Rumor had it that the Georgia species of harvester ant had a mellow, hospitable, Southern disposition compared with some wild western species, so we braced for what might come. Northern Louisiana is the eastern limit of the Comanche harvester ant, a generally unaggressive species that curiously sting-autotomizes; that is, it leaves its sting in human flesh, as do honey bees. The sting also hurts, like the sting of the Florida harvester ants in Georgia, only the pain lasts longer. In Texas, and fitting its proclaimed reputation for big, we ran into a larger harvester ant. This species, Pogonomyrmex barbatus, called the Texas agricultural ant by H. C. McCook, the famous popularizer of nature around the late 1880s, is also called the “red harvester ant,” an essentially meaningless name because, with a few exceptions, all harvester ants are red. This species builds impressive nests, with an entrance hole in the middle of large circles of barren earth. These formicid engineering experts clear and maintain the bare areas. Their size and color belie their true prowess as purveyors of pain. Though not fakes, their stings hurt less than those of the Maricopa harvester ant, a smaller and more delicate species. Their stings not only hurt less, the pain lasts a shorter time, and they do not sting-autotomize.
In the small and delightfully charming town of Willcox in southeastern Arizona, we found Maricopa harvester ants. These ants were the most impressive species of harvesters of the trip. They dominated low-stabilized sand dunes around the Willcox Playa, a usually dry lake in a small basin with no outflowing stream or river, a mini Great Salt Lake Basin, so to speak, only lacking the salt. Perhaps because of the high water table near the lake, the Maricopa harvester ants build enormous mounded ant castles for the 20,000-plus ants in the colony. Except during termite swarming times, these Willcox ants are placid collectors of seeds and generally do not readily sting. Swarming termites can be viewed as mobile “seeds,” packed with protein and fat and much easier to eat than hard dry seeds. The ants dramatically shift behavior during this time and become avid predators. Sandal wearing is not recommended. Don’t let the delicate, lithe body shape or unassuming demeanor of Maricopa harvester ants fool you. The stings of these ants really hurt. The throbbing pain can last 8 hours, decreases only slowly, and the ants readily autotomize their stings into humans or other unfortunate animals. These ants were the most painful stingers we encountered on the summer’s trip. To add veracity to their message, the venom of the ants at this particular location is the most toxic known ant, wasp, or bee venom, some 25 times more toxic than honey bee venom and 35 times greater than western diamondback rattlesnake venom.
WHY DO DEFENSIVE STINGS of venomous insects hurt so much? Why should some defensive stings be toxic, much less highly toxic? After all, isn’t the insect served simply by making the attacker release the insect and abort the attack? A first hint that points to an answer is that some insect venoms are highly toxic. High toxicity evolved independently many times in ants, wasps, and bees. Repetitive evolution of a similar property, especially when the molecules responsible for that property are different, indicates some function, not just random “mistakes” of nature. What possible function could venom toxicity have? The question becomes especially poignant when considering that the toxicity of harvester ant venom is 800 times more toxic to some predators than to an insect prey. The solution to this conundrum is revealed in the words “truth in advertising.” Pain is an advertisement that damage has occurred, is occurring, or is about to occur. Without enforcement, advertisement becomes a slick system of lies. Intelligent animals can see through lies, or learn to see lies as what they are, and the advertisement loses meaning. In the insect sting system, pain is the advertisement, and toxicity is the truth. Toxicity is truth because it is real damage or death. The toxicity truth becomes especially important for small vertebrate predators that are more susceptible to damage than large predators. Without toxicity, a smart predator learns the dishonesty of sting pain and can ignore the signal. When this occurs, for example, beekeepers learning that dozens of stings pose no real physical threat, continue to rob beehives; the stinging insects lose. In the case of a 20-gram shrew or mouse, in which four honey bee stings can be lethal, the message of sting damage rings crystal clear. Thus, a gradient of effectiveness of sting toxicity spreads across the predator field, helping stinging insects survive attacks by some predators more than others, overall providing a net benefit in the game of life. Even predators as large as a 50-kilogram (110-pound) beekeeper are at risk of death from 1,000 honey bee stings.1 Combined damage and lethality are crucial to the long-term evolutionary effectiveness of insect stings against intelligent predators.
The venom constituents that cause pain and those that are damaging or lethal are not necessarily the same. Selection pressure for pain came first. This we surmise based on its immediacy as a defense and the presence of painful stings in present-day wasps closely related taxonomically to the stinging ancestors. Examples of painfully stinging species whose venoms are essentially not toxic include some large ichneumon wasps (Megarhyssa) that are parasitic on wood-boring sawfly larvae, bethylid wasps, solitary parasitoids (insects whose young are parasites that eventually kill their hosts) of beetle larvae, velvet ants, solitary parasitoids of bees and wasps, and spider wasps, solitary parasitoids of spiders. The chemical nature of the painful venom components of these wasps with simple life histories is mostly guesswork, but likely, each type of wasp has a different chemical or set of chemicals responsible for the pain. The pain-inducing components for velvet ants include at least the biogenic amine serotonin (5-hydroxytryptamine), a known algogen when injected beneath the skin. Serotonin is also a pain-inducing component in a wide variety of social wasp venoms. Histamine, another biogenic amine, is widely present in venoms of yellowjacket wasps, paper wasps, hornets, honey bees, and some ants. Histamine primarily causes vasodilation of blood vessels resulting in swelling, warmth, redness, and some itching. It does not cause sharp pain. In that regard, histamine is not a strong agent of pain. Acetylcholine, a third biogenic amine, does cause sharp pain and is found only in hornets. These small molecules are not the important direct inducers of pain in insect venoms; that role falls to a variety of small peptides whose structures vary strikingly from one group to another. In honey bees, the painful component is melittin, a 26 amino acid peptide containing five basic amino acids. In wasps, the kinins, 9 to 18 amino acid peptides that cause heart pain among other activities, cause the intense burning pain. Harvester ant venoms contain barbatolysin, a 34 amino acid peptide that appears to cause pain. The pain-inducing agents in the various ant venoms are not known, though some species contain their own ant versions of kinins that likely cause pain and a wide variety of other peptides.2
Damage-causing venom components evolved subsequent to the early pain-producing agents. In most venoms, the identities of the actual toxic components are not known. In the best-studied insect venom, that of the honey bee, the most toxic component is the enzyme phospholipase A2, which causes no skin pain. The more abundant but less lethal component, melittin, is the other important toxin. Melittin is a heart poison that also causes hot, burning pain by destroying cell membranes, including those of nerves. We recently isolated and are characterizing a lethal component from harvester ant venom that induces all the skin and pain reactions of a sting.
The Pain Scale
AFTER WE RETURNED from the trip west with a car burdened with buckets of ants, urgent immediate and long-term questions arose. The immediate, and less-interesting, question was what to do with the ants. This was a less interesting question because the answer lay in dissecting and collecting massive quantities of venom for drying and freezing for future work. Massive quantities meant 5 milligrams or more, about of a teaspoon, of each type of harvester ant collected. At 40 ants needed to yield 1 milligram of venom, and at about 3 minutes to dissect each ant, gold is cheaper than harvester ant venom.
Long-term questions converged on determining the value of the venoms to the stinging insects themselves. The venoms evolved for t
he benefit of the insects, not for humans. What were these benefits, and how did they change the lives and biologies of the insects? To help answer these questions, the properties of the stings and venoms needed to be evaluated. Pain and toxicity are the two basic properties of each sting. To test hypotheses regarding pain and toxicity, each venom needed to be compared with the venom of other stinging insects. Then the lives of each species were compared to determine whether venom properties correlate with life histories. For toxicity comparisons, a variety of physiological and toxicological methods are available, each yielding a numerical value that can be compared to the numerical value of other venoms. In principle, toxicity comparisons are simple. But what about pain comparisons? There were no physiological or pharmacological methods to place accurate values on pain. Even today we lack reliable methods to insert electrodes into nerves or parts of the brain to measure pain, and then to understand the meaning of the electrode recordings. Likewise, interpreting the results of more advanced brain-scanning techniques vis-à-vis pain is unclear, though great progress is being made. Someday, we hope to measure pain quantitatively and cheaply. In the meantime, how could sting pain be measured numerically? What was the solution?
A pain scale was needed. The answer was simple; however, making the scale was not. A useful scale had to be reliable, reproducible, and indexable. There was precedent for a pain scale, albeit one designed for measuring human chronic pain. The McGill Pain Questionnaire, developed by Ron Melzack at McGill University primarily to measure chronic pain in patients, consisted of rating pain levels derived from patient questionnaires and caregiver evaluations of facial and body language.
Insect stings cause short-term, mainly ephemeral pain and a variety of nuances that relate to the person stung and the stinging insect. Sting pain induced by a single sting can vary depending on how much venom the sting delivered, where on the body the sting occurred (for example, stings to the nose, lips, or palms of hands hurt considerably more than stings to lower legs, arms, or top of the skull), the age of the insect, the time of day the sting was received, and other factors, including the sensitivity to pain of the individual. For reasons of consistency and reliability between different evaluations under different circumstances, only a few numbers were used in the sting pain scale. The scale ranges from values of 1 to 4 and is anchored by the value of a single honey bee sting (Apis mellifera), defined as pain level 2. The honey bee is a convenient reference point because honey bees exist nearly worldwide, are abundant, most people have been stung by a honey bee, and they are about midway within the range of pain intensities produced by wasp, ant, and bee stings. Also included on the scale is a trivial value of 0 for stinging insects incapable of penetrating human skin but possibly able to sting other animals. The criteria distinguishing between pain levels are that the pain of the lower level is substantially less than the pain in the upper level and that the evaluating person would clearly know that one sting hurt more than the other. When comparing species, the evaluator compares the current sting pain with memory of the pain of a previous sting by a honey bee or other species for which the pain was rated previously. In some cases, values halfway between whole numbers are assigned in which the pain appears distinctly greater than the lower level, yet distinctly less than the higher level. This evaluation system works remarkably well as witnessed by nearly identical ratings for stings by various colleagues. Chris Starr, a fellow graduate student colleague at the University of Georgia, and I spent innumerable hours discussing the topic. We also discussed ideas widely among others from the hotbed of Hymenoptera (ant, wasp, and bee) researchers at Georgia. Our primary goal was to evaluate the pain scale for accuracy and reliability. Stings of many different species have the same numerical value; this does not imply that they are identical in feeling, but that they fall into the same general range of painfulness and presumed effectiveness as predation deterrents. Pain that arises at, or near, the sting site hours or days after the initial sting pain has receded is not considered for this pain scale because it is caused by immunological or physiological reactions to the venom or its damage.
Once it was developed, the pain scale opened possibilities to delve into the secrets of the lives of stinging insects and to predict how their weaponry opened opportunities for them. Predictions operated both ways—we could predict the sting pain based on the appearance, behavior, and life history of a given insect, or we could predict lifestyles on the basis of the sting pain. For example, colorful solitary wasps and bees would be expected to pack a more painful wallop compared with more drab wasps and bees. The reasoning goes that, in evolving a colorful appearance, the option of inconspicuousness, a primary defense masterfully employed by most insects, was largely abandoned. Why should this time-tested defense be abandoned? Perhaps because the insect’s life history requires it. The cow killer, Dasymutilla occidentalis, illustrates the problem. Cow killers reproduce by locating the nests of other large wasps, entering their nests, and laying eggs in the cells of developing host wasps. The cow killer larva then feeds on the host to complete the life cycle. The problem to overcome is finding enough suitable hosts in the environment in which hosts are rare and usually sporadically distributed. The cow killer must devote much of its daytime activity to searching for these hosts. To make matters worse, female cow killers are wingless and must crawl to locate their hosts. As a result, cow killers are very long lived for an insect, living an entire summer to up to one and a half years. During this long life, cow killers actively crawl around during daytime in plain view of a host of lizards, birds, and other large predators. What chance would a tasty and undefended cockroach, cricket, or caterpillar have surviving for a season or a year under these conditions? Not much chance; those species likely would go extinct. Cow killers are far from extinct: ask any rural Southern inhabitant in the United States. We would predict that cow killers’ survival depends on delivering a powerfully painful sting. Indeed, they do, as witnessed by my student who got careless while feeding cow killers and ended up in the university student infirmary. Cow killer stings are not simply an attention-grabbing 2 on the pain scale, they are a solid, unforgettable 3.
In the Arizona Sonoran Desert, we witness amazing explosions of cactus bees, Diadasia rinconis. These drab grayish-brown honey bee–sized bees nest in huge aggregations, often by the tens of thousands, and race around collecting pollen from the abundant yellow, red, or magenta blossoms of various prickly pear and cholla cacti. They are hard to spot and blend effectively with their environment as they suddenly appear in a flower, only to be gone in the blink of an eye. Birds and other predators have difficulty seeing, tracking, and catching these cryptic lightning flashes. Their other effective defense is a short life span. They live only the few weeks during the cactus bloom. They need to evade predators only a short time, not months or a year as does the cow killer. Given this life history, we would predict that their stings are not really needed and would not be especially painful or effective. I know of no one who has ever been stung by a cactus bee, with the exception of entomologists who inadvertently pinch one between an insect net and vial while trying to capture it. Even then, it is hard to get stung. Many entomologists simply reach into the net, grab the bee, and, plop—into a jar it goes. Steve Buchmann, a leading expert on cactus bees who has earned the moniker “Buzzmann,” reports that the sting is pretty trivial, a 1 on the pain scale, hardly meriting discussion. My own experiences were similar to Steve’s. When I tried to cram scores of netted bees into a wide-mouth jar for venom collection, I received a couple of stings to the side of my index finger. A sharp but mild 1 on the pain scale.
A sting rating 4 on the pain scale is something to be avoided if possible: level-4 pain takes command of one’s body and sensory system, shutting down most self-control in the process. Excruciating is an understatement. Fortunately, few insects deliver level-4 stings. This level of sting pain will be described in more detail in the chapters on tarantula hawk wasps and bullet ants.
As more typ
es of stinging insects were investigated, more patterns fell into place. Sting pain, it appeared, really did affect the lives of stinging insects. It affected their lives because of the effect on both actual and potential predators. Predators, parasites, and diseases are the main driving forces in the lives of any animal. According to W. D. “Bill” Hamilton, the great twentieth-century English theorist and naturalist, predators, parasites, and diseases, along with variable environments, are responsible for the evolution of sex.3 Maybe we should be grateful to our predators and parasites. To an extent that they could think in such terms, stinging insects also should be paradoxically grateful to their predators. Without predators, many opportunities afforded stinging insects would disappear, inherited instead by other adventitious, nonstinging species. Predators opened niches in biological opportunities to those tough enough to exploit them, particularly stinging insects.
Insects, such as cow killers, can live openly and conspicuously because of their painful and effective sting. Stings are not the only solution for insects with conspicuous behaviors, but their stings are effective. Blister beetles, including the Spanish fly beetle, employ another solution. They produce deadly cantharidin, a chemical that blisters skin, mouth, and stomach on contact. Furthermore, blister beetles accelerate the delivery of cantharidin through reflexive bleeding in which beetle blood laced with cantharidin is hemorrhaged through preweakened membranes on the body. Most predators that taste blister beetle blood get the message quickly and reject the beetle unharmed. However the message is delivered—sting, toxic blood, or otherwise, life is always better if tangible message delivery can be avoided. Throughout the animal world, strong, dominant individuals communicate subtle and not-so-subtle messages to weaker individuals not to challenge.
The Sting of the Wild Page 6