Letters to a Young Scientist

by Edward O. Wilson

  There was no bravado in Corrie, no trace of overweening pride, no pretension. She was a quiet, serene enthusiast. As it turned out, she was also an open, helpful friend to fellow students and others around her. She’d come from New Orleans by way of San Francisco State University, and I took pride in her as a fellow southerner. I wanted her to succeed, and while I did not join as a collaborator, I found the funds to set up her laboratory. And why not? An effort like this celebrates imagination, hope, and audacity. And there was a fallback position for Corrie: if she fell short of the whole, she could use the part completed as a thesis. I even helped, a little, on the side. When I visited the Florida Keys on another project during the months that followed, I collected live ants of the genus Xenomyrmex for her, filling in a group difficult to obtain in the field. Along the way, she told me she needed to consult with an expert on some complex methods in statistical inference. I funded that also.

  At this point I was determined to see Corrie Saux to the end. I felt that she could actually accomplish what she envisioned.

  Her thesis was finished in 2007, read closely by her Ph.D. committee, and approved. On April 7, 2006, the core of her study was published as the cover article in Science, an achievement that would be considered exceptional even for a senior researcher. I admit I was nevertheless a bit tense when Corrie’s thesis went to the Harvard committee for review.

  Then I learned that the three-person team with the larger grant had also finished their work and planned to publish the results later in the year, allowing history to record that the two studies had been conducted independently and simultaneously. Of this I warmly approved, especially since each of the three was a highly regarded scientist. But it also meant that Corrie Saux’s research was about to be thoroughly tested. What if the two phylogenies didn’t match? That was a scenario I didn’t want to think about.

  To my great relief, however, the two phylogenies matched almost perfectly. There was a difference in the placement of one of the twenty-one subfamilies, the leptanilline ants, an obscure and little-known group. Even that variance in interpretation was later worked out through more data and statistical analysis.

  The story of Corrie Saux Moreau’s ambitious undertaking is one I feel especially important to bring to you. It suggests that courage in science born of self-confidence (without arrogance!), a willingness to take a risk but with resilience, a lack of fear of authority, a set of mind that prepares you to take a new direction if thwarted, are of great value—win or lose. One of my favorite maxims is from Floyd Patterson, the light heavyweight boxer who defeated heavier men to win and for a while hold the heavyweight championship. “You try the impossible to achieve the unusual.”

  Locations of the evolution of cichlid fish species in Africa. Modified from “Ecological opportunity and sexual selection together predict adaptive radiation,” by Catherine E. Wagner, Luke J. Harmon, and Ole Seehausen, Nature 487: 366–369 (2012). doi:10.1038/nature11144.

  Fourteen

  KNOW YOUR SUBJECT, THOROUGHLY

  TO MAKE DISCOVERIES in science, both small and important, you must be an expert on the topics addressed. To be an expert innovator requires commitment. Commitment to a subject implies sustained hard work.

  If you look beneath the surface of important discoveries to obtain a glimpse of the scientists who made them, you will soon see the truth of this generalization. Here, for example, is testimony from the theoretical physicist Steven Weinberg, who with Sheldon Lee Glashow and Abdus Salam won the 1979 Nobel Prize in physics for “contributions to the theory of the unified weak and electromagnetic interaction between elementary particles, including, interalia, the prediction of the weak neutral current”:

  I was born in New York City to Frederick and Eva Weinberg. My early inclination toward science received encouragement from my father, and by the time I was 15 or 16 my interests had focused on theoretical physics . . .

  After receiving my Ph.D. in 1957, I worked at Columbia and then from 1959 to 1966 at Berkeley. My research during this period was on a wide variety of topics—high energy behavior of Feynman graphs, second-class weak interaction currents, broken symmetries, scattering theory, muon physics, etc.—topics chosen in many cases because I was trying to teach myself some area of physics. My active interest in astrophysics dates from 1961–62; I wrote some papers on the cosmic population of neutrinos and then began to write a book, Gravitation and Cosmology, which was eventually completed in 1971. Late in 1965 I began my work on current algebra and the application to the strong interactions of the idea of spontaneous symmetry breaking.

  Obviously, Steven Weinberg did not just wake up one morning, reach for pencil and paper, and sketch out his breakthrough insights.

  Switching to a very different subject, X-ray crystallography, we have James D. Watson’s characterization of Max Perutz and Lawrence Bragg. It is in The Double Helix, arguably the best memoir ever written by a scientist, a book I recommend to any young person who wants to experience almost personally the thrill of scientific discovery. In it he describes what proved to be the essential step for solving the structure of the all-important coding molecule:

  Leading the unit to which Francis [Crick] belonged was Max Perutz, an Austrian-born chemist who came to England in 1936. [Perutz] had been collecting X-ray diffraction data from hemoglobin crystals for over ten years and was just beginning to get somewhere. Helping him was Sir Lawrence Bragg, the director of the Cavendish. For almost forty years Bragg, a Nobel Prize winner and one of the founders of crystallography, had been watching X-ray diffraction methods solve structures of ever-increasing difficulty. The more complex the molecule, the happier Bragg became when a new method allowed its elucidation. Thus in the immediate postwar years he was especially keen about the possibility of solving the structures of proteins, the most complicated of all molecules. Often, when administrative duties permitted, he visited Perutz’ office to discuss recently accumulated X-ray data. Then he would return home to see if he could interpret them.

  During nearly two decades, from 1985 to 2003, I brought to reality a dream that others before me considered inordinately difficult or even impossible. Fitted in between my classes at Harvard in the years before I retired, as well as other research and writing projects, I undertook the classification and natural history of the gigantic ant genus Pheidole. This is no ordinary group. It comprises by far the largest number of species of any ant genus, and further, it is among the largest genera of animals and plants of any kind. In many regions of the world, from desert to grassland to deep rain forest, it is also frequently the most abundant of all ants. What distinguishes Pheidole is the possession of two castes, slender minor workers and much larger big-headed soldiers. The possession of such variation within colonies adds to the biological complexity of these remarkable insects.

  So great was the species roster that the taxonomy of Pheidole when I started my revision was in a shambles. Most of the species recognized by earlier classifiers were unrecognizable from the brief descriptions given them. Worse, the collections of specimens accumulated over the previous century were scattered among half a dozen museums in the United States, Europe, and Latin America. By the time I picked up the task, Pheidole could no longer be ignored. Its many species are collectively among the major players in the environment. Ecologists trying to understand symbioses, energy flows, the turning of soil, and other basic phenomena were unable to name the species they were observing. Except for collection sites in North America, they were usually forced to report their specimens as belonging to “Pheidole species 1, Pheidole species 2, Pheidole species 3,” and so on to species 20 and beyond. This might work, at least roughly, for one researcher at one locality. But other biologists at other localities had their own independent rosters. Their Pheidole species 1, species 2, species 3, and so forth were by chance alone very likely different from the rosters of others, and the lists could be collated only if the researchers undertook the tedious task of bringing the specimens together. Better if from t
he start all writers used the same comprehensive list, comprising, for example, Pheidole angulifera, Pheidole dossena, Pheidole scalaris, and so on, each species having been defined earlier in a careful, formal manner and made universally convenient in the literature. When the taxonomy has been straightened out, biologists wishing to study the genus could identify the species to their single acceptable name. They could immediately collate their findings with those of other researchers, and pull from the literature everything previously known about every species of interest.

  Taxonomy is often spoken of as an old-fashioned discipline. Some of my friends in molecular biology used to call it stamp collecting. (Maybe some still do.) But it is emphatically not stamp collecting. Taxonomy, or systematics, as it is often called to spiff up its image, is fundamental to modern biology. In technology it is conducted with the aid of sophisticated field and laboratory research, using DNA sequencing, statistical analyses, and advanced information technologies. To take its place in basic biology, it is grounded in studies of phylogeny (the reconstruction of family trees) and in analyses of the genetics and geographical research devoted to the multiplication of species. The task of taxonomy drawing from these disciplines is made the more difficult, however, by the fact that most species of animals and microorganisms, together with a substantial minority of plants, await discovery.

  Ant taxonomists called the genus Pheidole the Mount Everest of ant taxonomy, towering arrogantly in front of us, seemingly too big to be mastered. There were many lesser but still important challenges on which others could build a productive career. I could face failure, I thought, so I took the job of ascending the ant Everest, at first in collaboration with my old mentor William L. Brown. When Bill’s health began to decline soon afterward, I soldiered on the rest of the way, starting with the Western Hemisphere, the biodiversity headquarters of the genus. I felt obligated to continue to the end, in part because I was located at the Museum of Comparative Zoology, with easy access to the largest collection and best library in the world suited to the task. But I also persisted partly for the challenge and partly because I thought of it as my duty. In the end, when Pheidole in the New World: A Dominant, Hyperdiverse Ant Genus was published in 2003, the book comprised 798 pages in which 624 species were diagnosed, 334 of them new to science, with everything known of the biology of every species cited, and all of the species illustrated, with a total of over 5,000 drawings I had made myself. Even as copies of Pheidole in the New World were being printed, new species continued to pour into the museum from collaborators in the field. It is likely that by the end of the century the total number of species will exceed 1,000, perhaps even 1,500, species.

  I planted our flag on the Pheidole summit, so to speak, but I am no Edmund Hillary or Tenzing Norgay. I had another goal in mind while encompassing the classification of the monster genus. One was to discover new phenomena in the course of giving thought to each species in turn. I was following the second of two strategies I gave you in an earlier letter: for each kind of organism there exists a problem for the solution of which the organism is ideally suited. One success in this correlative effort was the discovery of the “enemy specification” phenomenon. The principle behind its concept is simple. Every species of plant and animal is surrounded in its natural habitat by other species of plants and animals. Most are neutral in their effect upon it. A few are friendly, and at the extreme, there is the symbiotic level. In the latter case, two or more are dependent upon one another for their very survival or at least reproduction—for example, pollinator animals and the plants they pollinate. A few other plant and animal species are, on the other hand, inimical to a particular species, so much so in a few cases as to be dangerous to their survival. It is to the great advantage of individuals of that species to recognize dangerous enemies instinctively and to avoid or destroy them if possible.

  The principle sounds like common sense. But do species really evolve such an enemy specification response? I had never thought of it much one way or the other. Instead, I discovered it by accident. During the Pheidole project I cultured laboratory colonies of Pheidole dentata, an abundant species through the southern United States. I also kept colonies of fire ants (Solenopsis invicta). One day I was conducting one of my easy, quick experiments by placing other kinds of ants and insects next to the artificial nest entrances of the Pheidole dentata colonies just to see how they would respond. I was especially curious to see which ones would draw out the powerful big-headed soldiers.

  The response was usually tepid. Either the ants contacting the intruder retreated into the nest or, with a few other nestmates, engaged it in combat. But when I dropped just a single fire ant worker at the same spot, the reaction of the colony was explosive. The first forager to encounter the intruder rushed back into the nest, laying an odor trail as it ran, while frantically contacting one nestmate after the other. Both minor workers and soldiers then poured out of the nest, zigzagging and circling in a search for the fire ant worker. When they found it they attacked it viciously. The minor workers bit and pulled its legs, while the soldiers, employing their sharp mandibles and powerful adductor muscles that fill their swollen heads, simply chopped off the appendages of the fire ant to render it helpless.

  The fire ants are certainly enemies of the deadly kind. When, in the laboratory, I placed Pheidole and fire ant colonies close together, some of the fire ant scouts made it back home alive to report their find and recruit nestmates to the battle. The far larger fire ant colonies quickly destroyed and ate their opponents. Yet in some natural habitats, colonies of both species are abundant. It became apparent that Pheidole survive by building their nests a safe distance from the fire ant colonies and killing off fire ant scouts before they can report home.

  Later, in the Costa Rican rain forest, I found an even more remarkable response by another species (Pheidole cephalica) to rain or rising water that threatens to flood their nests. When I placed as little as a drop or two at the entrance of a nest, minor workers quickly mobilized the colony, and the whole emigrated within minutes to another location.

  Discoveries like these, whether minor or important—and who is to say at first which it will be?—can be made only rarely without a thorough advance knowledge of the organisms studied. This precondition is sometimes called “a feel for the organism.”

  Let me relate another story to reinforce this important principle. It occurred during an expedition I led in 2011 to the South Pacific. With me were Christian Rabeling, the ant expert and discoverer of the Amazonian “Martian” ant; Lloyd Davis, another ant expert and world-class birder; and Kathleen Horton, who was in charge of the complex logistics. We traveled during the austral spring of November and early December. Our destination was two archipelagoes, the independent island nation of Vanuatu and the nearby French possession of New Caledonia. In the process we visited localities where I had collected and studied ants in 1954 and 1955. I looked forward to observing changes in the environment that undoubtedly had occurred fifty-seven years later. I brought scanned images of my aging Kodachrome slides with me to make the comparisons exact. In particular I wanted to evaluate the condition of the wildlands and the reserves and national parks since 1955.

  What original discoveries we made, in particular with the ants we planned to collect and study, would depend entirely on the knowledge we brought with us. We were in fact well prepared. We discovered many new species, and kept notes on the habitats in which they were found. But that was only part of the plan. We had bigger game in mind: to clarify, if we could, phenomena in the formation of species and their spread from one island group to another across the intervening ocean gaps. If you look at a map of the South Pacific and make Vanuatu your focus, you see how plants and animals that colonized this archipelago could have come from any of three bodies of land: Australia and New Caledonia to the west, the Solomon Islands to the north, Fiji to the east, or some combination of all three. Ant colonists, although completely landbound, might have made the journey by floati
ng on the logs and branches of fallen trees or blown by storm winds. Queen ants capable of founding colonies might even have ridden in the feathers of far-ranging birds. We could not hope to determine how ants cross open water, but we did collect enough data to judge which island group contributed the most colonies to Vanuatu. It turned out, incidentally, to be the Solomon Islands.

  This discovery was important enough to justify the hard work in the field, but we devised another question to ask and perhaps answer. Leaving aside the Solomon Islands, whose ant fauna was still poorly explored, we were aware of a huge difference between Vanuatu and the two archipelagoes on either side of it, Fiji and New Caledonia. Both are ancient, having existed with a substantial land area for tens of millions of years. Vanuatu has been in existence for a comparable period of time, but only as a set of small, shifting islands. Only during the last million years has its land area been more than a tenth of what it is today. The antiquity of Fiji and New Caledonia is immediately apparent in the richness of their faunas and floras. In particular, each is occupied by a large number of species, some highly evolved, that occur nowhere else in the world.

  And what of relatively youthful Vanuatu? In November 2011 we were the first to take a close look at the ants on this archipelago. We knew that if it had a long geological history and large land area like New Caledonia and Fiji, we should expect to find a rich, highly evolved array of ants present. If, on the other hand, the current large area of Vanuatu had a relatively short history, as the geologists claimed, we should find a much sparser, distinctive array of ants there than occur on Fiji and New Caledonia. As it turned out, we found a smaller array, in accord with expectations from the record deduced by the geologists. But the ants of Vanuatu have not been inactive during their “brief” million-year tenure. We found clear evidence of new species in formation, and the beginning of the kind of expansion of biological diversity that is well advanced on the older archipelagoes. The ants of Vanuatu, to put the matter as succinctly as possible, are in the springtime of their evolution.