NESTING BEHAVIOR AND DESCRIPTION OF THE LARVA OF BOTHYNOSTETHUS DISTINCTUS FOX (HYMENOPTERA: SPHECIDAE)

Bothynostethus is a predominantly neotropical genus having a single described species., distinctus Fox, in America north o Mexico. The genus has been little studied, and its placement in the sphecid classification has been the subject .o some dispute. K.ohl (896) and Handlirsch (925) put Bothynostethus in the Nysso.ninae and related it to Alysson. In the Synoptic Catalog o Hymenoptera o America North o Mexico, Krombein (95) placed it in the tribe Miscophini o the subarnily Larrinae. More recently, Menke (968) grouped Bothynoste’thus, Scapheutes, I4illinbiella, and Bohartella in the tribe Bothynostethini o the subfamily Larrina.e. However, Evans (964) has questioned, on the basis o, larval structure, whether the Larrinae and Crabroninae deserve the status o separate subamilies, and he has cited Bothnostethus as a genus, in which the adults resemble certain Crabron’inae. Despite Menke’s (968) inclusion o the Bothynostethini in the Larrinae, he informs us (correspondence) that Bothynostethus is not a "typical" member o the subarnily, and that "it would not be difficult to derive a crabronid rom a Bothynostethus-like ancestor". It is clearly o interest to discover what light a study o nesting behavior and larval structure may shed on these problems. Cazier and Mortenson (965) ound B. distinctus nesting in the ground and preying upon a chrysomelid beetle, but otherwise the genus has

Mass., and the SUNY College 6 Forestry, Syracuse, N. Y.
NESTING BEHAVIOR Ecology. Our observations .on B. distinctus were made at two widely-separated localities. The first site, a small mound o sand overgrown with sparse grasses and other plants. (Fig. I), occupied a portion o a man-made sand pit near Colonie, Albany County, New York. The entrances to three nests we're located in the vertical sides o the mound, just beneath the overhang (Fig. 2), and were situated so that they did not receive direct rainfall or sunlight, fi_ ourth emale was captured on a sand cliff several meters away but her nest was not discovered. The second area, in Bedford, Middlesex County, Massachusetts, comprised a sand-bank with a slope oi: about 45 . One nest entrance was located in the side o the bank, near a projecting stone. _At both sites the entrances were let open by the wasps during the periods o observation, although the burrows could have been closed arther inside.
_At Colonie we observed emales provisioning nests between July 7 and I6, 97o, whereas at Bedford one o us (HEE) noted a emale nesting on Sept. 22, 97I. These collective dates, o observation suggest that B. distinctus may have two generations per year in the Northeast, although the possibility o a single generation with staggered emergence should not be overlooked. Cazier and Mortenson (965) ound B. distinctus nesting in "bare sandy soil" near Portal, Arizona.. One nest was located inside a vertical rodent burrow, the entrance being situated 5 cm below ground level on the "east wall." Burrow construction. B. distinctus emales apparently do not begin their burrows rom the sand surace but utilize appropriatesized nests o other insects, possibly those o solitary wasps., bees, or tiger-beetles. These burrows are renovated and moditqed accordingly. The "weak tarsal rake" of the ore'leg of the emale (Menke, 1968) and the variation in the burrow configurations and the positions o. the cells suggest that females o this species use pre-existing burrows or the initial portions o their nests.
Prey transport, entry, and exit. Two females (Note nos. H23, H23K) were observed taking a total of I4 prey into their nests during a 3-hour period. One wasp (H23) spent rom 5 to 30 (mean I2.5) minutes between provisions; the other (H23K) took longer I2-40 minutes mean 23). In every case the beetle was carried in flight, dorsal-side-upward and head-forward. The emale grasped the beetle's antennae with the mandibles, and held its body with the legs. During one entry in which the entrance had been somewhat disturbed, the provisioning wasp, upon landing, was attacked twice by a female satellite-fly of the genus Senotainia (Sarcophagida.e: Miltogramminae). She did not release the prey but entered in the usual manner.
Because o the concealed situation of the nest entrances, which were often behind overhanging plants and rootlets, the females were prevented from entering directly. One female (H23) invariably landed outside on the cliff-face before diving in, and the other (H23K), whose entrance was even more concealed (see Fig. 2), always landed on a plant to the left of the entrance and ran downward for a few centimeters before entering. The first wasp remained inside her nest, on the a.verage, 1 minutes before exiting to search for the next prey; the second wasp usually stayed inside 11/2-2 minutes before appearing in the entrance. These relatively long periods of time were undoubtedly related to. the considerable distances inside the nests to which the females had to proceed with the prey. The first female, apparently disturbed by our observations, made a 5-second orienta.tion flight after taking the first prey inside, followed by a I-2-second orientation flight after the second prey, but thereafter she did not orient. The second female, also seemingly disturbed by our presence, made a I-2-second orientation flight after the initial prey was taken in but did not orient thereafter. Cazier and Mortenson (1965) were unable to obtain information on the manner o prey transport o. the female of B. distinctus they observed in Arizona. They described one activity which may have been the female's orientation flight. The wasp they observed was much more active than the ones we studied, spending: rom only 35 seconds to 6 minutes in the field between prey. After entering the nest, this emale stayed inside 3o-18o seconds before exiting to search or additional prey. Although not specifically stated, her provisioning activities clearly increased in intensity rom late morning to late aternoon.
Nest structure and dimensions. We excavated two nests o B. distinctus at the Colonie, N. Y., site and one o us (HEE) dug a nest o. this species at Bedford, Mass. The two nests at Colonie were rather complicated in contqguration, containing a number o branching and curving side burrows leading rom a main tunnel. Because o the complexity o these nests, including much variation in the position and distance o the rea,ring cells rom the entrance, we suspect that the wasps were utilizing, in part, the abandoned burrows o solitary bees and constructing short side. burrows and cells therefrom. One nest (H23), still being provisioned when excavated, contained 7 ceils at distances o rom 32 to 45 cm rom the entrance (Fig. 4). The ceils were built in at least two distinct clusters, ceils A and B in one cluster containing larvae, and ceils C-F in an)ther with eggs, plus an additional isolated cell (G). Only 2-6 cm separated the ceils within a cluster, and the closest ceils in different clusters were 7-15 cm apart. The cells belonging to a cluster were oriented in one general direction and could ha.re been made at the ends o short, branching side burrows leading rom a single tunnel. None o the side burrows could be traced as they were/]lled with damp sand. _A short main tunnel and three proximal side burrows which ended blindly were open and could be ollowed.
The emale was ound in one o the side burrows, evidently extending this burrow in a different direction ater having sealed off the last cell. She had been inside the nest or 11 minutes ater entering with the last prey and before we began our excavation.
The second nest at Colonic (H23K), although more complex in structure than the irst, was more compact, i. e., the cells, were closer together and nearer the entrance (Figs. 3a, b). This nest contained a total o 8 cells and, likewise, was still being provisioned when excavated. Although the cells were oriented in different directions rom one another, they appeared to be constructed in three distinct clusters containing two different stages .o wasps. Cell A, holding a cocoon, was separated only 9 cm rom the entrance. Ceils B-F, also containing cocoons, were situated only 1-7 cm apart at distances o rom 16 to 22 cm rom the entrance. The two most recent ceils (G, H), containing larvae, were unearthed 9-1o cm rom the entrance on the other side o the burro,w rorn cell A. As  Psyche Ma reh-J une in the first nest (H23), the cells in this nest appeared to be built not in series but from short, branching side burrows leading from a single tunnel.
The nest at Bedford, Mass. (no. 230.6), was evidently in an early stage of development when excavated. The entrance, 4 mm in diameter, led to several subterranean cavities of which one contained a cicindelid larva. Upon further digging, a single cell containing prey and a 3-mm-long larva was unearthed at a distance of 5 cm from the entrance (Fig. 5). The temale was captured in another branch of the nest, 5 cm from the entrance. At Colo.nie, the depths of the seven cells in the first nest (H23) ranged from 22 to 44 (mean 32) cm beneath the surface, whereas those of the eight cells in the second nest (H23K) ranged from 2t to 24 (mean 22.5) cm. The relative uniformity in the latter set of figures is related to the fact that the nest was compact and the cells were constructed close to the edge of the mound, whereas in the first nest the cells were distributed from the edge of the mound to the top. There was no correlation between the depth of the cell or its distance from the entrance and the stage of the developing wasp. For example, in nest H23 the shallower cells nearer the entrance contained larvae and the deeper cells farther from the entrance held eggs. In nest H23K, on the other hand, the shallower cells nearer the entrance contained small la.rvae but the deeper cells tarther fro.m the entrance, mature larvae inside cocoons. The ovoidal cells were oriented either horizontally or sloped upward or downward in the soil at a 5-25 angle. Five such cells at Colonic averaged 6 (4-8) X II (IO-I2) mm. One cell at Bedford was 5-ram-wide.
One of the temales (H23K) had been observed taking in prey nine days before we excavated her nest, indicating that she took an average of more than one day to complete a single cell. This may have been due to poor weather conditions. However, even during optimal weather conditions, our data on the provisioning times suggest that a female constructs and completes not more than two cells per day. Such a wasp probably remains with the same nest tor several weeks, enlarging it and continuously adding new cells. Thus, a emale of B. distinctus might renovate and use only one or a few nests during her lifetime in the northeastern United States. Cazier and Mortenson (1965) noted that the female they observed in Arizona interrupted her rather continuous provisioning activities a total of four times in one day and spent exceedingly long periods each time inside the burrow, probably in ovipositing and filling the burrow leading to the old cell and in cnstructing a side burrow and new cell. I such is the case, this wasp completed our cells in a single day and probably would not have remained with a nest or more than several days under optimal weather conditions.. (Chrysomelidae: Galerucinae). Except or those beetles which had been devoured by the wasp larvae, prey collected from all three nests at both Colonic and Bedford were alive as indicated by the movements o the legs and abdominal segments.
A number o( beetles from the second nest were damaged during excavation and were not weighed. The two wasps associated with these nests each weighed o mg, or slightly less than an averagesized beetle. The total weight of prey per cell in these nests ranged rom 44 to 68 (mean 55.8) mg in our of the cells. The beetles were placed in the cells in a variety o positions. In cells with few prey (4 or 5/cell), the individuals were mostly headinward but either dorsal-side-up, ventral-side-up, or on the side. Some beetles were even placed perpendicular to the long axis ot: the cell. The fact that most individuals were head-inward is linked to the manner o( the female's transport of the beetle, i. e., head-forward. Egg. The female o B. distinctus did not lay her egg until the ull complement ot: prey had been put in the cell. In most cells the egg-bearing beetle was one of the innermost prey and was positioned ventral-side-upward and head-inward. In one nest (H23), four such beetles weighed -6 (mean 3.3) rag, or were slightly larger than the average-sized prey. The whitish, elongate, curved egg, .5 )< 0.4 ram, was attached by its cephalic end to the ventro-lateral portion of the thorax between a oreand midleg (Fig. 6). The exact sternite, suture, or membrane to which it was attached could not be discerned. Although the caudal end of the egg extended slightly obliquely backward and was raised ree above the opposite midleg, the position of the egg with respect to the long axis of the beetle's body was almost transverse. Eggs were attached to the left or right sides of the prey in about equal numbers. The larva began eeding at the site o the egg attachment.
One larva, 2-mm-long, was noted eeding between the beetle's foreand midcoxa, and another, 3-mm-long, was attached near the juncture of the pro-and mesosternum. Cocoon. The cocoon of this species is oval-elongate in shape and dark brown in color. The walls consist of silk and possibly other oral secretions, with various-sized sand grains embedded externally ( Fig. 7). Several beetle elytra were attached to the proximal end, i. e., the end nearest the entrance to the cell. The cocoon in Fig. 7 measured 9-5 mm long. DESCRIPTION OF THE I/IATURE LARVA Our description of the mature larva of Bothynostethus distinctus is based on two specimens taken from freshly made cocoons. These larvae had not yet passed their meconia, and, aside from the more compact body shape, they showed no obvious wear or other structural modifications. Body (Fig. 9). Length 7.5 ram; maximum width 2.8 ram. Robust, slightly curved; pleural lobes very prominent, especially on thorax; prothorax with a pair of additional finger-like lobes that extend forward laterad of head; apical abdominal segment rounded, anus distinctly ventral and pre-apical. Spiracles (Fig. I5) small, lightly pigmented; atrial walls lined with weak, irregular polygons, opening into subatrium simple, unarmed. Integument densely spinulose on thoracic venter, but dorsum barely so; thoracic segments each with a transverse dorsal row of IO to 12 setae (up to 30 / long); pleural lobes and accessory prothoracic lobes smooth, but each terminating in 2 or 3 short setae. Head (Fig. 8) Mouthparts. Labrum (Fig. o) truncate apically, bearing 8 setae across disc and 6 across apical margin, the margin with some minute, inconspicuous sensilla and rather weakly bristly; epipharynx strongly spinulose medially, with smaller and sparser spinules laterally; epipharyngeal sensory areas each with our .sensilla. Mandibles (Figs. , 2) slightly more than twice as long as their basal width, with three teeth on inner margin in nearly same plane; base with a single seta; upper surace roughened in area o teeth. Maxilla ( Fig. x3) protruding, rounded, without an evident lobe on inner margin, densely spinulose along mesal margin; palpus about 80 / tong, galea more slender, slightly more than hal as long as. palpus. Hypopharynx very densely spinulose. Labium (Fig. 4) with prementum densely and extensively spinulose; palpi about 60 /z long, slightly curved, much exceeded by spinnerets.
Comments on larval characters. In the articial key to genera ot: Sphecidae presented by Evans (959, P. 6), the larva o Bothynostethus runs to Entomognathus, a genus reconsidered in the paper that ollows (Miller and Kurczewski,972). In the table o subfamily characters (ibid, p. 68) it runs to the couplet separating the Larrinae and Crabroninae, agreeing with the Crabroninae except in having accessory finger-like lobes on the prothorax as in many Larrinae. Distinctly crabronine eatures o Bothynostethus larvae are as ollows: spiracular atrium weakly sculptured and opening into subatrium unarmed; mandibles slender and with teeth in about the same plane; mesal margin o maxilla, not lobed. Very similar 3-toothed mandibles occur in Lindenius pygmaeus (Rossi) (Grandi,928) and similar but 4-toothed mandibles are ound in L. tylotis Court and R. Bohart and in other crabronine genera such as Crossocerus and Entomognathus (Evans,959). In most Crabroninae the spinules on the labium occur in two separate patches, but Bothynostethus is like Entomognathus in having the entire oral surface clothed with spinules. In several other eatures the larvae o( these two genera are very similar" body setae and spinules are similar in size and distribution; the labrum is similarly shaped and with a similar distribution o setae and spinules; the orm o the palpi, galeae, and spinnerets is virtually the same; and the spinules on the hypopharynx, maxillae, and labium are similarly dense and abundant. Entomognathus has no lobes on the prothorax, more head setae, and, as mentioned above, our-toothed mandibles. [This corn-O0 Psyche [March-June parison is based on Grandi's (1927) description of the larva of E. brevis Lind, and on Miller and Kurczewski's (1972) (Evans,I964), it seems worthwhile to compare the ecologies, nesting behaviors, and cocoon morphologies of members of these two littleknown genera. Such a comparison is now possible because Miller and Kurczewski have recently studied the behavior of Entomognathus memorialis Banks in some detail and have reviewed the. literature on members of this genus in the following article.
Bothynostethus distinctus resembles Entomognathus memorialis and E. brevis Lind. in nesting in sand-cliffs and slopes, although the last species also nests in flat sand (Chambers,I949). These three species plus E. texana Cresson possibly utilize pre-existing burrows for the proximal portions of their nests (Adlerz, 1912;Cazier and Mortenson, 1965); however, at least E. memorialis, in addition, constructs nests from the sand surface. In either case, searching for nest-sites involves characteristic hovering flights interspersed with soil sampling or hole searching, a behavior prevalent in Mimesa and Trypoxylon. Although members of both genera transport the prey in flight in a head-forward position E. memorialis carries the beetle venter-up with the legs as in most crabronines (Pedal Type I, Evans, I962), whereas B. distinctus carries the beetle dorsum-up and, in addition to using the legs, grasps the prey's antennae with the mandibles as in many latrines (Mandibular Type 3). At least one species of Entomognathus, memorialis, stores prey in the burrow prior to their arrangement in the cell and oviposition. B. distinctus apparently takes the beetles directly into the cell. Many cells in the nests of E. memorialis are constructed in tandem series, whereas cells in the nests of B. distinctus are evidently built singly at the ends of short side burrows. The use of a common family of beetles by members of these two genera should not be overemphasized, as many examples of convergence in prey type are known in the Sphecidae. One noteworthy example is the frequent use of the tarnished plant bug, Lygus lineolaris (Beauvois) by the larrine Plenoculus davisi Fox and by the crabronine Atnacrabro ocellatus Packard (Kurczewski, 1968;Kurczewski and Peckham, 197o). It is entirely possible that the use of chrysomelids by Bothynostethus and Entomognathus was acquired independently and, furthermore, that the similarity in the position of egg attachment of these tw.o genera, was dictated by the morphology of the beetle [Compare Fig. 6 with fig. II in Grandi (1927) (E. brevis) and with Fig. 3 in Miller and Kurczewski E. memorial#) ].
The cocoons of species of Larrinae and Crabroninae are basically similar in shape and in having an outer covering of sand grains. That of Bothynostethus contains beetle elytra affixed to the proximal end and thus resembles cocoons of certain species of the crabronine genera Lindenius and Crabro, in which prey remains cover the outside. On the other hand, in latrine genera such as Lyroda, Plenoculus, and Tachysl)hex prey remains are not attached to the exterior of the cocoon and, in this respect, Entomognathus is much like a latrine [see Miller and Kurczewski,Fig. 4 (E. memorialis); Benoist,I915 (E. bre'vis) ]. The reliability of the exterior structure of the cocoon as an indicator of phylogeny can, however, be questioned. For example, Oxybelus, a genus of the tribe Oxybelini which is usually placed in the Crabroninae, has a cocoon that is essentially indistinguishable from that of many Larrinae.
Thus evidence on the relationship of Bothynostethus and Entomognathus based upon cocoon structure is equivocal, and that derived for certain behavioral components is difficult to interpret. Although larval structure suggests that the two genera are closely related, Bothynostethus does possess finger-like lobes on the prothorax, an important larrine feature. In spite of some similarities among these two genera in adult external morphology, Bothynostethus and Entomognathus or the most part retain their affinities with other Larrinae and Craboninae, respectively. Both Krombein (1951) and Menke (1968) concluded that Bothynostethus belongs in the Larrinae, while Entomognathus has consistently been placed in the Crabroninae.
Despite considerable similarity in larval morphology and several common ecological and behavioral features, we believe the two genera are sufficiently different to warrant their being placed in separate tribes. Therefore, we accept Menke's (1968) placement of Bothynostethus in the Bothynostethini, and support the retention of Entomognathus in the Crabronini. However, the resemblances among the larvae and cocoons of these and related genera lead one to ask again whether the larrine and crabronine wasps should not simply be considered as tribes within a single subfamily, as sug-