Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
1 Origins T he modern study of Earth is ultimately rooted QUESTION 1: HOW DID EARTH AND in humankindâs desire to understand its origins. OTHER PLANETS FORM? Although it was once assumed that intelligent life was unique to Earth, we have now gained an ap- One of the most challenging and relevant questions preciation that even though it may not be unique, about Earthâs formation is why our planet is the only the existence of advanced life on planets may well be one in the Solar System with abundant liquid water at uncommon. None of the other planets of the Solar its surface and abundant carbon in forms that can be System are presently suitable for the complex life forms used to make organic matter. This question is part of that exist on Earth, and we have yet to identify other a broader set: why the inner planets are rocky and the stars that have planets much like Earth. Although the outer planets are gaseous; how the growth and orbital odds are good that there is other life in our galaxy, this evolution of the outer planets influenced the inner Solar inference has not been confirmed. System; why all of the largest planets are so different Considering the apparent rarity of Earth-like life, from one another; and how typical our Solar System is it is natural to want to understand what went into mak- within the Milky Way galaxy. Although these questions ing Earth suitable for life and how life arose. Pursuing are longstanding, the answers are only now emerging these questions leads us to fundamental issues about from new insights provided by astronomy, isotopic how stars and planets form and evolve and to questions chemistry, Solar System exploration, and advanced about how the modern Earth works, from the inner- computing. And although we know in general how to most core to the atmosphere, oceans, and land surface. make a planet like Earthâstarting with some stardust This chapter presents three questions related specifi- and allowing gravity, radiation, and thermodynamics cally to originsâone regarding the origin of Earth and to do their partsâour answers often serve only to re- other planets and one regarding the origin of life. These fine our questions. For example, the details of Earthâs two questions are separated by a third that deals with chemical compositionâsuch as how much of the heat- Earthâs earliest history: the 500 million to 700 million producing elements uranium, thorium, and potassium years between the time of the origin of the Solar Sys- it contains; how much oxygen and carbon it contains; tem and the oldest significant rock record preserved on and how it came to have its particular allotment of Earth. During this early, still poorly understood, stage noble gases and other minor constituentsâturn out to of Earthâs development, tremendous changes must be critical to models of Earthâs geological processes and, have taken place, accompanied by myriad catastrophic ultimately, to understanding why Earth has remained events, all leading ultimately to a setting in which life suitable for life over most of its history. could develop and eventually thrive.
ORIGIN AND EVOLUTION OF EARTH How Do Planets Form Around Stars? small proportion of dust makes the disks opaque at visible wavelengths (Figures 1.1 and 1.2). Gas-giant We do not know how unique or unusual the Solar Sys- planets, such as Jupiter and Saturn in our system, are tem is, but observations of other planetary systems are believed to form in such circumstellar disks, but direct providing new ideas for how planets form and evolve. astronomical observations of planets forming have not Astronomical observations of star-forming regions and yet been made. young stars, together with hydrodynamic models of star Observations of planets around other nearby stars formation, support the conclusion that starsâinclud- with masses similar to the Sun indicate that planet ing the Sunâform by the gravitational collapse of a formation is a common outcome of star formation, but molecular cloud core composed of materials manufac- no star has yet been observed with a system of planets tured and reprocessed in many earlier generations of that looks anything like the Solar System. Over 200 stars. Because the typical molecular cloud is rotating at extrasolar planets have been discovered by several in- the time of collapse, the developing star is surrounded direct techniques (e.g., radial velocity of the host star, by a rotating disk of gas and dust. Most disks around stellar transit, and microlensing) (Butler et al., 2006; young stars, as viewed through telescopes, are approxi- <www.exoplanets.org>). Multiple planets are known to mately 99 percent gas and 1 percent dust, but even that orbit some two dozen stars. The vast majority of these FIGURE 1.1â Hubble Space Telescope images of four proto- FIGURE 1.2â Hubble Space Telescope WFPC2 image of planetary disks around young stars in the Orion nebula, located Herbig-Haro 30, a prototype of a young (approximately 1- 1,500 light-years from the Sun. The red glow in the center of million-year-old) star surrounded by a thin, dark disk and each disk is a newly formed star approximately 1 million years emitting powerful bipolar jets of gas. The disk extends about old. The stars range in mass from 0.3 to 1.5 solar masses. Each 6 Ã 1010 km from left to right in the image, dividing the edge-on image is of a region about 2.6 Ã 1011 km (400 AU) across and nebula in two. The central star is hidden from direct view, but is a composite of three images taken in 1995 with Hubbleâs its light reflects off the upper and lower surfaces of the flared Wide Field and Planetary Camera 2 (WFPC2), through nar- disk to produce the pair of reddish nebulae. The gas jets, shown row-band filters that admit the light of emission lines of ionized in green, are driven by accretion. SOURCE: Chris Burrows, oxygen (represented by blue), hydrogen (green), and nitrogen Space Telescope Science Institute; John Krist, Space Telescope (red). SOURCE: Mark McCaughrean, Max Planck Institute for Science Institute; Kare Stapelfeldt, Jet Propulsion Laboratory; Astronomy; C. Robert OâDell, Rice University; and the National and colleagues; the WFPC2 Science Team; and the National Aeronautics and Space Administration, <http://hubblesite. Aeronautics and Space Administration, <http://hubblesite. org/gallery/album/nebula_collection/pr1995045b/>. org/gallery/album/entire_collection/pr1999005c/>.
ORIGINS 215 Planets for why they form so close to the star (Butler et al., Msin i < 15 M JUP 0.8 2006). These hot Jupiters are thought to be telling us that large planets can drift inward toward their star as Orbital Eccentricity they form. Models also suggest that planets can under 0.6 some circumstances drift away from the star, so the ultimate location of the planets may have little to do 0.4 with where they originally formed. Extrasolar planets more than a few tenths of an AU distant from their host 0.2 star often have quite eccentric orbits, which contrasts with the Solar System where all of the planets except 0.0 Earth Mercury have nearly circular orbits. 0.1 1.0 Semimajor Axis (AU) How Did the Solar System Planets Form? Figure 1.3.eps FIGURE 1.3â Summary of known extrasolar planets sorted by The Solar System is composed of radically different distance from host star and orbital eccentricity. All of the planets in the Solar System have eccentricities of 0.2 or less. SOURCE: types of planets. The outer planets ( Jupiter, Saturn, Courtesy of Geoffrey Marcy, University of California, Berkeley. Uranus, and Neptune) are distinguished from the inner Used with permission. planets by their large size and low density. The outer planets are the primary products of the planet forma- tion process and comprise almost all of the mass held in planets are thought to be gas giants on the basis of their the planetary system. They are also the types of planet masses and densities. Presumably, more gas giants are that are most easily recognized orbiting other stars. The observed because they are large, and large planets are inner planets (Mercury, Venus, Earth, and Mars) are much easier to detect, leaving open the question of how composed mostly of rock and metal, with only minor many terrestrial planets remain hidden from Earth in amounts of gaseous material. There are âstandard mod- distant planetary systems. A few âsuper-Earths,â with elsâ for the formation of both types of planets, but they masses of several to 10 Earth masses, may be terrestrial have serious deficiencies and large uncertainties. planets, but no measurements of the radius or density According to the standard model for outer-planet of these objects has confirmed this. Gas-giant planets formation, the formation of giant planets starts with appear to be more likely with stars that have propor- condensation and coalescence of rocky and icy material tions of heavier elements (heavier than H, He, and to form objects several times as massive as Earth. These Li) as high as the Sun (Fischer and Valenti, 2005), solid bodies then attract and accumulate gas from the suggesting that heavy-element concentrations in the circumstellar disk (Pollack et al., 1996). The two largest circumstellar disk influence the rate or efficiency of outer planets, Jupiter and Saturn, seem to fit this model planet formation. reasonably well, as they consist primarily of hydrogen Measurements of the masses, orbital distances, and and helium in roughly solar proportions, but they also orbital eccentricities (Figure 1.3) of extrasolar planets include several Earth masses of heavier elements in provide clues about processes that may help determine greater than solar proportions, probably residing in a what the final planetary system looks like. A par- dense central core. Uranus and Neptune, however, have ticularly interesting class of planets, that of gas-giant much lower abundances of hydrogen and helium than planets in orbits extremely close to (less than 0.1 AU) Jupiter and Saturn and have densities and atmospheric their host starsâsometimes called âhot Jupitersââare compositions consistent with a significant component significant because models have been unable to account of outer Solar System ices. An alternative to the standard model is that the The rock and ice balls are not needed to induce the forma- astronomical unit, or AU, is a unit of length nearly equal to the semimajor axis of Earthâs orbit around the Sun, or about tion of gas-giant planets; they can form directly from 150 million km. the gas and dust in the disk, which can collapse under
10 ORIGIN AND EVOLUTION OF EARTH its own gravity like miniversions of the Sun (Boss, that evolution. Significant unknowns are how long the 2002). In this model the excess abundances of heavy process took, how solid materials were able to coagulate elements in Jupiter and Saturn would have been ac- into progressively larger bodies, and how and when the quired later by capture of smaller rocky and icy bodies. residual gas was dissipated. The time for centimeter- This model, however, does not account well for the sized solid objects to form at Earthâs distance from the compositions of Uranus and Neptune, which do not Sun, according to the standard model, might have been have very much gas. Other important questions about as short as 10,000 years. These small solid objects were the outer planets are when they formed and the extent highly mobile, pulled Sun-ward large distances by the to which they may have drifted inward or outward from Sunâs gravity as a result of drag from the still-present the Sun during and after formation. Where the outer H-He gas. Submeter-sized objects were also strongly planets were and when is important for understanding affected by turbulence in the gas. how the inner planets formed. A particular deficiency of the standard model is its The primary difference between the inner and inability to describe the formation of kilometer-sized outer planets (rock versus gas and ice) is thought to bodies from smaller fragments. The current best guess reflect the temperature gradient in the solar nebula. is that the dust grains aggregated slowly at first, and Temperatures were relatively high (>1000 K) near growth accelerated along with object size as small ob- the developing Sun, dropping steadily with distance. jects were embedded into larger ones (Weidenschilling, Near the Sun, mainly silicates and metal would have 1997). The aggregation behavior of objects greater than condensed from the gas (so-called refractory materials), a kilometer in size is better understood: they are less whereas beyond the asteroid belt, temperatures were affected by the presence of gas than are smaller pieces, low enough for ices (i.e., water, methane, ammonia) and their subsequent evolution is governed by mutual containing more volatile elements to have condensed, gravitational attractions. Growth of still larger bodies, as well as solid silicates. It was once thought that or planetesimals, from these kilometer-sized pieces as the nebula cooled, solids formed in a simple uni- should have been more rapid, especially at first. Gravi- directional process of condensation. We now know tational interactions gave the largest planetesimals that solids typically were remelted, reevaporated, and nearly circular and coplanar orbitsâthe most favorable recondensed repeatedly as materials were circulated conditions for sweeping up smaller objects. This led through different temperature regimes and variously to runaway growth and formation of Moon- to Mars- affected by nebular shock waves and collisions between sized planetary embryos. Growth would have slowed solid objects. Important details of the temperatures of when the supply of small planetesimals was depleted the solar nebula, however, are still uncertain, including and the embryos evolved onto inclined, elliptical orbits. such significant issues as peak temperatures, how long Dynamical simulations based on statistical methods they were maintained, and how temperature varied with and specialized computer codes are finding that a distance from the Sun and from the midplane of the number of closely spaced planetary embryos are likely disk. Defining these conditions is an important part of to have formed about 100,000 years after planetesimals understanding how the chemical compositions of the appeared in large numbers (e.g., Chambers, 2003). planets and meteorites came to be. The later stages of planet formation took much The standard model for the formation of the in- longer, involved progressively fewer objects, and hence ner planets is somewhat more complicated than the are less predictable (Figure 1.4). The main phase of model for outer-planet formation and is based largely terrestrial planet formation probably took a few tens on theory and anchored in information from mete- of millions of years (Chambers, 2004). The final stages orites and observations of disks around other stars were marked by the occasional collision and merger of (Chambers, 2003). The model strives to explain how planetary embryos, which continued until the orbits of a dispersed molecular cloud with a small amount of the resulting planets separated sufficiently to be pro- dust could evolve into solid planets with virtually no tected from additional major collisions. intervening gas and how the original mix of chemical Although there are four terrestrial planets, models elements in the molecular cloud was modified during suggest that the number could easily have been three
ORIGINS 11 Simulation FIGURE 1.4â Results of four representa- tive numerical simulations of the final stage of accretion of the terrestrial planets. The 1 segments in each pie show the fraction of material originating from the four regions 2 of the solar nebula shown by the shades of gray, and the size of the pie is propor- tional to the volume of each planet. In each simulation the largest planet has a size similar to Earthâs, but there can be either two or three other planets, and the sizes 3 vary. The planets typically receive material from all four zones, with preference for the 4 zones closest to their final orbit location. SOURCE: Chambers (2004). Copyright 2004 by Elsevier Science and Technology Journals. Reprinted with permission. or five, and they would have been at different dis- Figure 1.4.eps belt just as the Solar System was starting out. Thus, tances from the Sun (Figure 1.4). Tidal interactions they preserve significant clues about the state of the with nebular gas may have caused early-formed inner Solar System when the planets were forming (Figure planets to migrate inward substantially while they were 1.5). For this reason, studies of meteorites play a major forming, and several planets may have been lost into role in helping us understand Earthâs origin. the Sun before the gas dispersed (McNeil et al., 2005). One gift of meteorites is to reveal the age of the The fact that there are no rocky planets beyond Mars Solar System. Precise radiometric dating of high- is likely a consequence of the presence of the giant temperature inclusions within meteorites shows that planets, particularly Jupiter. The large mass and strong the first solid objects in our home system formed gravitational pull of Jupiter probably prevented the 4,567 million years ago (see Box 1.1). We also know formation of additional rocky planets in the region now that shortly thereafter planetesimals of rock and metal occupied by the asteroid belt by disrupting the orbits formed and developed iron-rich cores and rocky crusts of bodies in that region before they could form a large (see Question 2). Some meteorites are chemically like planet. Jupiter and Saturn also sent objects from the the Sun (for elements other than H, He, Li, C, N, O, asteroid belt either out of the Solar System or spiraling and noble gases), and some of these same meteorites into the inner-planet region where they became parts contain tiny mineral grains of dust that survived from of the planets forming there or fell into the Sun. The earlier generations of stars (see Box 1.2). Other mete- asteroids represent the 0.01 percent of material that orites are parts of small planetary bodies that experi- survived this process. enced early volcanism and that were later broken up by collisions. Beyond these clues, meteorites fall short What Do Meteorites Say About the Origin of of providing all the information needed to understand Earth? Earth, partly because most of them formed far from the Sun (the main asteroid belt is between Mars and Earth has undergone so much geological change that Jupiter), and the relationship between meteorites and we find little evidence in rocks about its origin or planets is not fully understood. The systematic collec- even its early development (Question 2). Many me- tion of well-preserved samples from Antarctica has teorites, on the other hand, were not affected by the greatly expanded the number of meteorites available for high-temperature processing that occurs in planetary study and has yielded rarities such as meteorites from interiors. They are fragments of, or soil samples from, Mars and the Moon. miniplanets that formed in what is now the asteroid Beyond what they tell us about Earth, meteorites
12 ORIGIN AND EVOLUTION OF EARTH moderately well from spectroscopic data. The planets, however, formed from the nebular disk, so it is impor- tant to know whether the disk had the same composi- tion as the Sun, and whether it was homogeneous or varied significantly in composition, perhaps with radial distance from the proto-Sun. The standard model for the composition of the solar nebula is based on stud- ies of a class of meteorites called chondrites (Figure 1.5). Chondrites, the commonest type of meteorites, are stony bodies formed from the accretion of dust and small grains that were present in the early Solar System. They are often used as reference points for chemicals present in the original solar nebula. The most primitive of these objectsâthose least altered by heat and pressureâare carbonaceous chondrites, FIGURE 1.5â The Allende meteorite, a carbonaceous chondrite, whose chemical compositions match that of the Sun for is a mixture of CAIs (calcium-, aluminum-rich inclusions; larger most elements. The relative amounts of elements and irregularly shaped light-colored objects) and chondrules (round their isotopes can be measured much more precisely light-colored objects) in a dark-colored matrix of minerals and compounds. The CAIs and chondrules are a high-temperature on meteoritic materials than by solar spectroscopy, component that formed and were in some cases reprocessed so chondritic meteorites play a special role in helping at temperatures above 1000°C. SOURCE: Hawaii Institute of to understand both Earth and nucleosynthesis in our Geophysics and Planetology. Used with permission. galaxy. Because chondritic elemental abundances look similar to those of the Sun, the disk likely had about the same composition as the Sun. provide a benchmark for understanding the composi- tion of the Sun and even the Universe as a whole. Most What Is the Chemical Composition of Earth? of the visible mass of the Universe, and almost all stars, is composed primarily of hydrogen and helium made The most critical question related to the formation during the Big Bang. The rest of the elementsâthe of Earth is why the planet has its particular chemical âheavierâ ones with more protons and neutrons in makeup. Although we know quite a lot about this is- their nucleiâwere produced by nucleosynthesis, or sue, a key unanswered question is the origin of Earthâs thermonuclear reactions within stars. Most nucleo- water. Earth, like other objects forming near the Sun, synthesis happens in big stars. These massive stars last is thought to have formed mainly as a relatively high- only about 10 million to 20 million years before they temperature partial condensate from a gas of solar explode as supernovae. The new elements they make, composition. The uncondensed gas containing water, before and during the explosion, are thrown back into carbon, and other volatile elements was swept away by space where they are later recycled into new stars. In the the early solar wind or by ultraviolet radiation pressure. approximately 10 billion years between the origin of the Much of the volatile elements that might have been Universe and the origin of the Solar System, hundreds incorporated into the early Earth is thought to have of generations of massive stars have exploded, and over been lost during the intense heating of the Hadean this long period about 1 percent (by weight) of the orig- Eon (Question 2). inal H and He has been converted to heavier elements. It has been suggested that the giant planets can Meteorites give us the most detailed information about pluck materials from the asteroid belt region and throw the abundances of these heavier elements. them in toward the Sun. Objects beyond Mars would Meteorites tell us still more about the forma- have formed in a cooler part of the solar nebula and tion of the Solar System out of the nebular disk. The hence would likely have contained more volatile com- abundance of heavy elements in the Sun is known pounds. Studies of asteroids indicate that meteorites
ORIGINS 13 BOX 1.1 Time, the Early Solar System, and the Age of Earth The initial events in the formation of the Sun, meteorites, and Earth and other planets took place in only a few million years, about 4,567 million years ago (Ma). Documenting early Solar System timescales is therefore a substantial challenge. Advances in geochronological techniques are beginning to enable the sequence of events to be discerned. The most primitive chondritic meteorites contain inclusions made up of minerals that condense at high temperature from a gas of solar nebula composition. These objects, called calcium-aluminum-rich inclusions (CAI), have recently been precisely dated using the decay of uranium to lead, where time is measured by the accumulation of the lead decay products formed at 4,567 (±1) Ma. This age is now generally accepted as âtime zeroâ for the Solar System. The U-Pb method gives the most precise and accurate ages for these ancient objects partly because the radioactive decay constants for 238U and 235U are precisely known. Once the absolute time is established using the long-lived radioactive isotopes of uranium, the sequence of events within the first few million years of the Solar System can be studied using isotopes with much shorter half lives (extinct radionuclides). These isotopes were present in the early Solar System because they had been produced in stars just prior to the beginning of the Solar System and were part of the molecular cloud that collapsed to form the Sun. Subsequently, virtually every atom of these short-lived radioactive isotopes that existed at time zero has now decayed to the daughter isotope. The isotopes used for this purpose are 26Al, 53Mn, 244Pu, 182Hf, 60Fe, and 129I and their corresponding decay products 26Mg, 53Cr, 136Xe, 182W, 60Ni, and 129Xe. The resulting sequence of events is summarized in the figure. How old is Earth? Although the start of the Solar System is well dated at 4,567 Ma, at that time and shortly after only the pieces that would eventu- ally come together to make Earth were present. About half or more of the planet was probably assembled by 4,550 Ma, and the Moon-forming impact, now generally thought to culminate the main phase of Earthâs formation, happened at about 4,530 Ma. Earth probably continued to accumulate small amounts of material, some of them perhaps quite significant chemically, until as late as 4,450 Ma. A short episode of renewed accretion may have oc- curred much later, at 4,000 to 3,900 Ma. Summary of recent geochronological data and models for the sequence and timing of events in the early Solar System. SOURCE: Adapted from Halliday (2006).
14 ORIGIN AND EVOLUTION OF EARTH BOX 1.2 Presolar Grains On the basis of characteristically anomalous isotope ratios (Lewis et al., 1987), we now recognize and can study âpresolar grainsââbits of stardust manufactured by individual stars before the birth of our Solar System that are preserved in primitive meteorites. Each of these grains contains chemical elements that were made or reprocessed by an individual star. How stars produce the heavier elements (from iron to uranium) was highlighted in Con- necting Quarks with the Cosmos as one of the 11 major science questions for cosmology in the new century (NRC, 2003a). Geochemists will play a key role in addressing this question because the relative abundances of elements and isotopes in the different types of presolar grains provide the most specific and detailed data for checking our understanding of how chemical elements are produced in different types of stars (Zinner, 2003). Electron microscope images of presolar grains representing materials that were manufactured by individual stars and condensed in the outflow of material marking the end of that starâs life cycle. Typical sizes are given in microns (µm), and typical abundances are given in parts per million (ppm) and parts per billion (ppb) by weight. SOURCE: Nanodiamond image courtesy of Tyrone Daulton, Washington University, Meteorite Magazine; graphite image courtesy of Sachiko Amari, Washington University; oxide image courtesy of Larry Nittler, Carnegie Institution of Washington. Used with permission. SiC image from Bernatowicz et al. (2003), copyright 2003 by Elsevier Science and Technology Journals. Reproduced with permission.
ORIGINS 15 that have little water are derived from the inner asteroid The chondritic model and the Solar Systemâs ap- belt (inward of 2.5 AU), while the volatile element-rich parent ability to sort chemical elements according to meteorites, some with as much as 20 percent water as their volatility have proven useful for understanding well as complex organic compounds, come from farther many aspects of planet formation. But our increasing than 3 AU. These objects from the asteroid belt region ability to probe the chemical and isotopic compositions may have been the source of Earthâs water and carbon. of meteorites and our planet is causing some serious There is also evidence that much later in the history rethinking of long-held models. Unanticipated com- of the Solar Systemâ500 million to 600 million years positional differences have been discovered between after its formationâa large but unknown amount of Earth and meteorites and between different types of rocky debris was flung into the inner Solar System, meteorites. Perhaps the most striking difference is bringing a last barrage of large impacts and finishing off that of the isotopes of oxygenâthe most abundant the major construction of the inner planets. However, element on Earth (Figure 1.6). Chondritic meteorites it is unlikely that this âlate heavy bombardmentâ added have a peculiarly variable proportion of the isotope enough material to significantly affect Earthâs overall 16O, and almost every class of meteorites has different composition. proportions of the three oxygen isotopes. Chondritic The aspect of Earthâs composition that is likely best meteorites, long thought to be the best model for the known is the proportion of refractory elements, which original Earth, are not like Earth with respect to oxygen form solids at the high temperatures thought to have isotopes. The one class of meteorites that is like Earth prevailed in the inner Solar System as the terrestrial in this respectâenstatite chondritesâwould probably planets were forming. Included among the refractory be no oneâs first choice for Earthâs main building blocks elements are most of Earthâs major componentsâSi, because they do not match Earth for most other ele- Mg, Al, and Ca. The relative amounts of refractory elements do not vary much among different classes of the benchmark chondritic meteorites, which is gener- ally taken as a strong argument that Earth is not much different from the meteorites. For the more volatile elements, which evaporate more easily, there are wide and puzzling variations throughout the Solar System. Oxygen is one example. Si, Mg, and Fe readily combine with oxygen to form SiO2, MgO, and FeO. On Earth almost all of the Si and Mg occur as oxides, but only about 20 percent of the Fe is combined with O; the rest is metallic Fe that resides in Earthâs core. The size of the core is therefore a rough measure of the amount of oxygen that Earth has. Most meteorites have different Fe/FeO ratios, and at least two of the other terrestrial planets have a different ratio of metallic core to silicate mantle. Elements of intermediate volatility also raise important questions of chemical evolution. Potas- sium, for example, is relatively volatile, and estimates FIGURE 1.6â Representation of the range of values of oxygen suggest that Earth has about 10 percent of what was isotope ratios on Earth, the Moon, Mars, and different classes of meteorites, including carbonaceous chondrites (CI, CK, CM, CO, available in the nebula. But exactly how much? The CR, CV); ordinary chondrites (H, L, LL); other chondrite groups answer is critical because the isotope 40K is radioactive (R); primitive achondrites (Acapulcoite [Aca], Brachinite [Bra], and provides 20 to 40 percent of the heat produced in Lodranite [Lod], Winonaite [Win], Ureilite [Ure]); Howardite, Eucrite, Diogenite [HED] achondrites; aubrite achondrites (Aub); the early Earth. This heat plays a role in powering the stony-iron meteorites (Pallasites [Pal], Mesosiderite [Mes]); and convection in the mantle that drives plate tectonics iron meteorites (IAB-IIICD irons). SOURCE: <http://www4.nau. (Questions 4 and 5). edu/meteorite/>. Used with permission.
16 ORIGIN AND EVOLUTION OF EARTH ments. Moreover, it has recently been reported that the for example, suggest that the nebular disk was not isotopes of neodymium, a lanthanide element that has entirely homogeneous. While this is a problem in one proven critical for understanding planetary processes sense, it is also an opportunity. If we can understand (Question 4), are also present in different amounts on how this heterogeneity arose or was preserved, and Earth and chondritic meteorites (Figure 1.7), as are the what its structure was, we can learn more about how isotopes of hafnium and barium. the materials of the nebula were sorted and gathered Although we have long assumed that the isotopic to produce the planets. compositions of the elements of the Solar System The Nd isotope discrepancy raises a different prob- were mostly homogeneous, and measurements have lem that has not yet been squarely addressed. Studies borne this out in large measure, improved sensitivity is of asteroids and meteorites show that the process of now showing small but significant differences between accretion, whereby small chunks of rock gradually co- various planetary bodies. The O isotope differences, alesce to form larger and larger bodies and eventually planets, is not one directional. When objects collide, they are almost as likely to blow each other apart as they are to coalesce. In addition, there is evidence that small accreting bodies become hot enough to melt, al- lowing crystals and liquid to separate. Thus, it was pos- sible to differentiate (make heterogeneous by internal processes) smaller bodies and then blast material off them that is chemically different from the bulk object. This process would create differentiated objects that could eventually become part of the planets (or be lost into the Sun or ejected from the Solar System). In this view we cannot expect even the refractory elements to be present in exactly the same proportions everywhere, and this would have enormous implications. For ex- ample, if we relax the requirement that Earth be exactly chondritic for the elements Nd and Sm, we reach a different interpretation of the subsequent evolution of Earthâs mantle and crust (Question 4). If the Hf/W ratio of Earth is not chondritic, the timing of forma- tion of Earthâs metallic core, as estimated by W isotope data, changes (see Question 2). We now know that even small bodies were able to partially melt and differenti- ate into core and mantle and that the mantle could potentially be removed from the core by an impact. So the timing and mechanism of formation of planetary metallic cores and the abundances of trace metals in planetary mantles have to be viewed in this context. FIGURE 1.7â Reported differences in 142Nd isotopic abun- dance between Earth, achondritic meteorites (Eucrites), and Was the Moon Formed by a Giant Impact? chondritic meteorites. The ε142Nd value is the difference in the proportion of 142Nd expressed in units of 0.01 percent. 142Nd is the radioactive decay product of the short-lived isotope 146Sm. More is known about the Moon than any terrestrial The differences may reflect deep sequestration of ancient crust planetary body other than Earth because of the rock formed in the early Earth or differences in refractory element samples collected by the U.S. and Soviet lunar missions ratios between Earth and chondritic meteorites. SOURCE: Boyet between 1969 and 1976. The peculiarities of these lunar and Carlson (2005). Reprinted with permission of the American Association for the Advancement of Science (AAAS). rocksâtheir great antiquity, their nearly complete lack
ORIGINS 17 FIGURE 1.8â Snapshots in a numerical simulation of the Moon-forming giant impact. Times are shown in hours and color scales with particle temperature in K; frames (a) through (e) are views onto the plane of the impact; particles with T > 6440 K are shown in red. Distances are shown in units of 1,000 km. Frame (f) is the final state viewed edge on; here the temperature scale has been shifted so that red corresponds to T > 9110 K. The large orbiting clump in (d) and (e) con- tains about 60 percent of a lunar mass. SOURCE: Canup (2004b). Copyright Elsevier. Reprinted with permission. of water and other volatile elements and compounds, the Moon. Another difference is that Earth has water, and the chemical complementarity of the dark lunar as well as other volatile species and oxidized (ferric) basaltic lowlands and the bright highland rocksâled iron; the Moon has virtually no water and all of its iron to enormous advances in theories of planet formation. is in the reduced (ferrous) state. Moon rocks provide one of the most persuasive pieces Studies of lunar rocks have helped persuade many of evidence that Earth and the Moon have a common geologists that the Moon was formed when a Mars- origin. The isotopic composition of oxygen varies dra- sized object collided with the still-forming Earth matically within the Solar System (Figure 1.6) but is about 40 million years after the formation of the Solar identical in Earth and the Moon. An important differ- System. This âgiant-impactâ hypothesis would explain ence is the size of their metallic coresâone-third of the the relatively large mass of the Moon relative to Earth, mass of Earth but only about 2 percent of the mass of the large amount of angular momentum in the Earth-
18 ORIGIN AND EVOLUTION OF EARTH Moon system, and the chemical similarities and dif- tions of the standard models for the composition of ferences between Earth and the Moon. None of the Earth and meteorites, and studies of presolar grains other terrestrial planets have a moon, except for the tiny are sharpening our understanding of stellar evolution moons of Mars, which are captured asteroids. and nucleosynthesis. Advanced computing capabili- The general features of the giant-impact hypothesis ties are enabling more realistic simulations of nebular were proposed in the 1980s, but new computer models disk evolution, the consequences of collisions between have provided a clearer picture of the requirements and planetesimals and planetary embryos, and the internal results (Canup, 2004a). A âMars-sizedâ object has a mass processes of proto-planetary bodies. about one-tenth of Earthâs, whereas the Moon has a But we still do not understand the composition mass about one-sixtieth of Earthâs. For the hypothesis to of Earth in enough detail to make sense of its subse- work, the impactor must hit Earth at a low angle and at a quent evolution. Among the most important remaining relatively low velocity (about 10 km/s). Models indicate questions are when and how Earth received its volatile that most of the impactor would become mixed with and components, how much of these components it still incorporated into Earth during the collision (Figure 1.8) contains, whether Earth is exactly the same as chondritic and the cores of the two planets would coalesce at the meteorites with respect to refractory elements, and what center of gravity of the combined system. The collision the absolute concentrations of heat-producing elements would eject a disk of molten rock and vapor into orbit are inside Earth. In a broader sense we need a better idea around the newly enlarged Earth, and a portion of that of the processes that formed planets during the first few disk would coalesce into the Moon. The energy of the million years of the Solar System, how much the plan- impact would have melted virtually the entire Earth and ets were influenced by late events (tens to hundreds of may have resulted in the loss of most of Earthâs volatile millions of years after the beginning), how the chemical elements (Question 2). composition and size of planets were determined by early The impactor event, coming late in Earthâs forma- Solar System processes, and the origins of the various tion, would have had an enormous effect. Many of forms of isotopic heterogeneity. Earthâs features may have been determined by the cata- Although theory and computation are essential strophic collision, which marked the conclusion of the tools, the starting point for posing and solving out- main phase of Earthâs formation. Any internal structure standing questions of Solar System evolution and that formed within Earthâs mantle up to that time planet formation remains observations and measure- would probably have been destroyed, and the intense ments of planets and other extraterrestrial objects. The heating could have homogenized large parts of the materials and processes of planet formation are so var- interior. If the impact hypothesis is indeed correct, it ied and complex, and the scales so immense, that new dispels any doubt that the earliest Earth was extremely breakthroughs in understanding will likely continue to hot. The next section resumes the story of the early follow real observations made by telescopes, spacecraft, Earth with the aftermath of the Moonâs formation. and sensitive Earth-bound analytical equipment. Summary QUESTION 2: WHAT HAPPENED DURING EARTHâS âDARK AGEâ (THE FIRST 500 Many lines of recent evidence have provided critical in- MILLION YEARS)? formation about how and when the Solar System began and how the planets formed. Astronomical observations Assuming that the Moon formed as the result of a giant from increasingly powerful telescopes have added a new impact, the impact would have erased the existing rock dimension to models of star and planet formation, as record, adding enough heat to turn Earth into a mostly have studies of asteroids, comets, and other planets via molten ball, probably to the very surface of the planet. spacecraft. There is increasing crossover between geo- The oldest rocks yet found on Earth are about 4,000 chemical studies and astronomical observations. With million years old, and there are precious few of them; improved mass spectrometric methods, new details of only about 0.0001 percent of Earthâs crust is composed meteorite isotopic compositions are forcing reevalua- of rocks older than 3,600 million years (Nutman, 2006).
ORIGINS 19 FIGURE 1.9â A speculative history of temperature, water, and CO2 during the Hadean. The Hadean begins with the Moon-forming impact (at time = zero in this figure). For 1,000 years Earth is enveloped in hot rock vapor. After the silicate vapor rains out, the atmosphere consists mostly of CO2. Water is gradu- ally lost from the magma ocean and added to the atmosphere. The green- 3000 1000 house effect and tidal heating maintain the magma ocean for 2 million years. Surface Temperature Liquid Water When the magma surface freezes over, 2000 surface temperature drops quickly and Reservoir Size (bars) 100 the steam atmosphere rains out to leave CO2 Temperature (K) CO2 a warm (~500 K) water ocean under 1000 ~100 bars of CO 2. This warm, wet Steam Earth lasts as long as the CO2 stays in 700 10 the atmosphere. This illustration shows 500 CO2 being removed on timescales of 400 20 million years (green solid curves) or 300 Runaway Greenhouse 100 million years (green dotted curves). Atmosphere 1 When the CO2 partial pressure drops 200 below about 1 bar, the oceans freeze Rock Magma over (blue region of graph). After the Vapor Ocean late heavy bombardment, CO2 is shown 100 0.1 returning to an arbitrary level of ~1 100 1000 104 105 106 107 108 109 bar, which allows the surface to be clement as required by geological data. Time (yrs) SOURCE: Zahnle (2006). Reprinted with permission. Figure 1.9.eps And most of those rocks are metamorphosed, some at allowing a clearer picture of that early fiery (and per- very high temperature and pressure, obscuring their haps sometimes frozen) Earth to gradually emerge. original form. Thus the period of time for which there is virtually no rock record on Earth extends from the How Did the Transition to Earthâs Current time of the putative Moon-forming impact ca. 4,530 Environment Occur? Ma to the age of the oldest rocks on Earth ca. 4,000 to 3,800 Ma. This period, about which we can discern Current models suggest that much of Earthâs rocky very little from Earth itself, is called the Hadean Eon. mantle was melted by the Moon-forming impact and The name of this eon is unusually graphic, for good that part of it was vaporized (Stevenson, 1987; Canup, reason. Earth during the earliest Hadean was probably 2004a). If this was the case, ������������������������� liquid mantle would have much less hospitable than even the grimmest represen- been present at Earthâs surface and the atmosphere tations of Hell. Yet somehow this inferno evolved into would likely have been mostly rock vapor, topped by hot a place not only suitable for life but welcomingâwith silicate clouds with temperatures up to 2500 K (Zahnle, abundant oceans as well as dry land, an atmosphere 2006). As Earthâs surface cooled, the silicate clouds dominated by nitrogen, and mostly comfortable tem- peratures. We have almost no idea how fast the surface K represents the Kelvin temperature scale, commonly used in environment evolved, how the transition took place, geology. The Kelvin scale is set so that zero degrees K is absolute or when conditions became hospitable enough to sup- zero, the temperature at which a substance has no remaining thermal energy. Zero K equals â273.15°C, and the two scales are port life. However, clues from Earthâs oldest minerals, otherwise the same, with one degree C having the same magnitude zircons, as well as from our Moon and other planets are as a one-degree increment in Kelvin.
20 ORIGIN AND EVOLUTION OF EARTH would have condensed and poured down as hot rain, perhaps at the torrential rate of about a meter a day. As BOX 1.3 Runaway Greenhouse Effect the silicates rained out, gaseous compoundsâespecially The runaway greenhouse effect is usually encountered as the CO2, CO, H2O, and H2 but also nitrogen, the noble culprit in textbook accounts of how Venus lost its water. In essence, gases, and perhaps moderately volatile elements, there is an upper limit on how much thermal radiation can be emitted such as zinc and sulfurâwould become increasingly by an atmosphere in equilibrium with liquid water. This upper limit prominent. is called the runaway greenhouse limit, and it is about 310 W/m2 How the transition from a hot, mostly molten for the modern Earth (Abe, 1993). If the planet absorbs less solar mantle to something more akin to Earthâs current energy than the runaway greenhouse limit, all is well: the climate settles into a stable balance between photons absorbed and photons structure happened����������������������������������� ���������������������������������� and how fast���������������������� ��������������������� are still matters of emitted. But if the planet absorbs more solar energy than the run- debate. The cooling of a âmagma oceanâ is a complex away greenhouse limit, the planet cannot balance its energy budget process, with significant uncertainties regarding the and its surface heats up. The heating continues until all the water, material properties of the molten and semimolten sili- including clouds, has evaporated. For Earth, total evaporation of all cates, the efficiency of gas exchange between a magma water would leave a deep H2O-CO2 atmosphere over a sea of magma ocean and the atmosphere, how much of which gases (Zahnle, 2006). Eventually, after some intervening photochemistry and a great deal of time, the hydrogen would be liberated from the were available, and the effects of tidal heating from the water and lost to space. This probably happened to Venus. As our Moon. We know from experiments that molten silicate Sun brightens, this too will be Earthâs fate. would start to crystallize when the surface temperature For the Hadean Earth a runaway greenhouse state could dropped to about 1700 K and would be completely solid theoretically coexist with a magma surface provided that sufficient at about 1400 K. According to one model (Figure 1.9), water (at least a tenth of the volume of our current oceans) is present t��������������������������������������������������� he surface magma could have cooled enough for crys- at the surface (Zahnle, 2006). The heat flow required to maintain a runaway greenhouse atmosphere (i.e., the maximum rate of cooling) tals to start forming after about 1,000 years and then would be ~150 W/m2. Heat flow on the modern Earth is an average become completely solid after about 2 million years of 0.087 W/m2. (Zahnle, 2006). During the cooling period, most of the water and CO2 held in solution in the magma ocean could have been vented to the atmosphere. According to this model, solidification of the away from Earth only as fast as it could deposit energy magma ocean would have taken as long as 2 million into Earthâs mantle by tidal heating. For the Moon to years because heat escaping from the surface would lose energy to Earthâs mantle efficiently, the mantle have triggered a ârunawayâ greenhouse state in the would have to be solid rather than liquid. Because atmosphere, slowing the rate of heat loss (see Box solidification likely took place slowly, the Moon could 1.3). Tidal heating of Earth by the Moon would also ���������������������������������������������� drift away from Earth only at an exceedingly slow rate. have slowed cooling of the magma ocean (Zahnle et This slow recession of the Moon would have allowed al., 2007). Just after its formation the Moon was much it to be captured into orbital resonances that gave closer to Earth (perhaps half the distance) and its tidal the orbit the inclination it now has. This seemingly force was much stronger than it is now. When the man- strange relationship between the Moonâs orbit and tle was still completely molten, the tidal heating would Earthâs mantle is produced by a fundamental property have been relatively weak, but because tidal heating is of planetary interiors�������������������������������� â������������������������������� the dependence of viscosity on concentrated wherever the mantle is solid, it would temperature����������������������������������������� â���������������������������������������� which is also critical to understanding have tended to prevent the mantle from freezing. why Earth is a geologically active planet (Questions The resultant slow cooling of the magma ocean 4, 5, and 6). could, in turn, have influenced the Moonâs distance from Earth, which would explain why the Moonâs How Did Earth Develop Its Oceans and orbit is tilted relative to Earthâs orbit around the Sun Atmosphere? (Touma and Wisdom, 1998). The relationship between the Moonâs orbit and the magma ocean is somewhat We do not know how thick the atmosphere would have complicated, but in essence the Moon could move been after the silicates vaporized by the Moon-forming
ORIGINS 21 impact condensed, largely because we are uncertain about how much gaseous material Earth contained at BOX 1.4 Zircons: Earthâs Oldest Minerals that stage (Question 1). The model depicted in Figure Earthâs oldest mineral grains are detrital zircons found in 1.9 assumes that both CO2 and H2O were abundant 3-billion-year-old quartzites in the Jack Hills of western Australia. and found primarily in the atmosphere rather than Zircons are zirconium silicate crystals renowned for their durabil- dissolved in the mantle. If this assumption is correct, ity. Because zircons incorporate uranium, their ages (a given grain the initial atmosphere would have been hundreds of may record several formation and metamorphic events) can be times thicker than the modern atmosphere, with about accurately determined from radioactive decay. Many zircons have 100 bars of CO2 (Zahnle, 2006). But once the surface been found that are more than 4 billion years old, and the oldest one is 4.4 billion years old (Cavosie et al., 2005). Their existence cooled to 500 K in this model, almost all the water suggests there were stable continental platforms on Earthâs surface would rain out of the atmosphere and cover Earth in the Hadean. That such zircons have been found in only one place with oceans, leaving the atmosphere made up mainly so far may suggest that such stable platforms were oddities rather of CO2. The abundant CO2 in the atmosphere would than the rule. partially dissolve in the oceans, making them acidic. The origin of zircons is also recorded in their oxygen isotopes. At this critical juncture the level of uncertainty is re- A mild fractionation of the oxygen isotopes suggests that many zircons formed in melts that incorporated rocks that reacted with doubled. The acid oceans should then chemically attack liquid water (Mojzsis et al., 2001; Wilde et al., 2001). The zircons the rocks of the seafloor, slowly turning the oceans into are silent on whether the water was 273 or 500 K, but they suggest a soup rich in dissolved solids. If the dissolved carbon that Earthâs oceans were in place by 4.4 Ga (billion years ago). A precipitated from the ocean as calcite, and if (a big if ) small number of old zircons are reported to have more strongly the calcite was removedâfor example, by dragging it fractionated oxygen isotopes (Mojzsis et al., 2001), which implies down subduction zones into the mantle (see Question that the melts incorporated sediments weathered by water to make something like a granite. These data are controversial. 5)âthe CO2 in the atmosphere would eventually be Zircons also incorporate hafnium, an element with a strong sequestered in the solid Earth, the greenhouse power chemical likeness to zirconium. A deficit of radiogenic Hf in Hadean- of the atmosphere would gradually diminish, and aged zircons suggests granitic (continental type) crust was already Earthâs surface temperature would drop. Removal of formed at 4.5 Ga (Harrison et al., 2005). CO2 from the atmosphere might not have stopped until the CO2 pressure was similar to todayâs value of about 0.0003 atmosphere. The drop in CO2 would have lowered the surface temperature to well below freezing and covered the oceans with a global ice sheet 10 to 100 m thick. At this stage, Earthâs mantle would still be exceedingly hot compared with the modern mantle, but Earthâs surface would be frigid. A key point, which links Earthâs story to astro- physical models of stars, is that the Sun was 30 percent fainter in the Hadean than it is now. Given the evidence for the presence of liquid water recorded in zircons (Box 1.4), Earth would have needed abundant potent greenhouse gases to keep its surface temperature above the freezing point of water (273 K). The only good candidate greenhouse gases are CO2 and CH4. If CO2 The mineral zircon is resilient to alteration and recrystallization was removed by weathering and carbonate subduction, and also contains high uranium content, which provides a means methane might work. But on todayâs Earth methane of dating individual crystals. The cathodoluminescence image of a zircon crystal shown here is 4.4 billion years old, the oldest known is made mostly by organisms, and the Hadean Earth mineral on Earth. Zircon is the only known survivor of the Hadean period on Earth. SOURCE: Courtesy of John Valley, University of Carbon sequestration by various means is discussed extensively Wisconsin. Used with permission. today as a means of mitigating greenhouse warming (Questions 7 and 10).
22 ORIGIN AND EVOLUTION OF EARTH likely had little or no life. Without methane a mecha- from meteorites, supplemented by Hf-W isotopic nism is required to keep more CO2 in the atmosphere. measurements (Figure 1.10), now clearly shows that Could the processes that regulated atmospheric CO2 core formation also happened in planetesimals that levels within a range that kept the surface temperature were much less massive than Earth (Question 1) and well above freezing almost all the time over the past hence too small to have been heated from within by several hundred million years (Question 7) also have U, Th, and K (Kleine et al., 2002). These planetesimal operated in the Hadean? cores also formed rapidlyâwithin a few million years How can we test models like the one depicted in of the beginning of the Solar System. Hence, it is now Figure 1.9? One useful observation is that the deep generally accepted that core formation on Earth began Earth still contains 3He, a primordial gas isotope that when the planet was still small and accreting and that must have been emplaced during accretion. This tells us core formation probably continued for many tens of either that the interior did not expel all its gases during millions of years as Earth grew. The state of the core at a magma ocean stage (perhaps because these gases are the time of the giant impact and the influence of this more soluble in mantle minerals during magma gen- esis than previously thought; see Parman et al., 2005; Watson et al., 2007) or that the present atmosphere was added after the Moon-forming impact. Measurements of Xe isotopes suggest that Earth lost about 99 percent of its original allotment of noble gases and that it did so at least 20 million to 40 million years after the giant impact (Ozima and Podosek, 1999; Halliday, 2003, and references therein). To satisfy both observations, Earth must have had an early, dense atmosphere of accreted solar nebula gas (mostly H2 and He) so that it could first gather large amounts of He and Xe and then later lose most of them. A hot and dense primor- dial atmosphere of nebular gas could have provided enough thermal insulation to maintain a magma ocean even before the heating of the Moon-forming event. This proto-atmosphere could have been lost when the Moon formed, or by hydrodynamic escape or ionization due to intense ultraviolet radiation from the early Sun. But since the available data are difficult to resolve with prevailing models, we are left with many uncertainties about the earliest evolution of Earthâs atmosphere. When and How Did Earthâs Metallic Core Form? FIGURE 1.10â Comparison of the 1.10.eps Figure tungsten isotopic composi- Early models for the formation of Earthâs core were tion of Earth rocks and meteorites. Epsilon units represent devia- based on a logical scenario, now known to be incor- tions in the 182W/184W ratio of Earth relative to the meteorites, rect, that Earth first accreted into a more or less ho- measured in parts per 10,000. The greater epsilon 182W value of Earth relative to chondritic meteorites indicates that Earthâs rocky mogeneous globe (a mixture of both silicates and iron portion formed when 182Hf was still alive, which produces 182W, metal), then gained heat from radioactive decay of U, but after most of the tungsten had been sequestered into Earthâs Th, and K. The heating gradually decreased the planetâs core. Iron meteorites (open circles) have even lower epsilon values and may be representative of Earthâs core composition, viscosity over hundreds of millions of years, which al- which is expected to be deficient in 182Hf, hence 182W. SOURCE: lowed the heavy metal to sink to the center, displacing Kleine et al. (2002). Reprinted by permission from Macmillan the lighter silicates toward the surface. But evidence Publishers Ltd.: Nature, copyright 2002.
ORIGINS 23 pothesis is that the mantle siderophile elements were added to Earth in a âlate veneerâ of meteoritic material after the core formed (see Palme and OâNeill, 2003, and references therein). If such a veneer was added (presumably sent in from the outer asteroid belt region as discussed under Question 1), it might also have included a substantial amount of volatile elements like water, sulfur, and carbon compounds. In this case, much of Earthâs water and CO2 could have been added long after the Moon-forming impact and might not have been present to form the blanketing atmosphere as- sumed in the discussion above. Another unresolved issue is whether Earth had a primordial atmosphere at the time the core was forming. If it did, the core might still have substantial amounts of H, He, and other gases because the thick atmosphere would have kept the gases dissolved in the mantle, and if the mantle had enough of these gases, the core should have gotten its share as well. To address this issue we need to know more about the physics of FIGURE 1.11â Possible core formation scenario during an Figure 1.11.eps early magma ocean in the early Earth. Small droplets of molten separating metal from silicate. Did separation occur metal sink to the base of a magma ocean, equilibrating as they when both components were molten (e.g., in a magma go, and pond when they reach the magmaâsolid rock interface. ocean) or by percolation of more easily melted metal- From there giant molten drops of metal (diapirs) sink through the lic liquid through solid rock? Experimental studies of solid but plastically deforming rocky lower mantle to reach the growing core. These diapirs do not equilibrate as they sink, so how metallic melts behave when mixed with silicates the overall pressure and temperature of metal-rock equilibration are beginning to shed light on this issue (Hustoft and is set at the base of the magma ocean. SOURCE: Wood et al. Kohlstedt, 2006; see Question 4). (2006). Reprinted by permission from Macmillan Publishers Ltd.: Nature, copyright 2006. How Did Earthâs Earliest Crust Form and What Became of It? A central question about the Hadean Earth concerns event on core formation and metal-silicate differentia- the nature of its crust and whether, in the absence of tion remain open questions. hard evidence, we can assess whether the crust had any Clues about how the core formed come from stud- similarity to the modern Earthâs crust. Most approaches ies of siderophile, or âmetal-loving,â trace elements to this question begin with evidence from other plan- (such as W, Pt, Os, and Pd). These elements are pres- etary bodiesâthe Moon, Mars, and Venusâand from ent in mantle rocks in the same relative proportions the oldest rocks and minerals found on Earth (see Box as in chondritic meteorites. This observation tells us 1.4). The results are so far inconclusive, but the nature that metal-silicate separation did not happen mainly at of the debate is rapidly changing as a result of new low pressure, where these elements would be strongly observations. fractionated relative to one another. One hypothesis to Earth today has two kinds of rocky crust, both of explain this observation is that the metal and silicate which are chemically different from the mantle (see last equilibrated chemically at the base of a magma Questions 4 and 5). Oceanic crust is relatively simple ocean, where higher pressures may cause these elements and is typically composed of solidified basaltic magma to enter the metal in the required similar proportions melted from the mantle. It forms by a well-understood (Righter et al., 1997; Figure 1.11). A competing hy- process at midocean ridges and returns to the mantle
24 ORIGIN AND EVOLUTION OF EARTH to effectively 0 million years. Like oceanic crust, conti- nental crust appears to be ârecycledâ to the mantle, but at an unknown rate. The surface of average continental crust stands about 5 km higher than the surface of the average oceanic crust, so Earthâs water is collected in the basins underlain by oceanic crust, and there is abundant dry land rather than a globe-encircling ocean. Crusts are widely variable throughout the Solar System and offer no clear insight about what Earthâs earliest crust was like. Samples returned by astronauts showed that the Moonâs light-colored highland crust is very old (ca. 4,400 million years) and probably formed from feldspar crystals that floated to the surface after the Moon-forming impact when it was largely molten (Figure 1.12). The crust of Mars appears to be variable in age, but most is extremely old (Frey, 2006). The crust of Venus is much less well known, but a large fraction is thought to be young (Hansen, 2005; Basilevsky and Head, 2006). The crusts of the larger moons of Jupiter and Saturn seem to resemble our conceptions of the early Earth in interesting ways. Jupiterâs moon Io, for FIGURE 1.12â The heavily cratered light-colored areas of the lunar surface, the lunar âhighlands,â reflect the intense rain of meteorites that occurred in the earliest history of the Solar Sys- tem. The highlands are composed of rock made mostly of a single mineral, plagioclase feldspar, which floated to the surface as the magma ocean crystallized, at about 4,500 Ma. The large, dark lunar âseas,â or maria, are huge impact basins that formed mostly between 4,000 and 3,900 Ma and are evidence of a late heavy meteorite bombardment that would also have affected Earth (see Box 1.5). The lunar maria are filled with dark lava flows of basalt that formed 3,900 to 3,300 Ma. The lower crater density in the maria indicates that the meteorite flux dropped off considerably by the time the lava flows formed. SOURCE: <http://www.nasa. gov/multimedia/imagegallery/image_feature_25.html>. by moving downward in subduction zones. Oceanic crust is thin (about 6 to 8 km), submerged under the oceans, and relatively young; its average age is about 60 million years, which is only 1.4 percent of Earthâs age. The continents, which are mostly above sea level, are underlain by a different kind of crust. Continental crust is a quilt of rocks of vastly different compositions, textures, and ages and forms by multistage processes FIGURE 1.13â Image of Jupiterâs moon, Io, from the National that are only partly understood. It is also thick (30 to Aeronautics and Space Administrationâs Galileo spacecraft. 80 km), more silica rich than basalt, and generally old. Io is a volcanically active miniplanet, with young crust and no The average age of continental rocks is about 2,000 plate tectonics. SOURCE: <http://www2.jpl.nasa.gov/galileo/ callisto/PIA00583.html>. million years, but they range from 4,000 million years
ORIGINS 25 example, which is rocky and about the size of Earthâs Moon, is thought to have a young crust (Figure 1.13), which it resurfaces rapidly by continuing volcanism. However, none of the rocky planets or moons have Earthâs crustal resurfacing mechanism of plate tectonics (see Question 5). One of the most obvious qualities of Earthâs early crust is that it no longer exists. Why did it all disap- pear? The type of crust most likely to be preserved is continental crust, since virtually no oceanic crust endures for more than about 200 million years before descending into the mantle at subduction zones (Ques- tions 4 and 5). One possibility is that the early Earth had only an oceanic-type crust and no continents. However, virtually all of the rocks preserved from the FIGURE 1.14â Photograph of exposures of some of Earthâs oldest sedimentary rocks (about 3,700 million years), from the period 4,000 to 3,600 Ma are continental (3.8 Ga eastern Isua supracrustal belt in West Greenland. Metacherts ophiolite in Greenland is an exception; see Furnes et (light gray) are interlayered with carbonate and calcsilicate al., 2007), and the only earlier materials are tiny zircon metasediments (dark gray). SOURCE: Friend et al. (2007). crystals that presumably also come from continental- Reprinted with kind permission of Springer Science and Busi- ness Media. type rocks such as granite. Isotopic evidence suggests the presence of pre-3.8 Ga continental crust, although the relative proportion varies with isotopic system (Nd isotopes suggest a greater proportion of ancient crust than Hf isotopes; Bennett, 2003). The fact that some of amount and distribution of water, we do not know the oldest rocks are water-deposited sediments (3,800- whether oceanic crust production was similar to that million-year-old rocks from Isua, Greenland) also on the modern Earth, whether plate tectonics was op- indicates that there was erosion and transport of sedi- erating, and how efficiently continental crust was being ment, which requires land standing above sea level at formed and recycled. The end of the Hadean, perhaps that time (Figure 1.14). At the average rate that Earth coincidentally, corresponds to the time of the âlate has been producing continental crust over the past 2 heavy bombardmentâ of the Moonâs surface, which billion years, we would expect one-fifth the mass of the produced the large lunar impact basins that were sub- present continental crust to have been produced in the sequently filled with basalt lava flows (Figure 1.12; Box Hadean. However, the total volume of rocks older than 1.5). Earth probably experienced this bombardment as 3,600 million years is very smallâabout 0.0001 percent well, but it is doubtful that such intense bombardment of the continents. The recent observation that every could cause the disappearance of a large preexisting Earth sample measured is enriched in 142Nd compared continental crust, given that low-density ancient crust to chondritic meteorites suggests very early formation is preserved on the Moon. Rather, vigorous internal of a crustal component enriched in incompatible ele- convection is more likely responsible for the demise of ments (such as the light rare-earth elements) and its Earthâs original crust. removal from the accessible portions of Earth (Boyet and Carlson, 2005). If this interpretation is correct, Summary Earthâs original crust may lie sequestered in the deep Earth today. The geological period called the Hadean, which ex- The uncertainties about any aspect of Hadean crust tends from the time of the Moonâs formation to the are large. Under the conditions of the Hadean Earth, time when the oldest Earth rocks were formed (~4.5 which was hotter, still being hit by meteorites in its to 3.9 Ga), is critical to our understanding of planetary waning stages of accretion, and bearing an unknown evolution. If we are ever to fully appreciate how our
26 ORIGIN AND EVOLUTION OF EARTH BOX 1.5 Late Heavy Bombardment A major scientific discovery that came out of the Apollo program is that at about 3.9 Ga the Moon was pummeled by several 100-km asteroids (or comets) and by hundreds of 10-km asteroids (Wilhelms, 1987). The craters they made carved the face of the Moon. Because Earthâs effective cross sec- tion is 20 times bigger than the Moonâs, Earth must have been hit 20 times as often. But not only was Earth hit by a hundred 100-km asteroids, statistics imply that it was also hit by a dozen bodies bigger than any that hit the Moon. The biggest would have been comparable to Vesta or Pallas, the largest asteroids now in the asteroid belt. Whether these impacts marked the tail end of a sustained bombardment dating back to the accretion of the planets or whether they record a catastrophic event, such as a sudden influx of planetesimals to the inner Solar System due to rapid migration of the giant planets (e.g., Gomes et al., 2005), is contentious but of great importance to the Hadean environment. Examples of both possibilities are shown in the figure. 104 50 Myr half-life Impact Rate (Relative to Today) 1000 multiple 100 Myr half-life 100 cataclysms 10 Single Cataclysm LHB 1 4.4 4.2 4 3.8 3.6 Time (Gyrs) Four models of the impact rate of the first billion years of the Moonâs life: a single figure.eps late heavy bombardment (LHB), multiple cataclysms Box 1.5 cataclysm with a throughout the Hadean, and sustained bombardments (denoted 50-Myr [million year] half life and 100-Myr half life). The single- and multiple-cataclysm curves are schematic representations, and the sustained bombardment curves are standard impact rates based on lunar crater counts and surface ages of the Apollo landing sites and impact basins. The 100-Myr half-life curve incorporates the age of the Imbrium impact basin and is more consistent with terrestrial and Vestan impact records than the 50-Myr half life curve, which incorporates the age of the Nectaris impact basin. SOURCE: Courtesy of Kevin Zahnle, National Aeronautics and Space Administration. Available data offer some support for a late cataclysm, but not for the enormous hidden impacts implicit in monotonic decline. The most telling argument against a huge unseen Hadean impact flux is that it does not explain anything else in the Solar System that needs explaining. By contrast, a cataclysm (or cataclysms) fits in well with current concepts of how a solar system might evolve. All that is required is a rearrangement of the architecture of the Solar System; such rearrangements are a natural consequence of the dynamical evolution of a swarm of planets (the Moon-forming impact provides a cogent example) and are expected to occur on every timescale (Gomes et al., 2005). Before the cataclysm, impact rates would have been higher than they are today, because there were more stray bodies in the Solar System. The impacts of the late heavy bombardment would have posed a recurrent hazard to life on Earth. Impacts by asteroids as big as Pallas or Vesta would have been big enough to boil away the oceans and leave Earth enveloped in 1500 K steam. The lunar impact record suggests that one or two of these struck Earth ca. 4.0 Ga. Conditions a few hundred meters underground would be little changed and life could have gone on (Sleep et al., 1989; Zahnle and Sleep, 1997). Later, smaller impacts may have boiled half the ocean and left the rest a scalding brine. It is this scale of event that suggests that life on Earth may have descended from organisms that either lived in hydrothermal systems or were extremely tolerant of heat and salt. It has been widely postulated that all life appears to descend from thermophilic organisms (Wiegel and Adams, 1998). Whether this means that life originated in such environments or that life survived only in such environments is debated. If the latter, the thermophilic root implies that life on Earth arose in the Hadean during the age of impacts.
ORIGINS 27 planet came to be the home of complex life, we must QUESTION 3: HOW DID LIFE BEGIN? be able to fill in this enormous gap in the geological record. At present we can construct plausible, but still The origin of life stands as one of scienceâs deepest and highly uncertain, models for the Hadean Earth, which most challenging questions. It is a historical problem are based on our present understanding of planet for- that emerged during a time with little recorded history, mation (Question 1), planetary interior processes and so it must be approached mostly through theory and ex- material properties (Questions 4 and 6), and climate perimentâimaginative efforts to re-create our planetâs (Question 7). These models are informed by obser- early conditions and establish plausible chemical routes vations of the Moon and other planets in the Solar to the emergence of life. The goal of understanding System, by measurements made on meteorites and the lifeâs beginnings has attracted scientists from geology oldest rocks and minerals on Earth and the Moon, and and from many overlapping disciplines, especially sub- by our geological understanding of how the modern fields of organic chemistry and molecular biology. In Earth works. A critical component of understanding an age of planetary exploration, the origin of life is also Hadean climate is our knowledge of atmospheric an astrobiological issue, currently investigated on Mars, processes, but despite the advanced state of models where a sedimentary record of earliest planetary history for the modern Earth atmosphere, our understanding is preserved, and potentially across the wider stretch of of radically different types of planetary atmospheres Universe where planets have been detected. is still rudimentary. Some of the most fundamental mysteries about the Recent studies have raised new hope of improving origin of life are geological in nature: From what mate- our understanding of the Hadean. New information rials did life originate? When, where, and in what form continues to be gleaned from precise measurement did life first appear? At its most basic physical level, life of the isotopic and chemical compositions of ancient is a chemical phenomenon, and because it arose billions zircons and their mineral inclusions. Observations of of years ago, geologists are intensely interested in creat- the Moon, Mars, Venus, and the moons of Jupiter and ing an accurate picture of the chemical building blocks Saturn have opened new windows for visualizing the available to early life. early Earth and for documenting what may have been happening in the early Solar System. Comparison of Top-Down and Bottom-Up Approaches meteorites with Earth rocks has led to better models of Earthâs early internal processes, including the forma- In The Origin of Species, Charles Darwin (1859) hy- tion of the metallic core, the implantation and loss of pothesized that new species arise by the modification of gaseous species from Earthâs interior, and the evolution existing onesâthat the raw material of life is life. Louis of the crust and mantle. Pasteur, Darwinâs great Parisian contemporary, went a The future is certain to provide additional break- step further. Pasteur decisively refuted the doctrine of throughs. Capabilities for microanalysis of geological spontaneous generation, the long-held view that life materials are improving, and hence the amount of in- can arise de novo from nonliving materials, declaring formation that can be extracted from even the tiniest instead that life springs always from life (Pasteur, 1922- samples of old rocks and minerals is increasing rapidly. 1939). These conclusions, among the most important With concerted effort, it is expected that many more of 19th-century science, require that forms of life ancient rocks and mineral samples will be found. More developed in an unbroken pattern of descent through precise isotopic measurements are revealing clues to time, with modifications, to produce the biological early planetary processes. Planned spacecraft missions diversity we see today. And indeed, students of fossils to the Moon and Mars will provide critical informa- have painstakingly traced such a pattern backward for tion about the nature of planets in the Hadean. There more than 3 billion years to the time of our planetâs is even a chance that pieces of Hadean Earth rocks infancy (Knoll, 2003). will be found on the surface of the Moon, sent there Before then, however, somehow and somewhere, by impacts on Earth in the same way that pieces of the the tree of life had to take root from nonliving precur- Moon and Mars have been sent here. sors. Scientists have tried to identify these precursors
28 ORIGIN AND EVOLUTION OF EARTH from both the top down and the bottom up (Penny, environments that have less hydrogen and therefore are 2005). Top-down approaches, favored by biologists, less strongly reducing (Kasting and Catling, 2003). In look at the complex molecular machinery of living cells contrast, Tian et al. (2005) have argued that less hydro- for clues about simpler antecedents on the early Earth. gen escaped to space from the early atmosphere than Bottom-up approaches, pioneered by chemists, inves- was previously assumed, which implies that while most tigate the pathways by which lifeâs chemical building carbon in the primitive atmosphere was in the form blocksâthe raw materials for top-down researchâ of carbon dioxide, hydrogen gas was also available for could have formed from simple inorganic constituents organic synthesis, with energy added by lightning. Im- of early environments. These bottom-up approaches re- pacts by iron-rich meteorites might also have transient quire the input of Earth scientists because they specify enrichment in compounds such as carbon monoxide physical setting, starting materials, energy sources, and that would have facilitated the synthesis of biologically chemical catalysts. Did life originate in what Darwin interesting organic compounds (Kasting, 1990). envisaged as a âwarm little pond,â perhaps a tidal pool repeatedly dried and refreshed? Or might life be rooted among hydrothermal vents? Could lifeâs origins even lie beyond Earth? Experimental approaches to prebiotic chemistry must be framed in terms of environments likely to have formed lifeâs incubators, and only Earth scientists can inform us about the physical and chemical characteristics of these settings. A Search for Clues in the Laboratory We have understood for more than half a century that modern laboratory experiments can shed light on our search for lifeâs beginning. In the classic Miller-Urey experiment, Stanley Miller (1953) ran an electric spark through a mixture of water vapor, ammonia, methane, and hydrogen gas, generating a complex array of or- ganic molecules, including amino acids, the building blocks of proteins (Figure 1.15). Intermediate products in amino acid synthesis included formaldehyde (from which sugars can be synthesized) and hydrogen cyanide (the starting material for abiotic synthesis of the bases that specify information in nucleic acids). In this experiment the spark serves as a proxy for lightning in the early atmosphere. The gas mixture approximates one hypothesis for atmospheric com- position. As it turns out, the success of Miller-Urey and other experiments in prebiotic chemistry depends critically on the relative amounts of gases present in the early atmosphere and oceans. The Miller-Urey mecha- FIGURE 1.15â Photograph of the experimental setup of the famous Miller-Urey experiment. An electric spark passes through nism requires more hydrogen than carbon (Miller a chamber containing hydrogen gas, ammonia, methane, and and Schlesinger, 1984), and Miller chose his starting water vapor; as the product of the resulting chemical reaction mixture to approximate the prebiotic atmosphere as cools, water condenses, carrying organic molecules to the flask at the bottom of the apparatus, where they can be sampled and envisioned by his mentor Harold Urey. But since then analyzed. SOURCE: Bada and Lazcano (2003). Reprinted with most atmospheric scientists have adjusted the model to permission from AAAS.
ORIGINS 29 The availability of gases such as hydrogen and car- The idea that metal ions, dissolved in early lakes bon monoxide in Earthâs early atmosphere is currently and oceans, might have catalyzed prebiotic chemi- the subject of vigorous debate among Earth scientists, cal reactions follows closely from our knowledge of and its outcome will determine how we think about biochemistry. Biological catalysts commonly depend environmental chemistry and the origin of life during on the action of a cofactor that contains a metal at its Earthâs early development. Whether or not amino acids functional heart. For example, a magnesium ion occu- and other organic molecules were widespread on the pies the center of the chlorophyll molecules that trap early Earth, they existed in some parts of the early Solar light energy and drive photosynthesis. Similarly, an iron System and reached our planet in the form of carbona- atom lies at the center of hemoglobin, the molecule ceous chondrites. These meteorites contain significant that transports oxygen in mammalian respiration. A abundances of biologically interesting compounds, as wide diversity of metals act as important catalysts for do some interstellar clouds. biological reactions, especially iron, manganese, mag- nesium, zinc, copper, cobalt, nickel, and iron-sulfur How Did Life Arise? clusters. Understanding the roles these metals might have played in prebiotic chemistry is a geological ques- Earth scientists are trying to answer this question by tion whose answer depends on how the metals were dis- combining field and laboratory studies of the planetâs tributed in primitive Earth environments. To find such oldest sedimentary rocks, laboratory simulations, and answers we need integrated data about (1) early crustal geochemical theory to define the environmental condi- differentiation and magma generation (see Question tions most likely to have nourished early life. A central 2), (2) the low-temperature chemistry of weathering, question, for example, is what combination of the basic (3) hydrothermal reactions in ancient seafloors, and conditionsânitrogen and phosphate availability, elec- (4) oxidation-reduction (redox) conditions in early trochemical and acid-base qualities of the environment, environments. Once we understand these conditions, and abundances of trace metals and mineralsâwere experiments in prebiotic chemistry can graduate from the most life enhancing? The challenge is to identify artificial media doped with single metal ions to complex and quantify every one of these conditions to actually ionic mixtures informed by Earth science. estimate the probability of forming life under primitive The same is true for mineral surfaces, long rec- Earth conditions. Because those conditions are today ognized as potentially important catalysts of prebiotic poorly preserved or absent, geologists must adapt tools chemical reactions (Schoonen et al., 2004; Figure 1.16). of many kinds to infer how life began. Clay minerals, for example, have been shown to cata- Essential to our understanding of how life emerged lyze the assembly of lipid micelles into vesiclesâtiny from prebiotic chemicals is accurate knowledge of spheroids that could have governed prebiotic-phase the kinds of catalysts present in the environment. A separation on the early Earth (Hanczyc et al., 2003). catalyst is a substance that increases the speed of a Clay minerals also catalyze the linkage of nucleotides chemical reaction, often dramatically. In every cell to form nucleic-acid-like polymers (Orgel, 2004), and the complex and coordinated chemical reactions that pentose sugars (including the biologically important support life require the action of catalysts, usually ribose) can be stabilized in the presence of calcium bo- enzymatic proteins. Many prebiotic reactions require rate minerals (Ricardo et al., 2004). A role in prebiotic catalysts as well, not only to support energy-yielding chemistry for iron sulfide minerals has been suggested reactions but also to permit the synthesis of the long- as well, most prominently in Wächtershäuserâs chemi- chain molecules such as nucleic acids and proteins that cally explicit theory of biogenesis around hydrothermal make up living systems. Some of the most essential vents (Wächtershäuser, 1988; see Hazen, 2005, for a catalysts used in experimental approaches to prebiotic discussion of recent experimental tests). Continuing chemistry are metal ions, which coordinate chemi- advances will require new experiments based on realistic cal reactions in developing metabolism, and mineral mineral catalysts, as well as constrained theory, experi- surfaces, which provide templates and catalysis in ments, and observations from Earth science (Schoonen synthesizing biopolymers. et al., 2004). In particular, we need to understand how
30 ORIGIN AND EVOLUTION OF EARTH FIGURE 1.16â Diagram show- ing the role of minerals in prebi- otic chemical reactions. SOURCE: Schoonen et al. (2004). Reprinted with permission. chemical reactions between water and the early crust logical processes such as photosynthesis tend to enrich shaped the chemistry of early environments. organic molecules in the lighter stable form of carbon (12C) relative to its heavier forms, we can estimate when When Did Life Arise? carbon began to be trapped by photosynthetic microor- ganisms. Carbon isotopic abundances in 3,500 million A second important question flows from the first: year old sedimentary rocks are similar to those found When did life arise on our planet? Paleontologists and in much younger deposits, suggesting that a biologi- biogeochemists have long agreed that the origin of life cal carbon cycle was established early in our planetâs preceded the deposition of minimally metamorphosed history. Indeed, highly metamorphosed rocks that are sedimentary rock deposited 3,500 to 3,400 Ma. Tiny nearly 3,800 million years old contain carbon isotopic fossils preserved in sedimentary rocks document mi- abundances suggestive of a still older carbon cycle. It crobial diversity in rocks deposited long before animals has further been proposed that the high concentra- evolved, and stromatolitesâsedimentary structures tions of organic matter in some of the earliest known formed by the interaction of microbial communities shales require primary production by photosynthetic and the physical processes of sedimentationâprovide organisms (Sleep and Bird, 2007). In light of these ob- independent evidence of widespread microbial life on servations, the close molecular similarity of all known the early Earth (Knoll, 2003; Figure 1.17). Because bio- species strongly suggests that all living organisms are descended from a common ancestor that lived nearly 4 billion years ago. Did Life Originate More Than Once? We cannot tell how many times life arose. Life may have originated many times on the young Earth, with the ancestor of present life persisting by good luck (chance survival of primordial mass extinctions) or good genes (outcompeting other early life forms). But experiments can help us understand whether there is more than one route to life. There is no reason these routes must all be terrestrial, and some scientists have speculated that terrestrial life was seeded from afar, most likely from Mars (Weiss et al., 2000). A mecha- nism certainly existsâseveral lines of evidence show FIGURE 1.17 2.76 billion year old stromatolite in Pilbara, that Earth receives a continuing stream of meteorites Australia. SOURCE: Ohmoto et al. (2005). Reprinted with permission. ejected to space from Mars by meteor impact and that
ORIGINS 31 some of these meteors could have delivered microbial eat? Can it move against gravity? Paleontologists have cargo to Earth. The obvious test is to learn by explo- a more difficult task, necessarily judging biogenicity ration whether Mars was ever a biological planet. At by shape, distribution, and chemistry. No sensible per- present we do not know, but exploration of ancient son would doubt that dinosaur skulls excavated from sedimentary rocks on Mars, guided by our geological Cretaceous sandstones constitute definitive evidence of and paleobiological experiences on Earth, may provide ancient life; no known physical processes can produce an answer. From orbital observations and the in situ ex- the complexities of a skull in the absence of biology. ploration by the Mars rovers Spirit and Opportunity, we Similarly, the preservation of cholestane (the geologi- know that Marsâunlike Earthâpreserves a sedimen- cally preservable form of cholesterol) in a Jurassic oil tary record of surface environments from its first 500 tells us that life existed when the oil deposit formed million years (e.g., Squyres et al., 2004). Thus, Martian because cholestane does not form abiologically. The rocks might preserve a record of prebiotic chemistry, problem gets harder when we go backward in time or even nascent life, if such records ever formed. Many beyond the first appearance of animals ca. 580 Ma. scientists have attempted to estimate the odds that life Some microfossils have complicated shapes clearly can emerge as a lucky accident, whether on a planet or related to living organisms (Figure 1.18a, b), and an elsewhere where environmental conditions are favor- unambiguous record of microfossils goes back some able. Experiments in prebiotic chemistry will nudge 2,500 million years. Older candidate fossils, however, us toward better answers, but what the question really tend to be poorly preserved and have simple shapes. requires is a second example of a living system. The tiny spheroid structure in Figure 1.18c is about In recent years, however, skeptics, stimulated in 3,500 million years old and is made of carbon. It is hard part by controversial claims about biological signatures to be sure this is a fossil because such simple structures in a Martian meteorite, have challenged the conven- might well form from physical processes. tional wisdom that terrestrial life arose on Earth prior The same uncertainties confound investigations of to 3,500 Ma. Explanations that do not involve biology larger scale features of sedimentary rocks that may have have been proposed for micron-scale carbon-bearing been imported by organisms, as well as molecular or structures previously interpreted as Earthâs oldest isotopic features of ancient organic matter that might microfossils, for stromatolites, and for carbon isotopic reflect biological processes. Stromatolites, for example, abundances in carbonate minerals and organic matter are commonly interpreted as the sedimentary products (e.g., Brasier et al., 2005, 2006). Vigorous defenses of of sediment accretion on ancient lake bottoms and sea- biological interpretations have been mounted (e.g., floors. Stromatolites formed by trapping, binding, and Schopf et al., 2002; Allwood et al., 2006; Schopf, 2006). cementing sediment particles have textures not easily At present the weight of evidence favors the hypothesis mimicked by purely physical processes, so they pro- that life existed 3,500 Ma, and likely existed back at vide reliable evidence for life in rocks more than 3,000 least 3,800 Ma, but much remains to be learned about million years old (Figure 1.19a). Other stromatolites the nature of early ecosystems. Only careful mapping form by mineral precipitation, however, especially in and stratigraphic analysis will tell us whether our planet the oldest sedimentary accumulations, and it is difficult preserves an earlier record of its biological (or prebio- to know what role, if any, life played in their accretion logical) history, and only innovative biogeochemical (Figure 1.19b). analyses set in the context of well-corroborated mi- The challenge of identifying the geological prod- crobial phylogeny will resolve uncertainties about the ucts of life becomes even more difficult when applied antiquity and nature of early microorganisms. to ancient rocks of Mars or other planets. We have no confidence that the diversity of life on Earth exhausts What Is Lifeâand What Is Not Life? all possibilities for living systems. Thus, the guiding question of paleo- and geobiological exploration of the In one way, at least, biologists have it easy: they can Solar System is whether a structure (molecular, micro- evaluate whether a structure is living by testing for scopic texture, or stromatolite) found during planetary evidence of metabolic activity. Does it breathe? Does it exploration can be explained adequately in terms of
32 ORIGIN AND EVOLUTION OF EARTH a b c FIGURE 1.18â (a) Fossil of a eukaryotic microorganism preserved in ca. 580 Ma phosphorite from the Doushantuo Formation, China. The fossil is 250 microns across. (b) Branching cyanobacterium preserved in ca. 800 Ma chert from the Upper Eleonore Bay Group, Greenland. The fossil is 500 microns from left to right. (c) Paired 4-micron-wide carbonaceous spheroids in ca. 3,500 Ma cherts from the Onverwacht Group, South Africa. Are these fossils? SOURCE: Courtesy of Andrew Knoll, Harvard University. a b FIGURE 1.19â (a) Stromatolite built by the trapping and building of sediment particles by microbial communitiesâ1,500 Ma Bilâyakh Group, Siberia. SOURCE: Courtesy of Andrew Knoll, Harvard University. (b) Stromatolites built of seafloor precipitate structures that are composed of calcium carbonate crystals without any obvious templating influence of microorganismsâ1,900 Ma Rocknest Forma- tion, Canada. SOURCE: Courtesy of John Grotzinger, Caltech. Used with permission.
ORIGINS 33 known physical processes. Some molecular and mor- Clay minerals in some of Marsâ oldest terrains may phological structures form only by biological processes signal that early in its history our neighbor was rela- (cholesterol, dinosaur skulls), while others clearly relate tively wet but less acidic (Bibring et al., 2006). Also, to physical processes (large quartz crystals, for exam- carbonate and sulfide minerals precipitated from fluids ple), and still others exist in a zone of overlap (2-micron flowing through crustal fractures document at least spheres, amino acids). We can never eliminate the zone transient subterranean environments neither strongly of overlap, but better understanding of the products of acidic nor oxidizing (McKay et al., 1996). Only fur- both biological and physical processes will better equip ther exploration, with Earth and planetary scientists us to pursue questions of lifeâs antiquity on Earth and working in partnership, will establish whether life on its distribution through the Solar System. Earth is unique in our Solar System or merely uniquely successful. Is There Life Beyond Earth? Summary Our understanding of our own origins remains sketchy, but it is expanding at an accelerating pace. Thanks to While synthetic organic chemistry and molecular biol- contributions from many fields and approaches, scien- ogy will continue to provide the experimental basis for tists are better prepared to approach a truly tantalizing probing lifeâs origins, Earth scientists will increasingly question: Are we alone, or has life also evolved else- specify the conditions under which laboratory experi- where? If life exists elsewhere, what forms does it take? ments are run. Stratigraphers, paleontologists, biogeo- With continuing planetary exploration, Earth scientists chemists, and geochronologists can provide sharper will be able to establish with greater certainty whether constraints on when life arose and the metabolic life could have originated elsewhere in our Solar Sys- character of early organisms. Geochemists focused temâand even whether organisms could have become on both crustal differentiation and low-temperature established on Earth by meteoritic transfer from an- reactions can build an improved sense of redox condi- other planet. Thanks to discoveries of the National tions, weathering reactions, and metal abundances on Aeronautics and Space Administrationâs rover Op- the early Earth. Modelers can use new data to provide portunity, we now know that around the time life took more sophisticated hypotheses about how our planetary root on Earth, at least regional environments on Marsâ surface operated in its infancy, setting the stage for surface were episodically wet (Knoll et al., 2005). But the intercalation of biological processes into the Earth they were also oxidizing and strongly acidicâserious system. And planetary scientists, now exploring Mars obstacles to many of the prebiotic chemical pathways and other bodies at a resolution previously limited to thought to have been important on Earth. Was early Earth, can provide comparable insights about environ- Mars arid, oxidizing, and acidic globally or just region- mental (and, at least potentially, biological) evolution ally, and when were such environments established? on other planets.