The Beak of the Finch: A Story of Evolution in Our Time

by Jonathan Weiner

  To get to know the fish and their diets, Dolph has had to learn to recognize on sight a whole new bestiary of animals and plants. The fish down at the bottom of the lake eat annelids, amphipods, gastropods, pelecypods, harpactacoid copepods, and chydorid cladocerans, among other tidbits. The fish that swim up and down the water column eat a whole different flora and fauna. There is almost no overlap in their diets.

  Dolph and the other stickleback watchers have measured the length and width of every one of these foods. The bits of fish food are so small that Dolph has to look at them through a dissecting microscope to measure length and width.

  With each and every pair of sticklebacks, the pairs are divergent, and every pair is divergent in the same way. One species in the pair always feeds on the bottom, the other always feeds in the water column, excluding the bottom. Moreover, in every lake, the benthics tend to be bigger and fatter, with bigger mouths and wider gapes but shorter and fewer gill rakers. The limnetics are smaller, thinner, with narrower gapes but longer and more numerous gill rakers. The benthics tend to eat the biggest prey, the limnetics the smallest prey.

  Five lakes, five pairs of species, and one pattern, repeated again and again. Meanwhile the sticklebacks that have a lake to themselves are intermediate in size, shape, gape, and gill rakers, and they range freely between the bottom and the water column, dining where they please, enjoying the best of both worlds.

  Even within the solitary species the pattern repeats. “In Cranby Lake, which is only a stone’s throw from Paxton Lake, there is only one species,” Dolph says. “But what’s amazing is when you look at individuals, there it is again.” The great majority of the fish that Dolph has netted in Cranby are specializing, taking at least 90 percent of their food either from the bottom or from the water column; their choice of specialty depends strongly on their build. The bigger ones with the shorter gill rakers are down at the bottom, and the smaller, slenderer fish with the longer, more numerous gill rakers are swimming up and around in the water column. So tiny variations in the dinnerware make a difference in the way the individual fish live their lives; the shape of the gill raker is as vital to them as the shape of the finch’s beak. The difference in the length of the gill rakers, the decisive difference that consigns a stickleback to one or the other of these two worlds within the lake, is only a third of a millimeter.

  “Variation is there already. It’s ever-present,” Dolph says. And as among Darwin’s finches these variations are passed down through the generations. Almost half the variation in gill-raker length is in the genes. “That’s very high,” he says. “In other words, there’s lots of genetic variability. The message being, these fish as we see them now are not stuck in a rut. If the environment changed they could respond. These populations are ready to respond should selection pressure arise.”

  So there is a simple trade-off here for a stickleback. If the fish specializes in the muck, it cannot compete in the open water; if it specializes in the open water, it is outclassed down in the muck. The fish is in much the same position as a finch in the Galápagos, where specializing in big seeds unfits you for the small ones, and specializing in the small seeds unfits you for the big ones.

  To Dolph all this evidence powerfully suggests that the colonists in these lakes have altered the course of each other’s evolution, just as the finches have altered each other’s courses in the Galápagos. They are a new case of character displacement, as neat in their way as the Galápagos finches, and Dolph found them right in his own backyard.

  The reason the pairs of species have divided their lakes the same way, over and over again, is apparently that natural selection creates the same intense pressures on the fish in every one of the lakes. In each lake, with almost no competitors around except the other sticklebacks, the quickest way to get away from the competition with the least sacrifice is to go up or down the water column. And that is what these pairs appear to have done, again and again, until they look almost as stratified as two layers of rock. But unlike layers of rock they are alive, and their impressive variability implies to Dolph that they are evolving now. The process of divergence is still going on today in those five lakes, for “natural selection continues to hold the species apart.”

  In principle, then, it should be possible to watch. If he is right, the adaptive landscape in the lakes of British Columbia has two big peaks with a valley between them, and where there are two species of sticklebacks in a lake, each takes one peak. Now Dolph wants to watch the sticklebacks evolve toward those peaks.

  “If there are two species of sticklebacks in a lake, and they really are in mutual repulsion, then we should be able to take one out and see if the other moves toward the middle,” he says. “These are rare species, and they are on Canadian threatened-species lists. So we don’t want to screw around with these lakes. But we can construct our own lakes, or ponds, and introduce the species, and see what happens.”

  This is Dolph’s answer to Connell’s “Ghost of Competition Past.” “Connell’s idea is, if you see two species that are different, you are always able to fabricate an argument that selection formed those differences,” Dolph says. “The idea in his mind is that what we see now is the consequence of selection that happened a long time ago.

  “But you look at stickleback populations, how young they are and how large their differences are, it’s quite clear that competition is no ghost. It affects populations today. And that’s what we hope to demonstrate with these experiments.

  “Connell’s idea about competition past is related to the more general idea that evolution took place a long time ago, that it is history. Whereas, in fact, natural selection is out there,” Dolph says, with that same mellow laughter in his voice. “One can actually see it happen within the time of a Ph.D. thesis. People never tried to look. They thought you would have to watch a population a thousand years. But that’s changing now.”

  The ponds are 70 by 80 feet, about 10 feet deep at their deepest. Dolph has built thirteen of them on campus. “If you go to a lake and you see two species and they are quite different,” he says, “the divergence argument implies that natural selection is maintaining those differences. Natural selection is acting all the time to maintain those differences. If you were to remove that pressure, they would relax toward the center. And this is what we intend to test.”

  While the Grants sort through their numbers in Princeton, Dolph is sorting through sticklebacks in Vancouver. He and the Grants still keep in touch, and he keeps them up-to-date about his evolutionary experiment. For now he is readying the ponds and breeding up batches of sticklebacks. “I can’t wait,” he says. “I’ve got fish coming out of my ears. The lab is full of them. They’re eating me out of house and home. I can’t wait to be rid of them.”

  In some of his ponds, Dolph will introduce limnetics; in other ponds, benthics. He predicts that in the solitude of its own pond, each species will evolve in the direction of the other. That is, the two species will converge over time, generation after generation, until their gill rakers are neither large nor small but medium-sized. Then all of the fish will be able to dine on the muddy bottom and also in the clearer waters above.

  At the same time, Dolph is collecting sticklebacks from different lakes, some limnetic, some benthic, some in between. In his laboratory he is crossbreeding them to create a hybrid swarm—a line of sticklebacks that is far more variable in size and shape than any line in the wild. “Because they are more variable,” he says, “they will allow more powerful and sensitive measurements of natural selection, as they are ‘sampling’ a broader reach of the adaptive landscape. I like to think of them as ‘selection probes’—instruments designed to measure selection.” He will stock a pond with these selection probes, add a line of benthic or limnetic competitors, and watch what happens.

  “I’m sure we’ll be able to measure selection pressures,” Schluter says. “We’ve done it in the Galápagos, we can do it here. It might be longer before we are able to detect an a
ctual evolutionary response. One year is one generation. I think we’ll be able to do it in ten, anyway!

  “Yeah, we’ve got our work cut out for us.”

  Chapter 13

  Fusion or Fission?

  The closer one looks at these performances of matter in living organisms, the more impressive the show becomes.

  —MAX DELBRUCK,

  A Physicist Looks at Biology

  Rosemary sits on her backless chair with her chin in hand, staring at the Macintosh. More martial columns of numbers parade down the screen. “It’s such a lot,” she says flatly, without taking her eyes from the screen. “And we’ve got to be so terribly careful. I mean, the amount of cross-checking, and double-checking, and triple-checking, to make absolutely sure there are no errors in the data …!”

  Her office is quiet. Cactus finches probe cactus flowers in the photographs on the walls. Guppies hover in an aquarium on the window-sill. They too are souvenirs of evolution in action, although no one is watching them now. They come from the famous fish tanks in John Endler’s laboratory. Rosemary’s daughter Nicola got them from some of the grad students. “She wanted them, but then I of course ended up with them,” says Rosemary.

  She and Peter have been climbing their mountain of numbers day and night, here in Eno Hall, and on a Macintosh in their house on Riverside Drive, a few minutes from the Princeton campus. They work together on the same desert island in the Galápagos and in the middle of civilization. She sometimes wonders if they could have done it when they were first married. But now they know how to work with each other, and they know how to work around each other. They interlock.

  This sabbatical they have carted much of their hybrid data across the Atlantic on visits to Uppsala, and to Arnside, her parents’ village in the Lake District, then back again to Princeton. They have combed through the hybrid data a dozen times and looked at the numbers from a dozen directions. Bit by bit, the heaps and mounds of data have grown tidier, shapelier, more manageable. On good days now it seems as if she has gotten up above the desert islet in a balloon, as if she and Peter have climbed a mountain above the mountain and they are looking down on all they have done and seen in the last twenty years.

  THEY CAN SEE in their numbers and computer-generated charts that since the flood, the crazy Niño of 1982–83, the adaptive landscape of the islands has changed dramatically. Tribulus has gone down, down, down. It was already in trouble before the first rains of El Niño. Rosemary and Peter suspect a fungus: some kind of rust that ate at its roots. Then of course the floodwaters drowned it, and green vines overwhelmed it, and Cacabus plants sprang up as if out of nowhere—great, sticky, hairy-leaved mats of Cacabus—and smothered it. After that came the droughts.

  Cactus on the island is down too. First the cactus trees took up too much water in the great flood, then too little in the droughts. The cactus trees toppled under tangles of vines and more Cacabus. By 1990 there were hardly any Tribulus or cactus seeds to be found anywhere on the island, even for finches flipping hours of pebbles with their beaks. The cactus may be just beginning to come back now.

  A pattern is emerging: a shift that the Grants had not seen clearly until they did the analysis. There have been fewer big, hard seeds on Daphne since the flood. But there have been more small, soft seeds—mostly from all that Cacabus. Rosemary and Peter have put these changes into hard numbers, and the changes are significant. In fact, the changes are enormous. For Darwin’s ground finches, life is seeds. If the pile of big seeds shoots down and the pile of little seeds shoots up, that is an upheaval, a catastrophe in the adaptive landscape, the fall and rise of alps. This is just what has happened on Daphne in the years since the flood. One adaptive peak has collapsed, and another peak has gone beetling skyward.

  These changes have been especially hard on the cactus finches. Cactus is their only home in the adaptive landscape (and in the ordinary landscape). If cactus falls, they fall too. The Grants have plotted the numbers of cactus finches on the island in relation to the numbers of cactus trees, fruits, and seeds. They can see that during each of the droughts since the flood, the cactus finches’ population on Daphne declined, as expected. By the start of this year, when Rosemary caught those two rogue finches on the north rim, there were only about one hundred cactus finches on Daphne, which is the lowest their numbers have dropped since the finch watch began.

  In spite of all this selection pressure, the cactus finches have not changed in the last ten years. By all the Grants’ measures, their beaks and bodies are the same now, on average, as they were before the flood. This too makes sense in terms of the adaptive landscape, because in evolutionary terms these birds have nowhere to go. “Flee as a bird to your mountain,” sings the psalmist. Cactus is this bird’s mountain. When this peak falls down, they have no peak nearby to which to flee. They are trapped on one falling alp.

  Since the big Niño, selection pressure has been strong on fortis too. Among the fortis that saw the flood, not quite one in three were still alive by 1987. But the Grants’ tables show that the fortis did not die at random. The survivors in 1987 were eating a much greater proportion of small seeds than large seeds. This was partly a change in behavior, since individual fortis are flexible in their choice of foods. However, they are flexible only up to a point. The Grants can see from their data base that it was the fortis individuals with significantly deeper, wider beaks, the birds that were committed by their anatomy to the pile of big seeds, the eroding peak, who were doing most of the dying. The fortis with significantly shallower and narrower beaks were doing most of the surviving. So the average beak of the fortis generation that was born after the flood—the baby-boom generation—was better adapted to the brave new landscape of the 1980s.

  In other words, while the cactus finches have gone down with their peak, the fortis have evolved. They have rolled with the adaptive landscape. The width of the fortis beak in the new generation, a generation of finches that is hopping around on the lava of Daphne Major at this moment, is measurably narrower than the beaks of the generation before them—down from 8.86 millimeters at the time of the flood to 8.74 millimeters now.

  That does not take fortis back where they were at the start of the Grants’ watch, but nearly so. The birds shifted toward large size in the first years of the study, and now they have shifted most of the way back. It is as if the whole island has dodged back and forth beneath its load of perching birds, as lost Spanish sailors once believed the whole archipelago could do, which is why they spoke of Las Encantadas, the Enchanted Islands. The adaptive peaks have slid to the east and slid to the west, and the birds have kept flying after them and perching on them again and again. The finches have to stay with their peaks because, as Peter has written, “valleys are steep, that is to say, the intensity of selection is great.” The finches have done a lot of flying to stay on their island. Fortis has done a lot of evolving just to stay in place.

  In these same years the Grants have seen a second oscillation, a change in the fate of the hybrids on the island. The mixed bloods were selected against in the first half of the watch, and they were selected for in the second half. Up until the flood, a male fortis that crossed with a female fuliginosa (a small beak) or a scandens (a cactus finch) was putting his young at a disadvantage. The hybrids did not prosper. Selection pressure was against intermarriage. But since the flood, selection has reversed. Now a cross with a fuliginosa or a scandens does the genes of fortis a favor.

  By putting the two oscillations together, the Grants can begin to understand what is going on with the hybrids. Their misfit data are beginning to fit.

  These two oscillations are driven by the same events. They are both governed by the same changes in the adaptive landscape. In an adaptive landscape that is wrinkling and rolling as fast as Daphne, a landscape in which the peaks are in geological upheaval, it can pay to be born different, to carry a beak 3, 4, or 5 millimeters away from the tried and true. Since the super-Niño, some of the old peaks have turned
into valleys, and some of the old valleys are peaks. Now a hybrid has a chance of coming down on the summit of a new peak. It can luck onto a piece of the new shifting ground.

  In this changing landscape the hybrids may have advantages not only because they are so variable in the dimensions the Grants are measuring. It is also possible that the influx of new genes that is the birthright of the hybrids could translate into a thousand subtle advantages too small for the Grants to measure: benefits that add up to greater physical vigor, even if the bird stays on the same adaptive peak as everybody else. “A hybrid could do all the things others do on an island,” Peter muses, “and just be a better piece of machinery generally.”

  Thoughts like these send Peter back to a paper that two evolutionists, Richard Lewontin and L. C. Birch, published in 1966, “Hybridization as a Source of Variation for Adaptation to New Environments.”

  We generally think of the adaptive landscape as being more or less fixed and constant, just as we think of the bodies and behaviors of animals as more or less constant. But what happens if the adaptive landscape changes dramatically? What happens, for instance, when a species leaves home and wanders into new territory? Lewontin and Birch suggested in their paper that the genetic changes that accompany the change in range must be “profound,” and “if a case could be found of a species rapidly expanding its ecological range, caught in flagrante delicto, it might be possible to study the genetic basis of such a change.”

  Lewontin and Birch found their case in a fruit fly, Dacus tryoni, close cousin of the notorious Mediterranean fruit fly, or medfly. Dacus tryoni once lived solely on fruits in Australian tropical rain forests. That began to change in the 1850s, while Darwin was writing the book that became the Origin of Species. In those same years, farmers “down under” began planting orchards in Queensland. The flies moved out of the rain forest and became a pest in the new apple, pear, and guava orchards. Within a hundred years the flies had expanded their range as far south as Victoria, with sporadic outbreaks in Adelaide, Melbourne, and Gippsland, and in the capital of Australia’s Northern Territory, on the Timor Sea, the port called Darwin.