Chapter 23: Most Toxic Insect Venom

W.L. Meyer
Department of Entomology & Nematology
University of Florida, Gainesville, FL  32611-0620
May 1, 1996

Insects in the order Hymenoptera were recorded as early as the 26th century BC as possessing a venom toxic to vertebrates. Harvester ants in the genus Pogonomyrmex have the most toxic venom based on mice LD50 values, with P. maricopa venom being the most toxic. The LD50 value for this species is 0.12 mg/kg injected intravenously in mice, equivalent to 12 stings killing a 2 kg (4.4 lb) rat. A Pogonomyrmex sp. sting produces intense pain in humans that lasts up to 4 hours.

A venom is a toxin that is injected into another organism using a specialized apparatus attached to a venom-producing gland. It may be used to immobilize or kill prey and/or to defend the delivering organism against attack by predators. Venomous insects are known from the orders Lepidoptera, Hemiptera, and Hymenoptera (Blum 1981). The method of delivery may be active, such as the sting apparatus of Hymenoptera (bees and wasps), and the mouthparts of Hemiptera (stylets), or passive such as the modified setae in some lepidopteran larvae (caterpillars) that are broken on contact and pierce the outer surface of the receiving organism.

Schmidt (1982) proposed that some insects in the orders Diptera, Neuroptera, and Coleoptera also possess oral venoms, but there is a problem with whether this constitutes a true venom or is a digestive fluid that is ejected. The biological activity of the venom can be classified as neurotoxic, hemolytic, digestive, hemorrhagic and algogenic (pain-producing). Venoms are chemically described as consisting of alkaloids, terpenes, polysaccharides, biogenic amines (such as histamine), organic acids (formic acid), and amino acids, but the majority are peptides and proteins (Schmidt 1986a; Blum 1981).

The first record of human death attributed to envenomation by a wasp or hornet was that of King Menes of Egypt in the 26th century B.C. (Waddell 1930). Toxicity of venoms is difficult to quantify in an unbiased manner and will vary among target species. It is also confounded by responses to the venom that are due to immune system disorders (such as hypersensitivity and allergies). For this reason, morbidity and mortality data may not be the best comparative method to classify venom toxicity (Schmidt 1986b). I will base my selection of the species of insect with the most toxic venom to vertebrates based on LD50 values using mice as the test organism.


Subscribing to the Entomo-L Listserv and posting a general inquiry about insect venoms was the most profitable first step in obtaining information about venomous insect species. Personal interviews with University of Florida and USDA-ARS staff provided often colorful information on people's "favorite" stinging bug. A wire story ("Killer Caterpillars," Gainesville Sun, January 16, 1996) apparently was widely distributed in newspapers and generated some discussion on the bulletin board. Searches on the WebLUIS Search System for information on literature in the University of Florida libraries retrieved some secondary literature such as the book by Blum (1981). Primary literature was identified using references obtained through Entomo-L replies and also by searching the AGRICOLA, Current Contents, and MEDLINE data bases available at University of Florida.


There were numerous insects suggested for the most toxic insect from personal interviews and the Entomo-L bulletin board replies, many of which were based on personal experience and descriptions of the reaction to being envenomed. Insects suggested included harvester ants (Pogonomyrmex; Hymenoptera: Formicidae), bees (Hymenoptera: Apidae), yellowjackets and hornets (Vespula, Dolichovespula; Hymenoptera: Vespidae), velvet ants (Hymenoptera: Mutillidae), puss caterpillars (Megalopyge opercularis; Lepidoptera: Megalopygidae), slug caterpillars (Sibene stimulea; Lepidoptera: Limacodidae), giant silkworm moth caterpillars (Lonomia sp. and Automeris io; Lepidoptera: Saturniidae) and assassin bugs (Rasahus sp.; Hemiptera: Reduviidae). However unpleasant the experience of being "stung" by ants, bees, wasps, and assassin bugs is, it is difficult to quantify pain responses objectively. Likewise, the perception of the toxicity or danger may be artificially inflated when death of humans or other vertebrates is the result of envenomation (Schmidt 1986b). LD50 values provide an unbiased method of comparing insect venoms.

Adult worker of Pogonomyrmex
Adult worker of Pogonomyrmex sp. [Photo courtesty of J.O. Schmidt].

Hymenopteran insects possess the most toxic venoms that have been characterized (Schmidt 1990; J.O. Schmidt personal communication). Table 1 lists the LD50 values for some of these insects that are known to most people, such as the honey bee, paper wasp, yellowjacket, velvet ant and harvester ants. The most toxic venom is found in a species of harvester ant, Pogonomyrmex maricopa with a mouse LD50 value of 0.12 mg/kg (Schmidt et al. 1989; J.O. Schmidt personal communication). Schmidt (1986a) states that for a 2 kg mammal only 12 stings are required to reach the LD50 dose. Other species of Pogonomyrmex also produce venoms with low LD50 values when compared with other Hymenoptera (Table 1).


Comparing LD50 values of a test organism (in this case, mice) can be a useful tool to objectively assess the toxicity of insect venoms; however, this method has its limitations. The values obtained in mice reveal a relative toxicity scale for different toxins in mice only. They do not reflect how the same toxins would rank for another species (such as humans). For example, the LD50 value of P. maricopa venom against a lizard, Phrynosoma cornutum, which is a predator of P. maricopa, was much higher than in mice (162 mg/kg). When one other lizard, Sceloporus jarrovii, was tested, the venom had an LD50 value of 28 mg/kg. These results suggest that P. cornutum has evolved resistance to the harvester ant venoms and can exploit the ants as a food resource (Schmidt et al. 1989). In another species of harvester ant, P. badius, there were high levels of an enzyme, phospholipase A2, which is also present in honey bee and wasp venoms (Schmidt & Blum 1978a). Although cross-reactivity to honey bee and wasp venoms may be involved in the response of humans to Pogonomyrmex envenomation, in those cases that have been studied cross-reactions to vespid and formicid venoms have not been found (Schmidt 1986b). Interestingly, the venom of P. badius is not particularly lethal against larval insects (Schmidt & Blum 1978b). Since harvester ants are non-predatory, it suggests that their venom has evolved from being used in prey capture as in other ant species (Schmidt 1986a), to defense against vertebrates; hence their power against humans and other vertebrates.


I thank Dr. Justin Schmidt (Southwestern Biological Institute, Tucson, AZ) for his suggestions on the most toxic insect, for providing a photograph, and for insights into the literature. I also thank Dr. Thomas Walker (University of Florida, Gainesville), Dr. Antonio CastiZeiras and Adrian Hunsberger, (University of Florida, Homestead), and Dr. Nancy Epsky (USDA-ARS, Gainesville) for critical reviews of the manuscript. Suggestions from two anonymous reviewers also improved the manuscript.

Table 1

LD50 values in mice for toxins found in Hymenoptera.

Family Species Common Name LD50 (mg/kg) Reference
Apidae Apis mellifera honey bee 2.8 Schmidt 1990
Mutillidae Dasymutilla klugii velvet ant 71 Schmidt et al. 1980
Vespidae Polistes canadensis paper wasp 2.4 Schmidt 1990
Vespidae Vespula squamosa yellowjacket 3.5 Schmidt et al. 1980
Formicidae Pogonomyrmex spp.* harvester ants 0.66 Schmidt 1990
Formicidae P. maricopa harvester ant 0.12 Schmidt et al. 1989

References Cited

  • Blum, M.S. 1981. Chemical defenses in arthropods. Academic Press. New York. [562 p.].
  • Cohen, S.G. & P.J. Bianchine. 1995. Hymenoptera, hypersensitivity, and history: a prologue to current day concepts and practices in the diagnosis, treatment, and prevention of insect sting allergy. Ann. Allergy Asthma Immunol. 74: 198-217.
  • Schmidt, J.O. 1982. Biochemistry of insect venoms. Annu. Rev. Entomol. 27: 339-368.
  • Schmidt, J.O. 1986a. Chemistry, pharmacology and chemical ecology of ant venoms, pp. 425-508. In T. Piek [ed.], Venoms of the hymenoptera. Academic Press, London.
  • Schmidt, J.O. 1986b. Allergy to hymenopteran venoms, pp. 509-546. In T. Piek [ed.], Venoms of the hymenoptera. Academic Press, London.
  • Schmidt, J.O. 1990. Hymenopteran venoms: Striving towards the ultimate defense against vertebrates, pp. 387-419. In D. L. Evans & J.O. Schmidt [eds.], Insect defenses: adaptive mechanisms and strategies of prey and predators. SUNY Press, Albany, NY.
  • Schmidt, J.O. & M.S. Blum. 1978a. The biochemical constituents of the venom of the harvester ant, Pogonomyrmex badius. Comp. Biochem. Physiol. 61C: 239-247.
  • Schmidt, J.O. & M.S. Blum. 1978b. Pharmacological and toxicological properties of the harvester ant, Pogonomyrmex badius, venom. Toxicon 16: 645-651.
  • Schmidt, J.O., M.S. Blum, & W.L. Overal. 1980. Comparative lethality of venoms from stinging Hymenoptera. Toxicon 18: 469-474.
  • Schmidt, P.J., W.C. Sherbrooke, & J.O. Schmidt. 1989. The detoxification of ant (Pogonomyrmex) venom by a blood factor in horned lizards (Phrynosoma). Copeia 1989: 603-607.
  • Waddell, L.A. 1930. Egyptian civilization. Summerian origin and real chronology. Luzac, London. [pp. 60-71]. [Not seen, cited in Cohen & Bianchine 1995, p. 201].

Return to top