Letters to a Young Scientist (18 page)
I was driven by two intense desires. I wanted to go on working on islands, whatever the excuse. And I wanted to do something radically new in biogeography. I reasoned that I might accomplish both if I chose an ecosystem small enough to be manipulated.
A solution then presented itself. Insects—my specialty—are almost microscopic in size compared to the mammals, birds, and other vertebrates that had been featured in earlier biogeographic studies. They weigh a few milligrams or less, where vertebrates are measured in grams or more. There are large numbers of tiny islands on which insects can live and breed for generations. Instead of just one or several islands the size of Cuba, Barbados, or Dominica, where birds and mammals can be studied, there are hundreds of thousands of islands around the world with an area of a hectare or less. Somehow, I thought, the insect, spider, and other invertebrate faunas of a few could be altered so that the rates of immigration and extinction onto them could be measured. From these data multiple tests could be devised to test hypotheses, to evaluate theory itself, and to discover new phenomena.
A new world opened in my imagination. I saw the islets of the world as the perfect model ecosystem. Now I sought a laboratory. It had to be a cluster of little islands, variously big and small, close and distant. Where might such an ideal micro-archipelago be? I scanned detailed maps of the eastern Atlantic and southern Gulf Coast of the United States, from the rocky prominences of Maine and the Harbor Islands of Boston to the barrier islands of the Carolinas, Georgia, Florida, and the Gulf states to the west. All could be reached in a day’s travel from Harvard University. It did not take long to settle on the multitudinous tropical islands of the Florida Keys and Florida Bay.
To conduct experiments that would yield what scientists like to call “robust” conclusions, I needed to have my islets start from zero—empty, harboring no insects at all. My attention fixed on the small, wave-washed islands of the Dry Tortugas, the outermost cluster of the Florida Keys. Except for Fort Jefferson at the very end, they are almost all desert islands, harboring only small patches of vegetation and relatively few species of insects and other invertebrates in residence. There was an advantage to their simplicity: whenever a hurricane crosses over them, they are swept clean of terrestrial life.
In 1965 I took a team of graduate students with me to the Dry Tortugas to look over the situation. We mapped every plant on several of the islands and recorded every insect and other invertebrate species we could find. During the next hurricane season, in 1966, not one but two hurricanes crossed the Dry Tortugas. We returned soon thereafter, and sure enough, the small islands were bare of plants and terrestrial animals.
It seemed that the main problem had been solved, but by this time I had begun to have doubts about using the Dry Tortugas. I believed that in order to conduct a high-quality experiment of lasting value, the kind that others could replicate conveniently, I needed a better laboratory. I wanted more islands than those that make up the Dry Tortugas. I needed to conduct the removal of the species myself, and not rely on random weather. It would also be best to use controls—islands closely identical to the experimental set, and treated the same but without removing the animals. Finally, I needed more biology. The faunas of the Dry Tortugas are so small and the life spans of the ecosystems so short as to reduce their faunas and floras to random number generators. I needed larger faunas more typical of natural ecosystems, and I needed less disturbed islands.
Before telling you how the goal was accomplished, I will pause to reinforce a point I made earlier: that successful research doesn’t depend on mathematical skill, or even the deep understanding of theory. It depends to a large degree on choosing an important problem and finding a way to solve it, even if imperfectly at first. Very often ambition and entrepreneurial drive, in combination, beat brilliance.
I was determined to solve this problem of biogeography, and was excited by the challenge of developing a new technology doing it. I found what I needed in the small mangrove islands of Florida Bay, just to the north of the Dry Tortugas. There are a lot of them: consider the implication of the archipelago at the northern end of the bay called the Ten Thousand Islands. The damage done to the entire Florida Bay mangrove system by removing the invertebrates from a dozen or so would be negligible, and soon repaired.
At this point I enlisted the collaboration of Daniel S. Simberloff, one of my graduate students with a strong background in mathematics. I quickly realized I had chosen a partner wisely. As with MacArthur’s work, Simberloff’s mathematics fitted my own natural history nicely. From this point forward, while facing the unknown together, we became more colleagues than teacher and student. Together, step by step, we worked out the method of removing all of the invertebrate animals from the mangrove islets without damaging the trees and other vegetation. Without detailing our failures and false starts to you here, we devised the simple and straightforward method of eradication: hire a pest control company to erect a tent over each island and fumigate it. That was not as easy as it sounds. Working as a team, we had to invent the right framework to be erected in shallow water and find the right kind and dosage of dispersible insecticide to use. We had to walk through gluelike muck, and convince the workers helping us that the ground sharks swimming in close to the islands at high tide were harmless.
Not least, Simberloff and I also had to create a network of experts on the various groups of invertebrates—beetles, flies, moths, barklice, spiders, centipedes, and so on—in order to identify the species correctly.
After two years of monitoring the immigrations and extinctions that followed, and to my great relief (and Simberloff’s also—he had to get a Ph.D. thesis out of his part of the work), the recolonization fitted the equilibrium model. We also learned a great deal about the colonization process itself. I found the whole of the adventure, from theory to experiment, one of the most satisfying experiences of my entire scientific life.
I hope that in your own career you will see one or more opportunities of this kind and, like Daniel Simberloff and myself, find the risk worth taking. We stood on the shoulders of giants and were able to see a little bit farther.
The U.S. National Medal of Science.
I
HAVE COME TO
the end of my counsel to you, and will now close these letters with advice on proper behavior in the conduct of your research and publication.
You are not likely to be directly pressed during your career on such largely philosophical questions as the propriety of creating artificial organisms or conducting surgical experiments on chimpanzees. Instead, by far the greatest proportion of moral decisions you will be required to make is in your relationships with other scientists. Entrepreneurial endeavor beyond the level of puttering creates difficulties other than the mere risk of failure. It will put you into a competitive arena for which you may not be emotionally prepared. You may find yourself in a race with others who have chosen the same track. You will worry that someone better equipped and financed will reach the goal before you. When multiple investigators create an important new field simultaneously, they often create a golden period of excited cooperation, but in later stages, as different groups follow up on the same discoveries, some amount of rivalry and jealousy is inevitable. For you, if successful, there will be gentle competitors and ruthless competitors. There will be gossip and some protective secrecy. That should come as no surprise. Business entrepreneurs suffer when competitors beat them to the marketplace. Should we expect scientists to be different?
Original discoveries, to remind you, are what count the most. Let me put that more strongly: they are
all
that counts. They are the silver and gold of science. Proper credit for them is therefore not only a moral imperative, but vital for the free exchange of information and amity within the scientific community as a whole. Researchers rightly demand recognition for all their original work, if not from the general public then from colleagues in their chosen field. I have never met another scientist who was not pleased—deeply pleased—by a promotion or award bestowed for original research. As the actor Jimmy Cagney said of his career in motion pictures, “You’re only as good as people say you are.”
The great scientist who works for himself in a hidden laboratory does not exist. Therefore, be rigorous in reading and citing literature. Bestow credit where it is deserved, and expect the same from others. Honest credit carefully given matters enormously. Recommending a colleague for awards or other forms of recognition is a relatively uncommon form of altruism when practiced among scientists. Even if it proves difficult, do not shrink from taking that step. On the other hand, granting it to a rival, especially one you do not like and at the risk of your own recognition, would be true nobility. It is not expected of you. Let others make the nomination. Instead just grit your teeth and extend your congratulations.
You will make mistakes. Try not to make big ones. Whatever the case, admit them and move on. A simple error in reporting or conclusion will be forgiven if publicly corrected. (At least one leading journal has a special erratum section.) An outright retraction of a result will not cause permanent harm if done graciously, and especially with thanks to the scientist who reported the error with evidence and logical reasoning. But never, ever will fraud be forgiven. The penalty is professional death: exile, never again to be trusted.
If you’re not sure of a result, repeat the work. If you don’t have the time or resources to do so, drop the whole thing or pass it on to someone else. If your facts are solid, but you’re not sure of the conclusion, just say as much. If you do not have the opportunity or resources to repeat and confirm your work, don’t be afraid to use words denoting timid uncertainty: “apparently,” “seemingly,” “suggests,” “could possibly be,” “raises possibility of,” “may well be.” If the result is worthwhile, others will either confirm or disprove what you think you found, and all will share credit. That’s not sloppiness. It’s just good professional conduct, true to the core of the scientific method.
Finally, remember that you enter a career in science above all in the pursuit of truth. Your legacy will be the increase and wise use of new, verifiable knowledge, of information that can be tested and integrated into the remainder of science. Such knowledge can never be harmful by itself, but as history has so relentlessly demonstrated, the way it is twisted can be harmful, and if such knowledge is applied by ideologues, it can be deadly. Be an activist as you deem necessary—and you can be highly effective with what you know—but never betray the trust that membership in the scientific enterprise has conferred upon you.
