Until two years ago, the lowest number of chromosomes, a commandingly minimal one pair, had been found for only a single organism—a nematode worm, appropriately honored in its subspecific name as Parascaris equorum univalens. This minimal complement had been discovered long ago, in 1887, by Theodor Boveri, the greatest cytologist (student of cellular architecture) of the late nineteenth century. Boveri (1862–1915) was a great intellectual in the European tradition—a complex and fascinating man who lived for the laboratory, but who also played the piano and painted with professional competence. His short life was scarred by fits of depression, and he died in despondency as the First World War enveloped Europe. Of Boveri’s many scientific discoveries, the two greatest centered on chromosomes. First, he established their individuality and shifted attention from the nucleus as a whole to chromosomes as the agent of inheritance (in years before the rediscovery of Mendel’s laws). Second, he demonstrated the differential value of chromosomes. Before Boveri’s experiments, many scientists had conjectured that each chromosome carried all the hereditary information, and that organisms with many chromosomes carried more copies of this totality. Boveri proved that each chromosome carried only part of the hereditary information (some of the genes, as we might say today), and that the full complement built the organism through a complex orchestration of development.
Boveri took great interest in his discovery of an organism that carried but one pair of chromosomes per cell—and therefore did place all its hereditary information into one package. But Boveri quickly discovered that P. equorum univalens, though no imposter in its claims to minimalism, was not entirely consistent either. Only the cells of the germ line, those destined to produce eggs and sperm by meiosis, kept all the hereditary material together in a single pair of chromosomes. In cells destined to form body tissues, this chromosome fractured several times during the first cleavage divisions of early embryology, leading to adult cells with up to seventy chromosomes!
Finally, in 1986, Australian zoologists Michael W. J. Crosland and Ross H. Crozier reported a remarkable new species within a closely related group of ants, previously united into the overextended species Myrmecia pilosula (see their article of 1988 cited in the bibliography). This name falsely amalgamates several distinct species sharing a similar body form, but carrying different numbers of chromosomes in their cells. Species with nine, ten, sixteen, twenty-four, thirty, thirty-one, and thirty-two pairs of chromosomes have been described. Obviously, this complex of forms has evolved some way of speciating in concert with substantial changes in chromosome number.
On February 24, 1985, on the Tidbinbilla Nature Reserve near Canberra, Crosland and Crozier collected a colony of winged males and females, plus a mated queen with pupae and more than 100 workers. All workers tested from this colony carried but a single pair of chromosomes in their cells—all of them, not just cells of a particular type. An unambiguous example of chromosomal minimalism had finally been discovered, almost exactly 100 years after Boveri found only one pair of chromosomes in the germ line cells of Parascaris.
But the story of M. pilosula is even better, deliciously so. If you were out searching for absolute minimalism, you would have to root for finding your single pair of chromosomes in an ant, bee, or wasp—for the following interesting reason: The Hymenoptera, and just a few other creatures, reproduce by an unusual genetic system called haplodiploidy. In most animals, all body cells contain chromosomes in pairs, and sex is determined by maternal and paternal contributions (or noncontributions in some cases) to a single pair. But haplodiploid organisms specify sex by a different route. Reproductive females usually store sperm, often for long periods. Genetic females (including the functionally neuter workers) arise from fertilized eggs, and therefore contain chromosomes in pairs. But males are produced when the queen fails to fertilize a developing egg with stored sperm (in most other animal groups unfertilized eggs are inviable). Thus, the cells of male ants, bees, and wasps do not carry chromosomes in pairs and bear only the single set inherited from their mother. These males have no father, and their cells contain only half the chromosomes of females—a condition called haploid, as opposed to the diploid, or paired, complement of their sisters. (The entire system therefore receives the name haplodiploid, or male-female in this case.)
Haplodiploidy implies, of course, that males of the Tidbinbilla colony of M. pilosula have a truly and absolutely minimal number of one chromosome per cell. Not even a single pair—just one. The only lower possibility is disappearance. Crosland and Crozier checked just to be sure. The males of their colony contained a single chromosome per cell.
If we have reached a limit in the search for less, the other extreme seems more open-ended. How many chromosomes can a cell contain and still undergo the orderly divisions of mitosis and meiosis? Can hundreds of chromosomes line up neatly along a mitotic spindle and divide precisely to place an equal complement into each daughter cell? At what point do things become so crowded that this most elegant of biological mechanisms breaks down?
Maximal numbers are most easily reached by polyploidy, or doubling of chromosomes. This process occurs in two basic modes with differing evolutionary significances. In autoploidy, a cell doubles its own complement, forming, initially at least, a cell with two sets of identical pairs. Thus, the new autoploid usually looks like its parent. Autoploidy is not a mechanism for rapid evolution of form, though the redundancy introduced by doubling does permit considerable evolutionary divergence afterward—as one member of the duplicated pair becomes free to change. On the other hand, alloploidy, the second mode of doubling, can produce viable hybrids between distant species and can serve as a mechanism for sudden and substantial changes in form. Hybrids, with different forms and numbers of maternal and paternal chromosomes, will usually be sterile because chromosomes have no partners for pairing before meiosis—the “reduction division” that produces sex cells with half the genetic information of body cells. But if the precursors of sex cells undergo polyploidy, then each chromosome will find a partner in the duplicated version of its own form.
Since polyploidy is so much more common in plants than animals, we should search for maximalism in our gardens, not our zoos. The numerical importance of polyploidy in plants can best be appreciated in a wonderful graph that I first encountered, when a graduate student, in Verne Grant’s The Origin of Adaptations. This graph is a frequency distribution for chromosome pairs in monocot plants. For ten pairs of chromosomes and higher, without exception, all peaks are for even numbers of chromosome pairs.
At first inadequate sight, this pattern doesn’t make sense in the deepest possible way. Biology is not numerology; its regularities do not take the form of such abstractions as “cleave to even numbers.” Such a graph will not be satisfying until we figure out a biological mechanism that, as a side consequence and not because evens are better than odds per se, produces an imbalance of species with chromosomes in pairs of even numbers. The resolution is elegantly simple in this case. Polyploidy is very common in plants, and every number, odd or even, when doubled, yields an even number. The peaks therefore indicate the prevalence of polyploidy in plants. Estimates range as high as 50 percent for the number of angiosperm species produced by polyploidy.
Since polyploidy can continue in cycles—doubling followed by redoubling—chromosome numbers, like the pot in a poker game with table stakes, can rise alarmingly from small beginnings. The champions among all organisms are ferns in the family Ophioglossaceae. The genus Ophioglossum exhibits a basic number of 120 chromosome pairs, the lowest value among living species. (Such a high number must, itself, be derived from earlier incidents of polyploidy among species now extinct. The basic number for the entire family, 15 pairs, may have been the starting point.) In any case, cycles of polyploidy have proceeded onward from this already large beginning of 120 pairs. The all-time champion, not only in Ophioglossum, but among all organisms, is Ophioglossum reticulatum, with about 630 pairs of chromosomes, or 1,260 per cell! (The tota
l need not be an exact multiple of 120, because doubling may be imperfect, and secondary gains or losses for individual chromosomes are common.)
Frequency distribution for the number of chromosome pairs in monocot plants. Note that all peaks are for even numbers of chromosomes. This occurs because so many plant species are produced by polyploidy, or doubling of chromosome number, and a doubling of any number, odd or even, produces an even number. FROM VERNE GRANT, THE ORIGIN OF ADAPTATIONS, 1963.
The very idea of a nucleus with 1,260 chromosomes, all obeying the rules of precise alignment and division as cells proliferate, inspired G. Ledyard Stebbins, our greatest living evolutionary botanist, to a rare emotion for a scientific paper—rapture (since Ledyard and I share a passion for Gilbert and Sullivan, I will write, for his sake, “modified rapture”—and he will know the reference and meaning): “At meiosis, these chromosomes pair regularly to form about 630 bivalents, a feat which to cytologists is as remarkable a wonder of nature as are the fantastic elaborations of form exhibited by orchids, insectivorous plants, and many animals” (see Stebbins, 1966, in the bibliography).
In fifteen years of writing these monthly essays, I have specialized in trying to draw general messages from particulars. But this time, I am stumped. I don’t know what deep truth of nature emerges from the documentation of minimal and maximal chromosome numbers. Oh, I can cite some clichés and platitudes: Quantity is not quality; good things come in small packages. I can also state the obvious conclusion that inheritance and development do not depend primarily upon the number of distinct rods holding hereditary information—but this fact has been featured in textbooks of genetics for more than seventy years.
No, I think that every once in a while, we must simply let a fact stand by itself, for its own absolutely unvarnished fascination. Has your day not been brightened just a bit by learning that a plant can orchestrate the division of its cells by splitting 630 pairs of chromosomes with unerring accuracy—or that an ant, looking much like others, can gallivant about with an absolute minimum of one chromosome per cell? If so, I have earned my keep, and can go cultivate my garden. I think I’ll try growing some ferns. Then I might take some colchicine, which often induces polyploidy, and maybe, just maybe….
10 | Planets as Persons
Prologue
The Voyager expedition represents the greatest technological and intellectual triumph of our century. The fact that this tiny, relatively inexpensive machine could explore and photograph every outer planet except Pluto (but including Neptune, now the most distant planet, as Pluto temporarily moves closer to the sun on its eccentric orbit) is not only, as the cliché goes, a triumph of the human spirit (not to mention good old American tinkering and know-how), but also a living proof that billions of bucks, bureaucratic immuring, and hush-hush military spin-offs need not power our space program—and that knowledge and wonder really could be the main motivation and reward.
Such a triumph must be celebrated by any writer in natural history. I have chosen my own idiosyncratic mode. Voyager’s results convey many messages. These two essays, with their common theme, embody my reading of the main lesson from the standpoint of an evolutionary biologist: Planets are like organisms in that they have irreducible individuality and must therefore be explained by methods of historical analysis; they are not like molecules in a chemical equation. Planets therefore affirm the larger goal of unity among sciences by showing that methods of one approach (biological-historical) apply to cardinal objects of another mode often viewed as disparate or even opposed (physical-experimental).
34 | The Face of Miranda
WHEN MIRANDA, confined for all her conscious life on Prospero’s magic island, saw a group of men for the first time, she exclaimed, “O, wonder! How many goodly creatures are there here! How beauteous mankind is! O brave new world, that has such people in’t” (the source, of course, for Aldous Huxley’s more sardonic citation). Now, almost 400 years after Miranda spoke through Shakespeare, we have returned the favor, gazed for the first time upon Miranda and found her every bit as wonderful—“so perfect and so peerless…she will outstrip all praise and make it halt behind her.”
Prospero used all his magic to import his visitors by tempest. We have seen Miranda, the innermost large moon of Uranus, through the most stunning feat of technical precision in all our history. Ariel himself, Prospero’s agent of magic (and also another moon of Uranus), would have been astounded. For we have sent a small probe hurtling though space for nine years, boosting it with the gravitational slings of both Jupiter and Saturn toward distant Uranus, there to transmit a signal across 2 billion miles and three light-hours, showing the face of Miranda with the same clarity that Prospero beheld when he gazed upon his daughter’s beauty and exclaimed, “Thou didst smile, infused with a fortitude from heaven.”
It is easy to wax poetic about this feat (especially with a little help from the Bard himself). Voyager’s data from Jupiter, Saturn, and now Uranus have supplied more scientific return for expended output than anything else that space exploration ever dared or dreamed. In the chorus of praise, however, we have not always recognized how much this new information has transcended the visually dazzling—how deeply our ideas about the formation and history of the solar system have been changed. This confluence of aesthetics and intellect must be celebrated above all—and I should like to record my delight by thoroughly repudiating an early essay in this series (March 1977) as an illustration both of our new understanding and of the vital generalization so obtained.
My story is the tale of an old and eminently reasonable hypothesis, proposed long ago and beautifully affirmed by the first explorations of other worlds—our moon, then Mercury, and finally Mars. Then, at the height of its triumph, the theory begins to unravel, first at the moons of Jupiter, then at the surface of Venus, and finally and irretrievably, in the face of Miranda.
The initial hypothesis sought to explain the surfaces (and inferred histories) of rocky planets and moons as simple consequences of their differences in size. Why, in particular, is the earth so different from the moon? Our moon is a dead world, covered with impact craters that have not eroded away since their formation, often billions of years ago. The earth, by contrast, is a dynamic world of relative smoothness.
This difference, we assume, is a result of historical divergence, not initial disparity. Billions of years ago, when the planets were young and our portion of space still abounded with debris not yet swept up in planets and moons, the earth must have been as intensely cratered as the moon. The current difference must therefore be a result of the moon’s retention, and the earth’s obliteration, of their early histories. Why the difference?
On earth, both internal and external “machines” recycle the landscape on a scale of millions of years. The atmosphere (external machine) generates agents of erosion—running water, wind, and ice—that quickly obliterate the topography of any crater. Yet even without rain and wind, the earth’s internal activity of volcanism, earthquakes, and ultimately of plate tectonics itself would eventually disaggregate and erase any old topography. Surfaces do not last for billions of years on an active planet. But neither machine works on the moon. With no atmosphere, erosion proceeds (even in geological time) at a snail’s pace. Likewise, the moon is a rigid body with a crust 600 miles thick. Moonquakes do not fracture the lunar surface and volcanoes do not rise from the tiny molten core.
The earth’s activity and moon’s silence are consequences of a single factor—size. Large bodies have much lower ratios of surface to volume than small bodies of the same shape, since surfaces (length × length) grow so much more slowly than volumes (length×length×length) as size increases. Our planet powers its two machines by low surface-to-volume ratios. The earth generates heat (by radioactivity) over its relatively large volume and then loses this heat through its relatively small surface—thus remaining hot and active enough to propel plate tectonics. The moon, by contrast, and by virtue of its higher surface-to-volume ratio, l
ost most of its internal heat long ago, and solidified nearly throughout. Likewise, the earth’s large mass generates enough gravity to hold an atmosphere and power its external machine, while the moon lost any gases once produced.
As planetary exploration began, this “size-dependent” theory of planetary surfaces and their histories received its first tests and passed elegantly. The first photos of Mercury showed nothing but craters—as expected for a body about the same size as our moon.
Mars posed a clearer and more crucial test. As a planet about midway in size between the earth and moon, it should preserve some of its early topography, but also display the action of weak internal and external machines. The Surveyor flyby and Viking landings affirmed this prediction. The surface of Mars is about 50 percent cratered. The remaining areas show abundant signs of erosion, primarily by winds today (dune fields and etched boulders) and by running water in the past (now frozen), and internal churning more limited than on the earth. Most intriguing are signs of incipient (but unrealized) plate tectonics—as though the Martian crust remains pliant enough to fracture, but too rigid to move.
At this point in space exploration, I felt confident enough to write an essay extolling the size hypothesis as a sufficient and elegantly simple explanation of planetary surfaces and their histories. Contrasting the earth with the smaller bodies then known, I wrote (in March 1977) that “the difference arises from a disarmingly simple fact—size itself, and nothing else: the earth is a good deal larger than its neighbors.”
The first test after my essay appeared would be Voyager’s photographic survey of the Galilean satellites of Jupiter—the four moon-sized rocky bodies that, by the size hypothesis, would surely be intensely cratered worlds, cold and dead. Thus, I waited with confidence as Voyager approached Io, the innermost moon of Jupiter. The first photos, distant and fuzzy, revealed some circular structures initially read as craters. Well and good. But the next day brought sharp photos, and evoked both wonder and surprise. The circles were not craters, but giant volcanoes, spewing forth lakes of sulfur. In fact, not a single crater could be found on Io, the most active satellite in the solar system. Yet, as a body smaller than our moon, Io should have been cold and cratered.