Return to Contents




This curious world which we inhabit

is more wonderful than it is convenient;

more beautiful than it is useful;

it is more to be admired and enjoyed

than used.

Thoreau, 1837 


The possibilities of utilizing the trap-nest technique while investigating the life history, behavior, and associates of solitary wood-nesting predaceous wasps and bees were brought to my attention by my esteemed friend Kenneth W. Cooper, now of the Dartmouth Medical School. His elaborate investigations of the bionomics of the vespid wasp Ancistrocerus antilope and its intricate relationships with the symbiotic mite Kennethiella trisetosa, and his ingenious and exquisitely designed set of ex­periments to investigate the orientation of larvae in the borings at the time of cocooning, have been an inspiration and challenge to all who have worked subsequently with this technique.

My study began in a modest way during the years 1953 and 1954 with a relative handful of some 70 nests. During the period 1955-1962 the number of traps set out in the field in­creased rapidly as the success of the technique was realized and as several collaborators were added. This activity reached a peak in 1961 with some 1,400 traps set out and a total of 9S6 nests received and processed. During 1963 and 1964 the project was sharply reduced with relatively few traps being set out (only in the Washington area and in Arizona) to ascertain some details not obtained in earlier nests.

Our numerous American solitary wood-nesting wasps and bees have not been the subject of previous intensive investiga­tion except locally by the Pcckhams in Wisconsin, by Medlen and his associates in the same State, by Cooper in New York, and by the Raus in Missouri. A long-term study such as mine, con­ducted in several different faunal zones, could be expected to yield a great amount of information. Such was the case. During my study I had the opportunity to examine and record the nest architecture and life-history data from approximately 3,400 nests from western New York, the metropolitan area of Washington, D. C; coastal North Carolina, south-central Florida, and the desert floor of southeastern and south-central Arizona. In addition, K. W. Cooper sent me a few nests from Rochester, N. Y., Oak Ridge, Tenn., and Gainesville, Fla.

Many other ecological niches and several faunal zones still await similar investigation.

Return to Contents


The several areas represented in this study have yielded important and largely unknown data on 75 different predaceous wasps and 43 nonparasitic bees. In addition, host relationships and life-history data were obtained for 83 parasites and predators associated with the host wasps and bees.

The information secured on nest architecture was quite de-tailed and thorough for almost all the nests. Regretfully, I can-not report the same for the life-history data on some of the wasps and bees and on the prey of some wasps. Two factors ac-count for the lack of information: First, occupants of nests mailed to me by collaborators had usually entered the prepupal stage by the time the nests reached me, so that I could not se-cure data on the duration of the egg and larval feeding stages or on the prey stored by the wasps except where there had been egg mortality in one or more cells; second, in the more active years of the project the sheer number of nests obtained was so great that most of my available time had to be devoted to initial processing of the nests and the killing, pinning, and labeling of the emerging wasps, bees, and parasites. 

A partial measure of the success of this project is afforded when one realizes that nesting and life-history data were published previously for only about a third of the wasps and bees, and their parasites and predators. Furthermore, a great deal of the previously published information consisted of scattered notes on fragments of the life history or prey preferences, which did not afford detailed knowledge on many facets of the bio­nomics of the species.

Some unexpected extra dividends accruing during the study were the new species and subspecies of wasps, bees, and parasites described from reared material. Among the wasps, it was necessary to describe no less than five new species from the Washington area alone; an additional new species was described from coastal North Carolina, and two new subspecies from southern Florida. A new subspecies of the common eastern carpenter bee was described from southern Florida. Among the wasp and bee parasites two new species of chalcidoid wasps were de-scribed, one species from the Washington area and one from Arizona; an additional new species from Arizona remains to be described. An amazing number of new mites, mostly symbiotic species, was encountered. It was necessary for E. W. Baker to describe no fewer than 2 new genera and 13 new species from these nests; and 4 additional species still await description! 

First instar larvae of the parasitic cuckoo wasps (Chrysididae) were preserved and also mature larvae of many other species of the solitary, free-living, and parasitic wasps and bees. Larvae of a number of species of Sphecidae, most of them not available from other sources, represented a substantial number of the species treated in H. E. Evans's recent monographic studies of the larvae of this family. The larvae from other families of wasps and bees still await taxonomic study. 

From the standpoint of host-predator relationships, this study emphasized again the considerable pressure exerted on populations of other arthropods by the predaceous solitary wasps. Consider, for example, the pressure exerted on the snare-building spider population from predation by such a wide-ranging sphecid as Trypargilum t. tridentatum. It stored an average of 23 spiders per cell, an average of 6 to 9 cells per nest depending on the locality; and I studied only a few more than a hundred nests from the tremendous total nest production of this wasp throughout its range. As another example, the Floridian vespid Euodynerus foraminatus apopkensis, which stored an average of 10 caterpillars of an unidentified species of Olethreutidae per cell, made an average of 8 cells per completed nest. I had nearly 250 nests of this species. 

Much of the information on nesting, prey preferences, and life history details will be useful eventually in helping to define the biological criteria which characterize the species and genera. One of the most fascinating developments in this area was the discovery that larvae of Trypargilum constructed such distinc­tive cocoons that the five North American species nesting in the traps could be identified from this feature alone, and this in spite of the finding by Evans that larvae of these same species were separable only with difficulty and on very trivial morpho­logical criteria. Furthermore, comparison of the prey prefer­ences of the three species of this genus in the Washington area showed that each had differences in this respect, one selecting entirely snare-building spiders, one about 90 percent snare-building and 10 percent wandering spiders, and the third only 20 percent snare-building and 80 percent wandering spiders. These differing percentages suggest that each species hunts in a different manner, one taking spiders directly from their webs, the second frightening them from the webs and seizing them on the ground where they also get a small number of wandering spiders, and the third hunting for its spiders on leaf litter or on vegetation where they obtain more wandering species. 

Another interesting taxonomic finding was the determination that Ancistrocerus catskill catskill and A. catskill albophaleratus, hitherto considered to be discrete subspecies, were actually only color phases of the same taxon which occurred together in 13 percent of the nests, with the typical yellow-marked form oc­curring alone in 60 percent and the white-marked form alone in 27 percent of the nests. 

In nests of the sphecid wasps belonging to Isodontia, subgenus Murrayella, I found a transition from elegans, which constructed individual cells separated from each other by thick partitions of grass stems and vegetable fiber, through mexicana, which sometimes made individual cells separated by flimsier, narrower partitions or else made just a larger brood cell in which several larvae developed without cannibalism, to auripes, in which there was always just a single large brood chamber. 

I discovered that the bee Megachile (Sayapis) policaris Say usually stored a large mass of pollen and nectar in a brood cell and laid eggs at intervals in pockets in this mass. These bee larvae also developed without cannibalism. This is the first megachilid bee known to construct such a brood chamber.

In my study of the vespid wasps I determined that species be-longing to genera primarily Holarctic in distribution, such as Ancistrocerus and Symmorphus, had shorter pupal periods than species of genera such as Pachodynerus, Monobia, Euodynerus, and Stenodynerus, which are primarily Austral or Neotropical in distribution. 

This study also disclosed a number of previously unknown facts about life history, behavior, and host associations of a great many parasites and predators. 

Almost all the mite-host associations were unknown previously. Two hypopial forms were discovered in the chaetodactylid mite Chaetodactylus krombeini, an apodous, encysted, over-wintering form which developed within the skin of the first nymphal stage and the ordinary, 8-legged active form which developed from the apodous encysted form. The occurrence of an apodous encysted form was considered to be of profound evolutionary significance because the presence of encysted forms in an abandoned burrow provided the mite species with an opportunity to parasitize other species of bees which also nest in abandoned borings.

Considerable evidence was gathered which established a presumptive association of the dermestid beetle Trogoderma orn-tum with spider-storing wasps of the genus Trypargilum. The newly hatched beetle larva appeared to be predatory on the wasp eggs or young larvae, and later larval instars fed on prey stored for the wasp larva or on wasp prepupae in adjacent cells.

A high parasitism rate of Euodyucrus foraminatus apopkensis by the stylopid beetle Pseudoxenos hookeri provided abundant material which enabled me to determine that the beetles do not exsert from the abdomen of the host until several days after eclosion of the adult wasps. This is contrary to the conjecture of earlier workers who supposed that the stylopids exserted while the host wasps were in the pupal stage. Furthermore, I determined that penetration of the host egg by the triungulinid larvae of the stylopid might be the cause of the abnormally high mortality rate of eggs of this wasp. 

Among the Bombyliidae Lepidophora lepidocera was unique in that the larva usually developed on the prey stored for the host wasp larva rather than on the resting wasp larva or pupa, as is normal in other genera attacking wasps and bees. Also, the Lepidophora larva usually had to feed on the contents of several cells in a series to develop to maturity. 

Associations of parasitic Coelioxys bees with their megachilid hosts permitted me to hypothesize that the shiny black Coelioxys species, of which dolichos is the only North American representative, probably parasitize species of Megachile, subgenus Melanosarus. Another such hypothetical association is that of the group of Coelioxys species having an upturned spicule at the apex of the last visible abdominal tergum with resin-using species of Chalicodoma, subgenus Chelostomoides. 

A final result of this project has been the acquisition for the collection of the U.S. National Museum of many thousands of specimens of wasps, bees, and their parasites in perfect condition. Some species were not represented previously and of others there were only a few specimens or series in relatively poor condition. Also preserved in the Museum collection are several hundred specimens of wasp and bee larvae for future taxonomic study, as well as several thousands of specimens of prey. All the reared and associated material bear identifying year, nest, and cell numbers so that they may always be associated with the original nest notes also on deposit in the Museum files.

Return to Contents 


During the many years of these studies, I have had the appreciated cooperation of several collaborators in the field, who set out empty traps, ran the trap lines on a weekly or biweekly basis, and mailed the completed nests to me for study. My father, Louis H. Krombein, M.D., set out traps near his home at Derby, Erie County, N. Y., from 1956 to 1961. Richard Archbold, director of the Archbold Biological Station, Lake Placid, High-lands County, Fla., ran similar series at the station during 1957 and 1959 to 1962. Mont W. Cazier, during his term as director of the Southwestern Research Station, Portal, Cochise County, Ariz., ran trap lines on the desert floor near Portal in 1960 and 1961; in 1963, while at Arizona State University, he set out traps at Granite Reef Dam, Maricopa County, and again near Portal. Richard S. Beal set out traps in several localities in Arizona in 1957 while he was with the U.S. Department of Agriculture; in 1961, when he was at Arizona State University, he ran trap lines at Scottsdale, Granite Reef Dam, and Molino Camp in the Santa Catalina Mountains. The valued cooperation of these men enabled me to obtain life-history data on a number of species for which I would otherwise have had no information.

Identifications of prey and parasites required the assistance of many specialists. Particular mention should be made of the help of two individuals who identified not hundreds but thousands of specimens of prey, almost entirely immatures and frequently in less than perfect condition. These are my colleague Hahn W. Capps, U.S. Department of Agriculture, who identified all the lepidopterous larvae used as prey by vespid wasps, and Benjamin J. Kaston, Teachers College of Connecticut, who determined almost all the spiders used as prey by Trypoxylon, Trypargilum, Dipogon, and Auplopus. My friend Edward W. Baker, U.S. Department of Agriculture, not only identified all the Acarina but also determined that almost all of them required description as new genera and species, a task which he also performed. Other specialists who furnished identifications of prey or parasites are as follows: My colleagues, past and present, in the Insect Identification and Parasite Introduction Research Branch, U.S. Department of Agriculture, D. M. Anderson and W. H. Anderson (coleopterous larvae), P. H. Arnaud (Bombyliidae), R. S. Beal (Dermestidae), B. D. Burks (Chalcidoidea), A. B. Gurney (Orthoptera), C. F. W. Muesebeck (Braconidae), K. O'Neill (Thysanoptera), L. M. Russell (Aphidae), C. W. Sabrosky (Milichiidae, Conopidae, Tachinidae), M. R. Smith (Formicidae), T. J. Spilman, (Rhipiphoridae), G., B. Vogt (Chrysomelidae), L. M. Walkley (Ichneumonidae), D. M. Weisman (Chrysomelidae), and W. W. Wirth (Bombyliidae, Phoridae); R. M. Bohart, University of California at Davis (Stylopidae); W. L. Downes, Jr., University of Illinois (Sarcophagidae); W. R. Enns, University of Missouri (Meloidae); W. J. Gertsch and W. Ivie, American Museum of Natural History (Araneae); C. G. Jackson, University of Richmond (Acarina); N. Marston, University of Kansas (Bombyliidae); R. B. Selander, University of Illinois (Meloidae, larvae); H. K. Townes, American Entomological Institute (Ichneumonidae); and H. K. Wallace, University of Florida (Araneae). P. W. Martin, Geochronology Institute of the University of Arizona, identified pollen from the nests of some southwestern bees. The identifications of wasps and bees are for the most part my own, although I have had help on a few difficult or puzzling species from the following: R. M. Bohart (Chrysididae, Vespidae); W. E. Ferguson, San Jose State College (Mutillidae); P. D. Hurd, Jr., University of California at Berkeley (Xylocopidae); A. Menke, University of California at Davis (Sphecidae); C.D. Michener, University of Kansas (Megachilidae); T. B. Mitchell, North Carolina State University (Megachilidae); and P. H. Timberlake, Citrus Experiment Station, California (Megachilidae).

Howard E. Evans, Museum of Comparative Zoology at Har­vard University, kindly sent me a few spider prey records from nests of several species of Trypargilum and Trypoxylon.

C. D. F. Miller, Canadian Department of Agriculture, courteously made available for my examination many of the wasps reared and reported on by R. E. Fye (1965a). Prior to 1965 I had seen only a few specimens of Euodynerus leucomelas (Saussure) and of Passaloecus ithacae Krombein from Fye's material.

I am grateful to Paul W. Oman and William H. Anderson, past and present chiefs of the Insect Identification and Parasite Introduction Research Branch, U.S. Department of Agriculture, for their support of this project during my employment with the Department, 1941-1965.

A special note of acknowledgment is due J. F. Gates Clarke, former chairman of the Department of Entomology, U.S. National Museum, for arranging the fabrication of some 3,500 traps in the cabinet shop of the Museum.

A generous grant from the American Philosophical Society enabled me to visit coastal North Carolina three times during the summer of 1955 to set out traps and retrieve the completed nests. A second grant from the Society permitted me to visit Arizona to study trap- and ground-nesting wasps and bees during the summer of 1959.

A final note of deep appreciation is due my wife for her enthusiastic interest in these life-history studies. Her patient and critical reading of the many manuscript pages in draft form and her cooperation in helping to check the final proofs have contributed substantially to the quality of the final report.

Return to Contents

Posted August 6, 2009