 322 pp. Koppen¨ W, Wegener A. 1924. Die Klimate der geologischen Vorzeit. Berlin: Borntrager.¨ 266 pp. Larsen HC, Saunders AD, Clift PD, Beget J, Wei W, Spezzaferri S, and ODP 152 Scientific Party. 1994. Seven million years of glaciation in Greenland. Science 264:952–55 Lazarus DB, Kotrc B, Wulf G, Schmidt DN. 2009. Radiolarians decreased silicification as an evolutionary response to reduced Cenozoic ocean silica availability. Proc. Natl. Acad. Sci. USA 106:9333–38 Lipps JH, Mitchell ED. 1976. Trophic model for the adaptive radiations and extinctions of pelagic marine mammals. Paleobiology 2:147–55 Lisiecki LE, Raymo ME. 2005. A Pliocene-Pleistocene stack of 57 globally distributed benthic δ18O records. Paleoceanography 20:PA1003 Luterbacher HP, Premoli-Silva I. 1964. Biostratigrafia del limite Cretaceo-Terziario. Riv. Ital. Paleontol. 70:67–128 MacAyeal DR. 1993. Binge/purge oscillations of the Laurentide Ice Sheet as a cause of the North Atlantic’s Heinrich Events. Paleoceanogr 8:775–84 Manabe S, Broccoli AJ. 1985. A comparison of climate model sensitivity with data from the last glacial maximum. J. Atmos. Sci. 42:2643–51 McGowran B. 2005. Biostratigraphy—Microfossils and Geological Time. Cambridge, UK: Cambridge Univ. Press. 459 pp. McIntyre A, McIntyre R. 1971. Coccolith concentrations and differential solution in oceanic sediments. In The Micropalaeontology of the Oceans, ed. BM Funnell, WR Riedel, pp. 253–61. Cambridge, UK: Cambridge Univ. Press Milankovitch M. 1930. Mathematische Klimalehre und Astronomische Theorie der Klimaschwankungen. Handbuch der Klimatologie, Bd. 1. Berlin: Borntrager.¨ 176 pp. Miller KG, Fairbanks RG, Mountain GS. 1987. Tertiary oxygen isotope synthesis, sea level history, and by Dr John Klinck on 07/15/11. For personal use only. continental margin erosion. Paleoceanography 2:1–19 Moore TC. 1969. Radiolaria: change in skeletal weight and resistance to solution. Geol. Soc. Am. Bull. 80:2103–8 Mountain GS, Katz ME, eds. 1991. Advisory Panel Report on Earth System History. Natl. Sci. Found., Washington, DC. 99 pp. Annu. Rev. Marine. Sci. 2011.3:1-34. Downloaded from www.annualreviews.org Muller RA, MacDonald GJ. 2000. Ice Ages and Astronomical Causes. Berlin: Springer. 318 pp. Munk WH. 1966. Abyssal recipes. Deep-Sea Res. 13:707–30 Muttoni G, Kent DV. 2007. Widespread formation of cherts during the early Eocene climate optimum. Palaeogeogr. Palaeoclimatol. Palaeoecol. 253:348–62 Olausson E. 1965. Evidence of climatic changes in North Atlantic deep-sea cores, with remarks on isotopic paleotemperature analysis. Progr. Oceanogr. 3:221–52 Parker FL. 1958. Eastern Mediterranean Foraminifera. Rep. Swedish Deep-Sea Exped., 1947–1948 8:219–83 Parker FL. 1962. Planktonic foraminiferal species in Pacific sediments. Micropaleontology 8:219–54 Parker FL, Berger WH. 1971. Faunal and solution patterns of planktonic foraminifera in surface sediments of the South Pacific. Deep-Sea Res. 18:73–107 Peltier WR. 1994. Physics of the ice age cycle. In Long-Term Climatic Variations, ed. JC Duplessy, MT Spyridakis, pp. 453–79. Berlin: Springer Petit JR, Jouzel J, Raynaud D, Barkov NI, Barnola JM, et al. 1999. Climate and atmospheric history of the past 420,000 years from the Vostok ice core, Antarctica. Nature 399:429–36

32 Berger MA03CH01-Prefatory-Berger ARI 17 November 2010 6:37

Pflaumann U, Duprat J, Pujol C, Labeyrie LD. 1996. SIMMAX: a modern analog technique to deduce Atlantic sea surface temperatures from planktonic foraminifera in deep-sea sediments. Paleoceanography 11:15–35 Phleger FB, Parker FL, Peirson JF. 1953. North Atlantic Foraminifera. Rep. Swedish Deep-Sea Expedition 1947–1948 7:3–122 Pisias NG, Moore TC. 1981. The evolution of Pleistocene climate: a time series approach. Earth Planet. Sci. Lett. 52:450–56 Pomerol C, Premoli-Silva I, eds. 1986. Terminal Eocene Events. Amsterdam: Elsevier. 420 pp. Prothero DR, Berggren WA, eds. 1992. Eocene–Oligocene Climatic and Biotic Evolution. Princeton, NJ: Princeton Univ. Press. 566 pp. Prothero DR, Ivany LC, Nesbitt EA, Eds. 2003. From Greenhouse to Icehouse, The Marine Eocene-Oligocene Transition. New York: Columbia Univ. Press. 541 pp. Purdy EG, Winterer EL. 2006, Contradicting barrier reef relationships for Darwin’s evolution of reef types. Int. J. Earth Sci. 95:143–67 Raup DM. 1985. The Nemesis Affair: A Story of the Death of the Dinosaurs and the Ways of Science. New York: Norton. 240 pp. Ryan WBF, Hsu¨ KJ, eds. 1973. Initial Reports of the Deep Sea Drilling Project. Leg 13. U.S. Gov. Print. Off., Washington, DC Ryther JH. 1969. Photosynthesis and fish production in the sea. Science 166:72–76 Savin SM, Douglas RG, Stehli FG. 1975. Tertiary marine paleotemperatures. Geol. Soc. Am. Bull. 86:1499–510 Schimmelmann A, Lange CB. 1996. Tales of 1,001 varves: a review of Santa Barbara Basin sediment studies. In Palaeoclimatology and Palaeoceanography from Laminated Sediments, ed. AES Kemp, vol. 116, pp. 121–41. Geol. Soc. Lond., Spec. Publ., London Schott W. 1935. Die Foraminiferen in dem aquatorialen¨ Teil des Atlantischen Ozeans. Wiss. Ergebn. Deut. Atlant. Exped. Vermess. Forschungsschiff Meteor 1925–1927, vol. 3, pp. 43–134. Berlin: Reimer Schulz M, Berger WH, Sarnthein M, Grootes PM. 1999. Amplitude variations of 1470-yr climate oscillations during the last 100,000 years linked to fluctuations of continental ice mass. Geophys. Res. Lett. 26:3385–88 Seibold E, Berger WH. 1996. The Sea Floor. Berlin: Springer. 356 pp. 3rd ed. Shackleton NJ, Berger A, Peltier WR. 1990. An alternative astronomical calibration of the lower Pleistocene timescale based on ODP Site 677. Trans. R. Soc. Edinburgh, Earth Sci. 81:251–61 Shackleton NJ, Kennett JP. 1975. Paleotemperature history of the Cenozoic and the initiation of Antarctic glaciation: oxygen and carbon isotope analyses in DSDP Sites 277, 279 and 281. Initial Rep. Deep Sea Drill. Proj. 29:743–55 Shepard FP. 1963. Submarine Geology. New York: Harper. 557 pp. 2nd ed. Shepard FP. Curray JR. 1967. Carbon-14 determination of sea level changes in stable areas. Prog. Oceanogr. 4:283–91 by Dr John Klinck on 07/15/11. For personal use only. Shin SI, Liu Z, Otto-Bliesner BL, Brady EC, Kutzbach JE, Harrison SP. 2003. A simulation of the Last Glacial Maximum Climate using the NCAR CSM. Clim. Dyn. 20:127–51 Simmons J. 1996. The Scientific 100. New York: Citadel. 504 pp. Soutar A, Isaacs JD. 1969. History of fish populations inferred from fish scales in anaerobic sediments off Annu. Rev. Marine. Sci. 2011.3:1-34. Downloaded from www.annualreviews.org California. Calif. Coop. Ocean. Fisher. Investig. Rep. 13:63–70 Steele JH, Turekian KK, Thorpe SA, eds. 2001. Encyclopedia of Ocean Sciences. San Diego: Academic. 6 vol., 3,399 pp. Suess HE. 1956. Absolute chronology of the last glaciation. Science 123:355–57 Summerhayes CP, Prell WL, Emeis KC, eds. 1992. Upwelling systems: evolution since the early Miocene, vol. 62. Geol. Soc. Lond., Spec. Publ., London. 519 pp. Sverdrup HU. 1953. On conditions for the vernal blooming of phytoplankton. J. Cons. Explor. Mer 18:287–95 Thierstein HR, Berger WH. 1978. Injection events in ocean history. Nature 276:461–66 Thierstein HR, Okada H. 1979. The Cretaceous/Tertiary boundary event in the North Atlantic. Initial Rep. Deep Sea Drill. Proj. 43:601–16 Thomas E. 2007. Cenozoic mass extinctions in the deep sea: What perturbs the largest habitat on Earth? vol. 424, pp. 1–23. Geol. Soc. Am., Spec. Pap., Boulder, Colo. van Andel TjH, Thiede J, Sclater JG, Hay WW. 1977. Depositional history of the south Atlantic Ocean during the last 125 million years. J. Geol. 85:651–98

www.annualreviews.org • Geologist at Sea: Aspects of Ocean History 33 MA03CH01-Prefatory-Berger ARI 17 November 2010 6:37

Vincent E, Berger WH. 1985. Carbon dioxide and polar cooling in the Miocene: the Monterey hypothesis. In The Carbon Cycle and Atmospheric CO2: Natural Variations Archean to Present, ed. ET Sundquist, WS Broecker, vol. 32, pp. 455–68. Am. Geophys. Union Monogr., Am. Geophys. Union, Washington, DC Warme JE, Douglas RG, Winterer EL. 1981. The Deep Sea Drilling Project: A Decade of Progress, vol. 32. Soc. Econ. Paleontol. Minerol., Spec. Publ., Tulsa, Okla. 564 pp. Wefer G, Berger WH. 1991. Isotope paleontology: growth and composition of extant calcareous species. Mar. Geol. 100:207–48 Wefer G, Berger WH, Siedler G, Webb DJ, eds. 1996. The South Atlantic: Present and Past Circulation. Berlin: Springer. 735 pp. Winter A, Siesser WG. 1994. Coccolithophores. Cambridge, UK: Cambridge Univ. Press. 242 pp. Winterer EL. 2009. Atolls. In Encyclopedia of Islands, ed. RG Gillespie, DA Clague, pp. 67–70. Berkeley: Univ. of Calif. Press Zachos J, Kump L. 2005. Carbon cycle feedbacks and the initiation of Antarctic glaciation in the earliest Oligocene. Glob. Planet. Chang. 47:51–66 Zachos J, Pagani M, Sloan L, Thomas E, Billups K. 2001. Trends, rhythms, and aberrations in global climate 65 Ma to present. Science 292:686–93 by Dr John Klinck on 07/15/11. For personal use only. Annu. Rev. Marine. Sci. 2011.3:1-34. Downloaded from www.annualreviews.org

34 Berger MA03-FrontMatter ARI 17 November 2010 7:6

Annual Review of Marine Science Contents

Volume 3, 2011

Geologist at Sea: Aspects of Ocean History Wolfgang H. Berger ppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppp1 Submarine Paleoseismology Based on Turbidite Records Chris Goldﬁnger pppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppp35 Natural Processes in Delta Restoration: Application to the Mississippi Delta Chris Paola, Robert R. Twilley, Douglas A. Edmonds, Wonsuck Kim, David Mohrig, Gary Parker, Enrica Viparelli, and Vaughan R. Voller pppppppppppppppp67 Modeling the Dynamics of Continental Shelf Carbon Eileen E. Hofmann, Bronwyn Cahill, Katja Fennel, Marjorie A.M. Friedrichs, Kimberly Hyde, Cindy Lee, Antonio Mannino, Raymond G. Najjar, John E. O’Reilly, John Wilkin, and Jianhong Xue pppppppppppppppppppppppppppppppppppppp93

Estuarine and Coastal Ocean Carbon Paradox: CO2 Sinks or Sites of Terrestrial Carbon Incineration? Wei-Jun Cai pppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppp123 Emerging Topics in Marine Methane Biogeochemistry David L. Valentine ppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppp147 by Dr John Klinck on 07/15/11. For personal use only.

Observations of CFCs and SF6 as Ocean Tracers Rana A. Fine pppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppp173

Annu. Rev. Marine. Sci. 2011.3:1-34. Downloaded from www.annualreviews.org Nitrogen Cycle of the Open Ocean: From Genes to Ecosystems Jonathan P. Zehr and Raphael M. Kudela pppppppppppppppppppppppppppppppppppppppppppppppp197 Marine Primary Production in Relation to Climate Variability and Change Francisco P. Chavez, Monique Messi´e, and J. Timothy Pennington ppppppppppppppppppppp227 Beyond the Calvin Cycle: Autotrophic Carbon Fixation in the Ocean Michael H¨ugler and Stefan M. Sievert pppppppppppppppppppppppppppppppppppppppppppppppppppp261 Carbon Concentrating Mechanisms in Eukaryotic Marine Phytoplankton John R. Reinfelder pppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppp291

vi MA03-FrontMatter ARI 17 November 2010 7:6

Microbial Nitrogen Cycling Processes in Oxygen Minimum Zones Phyllis Lam and Marcel M.M. Kuypers ppppppppppppppppppppppppppppppppppppppppppppppppppp317 Microbial Metagenomics: Beyond the Genome Jack A. Gilbert and Christopher L. Dupont ppppppppppppppppppppppppppppppppppppppppppppppp347 Environmental Proteomics: Changes in the Proteome of Marine Organisms in Response to Environmental Stress, Pollutants, Infection, Symbiosis, and Development Lars Tomanek ppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppp373 Microbial Extracellular Enzymes and the Marine Carbon Cycle Carol Arnosti pppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppp401 Modeling Diverse Communities of Marine Microbes Michael J. Follows and Stephanie Dutkiewicz pppppppppppppppppppppppppppppppppppppppppppp427 Bioﬁlms and Marine Invertebrate Larvae: What Bacteria Produce That Larvae Use to Choose Settlement Sites Michael G. Hadﬁeld pppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppp453 DNA Barcoding of Marine Metazoa Ann Bucklin, Dirk Steinke, and Leocadio Blanco-Bercial pppppppppppppppppppppppppppppppp471 Local Adaptation in Marine Invertebrates Eric Sanford and Morgan W. Kelly ppppppppppppppppppppppppppppppppppppppppppppppppppppppp509 Use of Flow Cytometry to Measure Biogeochemical Rates and Processes in the Ocean Michael W. Lomas, Deborah A. Bronk, and Ger van den Engh ppppppppppppppppppppppppp537 The Impact of Microbial Metabolism on Marine Dissolved Organic Matter by Dr John Klinck on 07/15/11. For personal use only. Elizabeth B. Kujawinski pppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppp567

Errata Annu. Rev. Marine. Sci. 2011.3:1-34. Downloaded from www.annualreviews.org

An online log of corrections to Annual Review of Marine Science articles may be found at http://marine.annualreviews.org/errata.shtml

Contents vii