How Not to Be Seen
Stealth Among the Leaves
The sycamore I planted in our side yard 30 years ago has grown into a magnificent, 100 ft. tall, spreading beauty. Delicately green in the spring, with row upon horizontal row of sapsucker pits on its massive trunk, way up into the crown region, it is my favorite tree, my baby. In late summer of some years, it hosts thousands of lacebug (Tingidae) families on the underside of its leaves, the lacebug mothers fiercely guarding their brood. Droppings on the leaf-litter below attest to the presence of caterpillars in the foliage above. It was hard to spot the caterpillars, but one year I found a slug caterpillar with the shape of a VW beetle, and another year, one with nine pairs of triangular projections, each studded with dozens of stinging hairs. Their Family Limacodidae has many bizarre caterpillars, all of them guaranteed to give you the pleasure of finding something strange.
But in 2018, there was something I hadn’t noticed before. The petiole of several of the partly eaten (but green) leaves on the ground had been snipped. These were not damaged leaves that the tree had shed. If they had been, the whole petiole would have been attached to each leaf because the shedding zone is a thin, corky layer where the petiole attaches to the twig. So, the snipped leaves were peculiar because many of them had only minor damage, and much remained to be eaten. Why had the caterpillar (and I was sure that it was one) not finished its food, and discarded it instead?
As I scanned the foliage above me, I spotted a few leaves that had parts missing, and it struck me that these were pretty conspicuous among the undamaged leaves. Any sharp-eyed bird would have no trouble seeing the damaged leaves and would search for the culprit nearby. This suggested that by snipping the remains of a meal and dropping it to the ground, the caterpillar was removing a clue to its whereabouts. An elongate, and perhaps striped caterpillar oriented lengthwise on a branch would be much harder to spot in the absence of a partly eaten leaf. So, this became my theory--- the Mystery Sycamore Snipper was hiding its whereabouts from visually searching predators.
The challenge for me was to do better than the birds and actually spot The Snipper. I spent hours looking up into the foliage, tracing branches, twigs and leaves, and searching carefully in the vicinity of damaged leaves. I repeated this with binoculars both in the day and at night, and even purchased a powerful UV flashlight on the chance that my snippers were fluorescent (many caterpillars are). My wife began to wonder about my sanity.
The summer of 2019 ended without success, but I had decided that I could at least determine the season during which the snippers were active. I did this by collecting and counting the number of snipped leaves during the summer days of 2020. By mid-April, the snippers were going at it with hammer and tongs, and by May/June, it was clear that there were two peaks of snipping activity, thus two generations of snippers.
A representative day’s snipping is shown below. The number of leaves suggested that, if the caterpillars did not engage in wanton defoliation there were probably a couple of dozen or more snippers in the tree overhead. If a leaf a day keeps the predators away, the number of leaves would be equal to the number of caterpillars, but I sometimes saw still attached leaves with similar damage, and I didn’t know whether a caterpillar snipped every leaf it nibbled, or even snipped more than one a day.
Still, with peak leaf snipping rates of 50 or 60 a day, I could bracket the number very roughly as between 30 and 100. My best guess was that there were 50 of them munching away above my head.
You would think that with that many caterpillars snipping away, it shouldn’t be all that hard to spot a few. But that’s only if you didn’t also think about how large this tree is. I estimated it had 100,000 leaves within an air space of around 200,000 cubic feet. What are my chances of spotting a few dozen caterpillars distributed among all these leaves and in this space? Here is how I made a rough estimate.
As an example closer to most human experience, consider that you need the ace of spades in a card game you are currently losing, and you must take the top card from a shuffled deck. Because there are 52 cards in the deck, your chance of drawing the ace of spades from the top of the deck are 1 in 52, or 0.0192. You set that card aside, so on the next draw, your chance of getting the coveted ace is 1 in 51, slightly higher than the first draw. And so on it goes until when you have drawn the last card, which if you haven’t gotten your ace yet, has a 100% chance of being the ace.
Let’s start with the simplest example: one caterpillar. Every leaf I look at is like the top card on the deck, only my deck has 100,000 “cards”. On the first leaf, I have one chance in 100,000 of finding a caterpillar, and I don’t check that leaf again . The chance of the caterpillar on the next leaf is 1 in 99,999, a tiny bit higher, and so on it goes for every leaf I inspect. When I have inspected all 100,000 leaves, my chance of finding the caterpillar is 100%.
Of course, my guess was that there were 50 caterpillars, not just one, so my chance of finding a caterpillar on the first leaf would be 50/100,000 or 0.0005. With each empty leaf, the chance on the next leaf would be a tiny amount higher. The trouble is, I can only see (and inspect) the bottom 20% or so of the tree, or 20,000 leaves. Although that sounds better, the number of caterpillars is also 20% of the total, or 10, and my chance of picking a leaf with a caterpillar is still the same (0.0005), i.e. not so hot. However, the total probablility of a caterpillar increases by 0.0005 for every inspection tour. By the end of a week, it would be 1 in 300, and my wife would be even less sure of my sanity. You can see what I was up against (the sycamore, not the wife).
Imagine my delight then, when I spotted three caterpillars chewing leaves on a branch that drooped down to shoulder height. Could this be the Mystery Snipper? To test this, I enclosed each caterpillar and all its nearby leaves in a loose mesh bag.
The next morning, the bag contained the caterpillar, a snipped, partly eaten leaf and a lot of caterpillar droppings (aka, “frass”). I had finally found my culprit, only two years after my first observation. I identified the Sycamore Snipper as the larva of the drab prominent moth (Misogada unicolor), family Notodontidae. David Wagner’s Caterpillars of Eastern North America reports that it is “closely associated with sycamore” and that “reports from cottonwood and other food plants may be in error.” Sounds right to me.
Having solved who was doing the snipping, it remained to ask what the snipper was doing after it had snipped. Most likely, it remained in the same vicinity, for if it moved a substantial distance from the scene of the crime, what would be the point of removing the evidence in the first place. Unfortunately, it is much harder to spot the stubs of snipped petioles, so I saw very few opportunities to search the vicinity for the caterpillar. On top of that, I would need to spot fresh stubs for a meaningful search. The best I could do at this point was to make a digital montage of a Misogada caterpillar on a sycamore twig, as below. Would that caterpillar be hard to spot? I am not sure, but you’d have to ask the bird, not me.
I gained a lot of sympathy for the bird because the bird’s probabilities were the same as mine. For me, this was a fun project. For the bird, it meant eating or not eating, and I expect that the bird has strategies that are not available to me, strategies that are the outcome of an evolutionary battle with the snippers who, by snipping, made the bird’s life harder and their own more secure.
Ah-ha! The case of the clever little snipper is solved!
Love the probability analysis and your dedication to solving the mystery!