The Homing Instinct

by Bernd Heinrich

  The silk wrapping the rotating bee was attached to the net at two ends, like a spindle, and the silk she shot out from her spinnerets was not a single thread like one from a spool. It looked like a triangular white sheet that spread out as it left the hind end and enveloped her victim. It was presumably a spray of liquid that solidified as it was extruded. Her hind legs were not only rotating her prey, but also spreading the spray of silk strands apart to make them into a sheet and to help wrap the prey in a straitjacket. Once the prey was fully wrapped, she released the bee from the net by biting through an end of the spindle, and then dangled the bee package by several centimeters of thread from her rear spinnerets. She then held the bee off the web with one of her stiffly held hind legs and ran up the web using her seven remaining ones to run. The bee package dangled on that one thread the whole way up to her lair (after watching several hundreds of this seemingly delicate maneuver, I saw no instance of a spider dropping its prey), where she attached it, and then turned around and bit into the bee. She held it by her front-most pair of legs and sat almost motionless for the next two hours, presumably injecting enzymes through her chelicerae as through the needle of a hypodermic syringe, to digest and then suck out the bee’s contents.

  Were these behaviors executed by blindly following a script? What would happen if the sequence were interrupted? What would she do if a second prey entered her web before she had finished the repertoire of handling a previous prey? Which “program” would she follow then? To find out I threw a big bristly (Tachinidae) fly into her net while she was still feeding on the bumblebee. She instantly released the bee (which stayed in place in her lair) to drop down into her web and grab and wrap the fly. But instead of now carrying the prey up into her lair as she normally did, she killed it in situ and left it secured in the web where she had caught it, to then walk back up to her lair and resume feeding on the bee. Would she now forget it? And how about another distraction, a third simultaneous potential prey?

  An hour later, while Charlotte was still feeding on the bee, I threw in a white admiral butterfly. It thrashed violently, but Charlotte did not budge, perhaps because the commotion was too vigorous or because she was no longer hungry. To test, I threw in another bumblebee of the same size as the first. It, too, was ignored. So, she was either not hungry or could not triple-task. The butterfly escaped.

  Charlotte finally finished her “breakfast bee” at noon and dropped its by-then dry husk. But instead of staying in her lair, as I had expected, since that is what she “always” did after finishing a meal, she casually walked down into the net to retrieve the dead fly. I was surprised because normally she stayed all day in her lair and came down into her web only to capture struggling prey in it. Since orb web spiders are practically blind, she could not have seen or felt the fly, which was past the point of jiggling the net. Since Charlotte had never spontaneously walked around in her net during the daytime, she must have remembered the fly that she had killed and left there earlier. (Wolf spiders, Lycosidae, and jumping spiders, Salticidae, build no webs and hunt by stalking and leaping on prey by vision alone. Jumping spiders have a set of two huge eyes directed forward, plus three more pairs pointing in different directions.)

  At 2:00 p.m. I threw still another bumblebee into the net. Charlotte instantly left the fly in her lair to get the bee but hesitated when she was close to this much bigger bee. It struggled for the next two minutes as the spider continued cautiously to palpate the net with her front legs. This bee eventually escaped, and Charlotte returned to her lair to continue feeding on the fly. The net, due to the struggles of the two large prey, was by now badly torn.

  The net was not repaired by the next day, August 7. I had read that a spider repairs its net every night. Charlotte had apparently not read the book, or the book is only about the very hungry spider. Is she still interested in capturing prey? I wondered. So—I threw in another bee to find out. Her answer was a resounding yes. She vaulted a meter or more down to capture it, wrap it, give it the bite of death, and immediately take it up to dine on in her lair. And she didn’t repair her net the next night, either. Her handling technique involving all eight legs was, as before, not only a theme of fascination for me but entertainment to the kids visiting me at the time.

  August 26. Charlotte was, as usual, in her lair. She had made a beautiful new symmetrical web of thirty spokes and sixty concentric catching threads the night before. In quick succession I tossed her two small bumblebees (the most common and easy-to-get insects at that time). She captured and silk-wrapped them one at a time but took only one of them up to her lair. Four hours later, while she was still feeding on it, I tossed a third one into the web. She ran down and tried to wrap it but twirled it only a few times and then left to resume feeding on the first bee up in her lair. This third bee escaped. But the second one from the morning, now five and a half hours later, was still securely in the net and moving now and then because Charlotte had neglected to kill it. Now she finally went down to bite and kill it, then climbed back again to resume her meal. It was a small bee, and she not only sucked out the juices but also chewed it to a pulp. In a few minutes she dropped the head of that first bee, and then left the rest of the carcass briefly, only to return to it and resume feeding on it. (She dropped all the empty silk-wrapped hulls onto the table below, where I was able to see the remains of her meals, both the hard parts that hadn’t been ingested, and the round liquid fecal spats of what had passed through her gut, imprinted on paper I had placed for record keeping under her lair.)

  The next morning the bees’ remains were on my desk, and Charlotte was motionless in her lair. I tossed a small grasshopper into the net. She rushed down into the middle of the net, then stopped as if lost. Unlike the energetic bees, trapped grasshoppers often do not budge. She gradually wandered all over the net, tweaking threads here and there, presumably trying to induce her prey to move to try to escape—a mistake that instead quickly results in their capture. The grasshopper still did not respond, but after three minutes she was finally within a couple of centimeters of it. This time when she tweaked the net, the hopper did move slightly and she was on it in a second. In about five seconds she rolled it in silk and then, with it attached by a thread from a spinneret on the end of her abdomen but held to the side off the net guided by the left hind leg, she headed back to her lair. But one of the hopper’s long hind legs was sticking out, and since it had not been held far enough away from the web, it got caught in it. Instead of continuing, she turned around and came back to reattach the hopper to a thread, and then resumed taking it up to her lair. But after ascending a short way, it again got caught in the web. But this time she was several centimeters from “home,” and she continued and went all the way in. Yet, a minute or two later she “remembered” that her hopper was missing, because she came down to bring it up.

  A big fly was handled differently. It also got caught in the web on the way up, but she continued all the way up to her lair and after a short pause came back down on a guy-string, bit two of the threads holding it, and attached another short one from her rear. Due to the angled web and the low position in the web, she and the fly then swung about ten centimeters away from the web where she dangled loosely, while dangling the fly. She then ran up the length of the one-meter vertically dangling thread to return to her lair, pulling along the huge fly five centimeters behind her.

  In late afternoon an approximately seven-centimeter-long dragonfly came into the cabin through the open door, flew to the window, and landed in Charlotte’s web. She ran down toward it immediately, stopped within about fifteen centimeters of it, and then retreated. No further reconnaissance was made as the dragonfly continued in its struggles, and after fifteen minutes it escaped. Was this prey too large? I tested by tossing an ant into a remnant of the now badly torn net, and after only a few seconds’ delay Charlotte came down from her lair, found it, and took it. I resumed testing her discriminations and tastes by tossing in a fuzzy caterpillar. She came down to i
nvestigate it, but then went back into her lair. Not hungry? I tossed in a bumblebee—it was taken.

  She rebuilt her web by the next morning, but with only thirty-six circular orbs, compared to the sixty on the day before.

  As the summer rolled on, Charlotte continued to be a constant presence a meter over my head at my desk. I fed her occasionally, and she fed herself on the cluster flies that were starting to come into the cabin in October. Most of the trees had by then shed their leaves, and temperatures in the cabin were at times low enough to form ice. Her routine had seldom varied through summer and fall, but one day—on October 29—I didn’t see her at her usual home made of silk under the ceiling boards. I felt the loss of her and presumed she had, according to the prevailing lore of orb web spiders, died. But on November 11, after I had been away a few days, I found her barely a meter from her old lair. She had come to rest on the underside of a ceiling beam where she was hidden by a loose piece of bark, and there she had molted; a cast “skin” (cuticle or exoskeleton) clung to some silk beside her.

  The cast skin was quite important: it meant that she had not yet reached maturity, despite all of the assisted hunting throughout the summer. As far as I could determine, she had not grown much during the summer, and she had been large to begin with. Was she already several years old when she settled by the window? That cast skin indicated that she still had some growing left to do.

  May 7, 2011. The winter was long past, and I was back at camp. The weather had lately been depressingly cold and rainy. Insects had just started to fly again. The birds were quiet and the trees leafless, but fiddleheads were starting to poke out of the ground, and to my great delight, but not surprise, Charlotte was still at the precise spot on the ceiling beam where I had last seen her in November. She had survived the winter; when gently poked, she moved. I was surprised, because I was aware of the spider lore that orb web weavers grow to adulthood in one season, as in E. B. White’s story.

  By May 19, 2011, Charlotte had left her hibernation spot and returned to her home, the exact lair in the right-hand corner of the second south window where she had lived the previous summer. She again built a symmetrical web against the window, also in the same place as the summer before. It was still early in the season and I saw neither flies nor any other insects at the window. Might she be hungry?

  Bumblebee workers were not yet available, but blowflies were aplenty just outside my door at a chipmunk carcass. I caught one and let it fly into Charlotte’s net, and the instant it hit the web and got snagged on one of the sticky threads, Charlotte dropped down faster than I had ever seen her move. Slam, bam, in a second she had the fly and, after a quick bite and rollup in silk, had it dangling below her on a thread and was running back to transport it into her lair. Then she began her slow “kissing” embrace, and two and a half hours later when she was done sucking the juices out of it, she also chewed it to a pulp.

  Her appetite was also expressed in other ways; as soon as it got dark, she came out of her lair and took up a position in the center of her web, where she stayed to “hunt” even while continuing to chew on the ever-tinier remains of her blowfly. While so engaged, one after another blackfly flew into the net (I had had the door open during the day). These tiny flies didn’t move once trapped, but she kept shaking the net to try to locate them. And after she did, one at a time, she added them to the same food bolus she was already feeding on. Since she was locating nonmoving prey, presumably by causing it to jiggle due to the momentum of the energy she had put into it, I needed to ask her something I had not asked before: Could she distinguish prey from dead non-prey?

  One at a time, I threw into her net pieces of bark, carrot, and a seven-centimeter sliver of wood. Charlotte did not respond to any of these immediately, but after a few minutes she did slowly walk down from her lair, jiggle the net, and locate the objects. Leisurely, she cut (bit) the threads holding them and dropped the bark and carrot chips out of the net. (The net was angled about ten degrees from the vertical, and the objects dropped below the net, not onto it.) But the long wood sliver flipped over and got caught again. She recut the strings holding it, and then it did drop. She then went to the center of the net, jiggled it once more, and with nothing left in the net, she returned to her lair on the ceiling.

  Charlotte’s hunting territory attached to her home is like the forest a wolf knows near its den. The wolf hunts by smell and by sight, paying special attention to movement. The spider hunts by feeling the movements of prey, triangulating, then pinpointing the location of potential live and even dead “prey.” She unerringly returns from this hunting territory at any time directly back to her lair, the tiny home where she spends most of her time in the day, but comes out into her web and stays there full-time all night.

  May 22, 2011. Charlotte’s new web was now of similar size to last year’s; today it had twenty-seven spokes and forty-three rings. Not unexpectedly after her seven-month fast, she continued to have a huge appetite. She slammed down to catch a blowfly, leaving a blackfly on which she had been feeding up in her lair. She wrapped the blowfly and took it up and later came down to catch another one before finishing her feed. She wrapped this one, too, but left it in the net to go back up to her previous meal. But later she calmly walked down to retrieve the one she’d left hours earlier. In the evening, she resumed her now-preferred night-hunting position—the center of the net. (In five others of her kind and size that I kept in the cabin two years later, all came down within several minutes of each other at dusk and returned again as synchronously at dawn. Net building occurred only at night.)

  June 3, 2011. I gave her a moth, and whereas she had hauled up her most recent prey guided on her right leg, this time she used her left as I’d seen her do many times. She was ambidextrous and flexible, using her left hind leg, right hind leg, and occasionally both legs, and even her mouth, to haul tiny prey.

  June 6, 2011. At 8:05 a.m. I threw a dead blowfly into the net. By 10:21 a.m. it was still not taken, even though I jiggled it three times to try to mimic a struggling insect. Was Charlotte not hungry? A minute or two later I threw in a live moth, and she came down immediately and grabbed it, but in the process the dead fly fell out. I threw it back in and got the same non-response from her. It appeared that she did not want this fly. Maybe a different one? So I put a live deer fly into her web, which brought her down to hunt immediately. She jiggled the net with the two flies in it and now grabbed the dead, previously rejected blowfly, apparently making a wrong inference, namely, that the net movements by the live deer fly had come from the location of the dead blowfly that she had rejected. But she wasn’t duped totally: she calmly cut the big blowfly out of the net, dropped it, and then located and took the deer fly.

  By early July, Charlotte’s upkeep of her web declined. She hardly reacted to any insects I offered. I destroyed the whole ratty mess of her web on July 8 to find out what she would do then. During the next night she installed seven meter-long lines down from her lair and attached them to the top of a Coleman lantern that was then on the table below. They appeared to be spokes of a web. Three days later, she still had not added circular concentric strands to her web. I again removed the existing strands and the lantern. On July 23, the web was rebuilt, but only crudely. Nevertheless, it did catch a big bristly fly, and she processed it. After another three days, she again strung a series of about a dozen threads in a funnel-like formation from the ceiling by the window in a radius of about forty degrees, converging into each other and attached directly to my table. In the middle of them she placed a rough orb, but it was no more than thirty centimeters in diameter, rather than the usual four-times-larger size. It looked more like a sloppy maze than a web. Charlotte seemed to have lost her touch, or was she perhaps overfed and doing only the minimum to get by?

  In September, when the first red was already showing on the maples in the swamps and it rained day and night, Charlotte was in her web at night and up in her lair on the beam in the usual place in the daytime
. She took a grasshopper. I then tested her tastes (on September 30) by throwing her a stinkbug, wondering how she might react to defensive chemicals. No problem: she rolled and wrapped the bug in silk but then left it in place. It looked like a rejection, but later that evening she came back to it, hauled it up for her supper, and sucked it dry. On October 4, 2011, I offered her a Polistes wasp, which has a long flexible abdomen with a reach and a stinger at the end of it. She came right to it but kept it at “leg’s length,” and her legs were long. After a lot of maneuvering, she did manage to silk-wrap the wasp so that it became unable to flex its abdomen far enough to contact her, after which she approached its front end and leisurely bit into its head. Throughout this process, the wasp’s long and flexible abdomen kept trying to reach around to her. After she killed it, she started to transport it up into her lair, but then just kept the wasp where it was and fed from it there in her web, which had again started to look like a ratty mess. Two days later when she refused a syrphid fly, an otherwise never-refused prey, I took down the tangled web, hoping to watch her rebuild it that night. But she didn’t work on it.