This chapter explores some of the techniques used
by earth scientists to unravel the unwritten histories of past oceans and
climate. In the absence of direct instrumental observations, changes in
these past environments need to be reconstructed from indirect information,
much of which is obtained from fossils and chemical properties in sea-floor
sediments. These sediments continually form at the sea floor because of
settling clay and silt, which is eroded from land and transported into
the oceans by rivers and as wind-blown dust. Another important component
of sea-floor sediments consists of skeletal parts of marine organisms.
In central oceanic settings, far from coastlines, accumulation of marine
sediments may be as low as 5 to 10 mm per thousand years, but deep-sea
values in the land-locked Mediterranean are typically about four times
higher than that. Within the Aegean Sea, with its countless islands, rivers,
and bays, values up to 150 to 300 mm per thousand years are not uncommon.
Earth scientists take samples from the sea floor, and examine a wealth of different characteristics of these sediments. Each of these characteristics informs us about a particular aspect of the environments in and around the basin. The information is, however, never a direct measure of – for example – temperature, but rather presents itself as something that is in some way affected by temperature. Changes in the parameter studied are then calibrated – with varying degrees of success – to absolute temperature variations, via comparative studies of the modern environment. Similar exercises are performed for other measurable parameters, calibrated to other environmental factors. In the best case, several methods are developed to independently estimate, for example, temperature. The different results can then be compared and mutually validated, which greatly enhances the confidence in past environmental reconstructions.
Thus, the earth scientist must learn how to approximate as accurately as possible the real environmental variations. To indicate an analytical parameter that approximates past temperature fluctuations, we use the term ‘temperature proxy’. In the ideal world, we would be able to develop reliable proxies for all major environmental factors, but in reality, many proxies depend on an interplay between several factors, so that a convoluted signal is received from the analyses of sedimentary sequences. It requires great ingenuity and continuous new proxy development and calibration to unravel these hidden messages about the past environment. In many ways, such earth science studies bear a strong resemblance to forensic science in criminal investigations, where the sequence, nature, and timing of events surrounding a crime is reconstructed from fragmented pieces of information at the scene of the crime.
The list of proxies in the earth sciences is very long indeed. Relatively few, however, have grown into such general acceptance that they now form part of a standard investigative approach. Research on Mediterranean anoxic events has been very intensive since the Swedish Deep Sea Expedition (1947-1948) for the first time brought up sea-floor sediment cores with distinct anoxic intervals. Virtually every imaginable proxy has been applied to the investigation of these intervals, but the majority of conclusive information has come from a fairly limited set of proxies, all of which now form part of the accepted ‘standard tool kit’. I will limit myself to these in the discussion of our investigative approach, even though this somewhat short-changes people who have invested considerable time in developing some very revealing work using other proxies. However, this book is not intended as a comprehensive review of everything that has ever been analysed, but rather as an explanation in general terms of the main environmental changes that led to the Mediterranean anoxic events, with specific emphasis on the latest event of 9.5 to 6.0 thousand years ago.
For deeper penetrations, corers are used that consist of much narrower metal tubes, propelled into the sediment by massive weights – often 1,500 kilograms or more. Purely weight-driven corers are called gravity corers. In the fairly stiff sediments of the Mediterranean, gravity corers with 9 cm diameter core barrels typically penetrate 3 to 6 m into the sea bed. The next step up in sophistication involves a piston-assisted system, where an ingeniously rigged piston is pulled up within the core barrel, just above the sediment-water interface, as the corer penetrates into the sediment. Such systems are called piston corers. In Mediterranean sediments, good piston corers with 9 cm diameter core barrels typically penetrate up to 9 m. For a sediment accumulation rate of around 5 cm per thousand years in this basin, therefore, piston coring allows recovery of sediments down to an age of about 180,000 years B.P. This can be extended by targeting specific areas with lower accumulation rates, or by ‘lucky’ penetrations of more than 9 m.
For recovery of longer cores, two major players have dominated the scene. One is the French Calypso corer, a giant piston corer deployed from the research vessel Marion Dufresne that has delivered many Mediterranean cores of 25-30 m length or more. The undisputed champion is operated by the international Ocean Drilling Project (ODP), successor of the Deep Sea Drilling Project (DSDP). Its specific deep-sea coring approach can penetrate several hundreds of meters into the sediment. As a consequence of intensive deployment of all these coring techniques, attracted by the region’s intriguing geological history, the Mediterranean sea floor has been studied in great detail.
* * *
It is worthwhile to briefly consider the cost implications. A typical day’s shiptime on an average research vessel costs roughly US$ 15,000. A typical duration for a Mediterranean research cruise is three to four weeks, which therefore costs upwards of US$300,000. A successful cruise recovers around 30 piston cores over such a period of time, but around five will commonly be damaged, bent, ‘pistonned’ (sucked by too vigorous action of the piston), lost (cable failure), etc., so that a rough estimate of 25 cores arrives on deck in good order. If we are really lucky, then about one fifth of these will be free of geological complications and offer exceptional records of the history of the sea floor through the entire time of their deposition. Let’s therefore say that about 5 cores will be of high value, while the others will be of limited value. The investment of a quarter of a million dollars – not counting the salaries of those on board or the laboratory consumables – will be repaid by five high quality cores of about 9 m length. The interpretation of this overview depends on one’s outlook. For an accountant, it works out to more than US$ 6,600 per meter of mud. For a marine geologist, it represents an extremely good deal and a highly successful sampling campaign, since each quality core offers an incredibly rich archive of information on past ocean and climate states, along with their myriad ecological and chemical responses. Because all marine geologists are aware of the considerable financial implications, the brunt of which is borne by national and federal research councils, exceptionally good sediment cores are treated with the utmost respect and care. Teaching and new technique development invariably centers on the lesser quality material.
Is it really worth the expense? Allow me to rephrase this into “Do we want to know what is happening to our climate, the oceans, and the ecosystems?” Research into the recent geological history of the oceans, intricately linked to climate, provides crucial insights into the natural variability. This must be understood before we can even dream about distinguishing any man-induced effects. Is the recent climatic warming perhaps a natural phenomenon? The recent geological history suggests that it may be to some extent, but also shows that both the rate and the magnitude of warming today appear anomalously rapid and great, relative to nature’s own experiments. There is no other way of obtaining such essential background information than by environmental geological research. Could this not be done cheaper on land, rather than in the oceans? Possibly, but not as completely. More than 70% of the world is covered by oceans, and the oceans form an integral part of the climate system. If we were to ignore the information about their past variability, therefore, we would not be able to properly understand past climatic changes. In addition, marine sediments are much less affected by erosion than sediments on land, and a wider array of ‘proxies’ can be studied from marine records than from terrestrial sequences. It is not a question of either/or, but in fact many high-quality marine and terrestrial records are needed before past (and therefore future) climate states can be understood. If we are genuinely worried about climate change, then YES, it is worth the expense!
Upon completion of the survey along a predefined grid, a careful inspection of the profiles identifies the most promising coring sites. The inspection takes into account the specific topographic features – avoiding steep-walled depressions that could be filled with submarine mudslide material – and any obvious layering. Organised horizontal layering is often a good sign of undisturbed sediments. The exact coordinates for the two selected sites are determined from the geophysical profiles and the recorded ship’s track, and the ship heads for the first target. Waterdepth at the core site was calculated already from the geophysical profiles, and is checked by echosounder as the ship gets on station. The vessel is held accurately on station by a dynamical positioning system that relies on computerised interaction between the continuous high-definition satellite navigation and the vessel’s multiple thrusters. On deck, technicians prepare the piston corer, and deploy it. When the water is very deep at the core site – over 3,000 meters could be reached in the Mediterranean, but in the open ocean it could be thousands of meters more – it can take about an hour for the corer to approach the sea floor. On gravity and piston corers, the main winch cable connects to a release mechanism. This release, operated by a trigger weight, is designed to let the corer drop in free-fall within a separate cable loop between the corer and the release mechanism. When the trigger weight touches the sea bed, the release is activated, and the corer hurls towards the sea bed. It penetrates the sediments, and then is ready to be pulled back out and up to the ship. There will have been some loss of the fluffy top layer of sediment because it gets blown aside by the pressure wave that builds up in front of the free-falling corer.
Once secured on deck, the corer is systematically dismantled.
Inside the steel core barrel resides a PVC liner tube, like a drainpipe.
The base of the steel core barrel, the ‘shoe’, is unscrewed. This part
holds the ‘core catcher’, a leaf-spring device that stops the cored sediment
from falling out of the corer when it is retracted from the sea bed and
winched up trough the water column. After removal of the shoe and catcher,
the PVC liner is extracted from the core barrel. As it is pulled out, the
inner tube that contains the sediment is cut up in lengths of 1.0 or 1.5
m. These core sections are carefully labelled, and either capped at each
end for cold storage, or moved to the core lab aboard ship for opening
and description.
Whether in a shore-based or the ship-board core lab,
the labelled core segments are eventually split lengthwise. Special saw
or cutter benches are used to split the PVC liner, taking care to only
saw through the PVC and not into the sediments inside the tube. Next, a
guitar string or fishing line is pulled through the sediment, guided by
the two opposite lengthwise cuts in the PVC liner. This splits the sediments
inside the barrel, allowing the researchers to separate the two half core
barrels with their contained sediments. The result is two identical lengthwise
core halves. One half is sealed and stored carefully as ‘archive’, and
the other (‘working’) half is ready for investigation and sampling.
Today, many laboratories, even aboard ship, follow a series of standard descriptive measurements, where the working half is first visually described, and subsequently run over a so-called multi-sensor track. This instrument provides a standard set of measurements in 1 to 3 cm steps, recording information on density and magnetic property variations in the sediment. Often, digital photographs are made of the opened core sections, since colour changes can provide useful first-order information by which cores from different sites can be related easily to one another.
Not always, but frequently, routine X-radiographs are made of working core halves, to identify any sedimentary structures that are invisible to the eye. In our Mediterranean cruise, this technique is applied to investigate the extent to which burrowing or ‘bioturbation’ has affected the recovered sediments. The X-radiographs show that intensive burrowing has taken place in the normal, well-oxygenated, beige-brown intervals, indicating that there was a thriving fauna on and in the surface sediments – the technical name for such fauna is ‘benthic fauna’. However, the core also shows several anomalous olive-green to black layers, which show no trace of burrowing, as evidenced by a well-preserved fine microscopic lamination. The absence of burrowing and consequent preservation of the original (often seasonal) sedimentary laminae can only be explained in terms of a total absence of benthic fauna. This is a very strong indication that there was no free oxygen around in the bottom water for extended periods of time. Such an observation from the X-radiographs thus offers a crucial working hypothesis for testing by subsequent investigation of the sediments for the presence or absence of fossils of benthic organisms.
One core may not sound like much, but recovering and processing it keeps the technicians, researchers and their assistants pretty busy for the best part of half a day. Let’s look at the processing of a piston corer that penetrated to a depth of about 8 meters. Between the moment the ship took up station and the return of the corer back on deck after deployment, some 3 hours had elapsed. While the technicians extracted the core liner with the sediments, the ship had already set out towards the next core site. Eight one-meter core segments were taken to the description lab, and opened. The visual description, running them on the multi-sensor track, and digital photographing of the eight sections took about two hours. During this time, the archive halves were carefully sealed and stored. The X-ray assessment added up to two more hours. Then, just as the team thought they’d finished, the ship arrived at the next station, and the circus started all over…..
Often, nowadays, we take samples immediately in high resolution, and perform the intial reconnaissance study by selecting every fifth or tenth sample. This limits the amount of times we have to return to the core, taking it out of its protective wrappings, and also reduces the physical damage done to the cores during initial sampling, which could interfere with a later high-resolution sampling. Geophysicists working on the sediments’ magnetic properties (‘palaeomagnetists’ – where palaeo stands for ‘past/old’) have devised a way to take a ‘U-channel’, a 2 cm deep and 2 cm wide plastic tube over the entire length of the core. The magnetic measurements are non-destructive, and when they are finished, the sediment within the U-channel can be cut up in a continuous series of 0.5 or 1.0 cm slices for further analyses. This provides a very good way of constant-volume sampling, as all samples will be 2x2 cm in the original horizontal dimensions, and 0.5 or 1.0 cm thick in the original vertical direction. Also, taking a U-channel is more precise and less destructive to the core, compared with scooping out individual samples.
When the discrete samples have been retrieved, they are shared between the various scientists, and they then prepare the material according to their own specific requirements. Commonly, researchers from several different disciplines work on shared samples, thus ensuring that the signals from the various proxies can be exactly related to one another.
* * *
In the phytoplankton, there are two main groups that produce fossilising remains. First are the coccolithophores, unicellular algae that form a 2 to 3 hundredth millimeter diameter sphere consisting of calcium-carbonate platelets. The sphere falls apart as the organism decays, but the individual platelets fossilise well. These platelets are extremely small, between 1 and 5 thousandth of a millimeter. As a result of their small size, coccoliths are often referred to as ‘calcareous nanofossils’, which is a bit of an exaggeration since a nanometer in fact is a millionth of a millimeter. Each coccolithophore species forms platelets that are distinct in shape and size to those of other species. Thus, by identifying the platelets, we can develop an impression of the composition of the coccolithophore flora, and this composition reveals certain aspects of the environmental conditions. Coccoliths can be extremely abundant in sediments – we then call them ‘coccolith oozes’ – as witnessed in most dramatic form by the white chalk cliffs of Dover, which are almost pure coccolith ooze.
The second main group of phytoplankton that produces fossilising skeletal remains are the diatoms. The diatoms are among the world’s most abundant phytoplankton. Their hard parts do not consist of carbonate, but of opal, which is hydrated silicon oxide. Diatoms span a very wide range of sizes, from 5 thousandth of a millimeter up to 5 millimeters. Several diatom species form colonies, which reach dimensions of many centimeters. Often, opal will dissolve in sediments, so that the diatom record is lost. In certain cases, however, the delicate opaline diatom skeletons are preserved, and we can then quantify them on the basis of size and shape arguments, to reconstruct past assemblages. As is the case with the coccoliths, different diatom assemblages give information about specific environmental conditions. In the Mediterranean, opal preservation is generally very poor, but in exceptional cases there remains enough to offer a rare and intruiging insight into the diversity and abundance distribution of species within this important group of phytoplankton.
During the last decade, a third promising phytoplankton group has worked its way into the marine geologists’ attention. This group comprises the dinoflagellates. These organisms form highly resistant organic-walled cysts, with good preservation potential, and a general size range of 30 to 60 thousandth of a millimeter. There is a great variety of dinoflagellates, and interestingly, they straddle the plant-animal boundary. Some dinoflagellate species are distinctly like plants, with photosynthesis and all, but other species act as consumers of other phytoplankton, and hence behave more like zooplankton (animals). Dinoflagellate research is a big thing in the petroleum industry, but its application to past environmental reconstructions is still in its infancy.
In the zooplankton, the main group for marine geological studies is that of the foraminifera. They are unicellular animals that form shells with species-specific shapes and wall structures. Adult specimens typically range between 0.1 and 0.6 millimeters in size, but occasionally size fractions down to 60 thousandths of a millimeter or less need to be taken into account. Some foraminiferal species only have an organic membrane, and do not fossilise well. Others, the so-called ‘agglutinants’, form their shells by pasting together sand grains. The most important group of foraminifera for the marine geologist consists of species that form their shell from calcium carbonate. Several species of carbonate-shelled foraminifera live free floating in the surface couple of hundred meters of the water column (as they do not actively swim, they are called ‘planktonic’, for ‘drifting’). Others live on/in the sediment surface, where they can occupy niches down to as much as 10 cm below the sediment-water interface (‘benthic’ foraminifera). The agglutinants always have a benthic lifestyle. Diversity and abundance variations of planktonic and benthic foraminifera have been the subject of intensive study, both globally and for the Mediterranean in particular, contributing vast amounts of knowledge about past changes in the water column, and at the sea floor, respectively. In addition, many chemical proxies, including high-definition radiocarbon dating, rely on extremely sophisticated analyses of hand-picked foraminiferal shells. It would be true to state that foraminifera have formed a focal point for the majority of research into the past environmental changes that drove the Mediterranean anoxic events, either by investigation of their assemblage compositions, or of their shell chemistry.
Another group of zooplankton that produces usable fossil remains comprises the pteropods. Pteropods are tiny representatives of the gastropod family, which makes them distant cousins of the sea snails. They form shells of a more easily dissolved variant of calcium carbonate, called aragonite, and their abundances in sedimentary records consequently is strongly determined by changes in the intensity of dissolution. In some basins, for example the Red Sea, pteropods provide a wealth of environmental information. In the Mediterranean, their applicability has been rather limited.
There are two further zooplankton groups that produce fossilising remains. One comprises the radiolaria, which form opaline skeletons that are slightly more resistant to dissolution than those of the diatoms. The other concerns the ostracods, tiny crustaceans (distant cousins of the crabs and lobsters) with two carbonate shields around their body. These fossilise well, have specific shapes and ornamentations per species, and can be interpreted in terms of past environmental conditions. Very little attention has been give to these groups in studies of circulation and climate in the Mediterranean during the recent geological past. For this reason, I refrain from further discussing them.
One of the most widely researched microfossil groups from Mediterranean marine sediments concerns distinctly non-marine material. Pollen and spores from land-plants are transported by wind and by rivers over vast distances. They consequently accumulate not only in lakes and soils, but also in marine sediments – especially when the basin of study is as landlocked as the Mediterranean Sea. Pollen and spore records are sensitive indicators of climate change and its impact on on the terrestrial vegetation, especially when considering times before human deforestation became important (ie., before ~3,000 years BP in the Mediterranean region). Because pollen and spore fossils occur in both marine and terrestrial settings, they offer an excellent way of relating climate impact studies between the two domains. This is very important, since there are only very few methods for establishing the exact temporal relationship between records from land and those from the sea. In addition, pollen and spore records complement ocean-climate reconstructions by offering insight into the contemporary land-climate conditons. Pollen and spore studies have played a very prominent role in the deciphering of the climatic conditions that were responsible for Mediterranean anoxic events.
* * *
We speak of stable isotopes when the various isotopes are not susceptible radioactive decay. By far the most widely used stable isotopes in studies concerned with the reconstruction of past environments are those of oxygen and carbon. There are three stable isotopes of oxygen: 16O, 17O, and 18O, with relative natural abundances of 99.76%, 0.04% and 0.20%, respectively. Because of the higher total abundances and greater mass difference between 16O and 18O, research using oxygen isotopes normally concerns the ratio 18O/16O. Carbon occurs as two stable isotopes: 12C and 13C, with relative natural abundances of 98.89% and 1.11%, respectively. Oxygen and carbon are both represented in carbonates (CaCO3), and stable isotope ratios for both are therefore obtained from one single analysis, which uses the CO2 (carbon dioxide) released upon acidification of carbonate. This technique has been refined to be applicable to samples of as little as 10 millionths of a gram of carbonate. Hence, stable oxygen and carbon isotope ratios can be measured on single foraminiferal shells. Today, state-of-the-art equipment for stable oxygen and carbon isotope analyses from carbonates costs around a quarter of a million dollars. Annually, about 5000 analyses can be expected from such a system. New technologies are sure to reduce the minimum material requirement, but such improvements are likely to affect the capital costs as well. There are many other isotopes that can be analysed, and a current focus of development concerns the stable isotopes of boron in foraminiferal shells. In addition, stable carbon and hydrogen isotopes are frequently analysed from organic matter, where preserved in the sediments. To date, however, the most fundamental isotopic contribution to reconstructions of past marine environments has been made by analyses of the 18O/16O and 13C/12C ratios in carbonate fossils – in particular of foraminifera.
Why are scientists so interested in these particular stable isotope ratios? Firstly, because the technique is very widely applicable. Foraminifera live virtually anywhere in marine waters, and are abundant in most sedimentary sequences. Hence, it is a very good group to study, both for species distribution, and for their chemical composition. Only very few foraminiferal shells are needed to ensure a successful analysis per sample, and each analysis kills two birds with one stone, giving both the oxygen and the carbon isotope ratios. Second, the two isotope ratios provide information about very different aspects of the environment. The oxygen isotope ratio reflects a complex interplay of factors that mostly concern the physical state of the basin, such as global ice volume, which determines global sea level, water temperature, and the freshwater budget (and therefore, indirectly, salinity). The carbon isotope ratio instead gives insight into more biochemical aspects of the environment, including productivity, respiration, and deep water ventilation. It will be obvious that few, if any, of these influences on oxygen and carbon isotope ratios are simple or straightforward. However, the research community has, over the past 50 years that these analyses have been performed, built up many detailed ways of achieving calibrations and deconvolutions of the complex signals. The analyses and their basic interpretations now play a pivotal role in studies of past marine environmental change.
* * *
The ratio of 14C to total C is extremely small, and only gets smaller as samples are older. Therefore, we need large quantities of datable carbon, or extremely sensitive equipment. As our sample availability is limited, most environmental reconstructions rely on age-dating with highly sensitive particle accelerator mass spectrometers (AMS). These require only a few thousandths of a gram of carbonate to return a reliable dating. The cost of these machines and their special laboratory requirements runs in the millions, driving generally high prices for AMS 14C datings (typically several hundred dollars per dating). Usually, however, this represents an invaluable investment, since AMS 14C dating provides an unsurpassed age control that is essential to understand past processes, environmental gradients, and the rapidity of climate change.
Other, uranium based, radioisotope dating methods are concerned with much longer half-life periods of 77,000 and 244,000 years. For the problem discussed in this book, these are of less interest than the radiocarbon technique, and I therefore refrain from discussing them any further.
* * *
Abundance profiles of bioaccumulated barium though the anoxic intervals mimic the organic carbon abundance profiles. The barium offers a bonus, though. Whereas organic matter can be remineralised (and so returned into the ocean water) after its incorporation in the sediments, the barium remains immobile. Therefore, barium profiles show us the original shape of the organic carbon profile. In some cases, this is identical to the observed organic carbon profile, but when (part of) the organic carbon has been respired at a later stage within the sediments, then the barium profile can be used to identify the original extent of organic carbon enrichment. Peaks of manganese can be used to identify the top of the original anoxic interval, ie. the level in the sedimentary sequence that corresponds to the re-oxygenation of the sea floor at the termination of the anoxic event. In addition, maganese profiles help us determine to what extent respiration of organic carbon within the sediments has been taken place when oxygen started to penetrate the top sediments following the re-ventilation of deep waters.
So-called redox sensitive elements are used to assess the extent to which the sedimentary environment has been well-oxygenated, poorly oxygenated, or anoxic. This allows particularly detailed assessment of the sea floor oxygenation when combined with studies of benthic foraminiferal abundance patterns at the same levels.
Abundance peaks of several elements, such as titanium,
non-biogenic silicon, and zirconium, are related to enhanced input of wind-blown
dust into the basin. These help in establishing the broad climatic changes
around the anoxic intervals, by contrasting dry, windy and dusty periods
with wet, vegetated, and less dusty intervals. Typically, the anoxic intervals
in the Mediterranean sedimentary sequence contain low concentrations of
these wind-blown dust indicators, suggesting humid climate conditions with
much vegetation that resists the uptake of dust by wind.
Chemical analyses of the organic compounds in the sediments
help in deciphering the origin of the organic matter in the anoxic intervals.
It is found to consist mostly of marine organic matter, but there also
seems to be enhanced preservation of organic compounds that have been transported
into the basin from land (eg., pollen, spores, and other vegetation remains).
Using specific markers, the organic geochemists are slowly beginning to
help evaluate the marine floras and faunas that were thriving and dying
at the time, producing the highly characterisitic organic matter that is
analysed.
A specifically interesting development in the organic geochemisty concerns ‘long-chain alkenone unsaturation ratios’. Variations in these biomarkers have been calibrated to sea surface temperature fluctuations with a reasonable degree of confidence. Hence, studies concentrating on this aspect of the organic geochemistry offer a useful independent temperature proxy.