Return to Contents


The sequence of sexes in multicelled nests is an important characteristic of species and even of genera and families. Insofar as possible I gave particular attention to determining this arrangement in all nests. There was no difficulty in determining the individual sexes from pupae in nests of the Vespidae, where the cocoons were so delicate that they usually split when the nest was opened. In nests of the other families, where the cocoons were usually opaque, and brittle or tougher, it was possible to make a large enough rip posteriorly in the cocoon to determine the sex. However, I usually put each cocoon of this type in a small glass vial with the associated nest and cell numbers. The sex of each wasp or bee could then be recorded when the adults emerged without the possibility of injury to the pupa by making a hole in the cocoon.

This phase of the study confirmed the already established fact that, in nests of Vespidae in which both sexes develop, females are always in the cells in the inner end of the boring and males are in the cells in the outer end. The deviations from this arrangement were so rare that I suspected them to be occasioned by failure of the sperm to fertilize an egg with consequent development of a male wasp in a cell in which a female should have developed, or that the original nestmaker had been superseded by another female of the same species.

    This pattern of females in the inner and males in the outer cells was usually found also in nests of the megachilid bees Ashmeadiella, Prochelostoma, Osmia, Chalicodoma, and the leaf-cutting species of Megachile. In 51 Osmia lignaria nests females were in the inner and males were in the outer cells in 75 percent of the nests. Under the specific heading I include an extended discussion of the arrangement in the other dozen nests in which I point out that the deviations from the normal sex sequence apparently fell into three groups. An occasional male cell in a sequence of female cells might have been caused by failure of the sperm to fertilize the egg; this hypothesis received some substantiation by the finding that these male cells actually were longer than normal male cells and that they had a larger store of food than was normal for male cells. In other nests, where there was a block of normal size male cells at the inner end or in the middle of the nest, their occurrence out of order might have been caused by the temporary "fatigue" of the muscles controlling egress of sperm from the spermetheca thus inhibiting release of sperm and resulting in deposition of unfertilized eggs. A third factor which perhaps accounted for the disarranged sequences in some nests might have been supersedure of the original nesting mother by another female of the same species, so that a sequence which started out as --- or --  might end up as ----♀-♀- or    ---♀-♀.

The megachilid bee Anthidium maculosum reversed the usual sequence by having males in the inner and females in the outer cells in all nests in which both sexes developed except in one where one male was out of place.

Wasps belonging to the genus Trypargilum demonstrated that sex sequence is an important taxonomic factor at the specific level. Except in a very few nests males were in the inner and females in the outer cells in both subspecies of tridentatum and in striatum. Females were in the inner and males in the outer cells in both subspecies of collinum in almost all nests.

    In the two closely related species johannis and clavatum the situation was confused. Females were in the inner and males in the outer cells in five mixed nests of johannis, the reverse was true in two nests, and there was a random arrangement in three nests. Females were in the inner and males in the outer cells in five nests of clavatum, the reverse sequence occurred in four nests, and a random arrangement was found in four nests.

Pompilid wasps belonging to the genus Dipogon usually had a random sequence of sexes in their nests. It is possible that the size of the single spider placed in each cell was the factor which determined whether a female or male egg was laid.

Of other genera insufficient nests were available to permit any generalizations regarding sequence of sexes. I have mentioned the sequence under the specific treatments whenever any data were available, so that conclusions may be drawn as to the probable pattern when more nests are at hand.

This is the appropriate place to mention a phenomenon which could be inferred from the previous discussion. It is the ability of the female wasp or bee to know in advance, before storing the cell with food, the sex of the egg which she will place in the cell prior to sealing it. With the Vespidae, the only family in which the egg is laid before the cell is provisioned, the wasp still controls and knows the sex of the egg she has laid, because she brings in a large number of caterpillars if it is to be a female or a small number if it is to be a male.

In the wasps and bees males develop from unfertilized eggs and females from fertilized eggs. For many years it has been known that the queen bee controls the sex of her offspring, laying fertilized eggs in the smaller worker cells and unfertilized eggs in the larger drone cells. Several other examples of sex determination by female Hymenoptera are: In Tiphia vernalis, a wasp parasite of Japanese beetle larvae, Brunson (1938) found that usually female eggs are deposited on the larger third-instar grubs and usually male eggs on the smaller second-instar larvae; and in the chalcid parasite Coccophagus ochraceus Flanders (1962) has described different oviposition postures which forecast the sex of the egg to be laid. In Osmia lignaria nests I found that almost without exception males were produced in 4.8-mm. borings, and both sexes were produced in the larger borings. I theorized that the movement of the mother's abdo­men in the smaller boring may have been so restricted as to prevent egress of the sperm from the spermatheca, thus insuring that only male eggs would normally be laid in these smaller borings. The few female eggs in the 4.8-mm. borings might have been laid by smaller than normal mothers.

    Several nests of the megachilid bee Megachile (Sayapis) policaris Say offered striking proof that a female can lay some female eggs and then some male eggs, followed by additional series of female and male eggs. This bee is unique in that it constructs a series of brood cells in large borings. Each of these brood cells contains a large store of pollen and nectar and an average of 6.5 eggs (range 2-16). The larvae develop without cannibalism. In two policaris nests I found that both sexes developed in at least two or three consecutive brood cells.

Jayakar (1963) proposed the terms protothelytoky and protarrhenotoky for the phenomena of all female-producing eggs being laid before all male-producing eggs or all male-producing eggs being laid before all female-producing eggs. He observed the latter phenomenon in an Indian vespid mud-daubing wasp, Eumenes esuriens Fabricius. He obtained only a single 13-celled nest containing both sexes and found that all five eggs which produced males were laid before the seven eggs which produced females. (The egg died in the sixth cell.) Jayakar hypothesized that in this species a female produced a series of unfertilized male eggs followed by a series of fertilized female eggs. He also cited a 9-celled nest of the sphecid mud-dauber Sceliphron madraspatanum (Fabricius) in which four male eggs were laid before five female eggs.

    Jayakar suggested that the phenomenon of protarrhenotoky could occur for three reasons: Delayed copulation so that some unfertilized eggs were laid first; the occurrence of a long interval between copulation and fertilization; and control of fertilization by a sphincter. His first suggestion is not substantiated by the large corpus of published observations, which indicate that mating takes place very shortly after emergence of the females and before there have been any nesting activities. I am not sure just what he means by his second reason, but he probably infers that the sperm may not be mature at the time of copulation; this hypothesis was not borne out by my nests of Euodynerus foraminatus apopkensis and Osmia lignaria, in which females developed in the innermost cells of some of the earliest nests stored by the mother wasps and bees. The third suggestion has already been proved for some of the parasitic and social Hymenoptera; there is no reason to suppose that it is not equally true in the solitary aculeates.

My observations show that females of all vespid wasps, and most sphecid wasps and bees, lay a series of female eggs before a series of male eggs in nests where both sexes are produced. Furthermore, although I have not worked with marked females, I am confident that a female is capable of alternating series of female and male cells. (See also my discussion several paragraphs above on sequence of sexes in the brood cell building Megachile policaris.) Consequently, during her life she may construct several nests each with females in the inner and males in the outer cells. However, the females of several sphecid wasps and bees, for example Trypargilum tridentatum and Anthidium maculosum, lay male eggs in the inner cells and female eggs in the outer cells. Here again, I do not doubt that a female can alternate these series so that several mixed nests built by the same female will each have males in the inner and females in the outer cells.

Jayakar stated that he could find no examples of complete protothelytoky, i.e., all female eggs laid before all male eggs, and I quite agree. However, I do not consider that he has established complete arrhenotoky on the basis of one nest each in Eumenes esuriens and Sceliphron madraspatanum. I believe that if marked females of these two species are kept under observation it will be found that in each of several nests constructed by a single individual there may well be first a series of male cells followed by a series of female cells.

Return to Contents 


In many species there was significant correlation between size of cell and amount of food stored in it with sex of the wasp or bee which developed therein. This correlation was especially noticeable in nests of almost all species of Vespidae and in nests of Osmia (fig. 84) and Ashmeadiella in the Megachilidae. In nests of these wasps and bees containing both sexes, the inner cells were longer and contained a proportionately larger amount of food than the shorter cells with less food in the outer section of the boring. Females developed in those longer inner cells and males developed in the shorter outer ones. 

For example, in 6.4-mm. nests of the vespid wasp Euodynerus foraminatus apopkensis, 418 cells at the inner end of the borings from which female wasps emerged had a mean length of 17.5 mm., whereas 308 cells at the outer end from which males emerged had a mean length of only 13.4 mm. The cells in which females developed contained an average of 14 caterpillars, whereas male cells held an average of 8. Osmia lignaria provided another excellent example of this correlation: In 6.4-mm. nests of this bee, 235 cells at the inner end in which females developed had a mean length of 14.3 mm., and 255 male cells at the outer end of the borings had a mean length of 10.8 mm.; the pollen-nectar masses stored in female cells averaged 8.2 mm., but they averaged only 5.5 mm. in male cells.

    This correlation was very rarely negative in vespid nests. It was negative in nests of Stenodynerus saecularis rufulus where male cells in both 4.8- and 6.4-mm. borings had a greater mean length than female cells. The sequence of sexes in mixed nests of this wasp was normal, that is females were in the inner and males in the outer cells. The anomaly occurred because the terminal stored cell in a nest frequently was considerably longer than any of the preceding stored cells, and most frequently males developed in these extra long cells.

    The species of Trypargilum again provided interesting and anomalous data. The male cells of both subspecies of tridentatum had a greater mean length than female cells. Presumably a larger amount of prey was stored in male cells because male cocoons were longer than female cocoons and adult male wasps were a little larger than females. The situation was reversed in johannis, which constructed female cells and cocoons that were substantially longer than those of males.

Return to Contents


    Spider wasps of the family Pompilidae stored a single spider in each cell (figs. 43, 47), and wasps of the family Ampulicidae placed only a single cockroach in each cell (fig. 50). Wasps of the other families, Vespidae and Sphecidae, stored several to many specimens of prey per cell (e.g., figs. 25, 52, 62).

    Most of the vespid wasps preyed on externally feeding lepidopterous caterpillars, chiefly leaf rollers and tiers (e.g., figs. 25, 39). However, the species of Symmorphus showed some diversity in that two of them preyed on externally feeding coleopterous larvae, Chrysomela species (fig- 22), while the third preyed chiefly on leaf-mining coleopterous or lepidopterous larvae (fig. 16).

    The sphecid wasps showed the greatest diversity in prey selection. Trypoxylon and Trypargilum preyed on spiders (fig. 52), mostly immatures, and the other genera preyed on insects. The sphecine wasps stored nymphal and adult Orthoptera: Cockroaches by Podium (fig. 62) and snowy tree crickets and occasional tettigoniids by Isodontia (fig. 60). Solierella preyed on nymphal lygaeids. The pemphredonines Passaloecus (fig. 67) and Diodontus stored aphids, mostly immatures. The crabronine wasps all preyed on adult insects: Tracheliodes on worker ants (fig. 68), Euplilis on chironomid midges, and an unidentified wasp, possibly Crossocerus, on a mixture of true flies and caddisflies.

    Spiders, whether stored by pompilid or by trypoxylonine wasps, were usually thoroughly paralyzed and exhibited only weak tremors of the mouthparts or tarsi. The aphids stored by Diodontus and Passaloecus were completely motionless; in addition to stinging them the wasps malaxated, or kneaded with their mandibles, the neck region of the aphids.

    The lepidopterous and coleopterous larvae preyed upon by vespid wasps were more lightly paralyzed. They defecated and wriggled their legs, and sometimes caterpillars thrashed around if they were not confined by pressure of other caterpillars. Occasionally, full-grown caterpillars went on to pupate, and rarely adult moths emerged from the pupae; however, these adults were unable to leave and perished in the cell. Sometimes the caterpillars became cyanotic, presumably as a result of envenomation.

 Return to Contents 


    I obtained nests over a consecutive period of years from three localities: Derby, N. Y.; Plummers Island, Md.; and Lake Placid, Fla. In Table 1 I have listed those species for which I obtained 10 or more nests and the number of nests for each species in each season. The species in the upper section of the table are from Derby, those in the middle section from Plummers Island, and those in the lower section from Lake Placid. An x indicates that traps were not set out early enough in the season to provide nesting sites for vernal species (the two Osmia species at Plummers Island in 1956 and 1957, and Euodynerus foraminatus apopkensis at Lake Placid in 1962) or that no traps were used at all (Derby in 1962, Plummers Island in 1954 and 1955, and Lake Placid in 1954-1956 and 1958). In each section the species are arranged in descending order according to the total number of nests stored.

    From year to year the number of nests for any one species fluctuated. Some fluctuation was caused by the number of traps available from year to year. However, some was caused by a particular species being more common during one year than during the next one; or, at least, it used relatively more traps one year than it did in another. Only one species, Euodynerus foraminatus apopkensis at Lake Placid, was the most fre­quent user of traps in each year that traps were available to it. The wasp nesting most frequently at Derby was Ancistrocerus a. antilope; it provisioned more nests than other species in most years except 1955 and 1960, when it was the third most frequent user. Trypargilum striatum used more traps at Plummers Island than any other species, but it was the most frequent user only in 3 years (1957, 1960, 1961), the second most frequent in 3 years (1956, 1958, 1959), and the third in 1962. Osmia lignaria ranked first at Plummers Island in 3 of the 5 years in which traps were available to it, and second and third in 1 year each.

    Most of the other species showed considerable fluctuation in their relative rank. For example, at Derby, the second most frequent user of traps, Symmorphus cristatus, had relative rank of 10, 3, 2, 2, 7, 10, 1, and 4 in the 8 years in which traps were available. At Plummers Island the third most frequent user of traps, Trypargilum clavatum, had relative rank of 1, 3, 6, 3, 8, 5, and 13. These rankings for the second and third most frequent species at Lake  Placid, Pachodynerus erynnis and Monobia quadridens, were 4, 3, 3, 2, 4, and 7, 2, 5, 4, and 1, respectively.

    Overtrapping was one other factor that caused some seasonal fluctuation. I was particularly aware of this condition at Plummers Island, where year after year I continued to set traps at the same or almost the same stations. I believe that over-trapping caused immediate decline of the Monobia quadridens population after 1957 and temporary reduction of Trypargilum striatum in 1958, 1959, and 1962; of Osmia lignaria in 1960; and of T. clavatum and T. collinum rubrocinctum after 1957.  

Return to Contents 


    Several other species of arthropods used these traps. The most serious of these from the standpoint of interfering with nesting by wasps and bees were Crematogaster ants (fig. 12). Colonies of these occasionally began in borings set on rotten wood. Although the ants nested in only one boring at first, their constant foraging into other borings in the same setting discouraged wasps and bees from attempting to nest. Ants belong­ing to Camponotus subgenus Colobopsis, which ordinarily nest in hollows in twigs and galls, nested in a few traps at settings on scrubby live and Spanish oaks on the barrens at Kill Devil Hills, N.C.

    A grasshopper, Melanoplus sp. probably punctulatus (Scudder), deposited egg pods just within the entrance of a few traps set on pine trees in open woods at Kill Devil Hills.

Sowbugs occasionally entered traps placed in moist, shaded situations on dead fallen tree trunks at Plummers Island, Md.

In wooded areas snare-building spiders occasionally took over borings as retreats and spun snares in them.

    Eleven caterpillars entered these borings to pupate, 9 of them at Kill Devil Hills, and 1 each at Plummers Island and Derby (fig. 13). Two moths were reared:   The noctuid Melipotis jucunda Hübner, from one of the Kill Devil Hills nests, and the phycitid Canarsia ulmiarrosorella Clemens from the Derby nest. One caterpillar destroyed the contents of inner cells of an Osmia ribifloris nest at Portal, Ariz. I also found a caterpillar tunneling in the moldy pollen-nectar mass of a cell of the leaf cutter bee Megachile mendica Cresson in a nest from Lake Placid, Fla.; presumably the egg from which this caterpillar hatched was on one of the leaf cuttings brought into the nest by the female bee.

 Return to Contents 

Posted October 30, 2009