“How inappropriate,” Arthur C. Clarke once observed, “to call this planet Earth when it is so clearly Ocean.”1 Most of our planet is covered with water, almost all of it salty. “Water, water everywhere, nor any drop to drink,” the Ancient Mariner complained. Yet it remains surprisingly difficult to explain why the sea is salty.
Most coastal cultures have a story that explains the origin of sea salt. These accounts typically suggest that the waters were fresh to begin with and salt was added later. In Norse mythology, the sea was made unpotable through an act of revenge by two enslaved giantesses, Fenja and Menja, who tended a magic mill that they alone were strong enough to turn. Held captive on a ship and forced to grind salt day and night by the sea-king Mýsingr, the giantesses kept on grinding until they produced enough to sink the vessel. The still-churning millstone fell into the abyss, creating a whirling maelstrom and forever mixing salt into the sea.
For Pythagoras, the sea represented the tears of Kronos, a Titan born of Gaia and Uranus. Empedocles suggested that seawater was the sweat of the earth. Aristotle wrote at length on the composition of seawater in his treatise Meteorology, sweeping aside earlier theories with scorn. “Metaphors are poetical,” Aristotle sniffed, “and so that expression of [Empedocles] may satisfy the requirements of a poem, but as a scientific theory it is unsatisfactory.”2 Aristotle observed that seawater is not merely salty but also bitter, concluding that it must therefore be an admixture of water and various earthy residues borne by rivers. He recognized that this idea posed a problem. “If it is maintained that an admixture of earth makes the sea salt,” Aristotle wrote, “it is strange that rivers should not be salt too.” He suggested that the action of the sun’s heat on seawater played a transformative role, and offered a comparison with the production of cinders and bodily waste:
What heat fails to assimilate becomes the excrementary residue in animal bodies, and, in things burnt, ashes. That is why some people say that it was burnt earth that made the sea salt. To say that it was burnt earth is absurd; but to say that it was something like burnt earth is true. … [E]verything that grows … always leaves an undigested residue, like that of things burnt.3
Aristotle further argued that the accumulation of salt in the sea was counterbalanced by rainfall, thereby keeping its composition in a steady state, an early articulation of a key concept in chemical oceanography.
More than three centuries later, Pliny the Elder, in descriptions of salt extraction methods around the Roman Empire, still espoused Aristotle’s views, although he downplayed the excretory analogies. Pliny added his own elaboration, suggesting that because salt precipitated in pools and pans at night, the action of the moon must also be significant.4 The continuing influence of Aristotle and Pliny into the Middle Ages is reflected in the writings of the tenth-century Arabic philosopher and historian al-Mas’udi.5 In his Muruj adh-dhahab wa ma’adin al-jawahir (Meadows of Gold and Mines of Gems), al-Mas’udi described the continuous chemical exchange between earth and ocean. Water flowing into the sea carries salt absorbed from the earth; heat from the sun and moon evaporate the “sweet portions of the water,” which later fall as rain. “This process,” he remarked, “is constantly repeated.”6
After two millennia, Aristotelian doctrine was finally overturned by Robert Boyle. In the opening paragraph of his Experiments and Observations upon the Saltness of the Sea, published in 1675, Boyle scolded Aristotle:
[H]is Authority, perhaps much more than his Reasons, did for divers Ages make the Schools and the generality of Naturalists of his Opinion, till towards the end of the last Century and the beginning of ours, some Learned Men took the boldness to question the common Opinion…7
Boyle’s empirical approach laid the groundwork for the practice of modern chemistry. He systematically analyzed seawater through hydrometer measurements, as well as evaporation, precipitation, and primitive titration experiments. Boyle disproved Aristotle’s hypothesis that some unspecified action of the sun’s rays on water left behind a residue of concentrated salts, suggesting instead that evaporation of riverine waters was sufficient to explain seawater chemistry.
By the nineteenth century, geology was emerging as a discipline in its own right, revealing increasing evidence that the earth had a far longer history than previously imagined. In 1899, John Joly suggested that the earth’s age could be estimated from the salinity of seawater. He assumed that the sea had started as fresh water and had grown more saline over geologic time. Joly gathered information about the composition of river water from around the British Empire to determine the average annual flux of sodium into the sea. He divided the total inferred volume of sodium in ocean water by this number and estimated that the age of earth, or at least its oceans, was between 90 and 100 million years.8
Joly’s work was criticized by the geological community, not least for his assumption that the salt content of the oceans had increased continuously over geological time. This seemed inconceivable to geologists steeped in the uniformitarian doctrine popularized by Charles Lyell in his Principles of Geology, which stressed that the earth’s processes and natural laws have remained constant.9 Moreover, geologists had also documented thick evaporite, or rock salt, deposits left by ancient marine waters in sedimentary sequences around the world. These clearly indicated that sodium and other elements did not accumulate continuously, but could also exit the sea in large volumes. In arguing for a steady state in seawater chemistry, both Aristotle and al-Mas’udi had been correct in a way that Joly was not.
As it turned out, Joly’s efforts to date the earth were eclipsed in the following decade by the development of radiometric dating techniques. But his result was not entirely without significance. Joly’s estimate of approximately 100 million years is close to the residence time of sodium in seawater: the average time a sodium ion stays in the sea before leaving via mineral precipitation, salt spray, or other sinks. Sodium is, in fact, part of a geochemical cycle—atoms do not simply take one-way trips into the ocean.
The emergence of a more nuanced understanding of ocean chemistry can be seen in a paper published by William Rubey in 1951.10 Rubey amassed data about the chemical composition of present-day river and ocean waters, ancient marine deposits, and igneous rocks to ascertain whether ocean chemistry could be explained solely by rivers conveying the dissolved products of rock weathering. In what can be recognized as an early effort to model global biogeochemical cycles, Rubey attempted to balance the budgets of the various constituents of seawater. Like Aristotle, he assumed that the composition of seawater had remained constant over geologic timescales. In his mass balance calculations, Rubey was unable to account for the profusion of volatile elements and compounds found in seawater, including chlorine, its most abundant ion. He concluded that there must be another source for some of seawater’s major constituents and that source might be deep-sea “volcanoes, fumaroles, and hot springs.”11 He was right.
Rubey’s exhalative vents were not identified until the advent of plate tectonics theory during the 1960s and the discovery of the global mid-ocean ridge system, a network of interconnected undersea mountain ranges formed at plate tectonic boundaries. Along the crests of these ridges, basaltic magma rises from the mantle, creating new oceanic crust. This process, known as seafloor spreading, was inferred from bathymetric and geophysical data in the 1960s, but was not directly observed until the following decade.12
In 1979, a tiny submarine carrying a pilot and two volcanologists descended to the Galápagos Rift, a ridge located off the coast of Ecuador at a depth of almost two kilometers. At a rocky fissure where incandescent basaltic lava was being discharged into frigid seawater, the team were astonished to observe two-meter-high rock spires spewing black jets of superheated fluid that resembled coal smoke “belching out of the smokestack of a steaming locomotive.”13 These hydrothermal vents, emitting brines at temperatures approaching 350°C, were the missing piece in the puzzle of seawater chemistry.
It soon became clear from the chemical composition of these brines that mid-ocean ridges are much more than spigots through which gases are released into the ocean. The temperature contrast between ice-cold seawater and fiery basalt fuels energetic convection systems that flush water through fractures in the rocks and back out the vents. As part of this process, the seafloor basalt takes in many land-derived elements from seawater including boron, magnesium, potassium, rubidium, and uranium; that is, some river-borne elements are actually removed from seawater at mid-ocean ridges. In exchange, the volcanic rock at the ridges gives up barium, calcium, lithium, silicon, and heavy metals.14 For some of these elements, the fluxes into the ocean exceed the contributions of rivers by a factor of one hundred or more. Estimates suggest that the entire volume of the world’s oceans flows through the rocks of the mid-ocean ridges in about eight million years.15
This is a profoundly different view of seawater than Aristotle, Boyle, Joly, or even Rubey entertained. The ocean is not merely a receptacle for inputs from rivers and volcanoes, but a dynamic medium whose chemistry is modulated by processes far more complex than previously imagined. The salt in seawater is now known to include more than 40 elements, ranging in concentration from gold at 0.00008 parts per million to sodium at more than 10,000 parts per million.16
Oceanographers classify these elements into three groups: conservative, scavenged, and recycled.17 Conservative elements, which include the major constituents of seawater, such as sodium and chlorine, have residence times on the order of millions to hundreds of millions of years. Their residence times are much longer than mixing times, the period over which they are distributed via waves, currents, and eddies. If an element’s residence time is greater than its mixing time, its concentration will be fairly uniform throughout the ocean, and it can be considered to have reached a state of chemical equilibrium in seawater.
Recycled and scavenged elements have much shorter residence times, on the order of hundreds or thousands of years.18 As a result, there is insufficient time for these elements to achieve equilibrium through mixing, and their concentrations vary according to geography and depth. Scavenged elements include aluminum, lead, mercury, and other metals. These elements are attracted to the surfaces of fine clay particles and are exported from shallow waters as the particles sink to the ocean floor. The chemical concentration of scavenged elements decreases with depth in the water column.
The recycled elements exhibit the most interesting and complicated behavior of all the groups. These are the ingredients essential for life, such as carbon, copper, iodine, iron, nitrogen, phosphorous, and zinc. In the uppermost region of the ocean where almost all sea life can be found, these elements are limiting nutrients, whose scarcity makes them coveted commodities in the marine biosphere. Any of these elements lost by one organism will be snatched up by another, keeping their concentration in the water itself low. Over time, a fraction of these nutrients leaks into deeper and relatively uninhabited waters, and so the concentration of recycled elements generally increases with depth.
Such leakage of carbon from the surface into the deeper ocean is, in fact, essential for climate regulation. Like land-based plants, phytoplankton use atmospheric carbon dioxide (CO2) to make sugars through photosynthesis. The carbon they fix is consumed by larger organisms whose remains and waste rain down to the deeper ocean and help to sequester carbon from the atmosphere. This process, known as the carbon pump, has partly offset anthropogenic increases in CO2 arising from the combustion of fossil fuels. But there are limits to this buffer. Some of the carbon in organic matter is reoxidized—that is, decomposed and converted to CO2 again—which is one factor causing the oceans to become more acidic. Excess CO2 in seawater can eat away at the shells of the tiny calcitic organisms that help to sequester carbon in mineral form. The carbon pump may also become stalled if other scarce elements, such as phosphorous and nitrogen, suddenly become abundant. In many coastal regimes this has occurred as a result of agricultural fertilizer runoff. This can spur prodigal blooms of microorganisms that gorge on these nutrients. Some of the excess organic matter will then sink to floor of shallow continental shelf areas and may soak up all the available oxygen in the process of decomposition. Waters that once hosted high levels of marine biodiversity become anoxic dead zones where, aside from sulfur-loving bacteria, virtually nothing can survive.
Paleo-oceanographers now recognize that ocean chemistry has not remained strictly constant over time. The onset of the frigid Pleistocene epoch around 2.6 million years ago is thought to be linked to an increase in the salinity of the Atlantic Ocean as a result of the formation of the Isthmus of Panama. The isolation of the Atlantic from the less salty Pacific profoundly affected the global thermohaline circulation system, of which the Gulf Stream is one part.19 Surface currents, such as the Gulf Stream, transfer warmth from the tropics to higher latitudes. As it flows northward, the water cools and is subject to evaporative loss, becoming saltier and increasing in density as a result. To the north of Iceland, the Gulf Stream waters are denser than in the ambient North Atlantic and sink to the seafloor, returning south as a mirror-image deep ocean current. Prior to the formation of the Isthmus of Panama, the Gulf Stream was able to flow further north into the Atlantic, carrying heat to the high arctic. Once the saltiness of the Atlantic was no longer being diluted by the Pacific, the more saline water started sinking further south. In the absence of the heat these waters brought to the far north, winter snows persisted through thousands of summers and ice sheets grew.20
Longer-term variations in the global salinity of seawater can be reconstructed from the sequence of minerals precipitated as rock salt deposits through the evaporation of marine waters; different ratios of ions will cause different minerals to become stable in a particular order. Even more precise paleo-salinity data can be gleaned from fluid inclusions, tiny pockets of ancient seawater trapped in the rock salt minerals. Both of these approaches indicate that salinity has varied between double and half the modern value over the past 500 million years. The ratio of dissolved salts to other elements has also fluctuated dramatically.21 Changing rates of seafloor spreading and water–rock ion exchange at mid-ocean ridges may explain these variations to some extent, but many aspects of seawater’s geologic history remain obscure.
Some profound changes in seawater composition are nonetheless clearly evident from the geological record. Rocks dated to around 2.4 billion years ago chronicle a revolutionary shift in ocean chemistry that has become known as the Great Oxidation Event. Evidence of this event can be seen in banded iron formations found in Brazil, Western Australia, and the Lake Superior region. In the preoxidation era, iron emitted by submarine volcanoes could remain dissolved in the open ocean, commingling with sodium, calcium, and other ions. But when oxygen, produced by the ancestors of modern planktonic photosynthesizers, began to accumulate in shallow waters, it bonded with the iron atoms, pulling them to the seafloor. The oceans were eventually purged of iron. In comparison to ancient oceans, present-day seawater is profoundly anemic: at 0.02 parts per million, iron is in such short supply that it has become one of the much-recycled limiting nutrients.22 The tremendous volume of iron formations, the source of all the steel ever forged, attests to the great abundance of iron in the Proterozoic oceans.
From the very beginning, the history of life on earth has been intertwined with that of seawater. All the mass extinction events evident from the fossil record have been linked to variations in ocean chemistry, such as widespread acidification, anoxia, and associated perturbations to the carbon cycle. The demise of the dinosaurs, for example, can be attributed in large part to oceans poisoned by the constituents of the carbon and sulfur-rich rocks vaporized by the Chicxulub impactor. For this reason, some of the changes in ocean chemistry observed during the Anthropocene ought to give pause. The magnitude of these changes are comparable to the Great Dyings of the geologic past.23 If not tears or sweat, seawater could perhaps instead be considered earth’s blood, its composition a proxy for the health of the planet.
