Sphex subtruncatus Dahlbom 1843

Sphex subtruncatus krombeini Vecht

This large Sphex (, 22–27 mm long) is described by J. van der Vecht in the appendix to this contribution. It is found only in the Wet Zone of Sri Lanka where the annual rainfall averages 1950–3900 mm and the altitude ranges from 100 to 600 m. It is easily recognized for it is the only Ceylonese Sphex having infuscated wings with violaceous reflections. In addition, the female is the only taxon in which the mid- and hind-femora and tibiae are red; the male hind femur is usually red, but occasionally the legs are entirely black in that sex.

COLLECTION LOCATIONS AND DATES.—CENTRAL PROVINCE. Kandy District: Udawattakele Sanctuary, Kandy, 26 Feb, 19–30 Mar, 18–20 Apr, 11 May, 3–5 Jun; Gannoruwa Timber Reserve, 4 Jun.

WESTERN PROVINCE. Colombo District: Labugama Reservoir Jungle, 14 Jul.

SABARAGAMUWA PROVINCE. Kegalla District: Kitulgala, Bandarakele, 3–7 Feb, 17–18 Mar. Ratnapura District: Gilimale, Induruwa Jungle, 2 and 7 Feb, 7–8, 13–15, and 26 Mar, 16–19 Apr, 17 and 19–22 Jun, 10 Oct; Weddagala, Sinharaja Jungle, 10 Feb, 19–21 Jun, 22–23 Sep.

SOUTHERN PROVINCE. Galle District: Kanneliya section, Sinharaja Jungle, 12–17 Jan, 12 Mar, 13–16 Jul, 13–16 Aug, 10 Sep, 2–5, 8–12 and 15–17 Oct.

Dates of collection in Gilimale, Weddagala, and Kanneliya suggest that the wasp may be active throughout the year except possibly during periods of little or no rainfall.

PREVIOUS RESEARCH.—Sphex subtruncatus Dahlbom is a wide-ranging, polytypic species of the Oriental Region with perhaps as many as a dozen subspecies ranging from the Indian subcontinent through southeastern Asia eastward to Sumba and Flores Islands, Indonesia, and northward into China, Korea, the Philippines, and Taiwan (van der Vecht, in litt.). It does not occur in Japan as stated by Bohart and Menke (1976:117).

Several authors have published biological notes on some of the subspecies. Henry (1932:232–233) contributed brief notes on the prey and nest of S. subtruncatus krombeini under the name S. nigripes Smith, a synonym of typical S. subtruncatus. Tsuneki (1963:41–48, fig. 9, tabs. 7–9) published extensively on the subspecies occurring in Korea and identified the wasp as S. haemorrhoidalis Fabricius. Tsuneki (1968b:54) and Tsuneki in Tsuneki and Iida (1969:1–3) published on the Taiwanese subspecies, calling it the Formosan race of haemorrhoidalis. Williams (1919:128–131, figs. 64–66) contributed extensive notes on the Philippine subspecies S. subtruncatus siamensis Taschenberg, misidentifiying his wasp as Ammobia mutica (Kohl).

Piel (1935:289–293, figs. 13–18) published notes on a Chinese wasp that he called S. haemorrhoidalis Fabricius, variety nigripes Smith. Sphex nigripes is a synonym of S. subtruncatus, but I am convinced that Piel misidentified his specimen and that it is some other taxon. He observed only a single nesting female that dug a nest on 24 August and stored it completely with four prey on the following day. This is a remarkable behavioral difference from the several subspecies of S. subtruncatus that Williams, Tsuneki, and I observed, all of which practiced progressive provisioning. Piel's wasp then began another nest nearby on 25 August and completed it on the next morning. Piel dug up this second nest to observe the architecture. The wasp then began a third nest nearby, completed it, and brought the first prey to it on the morning of the 27th. Adjacent to each of the three nest entrances the wasp prepared one or two empty accessory burrows such as I described earlier for S. sericeus fabricii Dahlbom. Furthermore, Vecht (in litt.) stated that the female of S. subtruncatus has more slender mandibles than Piel figured for his wasp. Neither Williams, Tsuneki, nor I observed the other subspecies of S. subtruncatus making accessory burrows, thereby substantiating the probability that Piel's observations were not made on S. subtruncatus. I have examined two pairs identified as “nigripes Smith” from the Piel collection now housed in the Academia Sinica, Beijing, and one pair collected and identified by Piel as “haemorrhoidalis v. nigripes” in the Paris Museum. All are Sphex subtruncatus Dahlbom, but one pair is from Ihing, Kiangsu Province, one female from Chenkiang, Kiangsu Province, and the others from Chusan, Chekiang Province. Piel's observations were made on No. 1206, a female from Shanghai, a specimen apparently no longer in existence if Piel did, indeed, collect it. The possibility remains, consequently, that his notes were not made on a specimen of S. subtruncatus. Two males from Shanghai in the Paris Museum are S. subtruncatus but they were not collected nor identified by Piel.

FIELD OBSERVATIONS.—Our most extensive notes on S. subtruncatus krombeini were made in the Kanneliya section of the southwestern Sinharaja Jungle and bear code numbers 31272 A, 81472 A–D, 81572 A–G, 10973 A, 11275 A–E and G–J, 11375 D, 11575 A, D, E, H–K, and M, 11775 B, and 18 color-coded wasps during the period 7–12 Oct 1973. We obtained a few notes from Weddagala in the northeastern Sinharaja Jungle coded as 62176 B and 21077 A–C. Some observations were made in the Induruwa Jungle at Gilimale on the southwestern slope of the Adam's Peak Sanctuary and were coded as 61776 E–G, 61976 E, 2777 A, and 31479 A. P.B. Karunaratne observed an aggregation in Udawattakele Sanctuary, Kandy, and code numbers 41875 A–C, 41975 A–G, and 51175 A–F were assigned to these wasps. Finally, we obtained one prey and nest record, 71478 D, in Labugama Reservoir Jungle.

The color-coded wasps were identified by letter and number, e.g., YT 3, GTW 5, standing respectively for the third wasp bearing a yellow spot on the thorax and the fifth wasp bearing a green spot on the thorax and wing.

My first extended observations on this wasp were made on 14 and 15 August 1972 along the edge of a logging road in Kanneliya. Two of the nests that we excavated, 81572 F and G, established conclusively that S. s. krombeini practices progressive provisioning. This behavioral trait is characterized by the wasp preparing a nest, capturing a prey, putting it in the cell and laying an egg upon it. After the egg hatches the wasp brings in additional prey on a day-to-day basis as more food is needed. Occasionally, several prey may be brought in before egg hatch, but in any case the wasp brings in additional prey after the larva begins feeding on the first prey. Nest F contained a small wasp larva, 5 mm long and 1.8 mm wide at the middle, feeding on the first prey brought into the nest, a katydid nymph, 15 mm long, Hexacentrus unicolor (Serville), and seven other katydid nymphs, five of a species of Phaneropterinae, 9–15 mm long, one of a species of Xiphidiopsis or related genus, 10 mm long, and another H. unicolor, 20 mm long. Wasp F returned to the nest while we were digging it but she carried no prey on this trip. Nest G contained two H. unicolor nymphs, each 8 mm long, one of them bearing a small wasp larva only 1.5 mm wide at the middle; the wasp flew in with a third H. unicolor nymph, 10 mm long, just as I finished digging the nest. Both Williams and Tsuneki independently established that the Philippine, Korean, and Taiwanese subspecies practiced progressive provisioning, although Williams did not recognize that he was witnessing true progressive provisioning.

The leisurely pace of provisioning of a few wasps in 1972 suggested the possibility that a wasp might be provisioning two or more nests simultaneously such as has been recorded for several species of Ammophila. Accordingly, on my next trip to Kanneliya, 7–12 Oct 1973, we color-coded a number of wasps with acrylic colors. There were several small aggregations scattered along several kilometers of a logging road, and four Ceylonese technicians and I spent many hours of the six days watching a number of marked individuals. We found no evidence whatsoever of a wasp maintaining more than a single nest, but we obtained a great deal of information on nest construction and architecture, temporary and final closures, orientation flights, and prey transport and placement in nest.

Nesting Sites: Almost all of the nesting sites at Kanneliya were located on relatively steep banks of 45° to 80° cut along logging roads in the rain forest. Usually the wasp selected a small, more or less level area of a few square centimeters on the bank to serve as a platform from which she could dig into the steep slope. A few wasps at Kanneliya nested on more gently sloping ground above a low cut. The ground here was a heavy, damp, light brown, clayey soil containing rocks. The few wasps observed at Weddagala selected similar sites for their nests, but here the soil was a damp sandy loam with quite a high sand content.

At Gilimale and Udawattakele the wasps nested in relatively level ground, although steep roadside banks were available at the latter locality similar to those preferred at Kanneliya and Weddagala. The nests at Udawattakele were on a level area of a former carriage road where the soil was a red clay mixed with coarse grains of quartz from deteriorated sandstone. All nests at Gilimale were on level or sloping jungle trails with compacted earth due to the constant foot traffic, or occasionally in sandy loam. Nests at Kitulgala were also in compacted earth along jungle trails but we did not excavate any of them.

It appears that S. subtruncatus normally nests in bare, level, or slightly sloping areas. Henry saw nests of the Ceylonese subspecies at Kitulgala along a jungle trail. Williams and Tsuneki found that the Philippine, Korean, and Taiwanese subspecies nested in flat areas on paths, along roads, or in sparsely vegetated fields. The atypical nesting sites at Kanneliya and Weddagala may have been an adaptation to avoid nesting at the bottom of the slope where there would have been frequent runoff from the rains. The level road surfaces at these two localities were not available because the logging lorries often churned much of it into a sticky mud.

Nest Construction: When the wasp selects a suitable site for a nest, she starts to bite out pieces of the heavy earth with her mandibles, buzzing loudly as she does so. As the burrow deepens she carries out the excavated soil in the basket formed with her forelegs and mandibles, and throws it backward beneath her body to form a spoil heap. Usually she brings out a load of soil at intervals of 15 to 50 seconds. Frequently she pauses on the ground to clean her antennae after dropping a load. Occasionally, she rakes the spoil heap backward to level it and spread it over a wider area. If she is nesting on a steep bank, much of the excavated soil falls to the bottom of the bank, but if she nests on level ground the spoil heap is low and elliptical, about 12.5 cm long and 7.5 cm wide.

The time required for the wasp to dig the burrow and the cell varies with the type of soil, sometimes with external factors such as humans or vehicles interrupting the work, or commensal flies lurking near the entrance, and, perhaps, occasionally when a wasp is less industrious than others in the nesting aggregation.

Wasp RT 1 began a burrow at noon and by 1205 the burrow was 12 mm deep. She was frightened off for periods of 8 and 27 minutes by my approaching too closely, and once she left the area for 2 minutes for an unknown reason. At 1425 she began to make a temporary closure at the burrow entrance, so actual digging of this nest required about 1frac34 hours.

We found wasp 81472 D digging a burrow at 0830. She must have been working no longer than 15 minutes for the burrow was only about 6 mm deep. We did not keep her under periodic observation, but sometime before 1400 she had made a temporary closure at the nest entrance and had left the area.

Wasp YT 1 began a new nest at 1010 after completing the old one nearby. She left the nest some time thereafter and returned at 1100 and resumed digging. She was not watched during the period 1145 to 1235, but at 1238 she completed the nest and began to make a temporary closure.

Commensal sarcophagid flies belonging to the Miltogramminae sometimes frustrate completion of a nest. We saw wasp RT 6 beginning a nest at 0850 on 10 October. She was observed digging at 0925 and 1012 but had left the nest before 1030. She returned to digging at 1035 and continued until at least 1155, but had left the nest before 1222. By 1321 she was back at her nest but had not resumed digging because two Miltogramminae were hovering or alighting near the entrance. She continued to guard the nest until at least 1525, but had left it by 1555 and had not returned by 1630. The nest entrance had a temporary closure at 0830 on 11 October but the wasp was not there, nor was she at the nest at 0845, 0910, 1015, and 1050–1110. At 1115 she was digging until 1130 when she began to close the entrance. She was still guarding the entrance at 1148 and at 1220, and there were miltogrammine flies hovering around a nest lower on the bank. The wasp had left by 1300 and was not seen at the nest during our six subsequent visits ending at 1605. The wasp was not at the closed nest at 0815 on 12 October but she was working at the entrance at 0837 and left at 0840; there were no flies there during that period. When we dug up the closed nest later that day, we found only a burrow 6 cm long and the wasp had not constructed a cell.

Tsuneki and Williams found the same kind of nest construction for the Korean and Philippine subspecies. Tsuneki commented on the loud buzzing (“dzi dzi” sounds) made by the wasp while digging.

The Ceylonese wasp occasionally uses an old burrow rather than digging a new one. Presumably a newly emerged female sometimes (always ?) uses the nest from which she emerged for her first nest. We observed wasp 11775 B digging in a nest in a 20° slope of gravelly loam. There was an empty cocoon near the entrance and we presumed that she was cleaning out an old nest. Later on this same date but at a different nesting area on slightly sloping gravelly earth we saw 8–10 emergence holes with remains of cocoons on the surface adjacent to some holes, suggesting that these nests were being cleaned out for re-use. Tsuneki considered that the Korean subspecies also utilized the old cell in which they were born.

Upon rare occasions we observed females fighting when two nest entrances were close to each other. This happened when one female apparently could not locate her own sealed nest entrance and tried to open the closed burrow of another. If the rightful owner returned during this event, she would dash at the other and the two would fly around each other in tight circles buzzing loudly until the interloper flew away.

Abortive Burrows: We noted at Kanneliya the occasional occurrence of empty burrows without closing plugs on the sloping banks in which most aggregations nested. We dug up 13 of the burrows without closing plugs and found that most had been abandoned after running into a rock. These burrows were 1.5–7.0 cm long and none terminated in a cell. We watched some of the wasps excavating the abortive burrows, and found that in some instances the wasps were simply frightened off by passing vehicles or humans and immediately began another burrow on the bank a few meters away.

Also at Kanneliya we obtained an anomalous abortive burrow. Wasp RT 1 began a permanent closure of a nest at midday on 9 October. She was not at the nest at 0900 on the 10th, but she was working on the closure by 0942 until 1000. She then flew about 2 m distant to another burrow that already was about 2.5 cm deep. Three minutes later she returned to the first nest and worked on the closing plug for half a minute and then flew off. She returned to the nesting site at 1025, flew to each burrow but did not work on either, and left the area at 1029. She flew back to the first nest at 1111, stayed there half a minute but did not work on the plug. She then rested motionless on the bank halfway between the two nests, and at 1116 began to make a temporary closure in the second nest about 2 cm from the entrance. She was frightened from the area by a passing lorry at 1120. At 1140 she flew back to the first nest and worked on that closure again. At 1150 she flew to the second nest and started to make a closure about 1.2 cm from the entrance. By 1159 she had compacted so much earth into the burrow that only a shallow rounded depression marked the burrow entrance. Immediately she flew to a spot halfway between the first and second burrows and started to dig a third burrow. At 1205 this burrow was 1.2 cm deep and she began to fill it with excavated earth from the spoil heap. She made an orientation flight around the area at 1207 and left. She returned at 1246, opened the third burrow, and soon was so deep inside that she could not be seen while she deepened the burrow. She was frightened off at 1254 and returned to resume digging in the third burrow at 1302. She continued digging until 1325 when she started to make a closure about 2.5 cm from the entrance. I frightened her away at 1327 by approaching too closely but she returned to resume digging at 1354, left for 2 minutes at 1356, and then continued digging. At 1425 she began a temporary closure and finished it at 1430. When we excavated the three closed nests of RT 1 on 12 October, we found a newly spun cocoon in the first nest, were astonished to find just a burrow 6.4 cm long without a cell in the second, and a burrow 7.5 cm long ending in an oblique ovoid cell without prey in the third nest. The second nest was obviously an abortive burrow although it did not terminate at a rock. It is puzzling that the wasp made a closure rather than leaving it open as was done in all other abortive burrows.

Tsuneki found some open abortive burrows without cells from 1 to 10 cm long in the Taiwanese subspecies of S. subtruncatus, but neither he nor Williams noted such burrows in the Korean and Philippine subspecies. Foot traffic may have disturbed Tsuneki's wasps and caused their inability to relocate the burrow, because he did not mention finding an obstruction at the end of any burrow. It is possible that some abortive burrows are abandoned because the wasp has not yet made an orientation flight to memorize the precise location of the burrow.

Nest Dimensions: The burrows were cylindrical at all localities and 12–13 mm in diameter. Tsuneki noted “10 mm or so” as the burrow diameter in the Korean subspecies; that taxon is as large as the Ceylonese subspecies. In nests constructed in level or gently sloping soil (Gilimale, Udawattakele) the burrows usually went into the soil at quite a steep angle ranging from 45° to 80° (average 67° for 15 nests). However, in nests constructed in a slope from 20° to almost vertical (Kanneliya, Weddagala) the burrows penetrated the soil at a more shallow angle ranging from 15° to 75° (average 41° for 20 nests).

Normally the burrows penetrated in a straight line to the cell that sometimes was just a widened terminal section along the burrow axis, and sometimes was placed at an angle to the axis. There was a significant difference in length of the straight burrows, those in level or gently sloping soil being longer (7.4 cm, range 3.8–10.0 cm in 14 nests) than those in banks of 45° or greater slope (5.4 cm, range 3.2–10.8 cm, in 15 nests).

Tsuneki's nests in Korea and Taiwan were on level ground and the burrows penetrated the soil at a steep angle, usually 70°–80° but occasionally as much as 90° or as little at 60°. The burrows in the Korean nests were straight for 6–8 cm and then turned gradually toward a horizontal position to terminate in the cell. The burrows in the Taiwanese nests were 7–10 cm long (average 9.6 cm for 8 nests). Williams found that in the Philippine subspecies “the burrow is at first vertical, or nearly so, to a depth of about three inches, when it extends horizontally for about two or three inches more, the terminal portion being somewhat enlarged to form a cell.”

Two nests at Kanneliya were anomalous in that the burrow made an angle before terminating in the cell. One went downward at 30° for 2.5 cm and then horizontally for 4.3 cm before the cell. The other went in at 35° for 2.5 cm, then turned at 30° and continued downward for 1.3 cm to the cell. Three nests at Udawattakele had burrows with an angulation before reaching the cell. Two burrows went downward at 80° for 5.1 cm, and one then angled at 60° for 5.1 cm, the other at 45° for 5.7 cm. The third was almost vertical for 3.8 cm and then horizontal for 2.5 cm. Such angulations in burrows of the Ceylonese subspecies may be caused by an obstruction, but I noted in the second Kanneliya nest that there was no obstruction at the angulation.

The cell is elongate ellipsoidal and in 20 nests it averaged 35.0 mm long (range 25–51) and 16.0 mm in diameter (range 13–28). In two of these nests the width was 19 mm and the height 13 mm, and this is the usual proportion, i.e., that width is slightly greater than height. Two nests at Kanneliya were unusual in that they were spherical with a diameter of 19 mm, and both contained a single prey with attached egg. It is unlikely that a wasp could develop successfully in so small a cell, so perhaps each of these females might have lengthened the cell as the larva increased in size. In steep banks the cell was just an enlargement at the end of the burrow along the same axis in 16 nests but at an angle to the burrow in 13 nests. In the latter group five cells were at an oblique angle to the burrow axis and sometimes at a lesser downward angle and sometimes horizontal. Eight cells were at right angles to the end of the burrow with the burrow sometimes at one end of the cell and sometimes in the middle. Of 10 nests in level or gently sloping soil the cell continued along the burrow axis in only one nest, and was at an angle to the burrow in the others. In two of the latter nine the cell was at right angles to the burrow, and horizontal or at a lesser downward angle in the remaining seven nests. Four cells at Udawattakele in level ground were 5.0–7.5 cm beneath the surface (average 6.0 cm).

Tsuneki (1963:44) noted that the cells in Korean nests averaged 40 mm long (range 35–55), 23 mm wide (range 17–28) and 20 mm high (range 15–25), but that the width always exceeded the height. Usually the burrow curved more or less toward the horizontal and terminated in a cell along the burrow axis, but he found that occasionally the cell was at an angle of 70°–90° to the burrow axis. The top of the cell was 3–13 cm below the surface (average 6.2 cm for 26 nests). In the Taiwanese subspecies Tsuneki found that five cells averaged 40 mm long (range 35–45), 27 mm wide (range 25–40) and 20 mm high (range 15–25). Three of these cells were 7–9 cm below the surface (average 7.7). Williams gave no dimensions for cells of the Philippine subspecies.

It should be noted that each wasp we observed making multiple nests constructed the second and later nests near the first nest, usually at distances not greater than 0.75–1.5 m.

First Temporary Closure: After the cell has been excavated the wasp makes the distinctive first temporary closure. Usually she constructs a thin mud plug about 12 mm from the burrow entrance. If there is a spoil heap of excavated earth beneath the burrow entrance, she scratches some of the loose soil backward beneath her body with her forelegs into the burrow and compacts the particles by pressing with her head, while buzzing loudly. Occasionally while compressing the earth, the wasp rotates around so that she is sometimes lying on her side or back. If the nest is on a steep bank and there is no spoil heap, she bites pellets of earth from above the burrow entrance or even from the upper end of the burrow, rotating within the burrow to maintain its cylindrical shape.

These pellets are then compacted by pressing with her head. Within five minutes she usually completes the closure by raking loose particles beneath her into the mouth of the burrow and scraping some around over the surface to conceal the entrance. One female (81472 C) behaved abnormally in making her first nest closure. She began the closure at 1153 in the manner noted above but interrupted closing activities frequently for no apparent reason. During the next 19 minutes she stopped closing activities a dozen times for periods ranging from a few seconds to 35 seconds, and once flew off into the jungle for as long as five minutes, perhaps to visit flowers for nectar. During the shorter absences she made brief irregular flights along the bank near the burrow or perched on vegetation nearby. She terminated the closure at 1212 in the usual manner by scratching some loose soil backward beneath her to fill the burrow and conceal the entrance. She returned three times to the burrow vicinity for a few seconds by 1258 but did not carry on further closure activities, and made an orientation flight only after the first visit.

Orientation Flights: We observed two initial orientation flights made by females just after excavation of a nest and construction of the first temporary closure. Each nest was the second and third, respectively, that we had seen each wasp construct. The first wasp began a series of short, low, irregular orientation flights along the bank on each side of and above the nest entrance. After each flight she alighted near the entrance for a few seconds before starting the next flight. Then she sat at the entrance for 15 minutes alternately flicking her wings and grooming herself, and then made three more brief flights. She was still sitting at the entrance when I left for another nesting site 10 minutes later.

The second wasp completed the first temporary closure at 1430 and spent the next minute making a series of irregular flights low over the bank. The flights began in the area about a meter around the entrance and gradually lengthened to a distance of 2.0–2.5 m on either side of the entrance. She then flew into the jungle at 1431, returned preyless at 1537, alighted briefly near the entrance, flew off and had not returned when I left at 1608.

Later orientation flights were basically the same except that there were fewer of them after provisioning was in progress and the wasp became more familiar with the precise location of the nest. Sometimes the flights were omitted when the wasp was thoroughly familiar with the area.

Usually wasps had no difficulty finding their nests. However, we confused some females by inserting a small twig in the entrance plug or covering the entrance with a bit of leaf so that we could determine whether the wasp had entered during our absence. Usually the wasp located the nest after a little searching, and sometimes she removed the marker with a loud buzz.

Subsequent Temporary Closures: After provisioning has begun later closures of the nest are different from the initial closure in that the wasp makes no attempt to conceal the burrow entrance by scattering mud pellets over a wider area. The upper part of the burrow is filled with compacted earth but this plug may end as much as 12 mm from the surface or it may fill so much of the burrow that only a dimpled depression is visible to mark the burrow entrance. Occasionally when the plug does not nearly fill the burrow, the wasp may place loose pellets of earth in the upper end.

Tsuneki reported that temporary nest closures in the Korean and Taiwanese subspecies were made of compacted earth and that the burrows were not filled flush with the surface. He did not mention that loose pellets of earth were placed in the upper end of the burrow in any of these nests. He stated that the closure made immediately after nest construction was the same as later closures, i.e., that the burrow entrance was not concealed by loose pellets of earth scattered around the area.

Final Nest Closure: We never observed a wasp making a final closure, but we dug up three nests that contained either a cocoon or a full-grown larva. It is clear from these nests that the final closure differs from the temporary closures. The wasp does not construct a mud plug either at the anterior end of the cell or about 12 mm from the entrance. Instead she fills the burrow with loose earth pellets and then constructs a firm but thin mud plug flush with the surface or not more than 5 mm below the surface. She does not place any loose pellets of earth above this plug. It is noteworthy that in these three nests there were scarcely any inedible prey remains such as tegmina, legs, or other heavily sclerotized parts. Apparently the wasp cleans such debris out of the nest before proceeding with the final closure.

Prey Transport and Provisioning: We were never so fortunate as to witness a wasp capturing prey, and neither Williams nor Tsuneki saw this aspect of wasp behavior. All of the wasps we observed flew off into the jungle after completion of the first temporary burrow closures and orientation flights. Most prey that we recovered from nests or captured with wasps were predominantly nymphs of Tettigoniidae (long-horned grasshoppers) although some adults were found. However, we recovered from one nest a paralyzed adult female cricket (Gryllidae) and five Tettigoniidae, and in another nest we found two adult females and a dismembered nymph of Tettigoniidae, and a dismembered adult male probably belonging to Gryllacrididae. Williams and Tsuneki reported that the prey of the Philippine, Korean, and Taiwanese subspecies were mostly nymphal Tettigoniidae (Williams called them by the incorrect name Locustidae), but Tsuneki said that some adults were used later in the season in Korea.

Only once were we able to ascertain the length of time that elapsed between completion of the first temporary closure and bringing the first prey to the nest. This wasp completed the closure at 1153 and returned with her first prey at 1345.

The wasp flies to the burrow with her first prey held beneath her, venter up and head forward. She alights near the burrow entrance, leaves the prey on the ground venter up, rapidly chews through the mud plug sealing the burrow, and scatters the loose earth beneath and behind her. Then she backs out of the burrow, turns around, grasps the prey by the head end and backs into the burrow dragging the prey in after her. If the nest is on a steep bank, the prey frequently falls to the bottom of the slope. After opening the burrow, the wasp searches for the prey, flying up and down facing the bank until she finds it. Then, she either flies back to the burrow with it and pulls it inside, or drags the prey backward on foot up the bank until she reaches the burrow and pulls in the prey.

The wasp places the prey on its back, head inward at the posterior end of the cell as we had ascertained by digging up 10 nests that contained only the first prey and egg. The wasp remains in the nest a few seconds, during which period she lays the egg. Then she emerges and makes the closure as described above (see “Subsequent Temporary Closures”). Tsuneki noted the same placement of the first prey in nests of the Korean and Taiwanese subspecies, but said that occasionally this prey was placed on its side rather than on its back.

In progressive provisioning the wasp normally opens the burrow occasionally after storing the first prey. This is to ascertain when the egg is about to hatch or has hatched. She then provides another prey, followed by additional prey at more frequent intervals as the wasp larva becomes larger and eats more voraciously. I presume that this is the normal sequence in Sphex subtruncatus krombeini, although our team could not obtain substantiating data by watching color-coded females. It is possible that the second prey is brought in before the egg hatches, for we dug up one nest that contained two prey, one of them bearing the wasp's egg. Perhaps we may have dug up the 10 nests that contained only a single prey before the wasp had time to supply a second. Tsuneki reported that six nests of the Korean subspecies contained only a single prey with egg, but that one nest contained four prey with an egg upon one.

Prey provisioning proceeds more rapidly after the wasp larva has grown some, probably to at least the second instar. In working with color-coded wasps we frequently inserted a small twig or grass stem into the closing plug. This marker would be removed when the wasp opened the nest to add another prey, although we noted that sometimes a preyless wasp would remove the marker if she visited but did not open the nest. For example, we found wasp No. 10 at her nest at 0845 on 11 October, placed a marker there at 0910, and this had been removed by 1015 when we revisited the area. We placed another marker, which the wasp removed when she brought a small nymph at 1120, closed the nest and left at 1130. Another new marker was gone by 1300, and still other new markers by 1340 and 1416. A final new marker on that date was still in place when we left the area at 1605. On 12 October we dug up the nest at 1115 and found a wasp larva about a third grown and whole specimens or remains of eight Tettigoniidae (five nymphs and three adult males), 10–14 mm long. Another wasp, GT 5, was noted closing her burrow at 0855 on 8 October after bringing in a tettigoniid nymph. This may have been her first prey because she continued to bring in prey and remove markers through 11 October. We saw her bring in a prey on 9 October and she removed three markers on that date. On 10 October we saw her bring in another prey, and she also removed three markers by 1630. We did not see her bring in prey on 11 October but she removed four markers. She did not visit the nest during the morning of 12 October and we excavated the nest early that afternoon. It contained a half-grown wasp larva, nine dismembered or whole tettigoniids, (five nymphs, two adult females and two adult males), 10–30 mm long, and an adult female gryllid, 10 mm long.

The prey never recover from the paralysis of the wasp's sting. Most uneaten prey found in the cells or captured with the wasps were capable of weak reflex movements of the palpi, antennae, and tarsal segments, and of excretion, respiration, and circulation as was noted also by Tsuneki for prey of the Korean subspecies.

It was not possible to ascertain the total number of prey stored per cell because the wasp removed most inedible fragments either before or at the time of final closure. Also, there was frequent considerable variation in size of prey within a single cell which would have affected the total number stored. In five cells, none of them completely stored, the range in length of prey was 10–14, 10–16, 10–18, 10–30, and 15–30 mm. The largest number of dismembered and whole prey found was 10 in the nest of GT 5 as summarized two paragraphs above.

Prey Identification: Listed below are the species of prey obtained from nests or wasps. Also included are the number of specimens, stage and sex where known, and the range in size. Measurements were taken from the head to the apex of the abdomen excluding the tegmina. The latter in adults may extend well beyond the tip of the abdomen but are not consumed by the larva. Including the tegmina in the length would give an erroneous impression of the bulk of the prey available for consumption. A number of prey from cells were too dismembered or fragmentary for positive identification and are not included in the tally. The prey were pale green or light brown according to the species except for the strikingly marked Gonatacanthus werneri Karny. This species is black with light yellow clypeus, labrum, and palpi, and a scarlet abdominal venter. All adult prey were fully winged except for G. werneri, which is flightless with very abbreviated tegmina.