Artful Diners - Summer World: A Season of Bounty - Bernd Heinrich

Summer World: A Season of Bounty - Bernd Heinrich (2009)

Chapter 8. Artful Diners

2 August 2006. TEMPERATURES ARE NOW REGULARLY IN the high eighties and low nineties, and as usual, the air is muggy and sweltering. The cicadas’ buzz is now the definitive summer sound. The house wrens are totally silent—a big change! The eggs of this clutch (their last) are about to hatch. Do they already know this? Is that why the male now no longer sings? In the trees I see a wiggle of leaves here and there, as birds search for caterpillars. I saw the first roadkills of the large caterpillars—mostly sphingids—about ten days ago; they are leaving their food trees to wander before pupating underground. Monarchs float by lazily, but they occasionally accelerate in a few wing beats. One of them flies in a more steady, stately manner, and then lands on a bush. It spreads out its wings, and I notice that another one (which has its wings closed) dangles from it, attached by its genitals. Extreme mate-guarding. Maybe I will soon see more of their white, yellow, black-striped caterpillars on the milkweeds. Viceroy butterflies, the monarch’s mimics, are making their first appearance. At night we see the flashes of distant lightning storms, and occasionally hear rumbles of thunder.

ONCE, WHILE STILL SEARCHING FOR CATERPILLARS, I SAW something that made my eyes pop. It was in midsummer, and I found, under tall forest trees, partially eaten green leaves on the ground. When I picked them up and examined them, it was obvious that they had not been shed by the tree in the usual way (at the junction between twig and leaf stem). They had been cut through the stem (petiole); caterpillars had fed on the leaves and had then snipped off the remainder. They were discarding the remnants of their meals, and they had used precious time and energy to chew through the very tough woody leaf stems. Since I had myself been using leaf damage as a cue in my caterpillar hunting, it seemed to me that caterpillars leaving feeding damage on leaves would effectively be leaving a “track” that caterpillar-hunting birds might use to find them. Obliterating those tracks would be a neat trick for “invisible” caterpillars, to keep predators at a distance. Like the blues’ caterpillars and their ants, the story of leaf-eating caterpillars and birds also involves an evolutionary arms race, and it is waged all day every day all summer long.

Birds have specialized behavior to capture insects, and insects have specialized behavior to try to avoid capture. Throughout evolutionary history, the field of battle in this arms race keeps shifting as each participant keeps up with the other. Those that don’t keep up will cease to exist. At no time is the race more intense than after the birds’ young hatch, when the parents’ real work of caterpillar hunting begins.

Most small northern forest birds attempt to raise four to six young in any one clutch. It requires an enormous daily investment in foraging to feed that many babies and bring them to adult body weight in about a week. To predators, baby birds are helpless gobs of tender meat whose noisy competition among themselves for their parents’ attention—in order to be fed—is nothing more than a convenient advertisement of where to find them. Thus the fledgling stage is the most dangerous time during a bird’s life, and there is a huge advantage in becoming capable of flight and getting out of the nest as soon as possible. To maintain the offspring’s phenomenal, sustained growth spurt, the adults must feed their youngsters every few minutes, and the food must be easily digestible protein. For most forest birds, that means caterpillars.

Much of what applies to baby birds also applies to caterpillars, of course, except that they necessarily feed on foliage, a decidedly low-protein diet. Caterpillars have few hard parts—no skeleton and usually no “fur.” They are easy to digest, and many of them require no preparation before eating: they can be swallowed as is. Like baby birds, they also need to grow fast, but because most (though not all) of them have a vegetarian diet, they don’t reach their full weight nearly as fast as baby birds. For many caterpillars survival requires a delicate balancing act—hiding versus feeding. This is a hard compromise, because the tree’s leaves are necessarily exposed to sunlight, where it is difficult to hide.

Birds, wasps, and flies have been preying on or parasitizing caterpillars (or both) for probably at least 100 million years. Year in and year out, the great majority of hatchlings of any one clutch of moth or butterfly eggs, consisting of perhaps two hundred, will be eaten. Such relentless pruning has necessarily left a deep imprint on the caterpillars.

Caterpillars are startlingly, dazzlingly diverse in their shapes, colors, and behavior. Some are protected by potent poisons. Others have greatly reduced their palatability by developing woolly hairs or sharp bristles. All the caterpillars that are conspicuous to us and to birds are relatively immune to wasp and bird predation; those that are hard to find tend to be the most prized as food by birds. Not surprisingly, therefore, the majority of edible caterpillars use a variety of strategies to keep themselves well hidden, if not to make themselves nearly invisible, and these are the birds’ main summer fodder. It might seem that if more than 90 percent of any clutch of moth or butterfly caterpillars will get eaten, then they are not well adapted to evade bird predation. But as one member in an evolutionary arms race gets better at hiding, the other gets better at finding. Birds are very good at finding.

I remember my own first finding of a caterpillar as an exquisite experience. I was in elementary school and was picking berries in the woods. I was startled to discover among the raspberries something very beautiful—a plump green body decorated with red tubercules sprouting short black bristles—and I have been enamored of caterpillars ever since. As a graduate student I chose to study how the tobacco hornworm caterpillar is programmed to handle leaves without moving from its attachment point at the base of the leaf, and without leaving any scraps. Avoiding being eaten is usually an even greater problem for a caterpillar than finding enough food. To discover how caterpillars might escape detection by birds, I first used students as stand-ins for the birds.

I got reacquainted with caterpillars during a summer in the late 1970s at the University of Minnesota Field Station at Lake Itasca, where I helped teach a field ecology course. Each of the three instructors designed “field projects” for our select group of about a dozen graduate students in biology. These field projects had to involve the local flora and fauna, and I was spending a couple of days getting oriented in the local woods, looking for potential projects. It was then that I found the clipped-off partially eaten leaf of a basswood tree.

In my previous fieldwork with bumblebees (in Maine) I had studied individual bees and found that they became specialists, developing individual skills for finding and working on specific kinds of flowers. In any one field with several kinds of flowers, one bumblebee might, for example, search out the clover flowers while ignoring most of the goldenrod. Meanwhile, another in the same field might visit the goldenrod and ignore the clover. The bees maintained their specialties independent of the abundance of other flowers. They developed “search images” of what flowers to look for. Birds (and humans) presumably also use search images to help them find specific caterpillars. It helps to know what you are looking for, but this knowledge usually comes at the cost of not noticing the remainder.

To demonstrate the significance of the search image to my students, I “planted” four stick-mimicking caterpillars (Geometridae) on a poplar sapling, and then gathered the students around it and asked them to search. I told them that there were four caterpillars directly in front of them, but gave no clue of what these looked like. (Two of the geometrids mimicked live green twigs, and two mimicked dead brown twigs.) No “successful foragers” were allowed to give clues to anyone else about what they had found or where they had found it.

I had expected that it might take seconds, or at the most a minute or two, to find any one of the caterpillars, which were only a few inches in front of our eyes. I was indeed surprised to discover that despite their earnest and continuous searching, few of these naive but eager hunters discovered a caterpillar within the first half hour. But those who did eventually find one then located the other, similar one in a minute or less. That is, as predicted, after the students knew what to look for, everyone’s performance increased enormously. This generalization has large implications. If birds search for caterpillars similarly to bumblebees and students, then there is a great advantage for each caterpillar species in having a different disguise from the next: that is, one not included in the existing repertoire of the predator’s search images. This implies that it helps to be rare and different.

There is an enormous variety of caterpillars among the 250,000 species of moths and butterflies. Some species look like leaves or parts of leaves; others resemble twigs, bird droppings, or debris; some blend into the background of bark that they choose to rest on when not feeding; others cover themselves with debris taken from their background. In a slide demonstration of caterpillars that I have given several times, I take the audience on a virtual caterpillar hunt featuring various showy caterpillars perched on their food plants. After a number of practice slides, to get the audience used to finding caterpillars in the projected images, I give a test: I show a picture with real hidden caterpillars, or things that look like them, or both. Invariably even professional entomologists have been fooled, either not seeing the real caterpillars that were in plain sight and magnified 1,000 times on the screen, or pointing out something they wrongly assumed to be a caterpillar. My own skill at finding caterpillars despite their various tricks involved looking for fresh feeding damage on leaves, in order to focus the search on a smaller area, since most (though not all) caterpillars don’t move far.

After I had found the first clipped-off partially eaten leaves in Minnesota, I searched and eventually found the inconspicuous stubs of leaf petioles still attached to a twig of the tree where the leaves had come from. The twig with its many remaining leaves looked un-grazed, and I would normally have passed by without a second look. But now I looked closely, and as expected found a large caterpillar (a big brown Catocala moth larva that was nearly invisible as it rested tightly pressed onto a nearby branch, mimicking the irregularities of the bark). I later watched and photographed the caterpillar and learned that it spent all day motionless in its disguise on the branch. In the evening it moved quickly out onto the branch, fed on a leaf that was far too large to consume in one meal, and after eating a portion of this leaf, backed down the petiole and chewed through it so that the leaf remnant dropped off. The caterpillar then walked away and resumed its position in its hiding place on a twig. I observed and photographed similar behavior in many species of caterpillars, but these were only those species that are “invisible” and hence have evolved to avoid predators that hunt by vision instead of scent.

These caterpillars were also the same ones that, while still feeding and before clipping, pared the leaves down so that they looked smaller rather than conspicuous because of tatters or holes. Some of these species, like the prominents (Heterocampidae), disguised their feeding damage by fitting their own bodies into the area at the leaf edge that they had consumed, and their bodies mimicked the leaf edge they had removed even to the extent of having fake leaf blemishes and leaf-edge patterns resembling those of the leaves of the tree they fed on.

Only those caterpillars that were routinely predated by birds had cryptic body markings, practiced leaf paring, held specific feeding positions on the leaf, and engaged in leaf clipping. The bristly or brightly marked ones that were not eaten by birds (but were still parasitized by wasps and flies) were “messy” feeders who did not clip leaves. These observations suggest that leaf clipping is part of a behavioral repertoire in the game of hide-and-seek with birds.

Although it seemed like a safe bet that birds hunt for the palatable “invisible” caterpillars, as I did, using leaf damage as a cue, nothing can be taken for granted. Any such idea needs to be tested, generally through long, tedious work that may take months or years to carry out, and that almost invariably leads to surprises. Here was a project that needed to be done, and I invited a friend and colleague, Scott Collins, to join me at my cabin in Maine for a summer’s work and fun, to do the critical tests and determine whether a bird might be capable of learning to hunt for caterpillars by using leaf damage as a tracking cue.

Scott and I decided to use chickadees as our subjects. They were abundant, tame, and easy to catch in mist nets we set out in the woods. We erected a screen-cloth aviary in a clearing we made in dense woods where chickadees would feel at home, and in that enclosure we set up two compartments: one was to hold our birds (an eventual six), and in the other we installed ten small birch or cherry trees every other day. Our results were clear—all of our subjects quickly learned to search preferentially on trees (whose locations we kept switching around) with leaf holes, if such trees had yielded food before. In subsequent tests we also determined that they could be trained to search at specific kinds of trees (such as birch versus maple or cherry), and also to use real caterpillar feeding damage as opposed to our experimentally damaged leaves, which had holes made with a paper punch.

Birds in the wild hunting for caterpillars face a considerably more difficult problem than our chickadees faced in our enclosure. If the natural situation in the surrounding woods were as simple as we had to make ours in order to answer one specific question, then birds would probably be behaviorally hardwired to be attracted to caterpillar feeding damage. They aren’t. Our chickadees learned to associate leaf damage on specific kinds of trees with food. However, to be indiscriminately attracted to feeding damage, as such, even on the proper species of tree, could be a liability in the field, because trees accumulate much leaf damage throughout the summer (or over a period of up to six years in the tropics), and that damage eventually ceases to be a clue to whether or not a caterpillar is still in residence. Early in the summer, when all the leaves are fresh, leaf damage can suggest that a caterpillar fed recently nearby; but in the fall a damaged leaf could mean that a caterpillar was there three months ago.

There is a second potential problem with using leaf damage as a tracking cue: the caterpillars that are the least palatable leave the most leaf damage. As already mentioned, caterpillars that are bristly or spiny (or both) and caterpillars that are poisonous, which are not routinely eaten by birds, are “messy” feeders—they make no attempt to hide their feeding tracks. They feed on the softest leaf tissues and often leave the tough leaf veins and the rest of the leaf hanging. Using leaf damage, as such, could therefore be a very misleading clue in hunting for palatable caterpillars. The contrasting behavior of the specific caterpillars that are less favored by birds thus provides independent evidence that parasitoids are probably not searching primarily by visual cuing on leaf damage.

I could distinguish whether a leaf had been fed on by a palatable or an unpalatable caterpillar, since unpalatable caterpillars ate a leaf into tatters, and palatable ones pared it down gradually to reduce tatters. I wondered if birds, to whom being able to make the distinction would matter, could also learn to differentiate leaves eaten by palatable versus unpalatable caterpillars. I talked about this problem with the animal psychologist Alan Kamil, who had recently used blue jays in lab studies to test these birds’ visual acuity in discrimination in finding cryptic moths. In his lab, jays undergo individual choice tests, after they are trained to peck a screen in response to specific pictures projected onto it. They get a food reward if they respond to the “correct” picture. I sent his lab a series of photographs of leaves eaten by palatable versus unpalatable caterpillars, and Pamela Real et al. conducted the experiments. To my satisfaction, but not to my great surprise, they reported that “the jays exhibited little or no difficulty” in distinguishing the pictures of leaves that these two kinds of caterpillars had fed from. Not only that: they generalized. The lab jays learned to peck at pictures on a screen of a leaf partially eaten by a palatable caterpillar, and to ignore pictures showing unchewed leaves or leaves fed on by unpalatable caterpillars.

29 MAY 1985. I WAS WALKING UP THE PATH TO MY CAMP IN Maine. The aspen leaves had unfurled pea-green leaves about a week earlier, and as I walked under their canopy I again found an interesting fresh leaf on the ground. But this leaf was neatly rolled up into a tube and carefully held together with silk. I picked it up, expecting to find a caterpillar inside when I unrolled it. And there it was indeed—a thin, pale microlepidopteran moth caterpillar. What surprised me was that the leaf roll was on the ground. Had it been discarded by the tree? Would a tree shed its leaves to get rid of any caterpillars rolled up in them?

I searched under the same tree and other poplar trees and within an hour or two had picked up 246 identically rolled-up leaves. Most of them had a similar caterpillar inside, about 0.3 to 0.4 inch long. Leaf-rolling by caterpillars is common, but finding rolled-up leaves on the ground under the tree is not. Every one of the rolled leaves that contained a caterpillar was missing most (though not all) of its stem—proof that the tree had not spontaneously aborted the leaves. Petioles are tough; they do not tear or break in a storm. The connection between the petiole and the twig would break first, and therefore it was not the tree that was getting rid of its caterpillars but vice versa.

The caterpillars had rolled themselves up by grabbing an edge of the leaf, attaching silk that stuck to the outer leaf edge, pulling in and attaching the other end of the (non-stretching) silk to the inner leaf surface, and repeating the process until they gradually rolled the leaf into a tube. After finishing that, they had reached through the top of the tube and gone to the considerable effort, and maybe risk, of sticking their neck out to chew through the petioles so that they and their leaf tube, or roll, would fall to the ground. They then stayed in their leaf rolls, and eventually pupated inside them. I kept each of the clipped-off leaf rolls with its enclosed caterpillar, which molted into a pupa, and in the first week of July little gray moths emerged.

Why did the leaf-rolling caterpillars clip off the leaves they were in? This leaf clipping was much different from what I had observed before. The other caterpillars stayed on the tree, where there were always other leaves to feed from. These, in contrast, were isolating themselves from the tree, and as a consequence seemed to be restricting their food supply. Why would that be advantageous? Are they safer on the ground than on the tree? To find out, I took 200 of the freshly dropped leaf rolls containing caterpillars, divided them into five groups, and distributed them to five different locations on the ground. One week later all the leaf rolls were still in place. So far, so good—the ground seemed safe. But maybe the caterpillars are equally safe on the tree. That would be difficult to know, though, if they normally don’t stay there. I wondered, however, what would happen if I unrolled them from their apparently safe little houses after they had become grounded and then put them back up into the tree.

There were two young poplar trees in the clearing by my cabin, and I released numerous of the unrolled caterpillars onto their branches. They seemed not well suited to hanging on, especially to poplar leaves, which vibrate wildly in even the slightest breeze as though designed to shake off caterpillars, and many did immediately fall off. However, others hung on, and within a day I found thirty newly made leaf rolls. That is, my caterpillars almost immediately made themselves new homes. Two days later, however, seven of the thirty rolls had been clipped off. I saw no sign of predation, but it seemed that using tree saplings in a field was not a fair test to examine predation when their natural habitat is the crowns of large trees in the forest.

Making the observations was fun, especially if I could continue them in the top of a tree. So I carried a supply of leaf rolls that I had gathered on the ground up into the crown of an old aspen tree, perched comfortably on the branches, made sure that they had no existing leaf rolls or petioles from previously clipped-off leaves, and then unrolled one leaf at a time and released caterpillars onto these cleared, marked branches. Two days later I climbed back up and found thirty-seven new leaf rolls (presumably made by the more than eighty caterpillars I had released). One caterpillar was being eaten by a shield bug; eight partly rolled (or unrolled) leaves were empty; twelve partly rolled leaves each held a caterpillar, but none of the leaf had been eaten; seven fully rolled leaves with caterpillars inside had some of the leaf eaten; and there were nine clipped petioles. Thus, apparently several of the caterpillars had been eaten; being up in the tree was risky. These were interesting observations, but they could be interpreted in several ways, and no firm conclusion was possible—not enough for a scientific publication. Meanwhile, the caterpillars’ season was over, and I thought about other things.

I did not get an opportunity to think about these caterpillars again until more than two decades later, when I started to write this book and found my notes squirreled away in a file. When I took a break and went for a daily jog during the next ten days in June 2006, I checked for freshly dropped leaf rolls on the gravel under the aspens along the road (in Hinesburg, Vermont). I found 208 of them. Twelve were freshly clipped and nine of these contained a mature larva. (As before, they pupated inside their rolls, and by the first of July the adults, the small gray moths who are fast runners as well as fliers, again emerged.) However, the remaining 196 leaf rolls had petioles (they had not been clipped off ). All but two of these were without a caterpillar. Therefore, caterpillars had left them to make another roll, and they had done so apparently in time, before the tree had shed them.

The leaf rolls I had picked that had petioles but no caterpillars contained piles of caterpillar frass (feces), indicating that a caterpillar had been in residence for a long time, all the while feeding and fouling its nest (or pantry?). The leaf tissue on the inside of the rolls had turned color (yellow) or become necrotic. In short, these rolls had ceased to provide food, and they had been shed by the tree, presumably as a mechanism of getting rid of nonfunctioning leaves. However, before that happened, the caterpillars had left their deteriorating leaf rolls to seek fresh leaves and make a new roll to feed and hide in. That explained why many of the freshly clipped-off rolls that I had found earlier had little feeding damage inside yet contained a large caterpillar.

Apparently caterpillars leave their roll when it fills with feces or becomes necrotic, and then make another roll and resume feeding. As a result, many “empty” rolls accumulate on the trees and these fall off, but late in the summer. Finally, when the caterpillar is nearly mature, it clips off the last roll that it is in, and then rides with it to the ground, where it remains inside, pupates, and then emerges as an adult.

By late August I started seeing leaf rolls of another sort, on young basswood trees. Basswood leaves are gigantic relative to poplar leaves. To a diminutive (microlepidopteran) caterpillar that needs to roll up a leaf to hide in while feeding, a large leaf presents a problem. But these caterpillars had solved it beautifully. Each caterpillar had made a cut into the leaf all the way through several large leaf veins and then up to but not through another major vein. As a result a large portion of the leaf had flopped down while the rest of the leaf stayed up, and the dangling portion of the leaf was then rolled up. The leaf vein from which the roll then hung would continue to supply the nutrients to the rolled-up leaf used by the feeding caterpillar.

I continued to jog almost daily along the same road, enjoying the summer and at the same time keeping track of other signs of caterpillar magic. On 10 September I found out something new about caterpillars that I had never seen or heard about. There was at that time a caterpillar outbreak on the maples (both sugar and red). This outbreak was not as eye-catching as that of the fall webworms (who may enclose whole trees in diaphanous veils of webs) or that of the forest tent caterpillars. Still, it was eye-catching enough. On some maple branches as many as a third of the leaves were folded over, much as one might fold a piece of paper and then glue it together to make an envelope. It was the typical work of one of the thousands of species of microlepidopterans, and I was tempted to let it go at that—I knew a little caterpillar would be inside and feeding. So what? But for some obscure reason I idly opened several of these folded-over leaves, and to my great surprise saw neither caterpillars nor masses of feces. I was puzzled, because caterpillar feeding damage was evident—these caterpillars had eaten the underside of the leaves. Many birds learn to open leaf “envelopes,” and these envelopes could have been easily opened. But if I had been a bird, I would have been disappointed—at least at first. I looked into many of them and wondered: where are the caterpillars?

No caterpillars were visible, but each leaf envelope contained a thin dark object an inch or more in length. At one end it was as thin as the lead in a mechanical pencil, and at the other end it was about one-tenth of an inch wide and had an opening: it was a long, tapering tube. I eagerly broke several of these rather dry and friable tubes open, and again at first found nothing. However, I eventually did find what I was looking for: the tiny, almost transparent caterpillars. They were near the bottom of the tube rather than at the top. It turned out that they retreat to the bottom of their tube as soon as a predator opens the leaf envelopes they are in. But where did the tubes come from?


Fig. 18. Small moth caterpillar that feeds on the undersurface of leaves and makes a tube house in the space created by folding the leaf into a sandwich.

The tubes were made by the caterpillars, and from their own fecal pellets. Thus, these caterpillars were not soiling and spoiling their food, but instead using their own waste to make a retreat to hide in. They silked their fecal pellets onto the mouth of the “door” of their house, to gradually build a wider, longer tube as they grew. The beauty of this caterpillar’s behavior is in the mind’s eye, but in still another caterpillar it also enhances the fall foliage show.

It is mid-October. The aspen leaves have turned a rich deep golden yellow. On a slight breeze after a light frost, the leaves come tumbling down. Unlike those of the red maples, for example, where random blotches of yellow, red, purple, and pink may intermingle, the poplar leaves are uniformly, unvaryingly gold. But under some trees are exceptions—many of the bright yellow leaves have a conspicuous pea-green spot near the petiole, between the mid-vein and the nearby subsidiary vein. The color, and the very specific spot on the leaf where it is always located, catches the eye—you look around and find another, and another. And each time, the green spot is in exactly the same place on the leaf. The singular green spots in otherwise rapidly aging bright yellow leaves took me by surprise. They had to be caused by an outside agent. And they were.

Examining the green spots under a microscope, I could see through the transparent leaf epidermis, and beneath, within the leaf tissue itself, was a little pale green caterpillar with a trail of black fecal pellets behind it. This “leaf miner” caterpillar also rides the leaves down to the ground and then feeds on them. But it is too small to be eaten by birds, and it is probably also too small to be able to chew through a leaf petiole. This caterpillar could surely do its eating and growing early in the leaf’s life cycle, but such a tiny caterpillar would very quickly dry out in summer heat and dry air. It would need to reach moist ground to pupate. It must leave the tree crown to do that, but for such a tiny larva to leave the leaf and enter the atmosphere in the dry summer heat could cause rapid death from desiccation. However, by shifting its caterpillar stage to late in the summer and early fall, when the weather is cooler and more moist and when the tree normally sheds its leaves, it can ensure itself a safe, moist harbor during a free ride to the ground. And just in case it still needs to feed a little more after it reaches the ground, as I assume it does, it persuades a part of the leaf to reduce its aging process, making it stay green. (See the last page of the color insert.) Color change signals leaf senescence and the beginning of fall. This caterpillar apparently has a chemical with which it manipulates the plant’s genes to extend the life span and thus the length of the leaves’ summer. Even though I never had a quarrel with fall, I think I could use a few drops of such a magic potion.