Return to Contents



The traps used in my study were fabricated from straight-grained pieces of seasoned white pine. Borings of several diameters and of two lengths were made in these sticks. Twist drills were used for the drilling. These drills left the inner end of the boring obliquely angled and relatively smooth; the use of regular wood drills would have left the inner end very rough.

The smallest size drill was 1/8 inch (3.2 mm.) in diameter. The sticks used for borings of this diameter were 20x20x75 mm. The boring was drilled along the longitudinal midline to a depth of 64-70 mm. This was the shortest boring I used in this study.

The traps used most commonly were those made with drills having a diameter of 3/16 inch (4.8 mm.) or 1/4 inch (6.4 mm.). These borings were drilled to a depth of 152 mm. in blocks of wood 20 x 20 x 165 mm.

A few traps used early in the study had 3/8 inch (9.5 mm.) borings drilled to a depth of 152 mm.

The largest size boring was 1/2 inch (12.7 mm.) in diameter, drilled to a depth of 152 mm. in blocks of wood 25x25x 165 mm.

    Early in my study I tried inserting pieces of glass tubing in some of the larger borings. These were accepted by the wasps, and the clear sides made it possible to observe the nest during its construction. However, the impervious walls caused such an increase in moisture within the cells that the contents usually grew moldy within a few days. Also, reflections from the surface of the glass interfered with photography of the nests.

    Other hollow tubular materials have been used by other investigators to attract these wasps and bees. Bamboo is excellent but has the obvious disadvantage of having a variable diameter so that cell sizes cannot be measured accurately. Soda straws also have been used, but one cannot split them open to make observations on the duration and activities of the immature stages.

    Medler and Fye (1956), and Fye (1965a, b) used borings similar to mine, but they were drilled in sumac (Rhus) stems. These worked very successfully in Wisconsin and Ontario; but because sumac is not too abundant around Washington, I have not tested it.

    Earlier, Medler used a domicile trap which consisted of a number of short boards bolted together with holes drilled along the interfaces. This trap attracted nesting individuals, but one could not readily make measurements and observations of an individual nest in this kind of setting. Most recently Nye and Bohart (1964) published plans for an improved domicile structure in which channels were routed out on boards and the latter then bound together to form many potential nesting sites.

    These structures are very useful for concentrating populations of desirable species of megachilid bees for transport and release of the progeny in areas where their pollination activities are needed; they are not satisfactory for the person who wishes to study individual nests, their occupants, and their development.

    Matthews and Fischer (1964) used traps like mine but sawed them lengthwise along the upper edge of the boring; the lower half of the trap, containing most of the boring, was then covered with transparent film (Saran Wrap), and the two parts were fastened together with rubber bands. This technique offered an opportunity to observe development after completion of the nest by periodic removal of the top half of the trap.

Return to Contents



Each trap was identified by a letter and number written on it with a black wax pencil. The same letter was used for all traps at a single locality in a given year. The numbers were assigned in sequence; thus, the traps used in a single season at Plummers Island, Md., for instance, might run from E 1 through E 190, whereas those at the Archbold Biological Station in Florida might run from B 1 to B 237 for the same season.

New traps were always set out unsplit. Frequently, I reused old traps when they split along the longitudinal axis of the burrow during the first season. The remnants of the old nests could be cleaned out with a narrow stiff brush and the two halves taped tightly together with friction tape. The early numbers on these old traps were scratched out and new ones assigned appropriate to the new season. I always sterilized these old traps before reuse by placing them in an oven at 200 F. for several hours.

Usually the traps needed no other preparation before being set out, except for those placed early in the season. During my first season of attempting to trap vernal bees, I experienced considerable loss from mold in the nests because of the frequent spring rains. I eliminated this difficulty in subsequent years by rolling the traps in a very shallow bath of melted paraffin so that the sides were coated with wax. This technique should be used on any early season nests in the United States. Only a shallow layer of melted paraffin should be used to prevent the wax from entering the boring itself.

Return to Contents


Usually I made up a bundle of six traps to set at each station, two each of the 4.8- and 6.4-mm. borings and one each of the 3.2- and 12.7 -mm. borings. Occasionally in special situations, such as when I trapped for vernal bees, I used only the 4.8- and 6.4-mm. borings. Sometimes I used only 3.2-mm. borings, as on an old cowshed wall that had had a heavy infestation of small anobiid beetles and a consequent abundance of small nesting Hymenoptera in the abandoned beetle borings.

In selecting stations in the field, I tried to set the traps where wood-nesting wasps and bees would be likely to search for nesting sites. Obvious sites for these were on or beneath dead branches, and on dead tree trunks or wooden structures that contained abandoned borings of other insects (figs. 3, 4, 7, 8). 

    In areas of open sandy scrub, such as Kill Devil Hills, N. C, and the Archbold Biological Station, Lake Placid, Fla., settings on live scrubby oaks bearing old cynipid galls were productive, as were settings on pine trunks with thick bark and beneath branches of live scrub hickory (figs. 1, 2, 9, 10). On the desert floor in Arizona the most productive settings were beneath dead branches of mesquite, palo verde, and desert willow; on dead agave and yucca stems; and on wooden fence posts (figs. 5, 6).

Usually settings in dense shaded areas in eastern woodlands were not satisfactory. Stations were more successful at the edges of woods or in areas of open woods where there was considerable sunshine during the day.

The traps were set so that the borings were in a horizontal position. Most of my traps were placed at elevations of 1-2 meters above the ground. Fye (1965c) noted the use of several devices to place traps at heights of 3-20 meters.

I found it preferable to check trap lines on a weekly basis whenever possible. Completed nests, signified by the boring entrances being plugged with mud or other substances, were picked up and empty traps were set out in their places. In the earlier years of the study, I picked up some traps as soon as it was evident that they were being used as nesting sites. Probing the borings delicately with a grass stem would identify those containing partially completed nests. Examination of incompleted nests provided data on egg size and hatch that were not usually available from completed nests.

Return to Contents


When I received the nests at my office, I tried to open the unsplit traps by inserting a jack-knife at the blind end of the trap and, by twisting the blade, to split the trap with the grain of the wood along the longitudinal axis of the boring. Frequently the grain was at a slight angle necessitating that two or more splits be made before the entire length of the boring could be inspected. Presplit traps were easily opened by cutting the friction tape along the line of the previous split on one side so that the top of the trap could be folded back.

I then entered details of the nest architecture on standardized, mimeographed data sheets. The architectural information recorded included the following: Presence or absence of a preliminary plug at or near the end of the boring and length of empty cell if present; thickness of cell partitions and closing plugs and materials of which they were constructed; length of vestibular cell and of any intercalary cells present. A separate section contained notations of prey used by wasps. A table at the bottom of the data sheet provided columns for each cell. In these were recorded cell length and cocoon length as well as dates of egg hatch, larval maturity, cocoon spinning, pupation and adult emergence, and sex of cell occupant. The cells were numbered in sequence, cell 1 being at the inner end. Obviously, I could not record many of these developmental data because the nests already had passed the early stages. Also, upon first examination nests in early stages of development required one or more subsequent examinations to ascertain some of the desired data. The plain back of the data sheet allowed space to record additional supplemental data such as cocoon description, activities of the rightful nest occupants or their parasites, and so forth.

I considered that the stored or provisioned cell consisted of the space in which the food was stored plus the partition capping this space. Likewise, I considered that the vestibular and intercalary cells consisted of the empty space plus the closing plug or partition which capped them. Consequently, the measurements given in the specific accounts which follow should be interpreted accordingly. The mean cell length included the partition thickness so that the mean partition thickness, as given, should be subtracted from the mean cell length in determining the mean length of that part of the cell containing the store of food.

After recording the nest measurements and development, I reassembled the two halves of the nest and wound a rubber band near each end to keep the halves closely apposed. When I desired to reexamine the nest, I removed the rubber bands and top half of the nest and recorded the development. Occasionally I wrapped a piece of Cellophane or transparent film (Saran Wrap) tightly around the half of the trap containing the cells. This enabled me to observe and photograph the activities of the newly eclosed adult during the 2-5 day period before it left the cell.

After occupants of the nest pupated, it was necessary to affix some kind of a net around the nest entrance with a rubber band to confine the adults as they emerged. I tried Cellophane dialysis tubing; but many kinds of wasps, bees, and parasites chewed through it without difficulty. Next, I had small sleeves made  from nylon organdy, about 85 mm. long and 65 mm. wide, with one end stitched shut. These kept in males and parasites, but females of many species were able to chew through them also. Finally, I settled on sleeves similar in size to those made from nylon organdy, but this time made from plastic mesh (Saran screening). These were objectionably stiff, but they kept the females from chewing through to the outside so they fulfilled the necessary function satisfactorily.

As adults emerged into these sleeves they were removed, killed, pinned, and labeled. It was necessary to examine the nests at frequent intervals daily once emergence began. Occasionally emerged adults would reenter the boring head first and fail to back out. If they were not removed by hand, these disoriented adults would block all further egress from the nest.

Frequently, when the cocoons were hard or tough, such as those of Trypargilum and Megachile, I placed each in a numbered glass vial to confine the adult upon emergence.

   I had to keep a constant check on nests in my office to discover and eliminate infestations by two parasites, which are the bane of laboratory cultures.  These parasites were the tiny eulophid wasp Melittobia chalybii and the grain itch mite, Pyemotes ventricosus. These parasites attack chiefly diapausing larvae and pupae of various insects in nature. In nature Melittobia apparently preys on solitary wasps and bees, but Pyemotes has a wider host range and attacks many kinds of terrestrial insects. I found both species in nests in the field and these were the likely sources of my laboratory infestations. Both parasites are slender, flattened animals. They had no difficulty in squeezing between the split halves of these traps, many of which did not fit tightly, or they entered the nests through the plastic mesh (Saran screen) emergence sleeves. Their development is so rapid that an infestation might spread through a number of nests if not discovered soon after it occurred.

Many nests, including all those stored later in the summer, had to be carried over the winter before adult emergence occurred. The nest occupants entered a prolonged larval diapause which could be terminated only after exposure to chilling temperatures for 2 or more months. In preparing the nests for overwintering I wound transparent adhesive (Scotch tape) around them lengthwise to cover the split. The purpose of this was to trap as many emerging Melittobia as possible and thus prevent their spread to other nests during warm periods in the late fall, winter, or early spring. The nests, except those from Florida, were then wrapped in bundles in manila paper and placed in a corrugated cardboard carton also wrapped in manila paper. This package was then hung outside an upstairs window at my home from mid-October or November through mid-March or early April. I treated the diapausing Florida nests differently only in that I exposed them to chilly weather for only 2 months and brought them indoors whenever temperatures went below freezing.

Return to Contents 


The trap-nesting technique is admirably adapted to still or motion-picture photography. A 35-mm. single-lens reflex camera with extension tubes and electronic flash can be mounted on a tripod to photograph the activities of the nesting females (figs. 14, 15, 70-77). After the traps have been opened, the same equipment can be used to record the nest architecture {e.g., figs. 22, 23, 25) or the developmental details (figs. 30-38).

Return to Contents


(Plate 1, Figures 1-4; Plate 2, Figures 5-8; Plate 3, Figures 9, 10)

The nests from Derby, Erie County, N. Y., came from a locality in western New York along the edge of Eighteen Mile Creek just a short distance from Lake Erie. Many of the settings were at the edges of wooded areas or along creek banks where the dominant trees were sycamore, willow, and elm. Other settings were made on the sides of wooden sheds or on window-sills, in crevices in a rock wall, and in piles of cut firewood.

Most of my nests from the metropolitan area of Washington, D. C, came from Plummers Island, Montgomery County, Md., the home of the Washington Biologists' Field Club. This island is in the Potomac River, a few miles northwest of the city and is 50-110 feet above sea level. Much of it is covered with deciduous woods, more or less open to sunlight at varying intervals during the day. The predominating trees are oak, hickory, elm, hop hornbeam, and sycamore. Many nests came from settings on dead branches or standing dead tree trunks in or at the edges of these wooded areas (figs. 7, 8). A large number also came from settings on the porch rafters (fig. 4) of the Club's cabin, Winnemana Lodge. These rafters contained abandoned borings of several different insects in which a number of species of solitary wasps and bees nested.

Elsewhere in the Washington area I ran a series of traps one year at Cropley, Montgomery County, Md., just a mile above Plummers Island and with similar habitats.

Most of the nests from Arlington County, Va., came from settings on the wall of an old wooden cowshed near my house. At one time the boards were badly infested with anobiid beetle larvae. A number of species of small wasps and bees nest in these old borings and some of them were induced to nest in 3.2-mm. borings in wooden traps. One year I ran a short series at Glencarlyn, Arlington County, in a rather densely wooded area of deciduous trees. Elsewhere in Virginia, I had a short series one year at Dunn Loring, Fairfax County, about 10 miles west of the District of Columbia, along the right of way of the Washington and Old Dominion Railroad.

At Kill Devil Hills, Dare County, in coastal North Carolina I trapped in two different habitats. One was the relatively barren, flat sandy area east of the line of dunes and parallel to the ocean. This was an area of sandy scrub characterized by scattered, scrubby live oak, Spanish oak, and bayberry (figs. 1, 2). The other area was on the lee side of the dunes where there were well-developed woods quite open to sunlight and with pine, oak, and hickory as the dominant trees (fig. 3).

All my Florida nests were from the Archbold Biological Station at Lake Placid, Highlands County. A few were from settings in an open concrete area beneath the laboratory building. Most of the nests were from the Highlands Ridge area of the Station at an elevation of about 200 feet. This is a sandy scrub area quite open to the sun and with oaks, scrub hickory, and pines as the principal trees (figs. 9, 10).

Most of the nests from Portal, Cochise County, Ariz., were from settings on the desert floor near the base of the Chiricahua Mountains at an elevation of about 4,000 feet (figs. 5, 6). There are no large trees in this area except for sycamores along the banks of intermittent streams. The dominant shrubs and plants are mesquite, acacia, yucca, agave, desert willow, and many kinds of cacti. The traps from Scottsdale and Granite Reef Dam in Maricopa County, Ariz., came from similar habitats.

 Return to Contents

Posted October 30, 2009