The Triumph of Seeds

Oh rats, rejoice!

The world is grown to one vast drysaltery!

So munch on, crunch on, take your nuncheon,

Breakfast, supper, dinner, luncheon!

—Robert Browning,

The Pied Piper of Hamelin (1842)

Appendix F of the International Building Code stipulates requirements for keeping rats and other rodents out of all habitable dwellings. These include two-inch (five-centimeter) slab foundations, steel kick-plates, and tempered wire or sheet-metal grating over any ground-level opening. Conditions for grain storage or industrial facilities can be even stricter, involving thicker concrete, more metal, and curtain walls buried two feet below grade. In spite of all this, rats and their relations still consume or contaminate between 5 and 25 percent of the world’s grain harvest, and regularly gnaw their way into important structures of all kinds. In 2013, a trespassing rodent shorted out the switchboard at Japan’s ill-fated Fukushima nuclear plant, sending temperatures in three cooling tanks soaring and nearly setting off a repeat of the 2011 meltdown. The story made headlines around the world, with journalists, bloggers, and TV commentators all wondering what makes rats so interested in electrical wires. But the real question isn’t about what rodents like to eat; it’s about how difficult it is to stop them. Why on earth should a rat be able to chew through concrete walls in the first place?

The name “rodent” comes from the Latin verb rodere, “to gnaw,” a reference both to the way rodents chew and to the massive incisors that help them do it so well. These teeth evolved in small mouse- or squirrel-like creatures approximately 60 million years ago. That’s approximately 60 million years before the invention of concrete, Plexiglas, sheet metal, or any of the other manmade materials that rats and mice now chew through. Experts still argue about the exact origin of rodents, but there is little doubt about what those big teeth were good for. While the family tree now includes oddballs like beavers, who chew wood, and naked mole rats, who use their teeth for digging, the vast majority of rodents still make much of their living the old-fashioned way: by gnawing seeds.

Before rodents came along, the ancestors of trees like oaks, chestnuts, and walnuts got by with little winged pips that offered scant protection from chewing. Fossils of these seeds look like lumpy flecks of chaff, insubstantial wisps designed to flutter a bit as they fell. Once the gnawing began, however, these plants and their rodent predators entered a virtual arms race: stronger teeth led to harder seed coats and vice versa, changing those ancient seeds into the acorns and thick-shelled nuts we’re familiar with today. (Other seeds responded by getting even smaller, in the hopes of being swallowed whole, or ignored altogether.) For the trees, rodents posed an evolutionary dilemma: the chance to get their seeds dispersed balanced against the risk of losing them entirely. For rodents, unlocking the nutrition in seeds turned out to be an evolutionary gold mine: they quickly became the most numerous and diverse group of mammals on the planet.

The notion of coevolution implies that change in one organism can lead to change in another—if antelope start running faster, then cheetahs must run faster still to catch them. Traditional definitions describe the process as a tango between familiar partners, where each step is met by an equal and elegant counter-step. In reality, the dance floor of evolution is usually a lot more crowded. Relationships like those between rodents and seeds develop in the midst of something more like a square dance, with couples constantly switching partners in a whir of spins, promenades, and do-si-dos. The end result may appear like quid pro quo, but chances are a lot of other dancers influenced the outcome—leading, following, and stepping on toes along the way. No one knows the exact sequence of events that gave us strong-jawed rodents and thick-shelled seeds; the story played out long ago and left only general clues in the fossil record. But few experts believe their sudden and simultaneous rise was mere coincidence.

In many cases, the relationships that developed became mutually beneficial—the gnawers got something to eat and dispersed a few of the plant’s seeds in the process. Hunger alone drives the rodent side of this equation, but for plants it’s like walking a tightrope. Their seeds must be attractive enough to be desired, but tough enough so that they can’t be devoured on the spot. A hard shell forces rodents to carry seeds away and gnaw them open later, in the safety of a burrow. Ideally, the rodents then forget where they’ve hidden things, or perish before they get around to eating them. Take the example of Beatrix Potter’s book The Tale of Squirrel Nutkin. Scholars think she wrote it as a commentary on Britain’s class system, but it’s also a story about seeds: if the squirrels on Owl Island gather and stash away nuts, and if Old Brown the owl attacks the occasional squirrel, then some of those nuts will go uneaten and the next generation of oaks and hazels will live on. (Nutkin managed to escape with only the loss of his tail, but we must assume that Old Brown is more successful on other attempts.)

Potter set her story in England’s Lake District, but if she had lived in Central America she would have put it right where I did my doctoral research, under the spreading boughs of an almendro tree. There, little Nutkin would have found not only plenty of squirrels to keep him company, but also other rodents: pocket mice, rice rats, climbing rats, and spiny rats, as well as pacas and agoutis, which look more or less like guinea pigs the size of small dogs. Like me, all of these species came to almendro trees in search of seeds. Unlike me, the rodents had been at it for thousands, if not millions, of years. (A dissertation only feels like it takes that long.) With so many gnawing creatures hanging around, it’s no wonder almendro developed a shell hard enough to challenge a graduate student. But the nuances of seed defense rarely stop at physical protection alone. The ecology of this one rainforest tree makes it clear why so many seeds are stony, and why it takes a lot more than concrete to stop a hungry rat.

FIGURE 8.1. The busy squirrels from The Tale of Squirrel Nutkin (1903), Beatrix Potter’s classic story about gathering (and dispersing) the acorns and hazelnuts of Owl Island.

An almendro seed measures two inches (five centimeters) long and slightly more than an inch (two and one-half centimeters) wide, with smooth sides and tapered ends that give it the appearance of a giant throat lozenge. Like the pit of a peach or plum, this seed includes an extra layer of stony shell, with the soft nut tucked safely inside. The surrounding flesh of the fruit is thin and brownish-green, but sweet enough to attract a wide array of monkeys, birds, and bats. At the height of the season, dozens of species gather around almendros, foraging in the canopy and feasting on the bounty that drops to the ground below. But among all these fruit-eaters, only one large bat carries its meal away from the tree. So if an almendro wants its young dispersed, it must also concentrate on the creatures that eat its seeds. And while it may be hard to think of trees as intelligent (at least outside of J. R. R. Tolkien stories), the system almendros have developed seems careful, calculating, and nearly perfect.

From a plant’s perspective, not all potential dispersers are created equal. When I collected almendro seeds, for example, I carted off large quantities and traveled great distances, but then systematically destroyed every one of them for my research. Even if I’d planned on sprouting the seeds, my laboratory was at a university in northern Idaho, hardly the right habitat for rainforest trees. At the other end of the spectrum, smaller rodents, like rice rats and pocket mice, lack the strength to move almendro seeds more than a foot or two. Invite them to the feast and the tree’s progeny would die without ever leaving home. Excluding small, ineffective seed predators and limiting the damage from large ones requires a shell with just the right level of defenses, one that optimizes what ecologists call handling time.

For almendro, the ideal shell turns out to be a woody husk that measures over a quarter of an inch (seven millimeters) thick at its widest point, twice the heft of a plum pit or a peach stone. The walls include additional protections: a layer of resinous crystals, much like the ground glass that exterminators add to concrete when they want to plug a rat hole. But in this case the seed isn’t trying to prevent gnawing entirely, just slow it down. For the average squirrel, chewing through the crystal-filled husk of an almendro takes at least eight minutes, and sometimes as much as half an hour. That’s a huge time investment for an animal that needs to locate and eat between 10 and 25 percent of its bodyweight every day just to survive. An almendro seed is worth the effort, but just barely. Spiny rats and smaller rodents rarely bother—not necessarily because they can’t, but because it’s not worth their while. The challenge and time involved would exhaust them to a degree that not even the reward of a large nut could repay. In this context, the strength and thickness of almendro shells seem perfectly adapted to reserving those nuts for squirrels, agoutis, and pacas—the large rodents most capable of carrying them away. Making them actually do so, however, lies beyond the control of the tree. That incentive must come from other players in the dance.

FIGURE 8.2. Almendro (Dipteryx panamensis). Seeds of the mighty almendro tree lie within one of the toughest shells in nature, a defense against the gnawing teeth of rodents. The shell is pictured at the top, partially cut away in cross section. An extracted seed is shown on the left, with a whole fruit on the right. ILLUSTRATION © 2014 BY SUZANNE OLIVE.

Once I perfected my mallet-and-chisel technique, I learned to cleave an almendro seed and neatly extract the nutmeat in less than a minute. That put me well ahead of squirrels, but it wouldn’t have seemed nearly fast enough if I’d been opening seeds in a dangerous setting—a crocodile pit, for example, or a pen full of hungry wolves. That’s the dilemma faced by rodents. Because as surely as an almendro tree attracts seed-eaters, it attracts the eaters of seed-eaters. I knew from experience that fer-de-lance hang around almendro trees, and so do other rodent-loving snakes like bushmasters and boa constrictors. I once watched a Semiplumbeous Hawk carry off something small and furred in broad daylight. If I’d stuck around until dark, I might have seen half a dozen different owls, as well as ocelots, margays, and jaguarundis, all of them attracted by the concentration of tasty prey. A friend of mine studying mammal communities once showed me a stack of photographs taken by remote cameras scattered throughout the forest—flash snapshots of surprised-looking jaguars, pumas, big weasels, and others. There were even a few hunters and their dogs. He asked me if anything seemed familiar, and then I saw it: time and again the backdrop included an almendro trunk, and the ground was littered with seeds. In the rainforests of Central America, the community of animals drawn to a good almendro crop doesn’t stop with fruit-eaters, seed-eaters, and predators. It includes all of the people who seek them: scientists, hunters, birdwatchers, and anyone else looking for a piece of the action.

In the face of all this commotion, much of it fanged and hungry, squirrels and other rodents usually treat almendro trees like a drive-through. They pick up a meal and carry it away before stopping to eat—forty feet (twelve meters), fifty feet (fifteen meters), and sometimes much farther. Agoutis, in particular, turn out to be vital dispersers. They not only move seeds a long way, they bury them for safekeeping in tidy little holes throughout their home range, a habit with the pleasing name of scatter-hoarding. From the tree’s perspective, this fits the bill nicely: a creature that moves seeds and plants them, and then stands a good chance of being killed off by one of the many predators lurking nearby.

This pattern repeats itself with different rodent and plant species all around the world, providing ample evolutionary incentive for nut-like seeds to develop their thick, hard shells. In fact, any trait that adds to handling time can be an advantage, which is probably why walnuts have that brainy, convoluted shape that’s so irritatingly difficult to remove in one piece. Rodents, too, have responded with more than just strong teeth, developing bulging, high-capacity cheek pouches to carry off numerous seeds at once, as well as an uncanny ability to sniff out and discard diseased or worm-infested nuts before bothering to gnaw them. Like so many evolutionary stories, the impact of rodents on seed defense is more than a bilateral arms race. It involves a whole suite of relationships and species, with give-and-take on all sides. For the almendro, my research showed not only how elaborate that system can be, but how quickly it can fall apart.

In a healthy rainforest, crossing the muddy ground near large almendro trees feels like walking on lumpy gravel—gnawed, split, and otherwise discarded shells carpet the ground underfoot. I counted them by the thousands and rarely found an intact seed, let alone a young tree. With such a concentration of gnawers around, the only seeds that sprouted and grew into saplings were those that got themselves dispersed far away. In patchy forest, however, hunting and other disturbances took a heavy toll on large rodents, and there was hardly any sign of gnawing or scatter-hoarding. Seeds simply germinated where they fell, leaving every adult tree ringed by a thicket of its own children. In the short term, this scenario meant bad news for the next generation—baby trees don’t fare well in the shade of their parents. From an evolutionary standpoint, it put almendro in a bind—removing its partners in the dance left it with a seed too hard for anyone left in the forest to chew.

Studying almendro taught me that plants defend their seeds with a complex calculus where protection is only one of the variables. But it also left an obvious question unanswered: Just how hard is an almendro shell? Harder than concrete? I found the answer while writing this chapter.

Though I finished my dissertation years ago, a person doesn’t invest that kind of time in a project without taking home a few souvenirs. Dried now to a rough honey-brown, the almendro shell I keep on my desk still bears the telltale grooves of rodent bites at one end. To test it against concrete, I simply stepped outside of my office and crawled under the porch. The Raccoon Shack rests on a foundation of concrete pier blocks, the standard variety with a built-in bracket available at any home supply store. I placed my almendro shell edgewise against a pier block—like a chisel—and gave the thing a good solid whack with a hammer. It didn’t surprise me in the least to see cracks form in the concrete: if rodents evolved to gnaw, and almendro responded with one of nature’s hardest seeds, then almendro shells should be very nearly as tough as rat teeth. A few more whacks produced a sizable concrete chip that flew loose and landed in the soil below. I reached down to pick it up, careful to avoid the less savory items lying around under the porch: droppings and tattered feathers from our chickens, and half a dozen empty rat traps. One look at those traps annoyed me, and I made a mental note to return that evening and re-bait them with nut butter.

“People won’t believe you,” Eliza warned, smiling when I told her what was happening under the Raccoon Shack. But as Oscar Wilde once observed, “life imitates art far more than art imitates life.” And the fact remains that while I was sitting at my desk, writing about rodent teeth and seeds, a version of that same drama was playing out directly beneath my feet. Attracted by the grain in our nearby chicken coop, an extended family of Norway rats had moved into the crawlspace under the Raccoon Shack. They gained entry by chewing a neat hole through a sheet of 23-gauge galvanized steel hardware cloth. Once inside, the rats had a cozy home base from which to stage raids on any nearby edibles. They soon discovered my pea bed, where I’d stupidly left the entire harvest for my Mendel experiment drying on the vine. By the time I figured out what was going on, the rats had decimated my Bill Jump Soup Peas and put a sizable dent in the Württembergische Wintererbse. The sad remnants that Noah and I picked through totaled a scant three cups, but luckily they included enough successful crosses to continue the experiment for another (better-protected) season.

Losing my peas to the rats turned out to be a valuable lesson—as Mr. Wilde also noted, “experience is the name we give our mistakes.” First of all, I gained another insight into the methods of that cagey old monk. Unless the Monastery of St. Thomas maintained an army of cats, Mendel must have built a safe place to dry his harvest. It wouldn’t surprise me if his lost journals and papers contained detailed plans for a rat-proof granary. More importantly, I learned that even in the artificial setting of my pea bed, with a domestic vegetable and a nonnative rodent, the same rules apply. When the rats sniffed out my vines, they did their job perfectly, using the precise logic found in any rodent/seed interaction. Bill Jump peas mature slowly and hadn’t fully dried yet, making them comparatively easy to chew. Those were eaten on the spot. When I tried biting into a winter pea, however, I nearly broke a molar. They required more handling time, and, according to theory, should have been hauled away for safe gnawing. Sure enough, when I opened up the crawlspace under the building, I found a huge pile of empty pods and winter-pea seed coats. (The opposite of a scatter-hoarder, the Norway rat stores all its seeds in one place and is known in biology circles as a “larder-hoarder.”)

In the weeks I spent baiting traps below the Raccoon Shack, I found myself wishing that rats had never evolved. But even in a world without rodents, something probably would have been after my peas. Once mother plants began packing lunches for their babies, everything from dinosaurs to fungi wanted a taste, and the evolution of seed defenses became inevitable. The relationships sometimes find a balance, but not always. Almendro trees appear to have the rodent situation figured out, but they must not have planned for peccaries, aggressive wild pigs whose massive molars can split and crush the seeds with ease. Even worse, the Great Green Macaw specializes on almendro, nesting in the trees and gorging on the seeds, which it can handily crack with a bill specifically adapted to the purpose. Among seed predators, birds have one of the longest evolutionary histories. They descend from dinosaurs, some of whom developed seed-crushing organs over 160 million years ago. Paleontologists know this from fossils containing telltale clusters of gastroliths, the distinctive little stones found inside gizzards. Modern birds still depend on grit to grind their food, and the strongest gizzards occur in seed-eaters—everything from chickens to canaries, grosbeaks, jays, and perhaps the most famous group of birds in the world.

FIGURE 8.4. This classic illustration by John Gould shows some of the diversity of beak shapes in Darwin’s Galapagos finches. Charles Darwin, Journal of the Beagle (1839). WIKIMEDIA COMMONS.

For Charles Darwin, the finches of the Galapagos Islands appeared like a gaggle of unrelated species, more notable for their tameness than anything else. As he recorded in the field, “Little birds . . . will alight on your person & drink water out of a basin held in your hand.” It wasn’t until his specimens reached ornithologist John Gould, who had worked on parrots and was very familiar with seed-cracking bills, that their close affinity came to light. As famously told in Jonathan Weiner’s Beak of the Finch, biologists have since learned that seasonal changes in seed abundance produce measurable evolutionary changes in the finches. Differences of less than half a millimeter in bill length determine which birds can crack the toughest seeds and which can’t. In times of scarcity, that distinction is a matter of life and death, and as a result, the bills of whole populations can change in a single generation. The fact that natural selection can play out so quickly helps to explain how one original Galapagos finch could have morphed into thirteen species, some with seed-crushing bills, some that sip nectar, and some that eat fruit or insects. There are also cactus-flower probers, and a finch whose beak can hammer bark like a woodpecker. Played out globally, the Galapagos scenario helps the mind grasp how specializing on seeds (or other foods) can have such evolutionary influence. According to one theory, overcoming the physical challenge of eating hard-shelled seeds may have even produced the distinctive shape of the human skull.

As a child, I subjected my skull to a typical variety of sports. Though I eventually settled on swimming, I passed several seasons playing soccer and baseball, and even briefly threw my small frame into the melee of American football. One similarity among all these activities was the healthful snack served during practices and games: fresh oranges sliced into wedges. And, given that snack, we young athletes would immediately stuff wedges into our mouths, skin side out, and run around hooting like chimpanzees. Try this now and you’ll find that it does create an undeniably ape-like impression. But it’s not the wide orange smile that does it. I spent two years studying mountain gorillas in Uganda, and while they expressed themselves in all sorts of ways, I rarely saw them leer. The orange-wedge trick works because it reshapes the skull, giving a forward, snout-like projection to the jawbone. All other apes, and most ancient hominids, share that structure. But in human ancestors, the face began to flatten, and that’s where the seeds come in.

“There was a radical shift around 4 million years ago,” explained David Strait, a professor of anthropology at the State University of New York. Modern human faces appear flat because our bones are small, he told me, probably an adaptation for eating soft, cooked foods. But it was another dietary shift that started the ball rolling. “The facial reinforcements,” he said, “the large cheekbones and muscle attachments, the size and shape of the teeth—all point toward producing and withstanding high loads.” Just the kind of “high loads” that come from cracking the shells of hard seeds and nuts.

For much of the past decade, Strait and his team have argued that habitually biting large, hard objects like nuts explains the changes in ancient skulls. Their computer models show the digitized facial bones of Australopithecus—an extinct hominid best known for the “Lucy” specimen—happily munching away, with the force of each bite distributed particularly above certain teeth. It’s a habit we maintain. To revisit the sports analogy, spectators at a game don’t eat orange wedges. Instead, they litter the stands with hot dog wrappers, drink cups, and, invariably, the empty shells of dry roasted peanuts. Next time you have a bag, notice which teeth you use to crack open the tough ones. Chances are you’ll position that nut on the side of your mouth, right behind the canines, where your skull best absorbs the force of the bite. Those are the premolars, and if Strait is right, using them for nutshells is a deep, evolutionary instinct.

“A lot of my colleagues don’t believe me,” he laughed, “and that’s fine!” Critics of Strait’s “hard-food” theory point to chemical analyses and patterns of tooth wear that suggest a diet dominated by grasses or sedges. But Strait doesn’t see this as a conflict. Ancient hominids might have gobbled up all sorts of things when food was abundant, but just like the Galapagos finches, what really mattered was getting through the tough times. “Nuts were the fallback food,” he said, and fallback foods can drive evolution because the stakes are so high. “Soft foods and fruits are wonderful and sweet,” he told me, with the ease of someone practiced at making his point, “but when those run out, you either have to move, eat something else, or die.” In those terms, it makes perfect sense for the hominid face to reorganize itself around a nut-chomping, premolar bite.

If the habit of eating hard seeds did influence our skulls, much in the same way it shaped finch beaks and rodent jaws, then how might human chewing have impacted seeds? Toward the end of our conversation, Strait hinted at an answer. He mentioned new research showing how the micro-structure of seed shells mirrors that of tooth enamel. In each case, the cells lie in tight rows of ray-like rods and fibers, as if both sides arrived at the same engineering solution to resist the impact of the other. He also passed along a paper about a Southeast Asian seed so strong it can barely germinate—the halves of its dense husk cling together at the very limit of what a growing shoot can split apart. Yet in spite of this, the seed still falls prey to beetles, squirrels, and the occasional orangutan. It’s a reminder that physical defenses can only go so far. From almendros to peanuts, the story is the same: no matter how strong a seed’s shell, there will always be a rat, parrot, or sports fan nearby who evolves a way to break through. Which is, of course, why shells are only the tip of the iceberg. If plants could successfully protect their babies just by building a better box, then there would be no point in drinking coffee, Tabasco sauce would be tasteless, and Christopher Columbus would never have sailed for America.