nena masthead
NENA Home Staff & Editors For Readers For Authors

 

Nesting Behavior of the Tachysphex terminatus Species Group (Hymenoptera: Crabronidae)
Frank E. Kurczewski

Northeastern Naturalist, Volume 16, Monograph 3 (2009): 1–88

Full-text pdf (Accessible only to subscribers.To subscribe click here.)

 



Access Journal Content

Open access browsing of table of contents and abstract pages. Full text pdfs available for download for subscribers.

Issue-in-Progress: Vol. 31 (1) ... early view

Current Issue: Vol. 30(4)
NENA 30(2)

Check out NENA's latest Monograph:

Monograph 23
NENA monograph 23

All Regular Issues

Monographs

Special Issues

 

submit

 

subscribe

 

JSTOR logoClarivate logoWeb of science logoBioOne logo EbscoHOST logoProQuest logo


2009 NORTHEASTERN NATURALIST 16(Monograph 3):1–88 Nesting Behavior of the Tachysphex terminatus Species Group (Hymenoptera: Crabronidae) Frank E. Kurczewski* Abstract - The nesting behavior of species in the Tachysphex terminatus group is presented in phylogenetic order following Pulawski (1988). Eight taxa are treated: T. clarconis, T. antillarum, T. alpestris, T. linsleyi, T. terminatus, T. similis, T. a. apicalis, and T. apicalis fusus. This paper includes a literature review summary for each species, Howard E. Evans’ unpublished field notes on T. linsleyi, T. terminatus, and T. similis, unreported insect museum prey records for T. terminatus and T. apicalis fusus, and long-term field studies on T. terminatus, T. similis, T. a. apicalis, and T. apicalis fusus. The field observations were made during the years 1960–2005 and include 929 nests excavated and examined at 52 locations in the central and eastern United States and southern Ontario. Emphasis is placed on: burrow excavation, tumulus leveling, temporary closure, orientation flight, hunting, prey capture, prey transport, nest entry, final closure, burrow design and length, cell depth, cell shape and size, number of cells per nest, cell sequential position, number of prey per cell, prey and wasp weights, total prey biomass per cell, position of prey in cell, and placement of wasp’s egg on prey. Introduction Tachysphex is a very large genus of ground-nesting solitary wasps, with 444 species worldwide (Pulawski 2008). Eighty-three species inhabit North America, Central America, and the Caribbean Region (Pulawski 1988). The majority of nearctic species is found in the western United States and northern Mexico in habitats with high temperature and low humidity (Pulawski 1988). Many of the species have two or more generations per year in the Austral and Tropical regions. Species of Tachysphex are largely cursorial and fast moving (Williams 1914), as the genus name implies (tachys = swift, sphex = wasp). The North American species excavate mainly short burrows and shallow cells in sandy, gravelly or, rarely, loamy soils. They store the cells with paralyzed orthopteroid insects, primarily nymphal grasshoppers (Acrididae) (Krombein 1979, Pulawski 1988). The Tachysphex terminatus species group is restricted to the New World, ranging from Alaska to Argentina (Pulawski 1988). Seven of the 11 species in the group are found in the United States: T. clarconis Viereck, T. ruficaudis (Taschenberg), T. alpestris Rohwer, T. linsleyi R. Bohart, T. terminatus (F. Smith), T. similis Rohwer, and T. apicalis W. Fox (Pulawski 1988). There are four extralimital species: T. antillarum Pulawski and T. quisqueyus Pulawski (West Indies), T. galapagensis Rohwer (Galapagos Islands), and T. peruanus Pulawski (Peru). *PO Box 15251, Syracuse, NY 13215; Fkurczewski@twcny.rr.com. 2 Northeastern Naturalist Vol. 16, Monograph 3 Species in the group are characterized morphologically by the presence of convexities behind the posterior ocelli (Bohart 1962, Pulawski 1988, Williams 1914), wide vertex (space between compound eyes at top of head; Fox 1894, Pulawski 1988), mainly long erect setae on the vertex (Pulawski 1988, Williams 1914), and unmodified apical tarsal segments (Pulawski 1988). Males and females resemble one another in size and coloration. Members of the species group are rather small in size (3.5–10.0 mm long) and often black with a red abdominal tip, hence the species names ruficaudis, terminatus, and apicalis. Intraspecific variation often includes abdominal coloration and wing infuscation (Pulawski 1988). More red on the abdomen and hyaline wings are characteristic of arid regions; black individuals with more infuscated wings are found in areas of higher relative humidity, cooler temperatures, and higher elevation. Species in the terminatus group have two or more generations per year throughout much of their range. They nest mostly in level or slightly sloping, barren or sparsely vegetated sandy soil. However, T. apicalis nests in sand cliffs and strongly sloping sandbanks (Kurczewski and Snyder 1968). Nesting behavior is characterized by several group-specific components. Except for T. apicalis, females spend much time on the sand surface leveling the soil removed during burrow excavation. After completing the burrow, they temporarily close the nest entrance with sand and then make an orientation flight above the area before searching for prey. The mostly small, prey grasshoppers are transported to the nest, usually in flight, and released headfirst in front of the entrance, followed by removal of the temporary sand closure, nest entry, exit, and re-entry with prey. In T. apicalis, leveling the sand and making and removing the temporary closure of the entrance are absent due to the limitations of the sand cliffs and strongly sloping sandbanks. The method of final closure in this species is group-atypical in accord with its unusual habitat (Kurczewski and Snyder 1968). Under optimal nesting conditions, a finished nest often contains two to five cells, with several or many paralyzed prey placed in each cell. The wasp’s egg is laid transversely across the pro- and mesosterna of a grasshopper, being affixed by the less tapered, distal end to the soft intersegmental membrane surrounding the base of a forecoxa. The terminatus group contains both sympatric (T. terminatus, T. similis) and allopatric (T. terminatus, T. alpestris; T. apicalis fusus, T. a. apicalis) forms with broad ranges, and relatively isolated forms (T. antillarum, T. quisqueyus) restricted to the West Indies. The specializations that enable the sympatric species to co-exist, often side-by-side in mixed aggregations, and those that restrict the isolated allopatric populations to particular geographic areas present interesting evolutionary phenomena for study. Although the morphological distinctions are sometimes subtle, ecological and behavioral differences offer reliable taxonomic characters for some of the species. Species of Tachysphex are a rewarding group for study because of their varied nesting behavior (Bohart and Menke 1976). 2009 F.E. Kurczewski 3 The purpose of this paper is (1) describe, categorize, and analyze the nesting behavior of species in the terminatus group, paying particular attention to species-specific components; (2) compare species behavior patterns in order to find similarities, contrasts, omissions, and additions; and (3) correlate species differences with pertinent morphological characteristics to make possible a functional interpretation of what would otherwise be obscure structural features. Methods This paper is an attempt to assimilate all pertinent information on the nesting behavior of species in the Tachysphex terminatus group. Such information is presented by species in phylogenetic order following Pulawski (1988): T. clarconis, T. antillarum, T. alpestris, T. linsleyi, T. terminatus, T. similis, T. a. apicalis, and T. apicalis fusus (Fig. 1). A specific order of information is presented under each species. First, the known literature on species’ nesting behavior is reviewed. Second, Howard E. Evans’ unpublished notes on T. linsleyi, T. terminatus, and T. similis are detailed. Third, previously unreported prey records from several insect museums are presented. Fourth, long-term field studies on T. terminatus, T. similis, T. a. apicalis, and T. apicalis fusus are described, categorized, and analyzed. Aggregations of T. terminatus, T. similis, T. a. apicalis, and T. apicalis fusus were located with the aid of state/province soil association maps, United States Department of Agriculture soil surveys, aerial photographs, and state road maps. The aerial photographs were especially useful in locating sizeable areas of bare sand and gravel such as abandoned and active sandpits, gravel pits, sandbanks, and sandblows. Figure 1. Hypothetical phylogenetic tree of species in Tachysphex terminatus group based mainly on external morphological characteristics (after Pulawski 1988). 4 Northeastern Naturalist Vol. 16, Monograph 3 Nine hundred twenty-nine nests were excavated and examined at 52 sites in the central and eastern United States and southern Ontario during the years 1960–2005. Site location, site description, site code number, and inclusive date(s) of observation are given for T. terminatus and T. similis in Tables 1 and 7, respectively. Site code numbers, descriptions, locations, and inclusive date(s) of observation for other species are included in the text. Site code number (TX-number) associates field notes with specimens and identifies the species of wasp and prey, site location, and date(s) of observation and collection. At least one female or male was collected from each site to serve as a voucher specimen. The specimens have been deposited in the Cornell University insect collection. Soil samples from nesting sites in New York (TX-1, 2, 20, 36, 76, 94), Kansas (TX-54, 56, 63, 67), and Florida (TX-4, 7) were analyzed at the State University of New York College of Environmental Science and Forestry, Syracuse, NY. Samples were screened to separate fine gravel, coarse sand, medium sand, fine sand, and silt/clay using the United States Department of Agriculture mm scale. Samples were sorted into particle size, weighed, and percent sample weight calculated. Soil color, hue, and value/chroma were delineated by comparison with the Munsell soil color chart. Emergence data on first and last generation T. terminatus (TX-56, 54, respectively) and first generation T. similis (TX-67) were obtained through direct observation of recently emerged individuals with full wings, complete foretarsal digging rakes, and clean integument. Recently emerged first-generation females of both species were hand netted, frozen, examined under a microscope for specific determination, and weighed (wet) on a Mettler balance. Tachysphex similis included both all-black and red abdominal apex variants. The weights of captured individuals were plotted against dates of emergence to determine if there was a size-emergence date correlation. Some live females were hand-netted, marked on their scuta with nontoxic paint, released at the site, and followed. Nests were marked with colored toothpicks that were dated and numbered chronologically. The wasps accepted the presence of the toothpicks and often used them for orientation. Nest density involved measuring the distance between adjacent nest entrances. Nesting aggregations were observed daily from 0800–0900 to 1700–1830 hours (Eastern or Central Daylight or Standard Time), weather permitting. Type of wasp activity was noted, described, and quantified. Some behavioral components were photographed in the field. Emphasis was placed on burrow excavation, tumulus leveling, temporary closure of the entrance, orientation flight, hunting, prey capture and transport, nest entry, and final closure. Wasps flying to and from nests were timed with a stopwatch between consecutive provisioning trips and from entry to exit. Tachysphex terminatus females in New York (TX-1, 20, 36 [34 individuals]) and Kansas (TX-54, 56 [28]) and T. similis females in Kansas (TX-55, 61, 63, 66, 67 [10]) and Florida (TX-4, 7 [16]) were marked, observed, and 2009 F.E. Kurczewski 5 their activities measured during provisioning sequences. Time intervals between consecutive prey items and duration to open temporary closure, turn around inside the burrow and reappear in the entrance, place prey in the cell, and temporarily close the nest entrance with sand were recorded using a stop watch. Time of day, temperature in degrees centigrade, and cloud cover were recorded concurrently with these data. Tachysphex similis and T. terminatus differed in their rates of activity at the same temperature. Two behavioral components selected for study exemplified this difference: (1) temporary closure of the entrance and (2) removal of the closure prior to nest entry. Wasps were timed with a stopwatch as they temporarily closed their entrances with sand and, later, when they returned with prey and removed the sand fill. Sand surface temperatures were simultaneously recorded from unshielded, narrow-range (range = 8 ºC), high-temperature mercury thermometers (ASTM RICCO) placed beside entrances. Both behaviors were examined over the range of sand surface temperatures at which these species nested (29–63 °C and 28–65 °C, respectively). The data were erratic and curvilinear below 35 ºC. Therefore, only measurements from 35° to 60 ºC were used for ease in comparison and analysis. An attempt was made to obtain at least 20 measurements for each species in each region for every degree centigrade in the 35–60 °C range. Tachysphex terminatus was studied in New York and Kansas and T. similis was examined in Florida and Kansas to account for geographic variation. Seven of the sites selected for comparison were chosen based upon equivalent soil texture values (Tables 2 and 6). For example, both TX-1 and TX-54 had >80% fine gravel and coarse sand, TX-36 and TX-56 had >70% coarse sand and medium sand, and TX-7 and TX-67 had approximately 70% fine gravel and coarse sand. The lower sand surface temperatures were obtained in the morning as the sun was rising to an overhead position. The higher sand surface temperatures were obtained mainly between noon and mid-afternoon when the sun was at zenith or just beyond zenith. No temperatures were recorded after mid-afternoon when the sun was declining and the sand was gradually and steadily cooling. The time lag between the thermometer temperature and the sand surface temperature would have introduced an error of 1–2 ºC under late afternoon conditions. Burrows and cells were excavated, examined, measured, and drawn. Burrow design and length, number of cells per nest, cell orientation and sequential position, cell depth, cell shape and size, number of prey per fully provisioned cell, position of prey in cell, and placement of wasp’s egg on prey were recorded in the field. Prey were removed from the cells, put in individual vials according to nest number, placed on ice in a cooler, transported to a laboratory, and weighed (wet) on a Mettler balance. The aggregate prey weight of each cell was summed. The grasshoppers were then transferred to vials with glycerol, code-labeled, and sorted to subfamily. The majority of prey was sent to US Department of Agriculture and National Museum of Natural 6 Northeastern Naturalist Vol. 16, Monograph 3 History orthopterists for identification. The wasps were placed in individual vials, put on ice, transported to a laboratory, weighed (wet), frozen, pinned, and code-labeled. The wasp species were tentatively identified by the author (F. Kurczewski) and, later, the identifications verified by one of three hymenopterists. Live wasp eggs were removed from the pedestal prey in the field and measured using a binocular microscope with an ocular micrometer. The contents of five fully provisioned cells in Kansas (TX-54) and 10 fully provisioned cells in New York (TX-1) were removed from an excavation, placed in individual rearing tins, and observed on a daily basis until adult emergence. Larval development and growth, beginning of cocoon construction, cocoon completion, and adult emergence were observed and recorded. Two-sample t-tests using independent samples were computed to compare means for the following T. similis and T. terminatus data: wasp generation and time spent between consecutive prey items; wasp generation and individual prey, aggregate prey per cell, and wasp weights; T. similis Florida and Kansas wasp weights; oviposition prey and other prey in cell weights; T. terminatus and T. similis all sites temporary closure and removal of sand fill duration; and T. terminatus and T. similis Kansas sites temporary closure and removal of sand fill duration. Least-squares regression line, Pearson’s linear correlation coefficient (r), and/or coefficient of determination (R2) were computed for the following T. similis and T. terminatus data: emergence date and wasp weight, and wasp weight and cell depth, T. similis Florida and Kansas wasp weight and cell depth, and mean duration of temporary closure and temporary closure removal. Statistical significance for all tests was assumed when P < 0.05. Bar graphs and standard error bars were computed for mean sand removal interval, mean sand removal distance, mean duration spent distributing sand, mean sand retrieval duration, mean sand retrieval distance, and mean sand retrieval interval for T. similis and T. terminatus. Tachysphex clarconis Literature review Evans (1970) reported on the nesting behavior of this species as T. sp. near linsleyi, later corrected to T. clarconis (Kurczewski and Evans 1986). Evans discovered two probably unfinished nests at Jackson Hole, WY. One nest had two cells “side by side,” about 1 cm apart, at a depth of 5 cm. One cell contained four “very small” grasshoppers and the other cell held five prey. The wasp’s egg was laid on the prosternum of a grasshopper transversely behind the front coxae. The grasshoppers, identified as early instars of Melanoplus sp., were “considerably smaller” than the “small” wasp and measured only 4 mm in body length. A small, immature Melanoplus sp. was collected later with a second wasp. A third female was observed flying with prey to her nest, holding the grasshopper with her mandibles and legs. She released it near the entrance, removed the temporary closure, entered the burrow, reappeared in the opening, and pulled in the grasshopper “by its antennae.” This nest had only one 2009 F.E. Kurczewski 7 cell at a depth of 3.5 cm. The cell held three immature grasshoppers, probably Chloealtis conspersa (Harris) (Acrididae), 6–8 mm long. The cell was likely unfinished as no mention of the wasp’s egg was made. Elliott and Kurczewski (1985) noted four females of T. clarconis nesting in sand at St. Anthony, Fremont County, ID. One female made 40 entries to remove soil from her burrow during excavation. An unidentified miltogrammine fly attended the excavating wasp. The amount of time the wasp spent between consecutive entries and exits onto the surface to level the removed sand increased from start to finish. After 44 min of excavation, the female made hovering flights above the area, flew off to rest on vegetation, returned, and temporarily closed the entrance with sand. Prey grasshoppers were then carried consecutively in extended flights directly to the nest and deposited near the entrance. One “rather large” grasshopper was brought to the nest in a series of short flights. With each prey, the female removed the sand closure, entered the burrow, reappeared headfirst, and pulled in the grasshopper by its antennae. Three cells contained three to six grasshoppers identified as Melanoplus sanguinipes (Fabricius) (8 examples) and M. foedus Scudder (1). A Hedychridium sp. (Chrysididae), a probable cleptoparasite on Tachysphex wasps, dug into one nest while the wasp was away. Tachysphex antillarum Literature review Tachysphex antillarum was observed nesting at four locations in Cuba (Genaro and Sanchez 1992). All sites consisted of bare friable soil, the largest area being 106.5 x 33 m. The soil selected for nesting ranged from “pure fine sand” to “black, scarcely friable soil.” Although nesting in bare ground, most nests were dug beside stones. This species was multivoltine and nested throughout the year. Females always leveled the soil removed from burrow excavation. After finishing a burrow, the wasp temporarily closed the entrance with soil. Prey grasshoppers were transported to the nest in flight, being held with the wasp’s mandibles and legs. The prey was released about 1 cm from the entrance in a ventral side upward position after which the wasp removed the temporary closure, entered, reappeared in the opening, and pulled in the grasshopper by the antennae or, rarely, a leg. Nests contained up to five cells. Burrow length averaged 5.9 ± 1.04 cm (range = 4.5–8.5, n = 28). Cell depth averaged 2.3 ± 1.14 cm (range = 1.0–5.2, n = 28). Six to 11 grasshoppers (mean = 8.2 ± 1.86, n = 20 cells) were placed in the cells head inward and often ventral side upward. Prey averaged 4.9 ± 1.15 mm (range = 3.0–9.0, n = 208) in body length. The nymphal grasshoppers were identified as Parachloebata (= Orphulella) scudderi Bolivar (148 examples), Chortophaga cubensis Saussure (35), and an unidentifiable species (22). The wasp’s egg was placed on the innermost prey in the cell, usually one of the largest individuals, and affixed in the genus-typical position. Wasp cocoons averaged 7.2 ± 0.70 mm (range = 6.0–8.1, n = 21) in length. 8 Northeastern Naturalist Vol. 16, Monograph 3 More than half the cells (53.6%) were infested by maggots or puparia of two species of miltogrammine flies, Senotainia (= Sphixapata) rubriventris (Macquart) and Senotainia (= Sphixapata) trilineata (Van der Wulp). Some provisioning wasps altered their flight patterns or did not enter their nests upon the presence of such flies. Some females flew at the flies in an attempt to drive them away. Wasps often inspected prey with their antennae and mouthparts following fly intervention. Tachysphex alpestris Literature review The taxonomic status of T. alpestris is currently in question. Elliott and Kurczewski (1975) treated the taxon as a western population of T. terminatus. Pulawski (1988) elevated the form to specific status, although he indicated that it might be a geographical race of T. terminatus. Buck (2004), after analysis of specimens where the two taxa intergrade on Manitoulin Island, ON, Canada, believed that T. alpestris is merely a subspecies of T. terminatus. Regardless of its status, the species or subspecies is an inhabitant of psammophilous sites. Buck (2004) reported it from sandy areas, sand dunes, dunes, dune trail, beach, sandy forest edge, old sandpit, and sandy field. Evans (1970) observed T. alpestris, identified initially as T. terminatus (Kurczewski and Evans 1986), nesting at Jackson Hole, WY. He reported this species making a temporary closure of the nest entrance, excavating one- and three-celled nests, preying on small acridid nymphs, and stocking several prey per cell. Cell depths were 3.2 cm for the one-celled nest and 4.0–4.5 cm for the three-celled nest. The wasp’s egg was placed on a prey in a cell in the position typical of the genus. Twenty prey grasshoppers from the nests were all identified as immature Chorthippus curtipennis (Harris) (Acrididae), 5–9 mm (mean = 6.5) long. F.E. Kurczewski, using Evans’ field notes, added that prey transport was by flying and the three-celled nest contained three (no egg), two (egg), and 13 (egg) prey per cell (Kurczewski and Evans 1986). Burrow lengths were 5 cm for the one-celled nest and 8–10 cm for the three-celled nest (Kurczewski and Evans 1986). Elliott and Kurczewski (1985) noted a female from Morro Bay, San Luis Obispo County, CA pinned with a nymphal acridid, Oedaleonotus sp. Pulawski (1988) reported three females pinned with prey Acrididae: (1) no locality label (actually the wasp, 8.5 mm long and determined as T. terminatus by F.E. Kurczewski, was from Santa Cruz, Santa Cruz County, CA), “two young nymphs” of Melanoplus sp., 7.0 and 8.5 mm long; (2) Sierra Valley, CA, “four young nymphs” of Melanoplus sp., probably sanguinipes; and (3) Arcata area, Humboldt County, CA, nymphal Microtes sp., probably helferi (Strohecker). 2009 F.E. Kurczewski 9 Tachysphex linsleyi Literature review Krombein (1967) recorded unidentified acridid nymphs as prey of this species. A female from Rodeo, Hidalgo County, NM is pinned with an immature Eremiacris (= Paropomala) pallida (Bruner)(Acrididae)(Pulawski 1988). H.E. Evans’ field note H.E. Evans’ field note no. 1867 (1 mile w. Lajitas, Presidio County, TX; April 28, 1963) refers to T. linsleyi (det. R.M. Bohart). The female, 7.0 mm long, excavated a nest at 1000–1100 hours in “soft, powdery clay-sand” in the side of a footprint. She temporarily closed the entrance at about 1100. By 1500, the wasp had entered, exited, and temporarily closed her nest several times. At 1605, she entered her burrow without prey, exited, made a temporary closure, and left. The female was observed making a final closure at 1715, coming out frequently and backing into the burrow while raking soil backward. The rather straight burrow was 4 mm in diameter. It was 9 cm long and led to a cell, 6 X 11 mm, at a depth of 6.5 cm including cell height. There were seven acridids inside, with the wasp’s egg affixed near a front leg of a grasshopper at the back of the cell. The immature grasshoppers were identified as Psoloessa sp., possibly texana Scudder (Gomphocerinae) (6 examples, 4.5–6.5 mm long) and Clematodes larreae Scudder (Melanoplinae) (one, 9.5 mm long). Tachysphex terminatus Literature review Ashmead (1894) was the first to report on aspects of the nesting behavior of T. terminatus. Citing C.V. Riley’s manuscript notes, he recorded this species preying on the acridid Chortophaga viridifasciata (DeGeer). Rau (1926) also found T. terminatus preying on C. viridifasciata. Williams (1914) collected a female of this species in Kansas with a “young tryxalid” (= Gomphocerinae). He noted another female nesting in sand in Graham County, KS and making a temporary closure of the entrance. The wasp later finally closed this nest, smoothed over the area, and took flight. The short burrow led to a single, fully provisioned cell that contained “several acridians of very small size.” Rau and Rau (1918) found T. terminatus at two sites in/near St. Louis, St. Louis County, MO. Females at the first site nested in sandy soil, those at the second site in “clayey” soil. Wasps that dug in clayey soil made numerous false starts, whereas those digging in sandy soil usually finished the excavation, unless disturbed. In digging a burrow, the female flung sand backward beneath her body with her forelegs. Periodically, she backed from the burrow to rake away the accumulated sand in front of the entrance. After excavating a burrow, the wasp closed the entrance with loose sand. She always filled the entrance completely, although the closure was “loosely packed.” One female made no orientation flight—presumably because she was already familiar with her surroundings. 10 Northeastern Naturalist Vol. 16, Monograph 3 Rau and Rau (1918) noted that females with prey flew directly to their nests, holding the grasshopper beneath the body. One wasp released her prey on the sand, removed the temporary closure, went inside, turned around, came out, and pulled in the grasshopper. She reappeared in the opening, remade the closure, and flew away. After stocking the cell, the female filled the burrow with loose sand and attempted to conceal the entrance by covering it. A burrow in sandy soil comprised a short, curving channel “about three inches long” terminating “about an inch and a half below the surface.” Rau and Rau (1918) erroneously reported that there was no cell. Rau (1926) noted that nests dug in loose dirt on the top of a “clay bank” were so shallow that they could be readily exposed by a person blowing on them. Five to nine prey were “usually used to provision a cell.” Rau and Rau (1918) mentioned that one larva fed on the ventral side of the grasshopper between the second and third pairs of legs. Prey included small and mediumsized nymphs of Arphia sulphurea (Fabricius), Syrbula admirabilis (Uhler), Dichromorpha viridis (Scudder), Melanoplus sp., and an unidentifed Oedipodinae (Rau 1946, Rau and Rau 1918). Strandtmann (1953) noted T. terminatus nesting in loose soil under a farm implement shed near Maxwell, Caldwell County, TX. He believed the wasps nested from May through October, allowing five generations of wasps per year at the rate of one generation per month. Burrows sloped evenly at about a 45° angle for 7–10 cm, ending in enlarged cells about 3 cm deep. “Two cells contained seven small hoppers each.” The number of grasshoppers per cell “was inversely proportional to their size.” An “average of four” grasshoppers per completed cell was his “conservative estimate.” Grasshoppers removed from the cells comprised first and second instars of Melanoplus and Tryxalus. Tryxalus is not a New World genus (Finnamore 1982). Strandtmann (1953) figured one wasp atop its prey in an atypical dorsal side upward position. Kurczewski (1966a) reported on the extremely rare use of bush katydids (Tettigoniidae) as prey in three single-celled nests of T. terminatus in a sandbank at Groton, Tompkins County, NY. The three cells each contained a single tettigoniid along with two, five, and seven acridids (Melanoplus bivittatus [Say]). A fourth bush katydid was abandoned on the sand surface beside a nest entrance. This nest was later stocked entirely with nymphal M. bivittatus. Wasps provisioning with bush katydids exhibited no atypical prey transport or storage behavior. The wasps’ eggs were not laid on any of the bush katydids, possibly because of their small sizes. Initially determined as Phaneroptera by A.B. Gurney, the specimens were later re-identified by D.A. Nickle as Scudderia sp., probably furcata Brunner von Wattenwyl (Pulawski 1988). Kurczewski and Harris (1968) noted 12 cells of T. terminatus in two 0.5- m2 plots at the same site in Tompkins County, NY. They found T. terminatus to be the second most abundant solitary wasp at the site in terms of nest density, exceeded only by Oxybelus bipunctatus Olivier. 2009 F.E. Kurczewski 11 Kurczewski and Snyder (1968) summarized the nesting behavior of T. terminatus in a paper on the evolution of cliff nesting in digger wasps. They reported that a female excavated a shallow slanting burrow in sand, then hunted and captured small grasshoppers of her own size or smaller that she stored in an oval cell at the end of the burrow. When provisioning was finished, the wasp laid an egg on one of the prey and sealed off the cell with sand. Before filling the entire burrow with sand and making a new nest, she commonly made and provisioned a second and third or even a fourth and fifth cell at the ends of short side burrows leading from the main burrow. The female closed the nest entrance with loose sand each time she went afield to search for and capture a grasshopper. When she returned with prey it was necessary for her to release the grasshopper on the sand near the entrance, remove the sand fill, enter the burrow, turn around inside, come out, and re-enter backward pulling the grasshopper in by its antennae. The prey thus lay exposed on the sand for several seconds unprotected from the attacks of satellite-flies. In the presence of the flies, a female sometimes raised the legs of the grasshopper with her mandibles in an attempt to search for fly maggots. One female nesting in a steep slope was relegated to making an incomplete temporary closure. She entered her nest with a very small prey directly without releasing the grasshopper on the sand. Miller and Kurczewski (1973) reported on the speed at which T. terminatus females opened and closed their nest entrances as sand surface temperature increased. As temperature increased, the wasp dug more rapidly and made and removed the temporary sand closure increasingly more quickly. At 57 ºC, the female made several brief hovering flights to avoid the hot sand surface. At 62 ºC, she made only a token effort to close the nest upon leaving, thereby leaving the entrance partly open. Some, but not all, females ceased provisioning during the hottest part of the day and remained inside their burrows. Spofford et al. (1986) reported on the cleptoparasitism of T. terminatus eggs and larvae by three species of Miltogrammini (Diptera: Sarcophagidae) in two sandy areas of central New York near Auburn, Cayuga County and near Chittenango, Madison County. Phrosinella aurifacies Downes larviposited primarily in temporarily closed entrances, the maggots burrowing through the loose sand fill and down the burrow. Senotainia (= Sphixapata) vigilans Allen maintained surveillance perches near wasp entrances and larviposited on the prey while the female was preoccupied with provisioning activities. Senotainia (= Sphixapata) trilineata was initially attracted by the wasp’s provisioning flight and larviposited on the prey in mid-air or on the ground. Females of T. terminatus attempted to prevent fly larviposition through face-offs, chasing flights, and diversionary flights. Wasps also attempted to remove maggots from the prey with their mandibles. Frequency of successful cleptoparasitism at the two sites was 30.6% and 57.9%, respectively. Spofford and Kurczewski (1990) documented the miltogrammine fly cleptoparasites and frequency of cleptoparasitism of T. terminatus at 12 Northeastern Naturalist Vol. 16, Monograph 3 many sites in the eastern United States. Phrosinella aurifacies, Senotainia (= Sphixapata) trilineata, and Senotainia (= Sphixapata) vigilans were the predominant cleptoparasites of this wasp species. Spofford and Kurczewski (1992) reported further on the species of Miltogrammini associated with T. terminatus females and their nests, cataloguing several pre- and postlarvipositional, counter-cleptoparasitic behavioral tactics employed by this species in counteracting fly larviposition. Kurczewski (2000) examined 2 three-celled and 3 two-celled nests of T. terminatus in recently cleared nearly level sand in Brunswick County, NC. The 12 cells were unearthed at depths of 27–48 mm (mean = 34.3) beneath the surface, including cell height. The fully provisioned cells held three to nine (mean = 6.4, n = 12) nymphal acridids identified as Melanoplus femurrubrum (DeGeer) (46 examples), Dichromorpha viridis (3), Orphulella pelidna (Burmeister) (24), and Dissosteira carolina (Linnaeus) (4). The individual grasshoppers weighed (wet) 4–19 mg (mean = 11.3, n = 77). Two of the wasps weighed 15 and 17 mg. Prey were placed in the cells head inward and ventral or dorsal side upward or on the side. The wasps’ eggs were <2 mm long and affixed to the right (7) or left (5) forecoxal coria of the prey. Females exited their overnight burrows in the morning to search for prey as early as 0910 (EDT) at air and sand surface temperatures as low as 18 ºC and 25 ºC, respectively. H.E. Evans’ field notes H.E. Evans’ field note nos. 53, 56, 67, 76, 114, and 149 refer to field studies of T. terminatus (det. R.M. Bohart) at 7 miles east of Manhattan and Blackjack Creek, both sites in Pottawatomie County, KS. They include 5 one-celled nests. The 5 burrows entered the soil at 40–50° angles with the surface and were 2–6 inches long. The cells were situated 1–3½ inches beneath the surface, including cell height. They held 3–12 nymphal acridids (n = 5). The largest prey was 13 mm long, but most others were less than 8 mm long. The grasshoppers comprised Gomphocerinae and Melanoplinae, one prey being identified as Melanoplus sp. Many of the grasshoppers were put in cells head inward and ventral-side upward. A wasp’s egg, 2 mm long, was laid at the base of a foreleg and extended transversely across the prothorax between the fore- and midlegs. The grasshoppers were carried in flight by the wasps, being held underneath with the legs and mandibles. Each prey was released on the sand surface as the female removed the temporary sand fill from her entrance. The wasp then entered the burrow, came out, grasped the grasshopper by its antennae, backed in with it, came out again, and refilled the entrance with sand. Final closure involved the wasp raking sand backward into the burrow as she backed inside. Evans’ field note nos. 701, 1105, 1139, 1200, 1477, and 1554 pertain to studies of T. terminatus (det. H.E. Evans) made in Oswego (Granby Center), Tompkins (Six Mile Creek, SQR Sand Pit, Ithaca), and Wayne (Ontario) counties, NY. The wasps’ burrows were 2–10 cm long with cells 1.5–6.5 cm 2009 F.E. Kurczewski 13 deep, including cell height. Three to seven nymphal acridids were placed in five of the cells. The wasps’ eggs, 1.5–2.0 mm long, were attached as described above. Prey transport and nest entry were as described above. Evans’ field note nos. 1840, 1898, and 1900 refer to four nests of T. terminatus (det. H.E. Evans) excavated and examined in a sandbank at Lexington, Middlesex County, MA. The four burrows were 5.0–9.5 cm long, with cells situated 3–5 cm beneath the surface, including cell height. The number of prey per fully provisioned cell ranged from two to five. The grasshoppers were identified as nymphs of Chloealtis conspersa (3 examples), Dissosteira carolina (1), and Melanoplus sanguinipes (7). Nesting behavior and position of the wasp’s egg on the grasshopper were as described above. Insect museum specimens A T. terminatus female, 8.0 mm long, in the Museum of Comparative Zoology, Harvard University (H.E. Evans, Biol. note no. 1840) is associated with two juvenile Dichromorpha viridis, 6.5 and 5.5 mm long, and one juvenile Orphulella speciosa (Scudder), 9.5 mm long. Prey body lengths from H.E. Evans’ Biological note nos. 1898 and 1900 specimens in the Museum of Comparative Zoology were as follows: Chloealtis conspersa, 9.0, 8.0, 11.5 mm; Dissosteira carolina, 7.5 mm; and Melanoplus sanguinipes, 11.0, 7.5, 6.5, 11.0, 4.5, 6.0, 6.5 mm. A female from Colorado (det. K.V. Krombein), 9.0 mm long, is pinned with two juvenile Melanoplus sp., 9.0 and 8.5 mm long. A female T. terminatus (det. K.V. Krombein) from Clifton, Prince William County, VA (“ex. 1 nest”) was collected with four juvenile Chortophaga viridifasciata, 9.5, 9.5, 7.5, and 9.5 mm long, and one juvenile Syrbula admirabilis, 7.5 mm long. F.X. Williams collected four females from Boston, Suffolk County, MA (July 1913), 8.5–9.0 mm long, with two juvenile Melanoplus sp., 7.0 and 8.0 mm long, one juvenile C. viridifasciata, 7.0 mm long, and one juvenile Arphia sp., 7.5 mm long. Four females from Ithaca, NY, the oldest specimen being dated August 10, 1886 (J.H. Comstock), 8.0–10.0 mm long, are pinned with two juvenile Melanoplus femurrubrum, 8.0 and 8.5 mm long, one juvenile Arphia xanthoptera (Burmeister), 8.0 mm long, and one juvenile genus unknown (Acrididae), 7.5 mm long. A female from Hamilton, Madison County, NY, 7.5 mm long, is pinned with a juvenile Melanoplus sp., 6.5 mm long. A T. terminatus female from Barton County, KS, elevation 559 m (1816 ft.) (June 22, 1912; F.X. Williams), 8.0 mm long, is pinned with a juvenile Mermiria sp., 8.5 mm long. Two females from the Fridley Sand Dunes, Anoka County, MN, 8.0 and 8.5 mm long, are pinned with a juvenile Melanoplus sp., 8.5 mm long, and a juvenile Dichromorpha viridis, 9.0 mm long. A female from a gravel pit at Forks of the Credit, Regional Municipality of Peel, ON, Canada is pinned with a 4th-instar Chorthippus curtipennis (M. Buck, University of Guelph, Guelph, ON, Canada, 2007 pers. comm.). Field studies Nest location. Tachysphex terminatus nested in bare or sparsely vegetated, moderately dry, loose-to-firm fine gravel, sand, or loamy fine sand. 14 Northeastern Naturalist Vol. 16, Monograph 3 Nesting sites were usually level, but some sloped as much as 25–50° with the horizon (Fig. 2). In some areas, the soil was loose and dry on the surface, but moist, gravelly, or compact underneath. Most sites were surrounded or bordered by vegetation—a necessity to provide a source of prey. Nests occupied a variety of man-made or otherwise disturbed substrates (Table 1, Fig. 2). Five rather large aggregations (TX-1, 20, 36, 54, 56) contained Table 1. Tachysphex terminatus site code numbers, site locations, site descriptions, and inclusive dates of observation. Inclusive dates Site No. Site location Site description of observation TX-1 Groton, NY Sandbank 5 Aug–1 Sep 1960, 13 Jun–29 Aug 1961, 29 May–1 Aug 1962, 6 Jun–24 Jul 1963, 6 Jun–21 Jul 1964, 13 Jun 1967 TX-2 3 km E Auburn, NY Sandbank 16 Jun–29 Aug 1962, 18 Jun–23 Aug 1963 TX-10 Kill Devil Hills, NC Sand barren 23–24 Jul 1962 TX-19 2 km W Sennett, NY Sandy field 19 Jun–22 Jul 1963 TX-20 2 km S Auburn, NY Sandpit 22 Jun–23 Aug 1963, 11–28 Jul 1964, 23 Jun–6 Oct 1966, 14 Jun–15 Sep 1967, 7 Jul 1968 TX-24 Weymouth, NJ Sandy trail 29 Jun 1963 TX-26 Scullville, NJ Sandy field 30 Jun 1963 TX-27 Oyster Creek, NJ Sand pile 1 Jul 1963 TX-31 Five Points, NY Gravel pit 7–18 Jul 1963 TX-35 Cape May, NJ Sandy beach 16 Jul 1963 TX-36 Ithaca (Inlet Valley), NY Sandy field 4–9 Aug 1963, 12–26 Jul 1964, 9–29 Jun 1966 TX-38 Presque Isle State Park, PA Sandblow 27 Jul–17 Aug 1961, 26 Jul 1963, 3–10 Jul 1964 TX-41 Vincentown, NJ Sandy field 11–15 Aug 1963 TX-44 Colonie, NY Sand barren 17–18 Aug 1963 TX-48 Port Burwell, ON Sandy field 1 Jul 1964 TX-50 Lakeview, KS Sandy lakeside 21 Aug-14 Sep 1964, 13 May–4 Jul 1965 TX-54 6 km SE Lecompton, KS Sandy RR embankment 3 Sep 1964, 9 May–27 Oct 1965, 5–30 May 1966 TX-56 8 km NW Lawrence, KS Sand flats 22 May–13 Sep 1965, 5–30 May 1966 TX-60 8 km E Manhattan, KS Sandy draw 16 May–15 Aug 1965 TX-69 4 km N Guthrie, OK Sand flats 12 Aug 1965 TX-86 2 km SE Erie, PA Gravel pit 2 Jul–29 Aug 1983 TX-95 Allegan State Game Area, MI Sandy trail 11–13 Jun 1993 TX-97 Canadian Forces Base Borden, ON Sandy trail 27 Jul 1996 TX-99 Cape Henlopen State Park, DE Sandy parking lot 11–13 Jun 2005 2009 F.E. Kurczewski 15 50–100 females nesting simultaneously (Table 1). Tachysphex terminatus and T. similis occurred in mixed aggregations at 17 locations, but some of the sites were not given code numbers because of the low number of wasps found and minimal information collected. Soils. Females nested in brown, tan, white or gray fine gravel, coarse sand, medium sand, fine sand, and loamy fine sand based on an analysis of soils from all nesting sites (six of which are described in Table 2). Color designation, hue, value, and chroma were not as important as soil particle size in nesting-site soil characterization (Table 2). Activity conditions. Females in central New York nested from May 29 (1962) to October 6 (1966) (Kurczewski and Acciavatti 1990). The latest Figure 2. Man-made opening in sandy field, Inlet Valley, Ithaca, Tompkins County, NY (TX-36). Tachysphex terminatus nested in level sand at bottom of slope. Table 2. Soil characteristics at some Tachysphex terminatus nesting sites. Site Soil particle size (%) Value/ No.1 Fg2 Cs3 Ms4 Fs5 Color Hue chroma TX-1 34.1 46.9 13.7 5.3 Gray 5 Y 5/1 TX-2 1.1 16.0 56.1 26.8 Grayish brown 10 YR 4/2 TX-20 0.2 2.5 33.9 63.4 Pale brown 10 YR 6/3 TX-36 2.5 38.1 36.5 22.9 Grayish brown 2.5 Y 5/2 TX-54 55.1 40.7 3.1 1.1 Reddish yellow 7.5 YR 6/6 TX-56 9.7 60.0 23.6 6.7 Brown - - 1Table 1 provides a key and brief description for TX- sites. 2Fine gravel. 3Coarse sand. 4Medium sand. 5Fine sand + silt/clay. 16 Northeastern Naturalist Vol. 16, Monograph 3 emergence date at a central New York site (TX-20) was June 17 (1968). There were two adult emergences annually in this region in late spring and mid-summer. In eastern Kansas, females dug burrows from May 5 (1966) to October 27 (1965). Nest provisioning was observed in eastern Kansas from May 6 (1966) to October 18 (1965) and included several successive generations of wasps. Wasps were most active on bright sunny days when the air temperature in the shade was 27–32 ºC. Females stopped nesting and rested on the sand at the onset of moderate to dense cloud cover. Females exited overnight burrows in Kansas as early as 0855 (CDT) at cell depth and sand surface temperatures as low as 25 ºC and 28 ºC, respectively. Wasps hunted for prey on vegetation and provisioned nests at sand surface temperature as low as 29 ºC. Females in Kansas sometimes nested as late as 1830 (CDT). In New York, females exited overnight burrows at cell depth and sand surface temperatures as low as 21.5 ºC and 24.5 ºC, respectively. Subsequent provisioning and temporary closures occurred at sand surface temperature as low as 28 ºC. Last generation females in Kansas (TX-54) were observed nesting or attempting to nest through much of October 1965. Wasps excavated burrows, leveled the removed sand, and made temporary closures at sand surface temperatures of 33–37 ºC. One female was observed October 27, walking on the sand and making low flights at a sand surface temperature of only 26 ºC. In Kansas, more females provisioned nests at a sand surface temperature of 47 ºC than any other temperature. Two females in Kansas provisioned nests at sand surface temperatures as high as 64–65 ºC, while three wasps in upstate New York brought prey to nests at sand surface temperatures as high as 62–63 ºC. In New York (TX-36, 20 June 1966), 10 females entered their nests and stayed inside while eight wasps continued to bring in prey at sand surface temperatures of 58–63 ºC. The females that entered burrows resurfaced when the sand surface temperature decreased to 57 ºC. Emergence. First generation females at one Kansas site (TX-56) began emerging May 5, 1966. Emerging wasps were collected, frozen, and weighed throughout the remainder of May. Recently emerged, first generation females weighed, on average, 14.8 ± 3.2 mg (range = 11–21, n = 28). There was weak correlation between date of emergence and size (weight) of first generation wasps (r = 0.1317). A last generation of wasps at another site in Kansas (TX-54) emerged in early fall of 1965. Individuals collected 22 September–1 October had full wings, unworn foretarsal digging rakes, and clean integument. Recently emerged females rested on the sand surface and cleaned the sand grains from their wings, antennae, and abdomen with their fore- and hindlegs. Within minutes, males had located, hovered above in flight, mounted, and copulated with them. Some females were mounted and copulated with twice during a few minutes-long period. Burrow excavation. A female searching for a place to dig a burrow walked rapidly forward moving from side to side in a zigzag manner, held her wings 2009 F.E. Kurczewski 17 flat on her dorsum, and tapped her antennae on the soil surface. She pivoted frequently, sometimes stopped, cleaned herself, resumed walking, or made a low rapid flight to a new area. Females paused in certain spots, circled excitedly, rapidly tapped their antennae alternately on the ground, and sampled the soil with their mandibles. Their antennae were forked ventro-laterally, their mid- and hindlegs were braced ventro-laterally against the substrate, and their wings were held flat on the dorsum. Some wasps backed repeatedly from the excavation, walked around the hole, and continued to dig with their mandibles. During high sand temperatures (>54 ºC), females made brief hovering flights, 5–8 cm high, turned one or two circles in mid-air, landed facing the opening, and continued to dig. As the mandibles loosened the soil, the forelegs were used in unison to rake it backward beneath the synchronously lifting abdomen. The front tarsi bear a series of long spines laterally and are well adapted for this purpose. Females threw sand backward at the rate of 6–7 scuffs per second in direct sunlight at 54 ºC sand surface temperature, as estimated from motion pictures. The mid- and hindlegs were used only for walking. The wings were held flat on the dorsum. The antennae were bowed downward, the tips alternately tapping the sand surface. In certain spots, a female dug down a few millimeters or centimeters, using her mandibles and forelegs, but soon abandoned the excavation and started to dig elsewhere. Three time consuming false starts were 31, 37, and 43 mm long. Such excavations were abandoned because the surrounding soil caved in, the wasp encountered pieces of wood, stones, clay strata, hardpacked soil or the nests of other insects, particularly ants, or perhaps she did not receive appropriate stimuli to continue digging. Upon abandoning, some females flew off directly or they turned away from the excavation and left in flight or by walking (n = 16). Some wasps turned away from the opening and threw sand backward into it with their forelegs before leaving (n = 16). Females exhibiting the latter behavior spent 20–36 sec flinging sand into the opening. Some wasps practiced both types of behavior upon abandoning successive excavations. False starts were made 2–60 cm apart. Forty-one females made 1–9 false starts before they remained in one place and finished a burrow. They dug for 21 sec to 16.5 min (n = 34) and removed 2–32 loads of sand from an excavation (n = 30) before leaving. Abandoned excavations were 4–43 mm long (n = 11). Duration and length of excavation were not correlated with whether or not a female flung sand into the opening upon leaving. Not all wasps dug trial excavations before completing a burrow. Some females remained in one place and finished the burrow they started. At intervals, a wasp backed from her excavation and removed the loose sand that accumulated in the burrow and entrance (Fig. 3). The intervals, as measured by the duration a female was inside her burrow and out of sight, usually increased as she dug deeper (Fig. 4). The distances to which females removed loose sand from their burrows increased as the excavation advanced (Fig. 5). As sand accumulated in front of an entrance, the wasp distributed it increasingly farther from the opening. 18 Northeastern Naturalist Vol. 16, Monograph 3 During later stages of an excavation, females extensively leveled the accumulation in front of the entrance by walking in various directions and flinging sand backward with their forelegs. Wasps moved laterally, Figure 3. Female Tachysphex terminatus excavating burrow using her mandibles and her forelegs in unison (TX-54). Her abdomen lifts synchronously as her forelegs fling sand backward. Her antenna tips are held downward touching soil. Figure 4. Mean interval (in seconds) between successive trips backward to remove sand during burrow excavation plotted against sand removal number in Tachysphex terminatus (bar graph averages in sets of five with standard error bars; TX-1, 20, 54). 2009 F.E. Kurczewski 19 diagonally, backward, and forward with respect to the direction of the burrow. Standing sideways, females walked forward to the periphery of the accumulation throwing sand back toward the center. Upon reaching the edge of the accumulation, they turned around and moved toward the center flinging sand in the direction of the periphery. Females also backed to the periphery throwing sand and then moved forward to the center of the accumulation flinging sand without turning around. During very late stages of leveling, females stood several centimeters away with their abdomen pointed toward the entrance and threw sand toward it. Regardless of the directions of the leveling paths, the complexity of the patterns increased as a female neared completion of her burrow (Fig. 6). After sometimes spending several minutes leveling the sand accumulation in Figure 5. Mean distance (mm) traveled to remove sand during burrow excavation plotted against sand removal number in Tachysphex terminatus (bar graph averages in sets of five with standard error bars; TX-1, 20, 54). Figure 6. Leveling path for single sand removal episode near end of burrow excavation in Tachysphex terminatus (TX-1). 20 Northeastern Naturalist Vol. 16, Monograph 3 front of an entrance, a wasp moved forward into her burrow, often in a straight line, flinging sand backward with her forelegs. Extensive leveling where a female stood sideways and flung sand behind her with her forelegs while moving in various directions was not noticeable until about midway or later through an excavation. Fifty-two of 58 (89.7%) burrow excavations included extensive leveling in the latter half of the process. One wasp (TX-1), digging in very damp coarse sand following a rain, did not level the accumulation removed from burrow excavation. The first extensive leveling was observed following 30–72 sand removal episodes, 9–37 min after the start of an excavation (n = 58). The duration spent leveling the accumulated sand in front of an entrance increased as the complexity of the patterns intensified (Fig. 7). Near the end of an excavation, some females spent 1 to 4.5 min for an individual leveling session on the sand surface following a period of sand removal. Females often interrupted distributing the loose sand, made a 5–30-cmhigh hovering flight above the area, turning around once or twice in mid-air, and landed facing the entrance. They then walked forward into the burrow throwing sand backward. Such flights perhaps served as orientation to facilitate future returns to the nest with prey. One excavating female made 50 trips backward with loose sand from the burrow, twice made a hovering flight without turning, but turned completely around in mid-air while hovering 39 other times. Some wasps stopped leveling the accumulation, flew to nearby vegetation, rested for several seconds, flew back to the area, and recommenced distributing the loose sand. Figure 7. Mean duration (in seconds) spent on surface distributing sand removed from burrow excavation plotted against sand removal number in Tachysphex terminatus (bar graph averages in sets of five with standard error bars; TX-1, 20, 54). 2009 F.E. Kurczewski 21 Females turned away from the entrance and threw sand toward it following the last or last few sand removal episodes. The wasp later used this loose sand for fill in making her temporary closure. Females periodically stopped flinging sand backward and walked to the entrance up to 14 times per excavation to examine the fill. If an entrance still remained open, wasps turned away and threw sand toward it. After partly or completely filling the opening, females reopened the entrance with the forelegs and entered the burrow. A wasp then disappeared for 1.8–8.9 min (mean = 3.6 ± 1.8, n = 52), pushed moist sand into the entrance from below, sometimes closing it, while apparently shaping a cell at the end of the burrow. Females digging in damp sand backed from an entrance up to 14 times to clean their antennae by pulling them through the tibio-tarsal cleaning apparatus of their front legs. Such cleaning probably ensured that the sensory receptors on their antennae would function optimally. Females demonstrated more antennal cleaning and much less leveling behavior when digging in damp sand. Wasps removed from their burrows stones or pieces of agglutinated soil up to 3 mm in diameter with their mandibles. They deposited this debris on the surface at distances of 15–45 mm from the entrance. Stones were removed from an excavation 1–37 min after the start of digging. Up to 13 stones were removed from a burrow during an excavation. After making the cell, females appeared headfirst in their entrances with their wings held flat on their dorsum. They walked from the opening throwing sand backward into it using their forelegs in unison. Some wasps even used the tip of the abdomen to pack the loose sand in place. This comprised the temporary closure of the entrance. Two females (TX-1) dug on a steep slope, and the loose sand they threw backward rolled down the incline and out of reach. Both wasps, when making their temporary closure, obtained other sand that they loosened with their mandibles from the hard-packed incline and flung with their forelegs into the opening. Females on level sand proceeded as far as 30 mm from an entrance and spent 3–26 sec for closure at various sand surface temperatures. Some wasps interrupted their closure, turned around every few seconds, and examined the fill, before turning back to throw more sand into the entrance. After finishing the closure, some females turned toward it and then took flight. Other wasps made circular orientation flights above the area without turning around. During such a flight, females made 1–12 circles of increasingly larger radii, at first flying 2–3 cm above the sand in circles 20–30 cm in diameter and later flying in circles 4–5 m wide. The wasps frequently interrupted the flights to land on the sand surface or they flew to nearby vegetation and rested momentarily. The total number of sand removal episodes for excavation of a main burrow and first cell was 28–76 (mean = 50.5 ± 12.5, n = 52; TX-1, 54). The duration spent excavating a main burrow and first cell was 15–51 min (mean = 32.1 ± 11.1, n = 60; TX-1, 54). 22 Northeastern Naturalist Vol. 16, Monograph 3 Additional cells. Females often excavated a side burrow and second cell off the main burrow after stocking the first cell with grasshoppers and ovipositing on a prey. Later, they sometimes dug additional side burrows and up to five cells in the same nest. They packed most of the soil from the new dig into the previous open burrow, working mainly beneath the surface and out of sight. There were far fewer sand removal episodes per excavation for an additional burrow and cell than for the main burrow and first cell. Wasps appeared in their entrances 0–27 times (mean = 11.4 ± 9.0, n = 44) to remove accumulated soil from the side burrow and cell. An orientation flight was seldom made and, if so, it was brief in duration and never elaborate. Apparently, the wasp retained enough information from the orientation flight(s) following the excavation of the main burrow and first cell to enable her to readily locate her entrance upon return with prey. The initial indication that a female was digging a side burrow and second cell was her backing into the entrance removing sand or, rarely, her emergence headfirst to rake sand into the main burrow. The latter behavior served to supplement the amount of soil needed to fill her first excavation and occurred 2–17 min (mean = 9.9 ± 4.7, n = 46) after taking in the last prey for the first cell. Females excavating a third burrow and cell backed from an excavation removing sand for the first time an average of 13 min from start. Wasps backed out with sand for the new excavation at intervals of 0.5–6.5 min. Midway through this excavation, a female usually spent a lengthy interval (2–7.5 min) inside transferring sand from the new burrow to the main burrow leading to the first cell. The last interval inside the nest was probably spent making a cell at the end of the side burrow, as females were below ground and out of sight for 1.8–5.7 min (mean = 3.0 ± 1.0, n = 33). Females walked backward 0–32 mm (mean = 16.3 ± 8.0, n = 43) when removing sand from the new side burrow and cell. Only three of 51 (5.9%) excavations of additional cells included a semblance of leveling the slight sand accumulation on the surface, and this leveling was seen only 2–5 times per excavation. Five of 45 (11.1%) females turned to check the amount of fill in the entrance 2–5 times per excavation. Antenna cleaning was rarely observed for excavation of this later burrow and cell. Following the excavation, a wasp made a 4–27-sec-long (mean = 9.6 ± 4.6, n = 36) temporary closure of the entrance, throwing sand backward into it with her forelegs at varying sand surface temperatures. Only four of 45 (8.7%) females turned to examine the amount of fill in the entrance, and they made brief flights—perhaps for orientation—above the area before leaving. The amount of time spent making the second burrow and new cell was 10–49 min (mean = 20.6 ± 8.3, n = 42). One wasp (TX-54) took 87 min to excavate a third burrow and cell at a sand surface temperature of 65 ºC, staying inside for long periods of time to avoid the hot sand. Females in Kansas (TX-54) and New York (TX-36) completed up to four fully provisioned cells per day in June, but only one per day in September– October (TX-20, 54). 2009 F.E. Kurczewski 23 Rain burrows. At the approach of rain, some females quickly excavated a short simple burrow in the sand with their mandibles and forelegs in unison rather than remaining in a nest with prey. They closed themselves inside this burrow in a head outward position after pushing damp loose sand upward into the opening. This temporary burrow did not end in an enlarged ovoid cell. The burrows were abandoned following the end of a rain and the reappearance of optimal weather. Two such burrows entered the sand at about a 45° angle to the surface and were 37 and 41 mm long. They took 5–6 min to excavate and included 17 and 18 sand removal trips, respectively. Hunting. Females flew to a nearby source of vegetation to hunt for prey after making an orientation flight. Some wasps hunted as far as 20 m from the bare sand in which they nested. The females were able to hunt at such distances from their nests because the relatively small grasshoppers they captured were transported relatively quickly in flight. Wasps often made consecutive provisioning flights to and from the same hunting areas. Females searched for prey by hovering next to upright plants, flying slowly 1–2 cm above and beside the leaves. After hovering around several plants in succession without locating prey, a wasp landed, rested, and often cleaned herself. Some females landed on plants and walked on the stems and leaves tapping their antennae constantly. Frequently, they paused at the petiole of plants with compound leaves and tapped their antennae rapidly. If unsuccessful in capturing prey, females periodically returned in flight to the nest, landed in front of the entrance, tapped their antenna tips on the closure, turned away, and took flight, or they removed the sand fill, entered the nest, exited several seconds later, re-closed the opening, and flew away. One wasp returned without prey six times at 1–60 min intervals. Some females without prey spent as long as 1–3 hours between consecutive returns to the nest. Grasshoppers were captured on plants, 7–11 m from the nests. From 1–3 stings were administered to the ventral thorax of a grasshopper near its leg bases after the wasp pulled it from vegetation and, holding it, flew to the ground. The paralyzed prey was then often transported directly to the nest in flight without interruption. Prey transport. A female straddled the grasshopper headfirst and usually ventral side upward after stinging and, rarely, mandibulating the prey and cleaning herself (Fig. 8). Infrequently, a wasp straddled her prey dorsal side upward or on the side, but often quickly switched the grasshopper to the ventral side upward position. Atypical prey transport in the dorsal side upward position was probably related to the small sizes of the grasshoppers. Several wasps brought tiny prey to the proximity of the nest headfirst and on the side. A female grasped her prey by the bases of its antennae if ventral or dorsal side upward or by the base of one antenna if on the side (Fig. 8). The grasshopper was held by the antenna bases rather than farther out because of its small size. Embracing the prey with all her legs and holding it tightly underneath, a wasp took flight toward her nest. One wasp brought a very small grasshopper to her nest in short low flights abdomen, not head, first! 24 Northeastern Naturalist Vol. 16, Monograph 3 Provisioning flights were often made straight to the nest, but some flights included infrequent pauses and rests. Several females landed on soil or vegetation, retained their mandibular grasp of the grasshopper’s antennae or one antenna, but let loose its body and stood on all legs. The wasp’s abdomen moved rapidly upward and downward as the individual segments moved continuously back and forth breathing. The mean ratio of weight of prey to wasp for flight transport was 0.7:1 (range = 0.20–2.31, n = 314) in first-generation New York females and 0.9:1 (range = 0.25–2.36, n = 277) in second-generation New York females. Nineteen females with slightly larger than normal-sized prey, 10–11 mm long, alternated flying with running on the ground, but always arrived at the nest in flight. Nine wasps with grasshoppers, 12 mm long or longer, carried the prey forward on the ground to the nest. These prey were probably too heavy to be transported in flight, some weighing as much as 2.54–2.60 times the wasp’s weight. Three females weighing 17, 18, and 21 mg transported larger grasshoppers weighing 46, 50, and 58 mg (ratios: 2.7, 2.8, and 2.8:1), respectively, from the prey capture sites to their nests entirely on the ground. First-generation (May–June) wasps at one Kansas locality (TX-54, 1965) spent significantly less time obtaining consecutive prey items than second- and later-generation (July–October) wasps from the same site: mean = 10.4 ± 9.0 min, range = 1–48, n = 103 vs. mean = 46.2 ± 29.5 min, range = 3–189, n = 130 (P = 0.0134). First-generation (June) New York wasps at one site (TX-1, 1964) averaged less time (mean = 11.4 ± 8.3 min, range = 1–33, n = 39) between successive prey items than Figure 8. Female Tachysphex terminatus resting on sand following lengthy flight transport (TX-54). Wasp grasps the bases of grasshopper’s antennae with her mandibles and, in flight, holds the prey’s body with her legs. 2009 F.E. Kurczewski 25 second-generation (mid-to-late July) wasps from that locality (mean = 39.3 ± 21.7 min, range = 10–61, n = 13) (P = 0.0994). Nest entry. A female with prey usually landed in front of the entrance. She placed the grasshopper ventral side, rarely dorsal side, upward or on its side with its head near the fill. Such a placement facilitated the wasp’s pulling the prey into the nest by an antenna. The heads of 337 of 343 (98.3%) grasshoppers at one site (TX-1) were positioned 0–40 mm (mean = 6.2 ± 5.4) from the bottoms of the entrances. The prey was nearly always placed in a straight line or at a slight angle to the direction of the burrow. Only 25 of 878 (2.8%) grasshoppers were released to the left or right of the entrance at roughly right angles to the direction of the burrow. Three females each brought in two consecutive grasshoppers and placed both to the left of the entrance at a right angle to the direction of the burrow. One prey was transported to the nest headfirst and dorsal side upward, released on the sand surface near the entrance, re-stung in its ventral thorax twice, and dragged into the burrow in the usual manner. Seven grasshoppers became wedged temporarily in entrances during prey entry into the burrow. In all cases, the wasps were able to excavate around the grasshopper and eventually take the prey into the nest. After releasing the prey, a wasp walked to the closure and tapped her antenna tips on the fill (Fig. 9). She then used her forelegs in unison to rake open the filled entrance, throwing the loose sand backward beneath her abdomen (Fig. 10). At high (55–65 ºC) sand surface temperatures females paused, made 6–10 cm high hovering flights above the partly opened entrance turning completely around in mid-air, landed facing the fill, and continued removing Figure 9. Female Tachysphex terminatus, having released a paralyzed grasshopper on the sand surface, searches for temporary closure with her antennae (TX-54). 26 Northeastern Naturalist Vol. 16, Monograph 3 the loose sand with their forelegs. The hovering flights were repeated a few to several times, interspersed with bouts of raking sand backward, before a wasp removed enough sand to enter her burrow. The female entered the burrow headfirst, turned around inside (Fig. 11), walked up the burrow, and Figure 10. Female Tachysphex terminatus begins to remove temporary closure using her forelegs in unison (TX-54). Figure 11. Female Tachysphex terminatus is inside her burrow. A paralyzed grasshopper lies on the sand surface (TX-54). 2009 F.E. Kurczewski 27 appeared headfirst in her entrance 2–9 sec (mean = 4.0 ± 1.9, n = 1074) after entering. She walked to the grasshopper, usually grasped it by the end of an antenna with the mandibles (Fig. 12), and backed into the burrow dragging the prey behind. The removal of temporary closure, entry without and with Figure 12. Female Tachysphex terminatus grasps grasshopper by antenna with her mandibles and begins dragging it backward toward open entrance (TX-54). Figure 13. Female Tachysphex terminatus makes temporary closure, using her forelegs in unison, after placing grasshopper in nest (TX-54). 28 Northeastern Naturalist Vol. 16, Monograph 3 prey, exit, and new temporary closure (Fig. 13) averaged about 30 sec, with 4–36 sec (mean = 9.5 ± 4.7, n = 974) spent to place the prey in the cell. Six females exited their entrances, straddled the grasshopper headfirst grasping it by one or both antennae and, with it, walked forward instead of backing into the open burrow. The small size of the prey allowed the wasp to enter the burrow atop the grasshopper. Three females, after turning around inside and exiting, grasped the prey by a hindleg and pulled it backward into the burrow. One wasp grasped the grasshopper by the end of its abdomen and backed into the burrow. All of these grasshoppers were positioned in the cell head outward, i.e., in the manner in which they were taken inside. Wasps sometimes provisioned continually from mid-morning to late afternoon regardless of sand surface temperature. Seven Kansas females (TX-54) consecutively brought 8–16 grasshoppers to nests in 2–3 hours at sand surface temperatures ranging from 40 to 65 ºC. Six New York females (TX-36) consecutively arrived at nests with 8–23 prey over a 2–5 hour period at sand surface temperatures ranging from 43 to 63 ºC. The cumulative hours per female did not include the duration spent excavating additional burrows and cells in nests. Removal of the sand fill prior to nest entry and subsequent temporary closure were made increasingly faster in a straight-line relationship as sand surface temperature rose from 35 to 60 ºC (Figs. 14, 15). There was no difference in removal of the sand fill or temporary closure activity rates and duration of behavior at the same sand surface temperatures between New York and Kansas females (Figs. 14, 15). Wasps made brief hovering flights Figure 14. Mean duration of temporary closure removal (in seconds) plotted against sand surface temperature (in degrees C) in Tachysphex terminatus (TX-1, 20, 36, 54, 56). 2009 F.E. Kurczewski 29 into cooler strata of air to avoid the hot sand from 55 ºC to 63 ºC (New York)–65 ºC (Kansas), the mean number of such flights increasing gradually up to 59–61 ºC (Fig. 16). Concurrently, wasps in Kansas and New York left their entrances increasingly more open as they spent less time making the Figure 16. Number of hovering flights per nest entry plotted against sand surface temperature (in degrees C) in Tachysphex terminatus (TX-20, 36). Figure 15. Mean duration of temporary closure (in seconds) plotted against sand surface temperature (in degrees C) in Tachysphex terminatus (TX-1, 20, 36, 54, 56). 30 Northeastern Naturalist Vol. 16, Monograph 3 closures at higher sand surface temperatures. Leaving the entrance partly open culminated at 59–65 ºC with gaps, 2–4 mm wide, at the tops of entrances 4.0–7.5 mm wide (Fig. 17). In cases where an entrance had been left half or more open during a previous closure, the wasp simply ran inside in <1 sec, with her forelegs feigning removal of the fill. On very windy days, several prey were blown away from where they were released near the entrance by the wasp. Other grasshoppers were accidentally moved to an atypical position by the wasps. In each case, the female flew or walked to the prey, straddled it, then walked or flew back to the entrance, and repositioned the grasshopper with its head near the opening. She then usually took it inside in the usual manner. However, some wasps entered directly with the prey instead of going into the burrow and turning around inside. Many females, upon removing the fill from an entrance, threw sand atop the grasshopper. Having gone into their burrows, such wasps came completely outside, stood over the prey, and, using their mandibles and forelegs, removed most of the sand grains before reentering the nest, coming out, and backing in with the grasshopper. After placing the prey inside, the female reappeared in her entrance headfirst and temporarily closed the entrance with sand. Females nesting in close proximity sometimes laid their prey near the entrance, went inside, came out, and found another female trying to steal the grasshopper. The intruder often flew away at the sight of the resident female emerging from her entrance. The resident then picked up her prey with her legs and mandibles and flew in circles over the area for a few Figure 17. Number of observations of various nest entry behaviors plotted against sand surface temperature (in degrees C) in Tachysphex terminatus (TX-20, 36). 2009 F.E. Kurczewski 31 seconds. She landed, placed her grasshopper in front of the entrance, and went through the sequence of placing it in the nest as described above. Several instances of prey stealing were observed where females nested in close proximity, especially when two wasps shared a common entrance. Grappling between a resident and intruder using their mandibles and legs was commonplace, during which the grasshopper was passed back and forth among the two wasps. This behavior sometimes culminated with the usurper quickly taking the stolen prey into her nest. Following the loss, the resident female searched for many minutes by running excitedly on the sand and making rapid short circular flights above the area of the skirmish. Final closure. A few females, before closing, appeared headfirst in the entrance several minutes after taking in the last prey, turned around outside, and reentered and re-exited the burrow from one to four times in succession. A wasp began her final closure inside the nest, sometimes producing a buzzing sound as she pulled down soil from the top and sides of the burrow with her mandibles. Keeping her abdomen pointed toward the cell, she flung the loose soil backward with her forelegs in unison beneath her synchronously lifting abdomen. The soil was tamped into place with the rapidly vibrating end of her abdomen. An audible buzzing sound often accompanied this activity. Her mid- and hindlegs were braced against the lower sides of the burrow. Her wings were held flat on her dorsum. Her antennae were spread ventro-laterally and sometimes held against the burrow walls, perhaps aiding in centering the blows from the end of her abdomen. Once the female filled the burrow, preventing her from backing inside, her antennae were no longer held against the burrow walls while tamping the fill with her abdomen. A wasp appeared in her entrance for the first time to obtain sand from the surface 0.5–20.0 min (mean = 6.3 ± 4.6, n = 33) after taking in the last prey. As she walked onto the surface, her forelegs, working in unison, continuously raked sand backward beneath her synchronously vibrating abdomen. Her apical foretarsal segments were bent medially to expose more tarsal spines for catching and moving the sand grains. Most females walked forward onto the sand surface in a straight or nearly straight direction. As the wasp raked sand backward into the burrow, her antennae were held downward, away from the head, the tips tapping the floor. She held her wings flat on her dorsum. At first, females stayed outside only a second or two getting sand and then backed rapidly into the burrow throwing the loose sand backward with their forelegs. Later, they remained on the surface for longer periods of time flinging sand backward (Fig. 18). Females walked from their entrances increasingly farther to get sand as the closure went to completion and loose soil near the entrance was depleted (Fig. 19). The intervals at which a wasp came outside, as measured by the duration she was in the burrow, decreased between consecutive trips for soil (Fig. 20). As the wasp filled her nest, she did not have to back down or move up the burrow as far to obtain sand. 32 Northeastern Naturalist Vol. 16, Monograph 3 Wasps exited their burrows to get sand for the closure 6–48 (mean = 17.4 ± 8.7, n = 31) times. Cleaning the antennae and mandibles once or twice in Figure 18. Mean duration (in seconds) spent on surface raking sand backward into burrow for final closure plotted against sand retrieval number in Tachysphex terminatus (bar graph averages in sets of five with standard error bars; TX-1, 20, 54). Figure 19. Mean distance (mm) to which wasps walked to get loose sand for final closure plotted against sand retrieval number in Tachysphex terminatus (bar graph averages in sets of five with standard error bars; TX-1, 20, 54). 2009 F.E. Kurczewski 33 the entrance or on the surface was seen during 5 of 31 (16.1%) closures. Females turned toward the entrance to examine the fill 1–19 (mean = 5.7 ± 4.5, n = 27) times during a closure. Seven of 31 (22.6%) wasps, while filling burrows, made 1–10 hovering and turning flights above the area during periods of high sand surface temperature (>54 ºC). They turned around completely once or twice in mid-air, landed facing the entrance, turned away, and raked sand backward. Near the end of the closure, females walked in various directions on the surface and, keeping their abdomens pointed toward the entrance, threw sand backward onto the fill for 0.5–3.5 min (mean = 1.5 ± 0.8, n = 30). Atypical final closures were seen infrequently. A few females walked onto the surface to get sand, turned around, and, instead of backing into the burrow, entered headfirst flinging sand backward up to seven times per excavation before closing in the usual manner. Females that nested in steep sandbanks were unable to obtain loose soil from near the entrance because it had fallen out of reach. This situation required the wasp to use her mandibles nearly continuously to break down the top and sides of the burrow to get soil for the closure. She pulled the fallen loose soil backward with her forelegs and tamped it in place with the vibrating end of her abdomen. The duration of final closure varied with different wasps and their nesting situations, ranging from 6 to 34 min (mean = 13.9 ± 7.9, n = 27). Nest structure and dimensions. Females excavated short, shallow one- to five-celled nests that entered the soil obliquely at an angle with the surface Figure 20. Mean interval (in seconds) between successive exits to obtain sand for final closure plotted against sand retrieval number in Tachysphex terminatus (bar graph averages in sets of five with standard error bars; TX-1, 20, 54). 34 Northeastern Naturalist Vol. 16, Monograph 3 ranging from 14–41°, except for nests dug into steep slopes. Three-celled nests were the rule under optimal nesting conditions, but some wasps made four- and five-celled nests when conditions were appropriate (Fig. 21). Some nests with only one fully provisioned cell were closed prematurely at the onset of rainfall, extremely cloudy weather, and dusk. Eight pairs of females at one site (TX-1) shared common entrances due to limited space and high density of wasps. Figure 21. Representative configurations of Tachysphex terminatus (top), T. similis (middle), and T. a. apicalis (bottom) bi- and multi-cellular nests. Numbers indicate order in which cells were made. Scale refers to all nests. 2009 F.E. Kurczewski 35 With abundant prey, nesting space, and optimal weather, a second and, often, third or fourth burrow were excavated off the main one, each terminating in a cell. Side burrows leading to second cells left the main burrows at distances of 8–90 mm (mean = 32.7 ± 20.2, n = 63) from the entrances. Side burrows leading to third cells left the main burrows at distances of 13–64 mm (mean = 41.9 ± 19.0, n = 10) from the entrances. Distances between first and second cells were 22–74 mm (mean = 41.6 ± 13.5, n = 53), between second and third cells were 22–66 mm (mean = 39.6 ± 15.2, n = 11), and between third and fourth cells was 24 mm (n = 1). In attempting to excavate a side burrow and cell, a wasp sometimes encountered clay, stones, roots, pieces of wood, and other debris resulting in an incomplete burrow or spur. During inclement weather, a female sometimes turned around and rested head outward inside the spur. Spurs were exceedingly rare in bare sand where there was little or no subterranean debris. Spurs were found in 28 of 279 (10.0%) nests at two sites (TX-1, 54). Spurs were 13–36 mm long (mean = 24.8 ± 8.4, n = 28). Most burrows were straight or bent slightly as they penetrated the soil (Fig. 21). Of 456 burrows examined for lateral curvature, 185 (40.6%) bent to the left or right at angles greater than 22.5°. Two hundred seventy-one (59.4%) burrows were straight or exhibited a lateral curvature of less than 22.5°. Different aggregations showed much variation in burrow curvature. Only 12 of 82 (14.6%) burrows excavated in 1961–1962 at one site (TX-1) were bent more than 22.5°. However, 69 of 95 (72.6%) burrows dug at two other locations during 1962–1963 (TX-2, 20) exhibited lateral curvature of more than 22.5°. Burrows were 3.5–5.5 mm in diameter, enlarging at the entrances to 4.0– 7.5 mm. The smaller measurements were from hard-packed soil and/or smaller females, and the larger measurements from loose coarse sand and/or larger females. Main and side burrows usually terminated in cells that were oval or elongate-oval in shape. Cell height, width, and length were rather uniform at the various locations. One hundred eighty-four of 190 (96.8%) cells were 6–9 mm high and wide and 10–15 mm long. Variation in cell size was related to differences in soil texture and moisture, perhaps influenced by generation of wasp, size of female, number of prey, and aggregate prey weight. The longest and deepest nests were located in undisturbed, loose coarse sand in southern New Jersey, eastern Kansas, and Presque Isle State Park, PA. The shortest and shallowest nests were found in compact, loamy fine sand of man-made sandpits and sandbanks in central New York (Table 3). Longer burrows and deeper cells characterized first generation (spring) nests, and shorter burrows and shallower cells were typical of second- and later-generation (mid-late summer) nests at nearly all sites in New York and Kansas (Table 3). There was a certain cell order in some three-celled nests, but no distinct cell order in sequential positioning or cell depth within a nest in four- and five-celled nests (Fig. 21). There was no correlation between wasp size (weight) and cell depth (r = -0.1800). 36 Northeastern Naturalist Vol. 16, Monograph 3 Prey. Females preyed exclusively on small nymphal Acrididae and, very rarely, small nymphal Tettigoniidae (Kurczewski 1966a). Most wasps were rather unselective of the genera and species of Acrididae as long as it was of appropriate size. Sixteen genera and 22 species of acridid prey were identified from 2367 specimens (Table 4). Another 1000 or so specimens Table 3. Burrow length and cell depth in Tachysphex terminatus nests.3 Site Burrow length (mm) Cell depth (mm) TX-year4 n Range Mean ± SD5 n Range Mean ± SD 12-1960 25 41–86 62.0 24 25–61 43.0 11-1961 13 42–110 80.0 13 25–65 50.0 12-1961 30 35–90 62.0 44 16–54 40.0 11-1962 87 45–105 72.0 146 22–75 42.0 12-1962 7 52–67 58.7 ± 5.65 8 28–35 32.0 ± 2.45 11-1963 10 50–65 56.8 ± 5.45 16 35–49 43.1 ± 3.61 11-1964 30 40–108 65.3 ± 12.56 34 29–59 44.8 ± 8.74 11-1967 1 54 54.0 21-1962 5 79–101 93.0 5 38–50 46.0 22-1962 12 32–67 45.0 12 16–29 21.0 21-1963 5 47–98 72.2 ± 21.08 8 16–55 40.3 ± 15.06 22-1963 12 33–72 56.6 ± 11.13 14 15–36 25.5 ± 6.30 102 2 62–76 69.0 ± 9.90 2 34–48 41.0 ± 9.90 191 2 65–69 67.0 ± 2.83 2 39–45 42.0 ± 4.24 192 11 43–105 61.4 ± 17.34 12 25–59 34.4 ± 11.08 201-1963 18 34–118 64.5 ± 23.78 19 18–41 26.1 ± 7.75 202-1963 62 38–80 52.5 ± 10.06 67 17–39 25.6 ± 4.60 20-1964 40 47–114 74.4 ± 18.93 40 19–63 38.1 ± 10.84 202-1967 5 15–20 17.6 ± 2.07 241 1 82 82.0 2 42–52 47.0 ± 7.07 261 3 93–106 99.3 ± 6.51 5 35–54 46.9 ± 6.98 271 2 66–68 67.0 ± 1.41 2 51–55 53.0 ± 2.83 311 4 74–79 76.3 ± 2.63 4 44–60 51.3 ± 7.18 351 1 102 102.0 1 58 58.0 362-1963 20 45–99 59.5 ± 17.09 21 18–39 27.7 ± 5.91 36-1964 6 57–104 78.2 ± 17.01 6 24–47 33.0 ± 9.40 382 3 105–129 119.3 ± 12.66 3 62–69 65.7 ± 3.52 412 7 49–69 59.7 ± 6.78 9 25–41 32.4 ± 5.53 442 14 44–72 55.6 ± 8.51 15 22–34 28.6 ± 3.94 481 4 56–85 70.5 ± 12.66 5 30–44 37.2 ± 5.72 502 13 51–73 65.1 ± 6.86 13 19–51 34.0 ± 8.40 542-1964 3 49–90 73.3 ± 21.55 3 19–34 28.7 ± 8.39 541-1965 33 65–165 121.3 ± 26.00 33 36–76 53.9 ± 11.89 542-1965 41 37–73 54.2 ± 9.73 41 18–37 24.1 ± 4.12 561 4 58–84 70.8 ± 11.87 4 29–56 38.3 ± 12.18 601 6 78–126 99.7 ± 19.83 6 31–65 43.0 ± 13.16 692 2 64–66 65.0 ± 1.41 2 49–54 51.5 ± 3.54 951 1 49 49.0 1 26 26.0 991 3 59–67 62.7 ± 4.04 3 27–33 30.0 ± 3.00 1Spring generation. 2Summer generation(s). 3Includes cell length and cell height, respectively. 4Table 1 provides a key and brief description for TX- sites. 5SD values given where available. 2009 F.E. Kurczewski 37 remain unidentified, so the number of prey genera and species might be half again higher than that number. The subfamilies of Acrididae taken as prey were Cyrtacanthacridinae + Melanoplinae, 1661 specimens (70.2%); Gomphocerinae, 396 (16.7%); and Oedipodinae, 310 (13.1%). As many as three subfamilies and four genera and species of grasshoppers were found in a single cell. There was a distinct difference in the acridid species from first and second generation New York nests. Chloealtis conspersa, Chorthippus curtipennis, Dissosteira carolina, Melanoplus bivittatus, M. keeleri, and M. sanguinipes were prevalent in June–early July nests, but conspicuously absent later in the summer. Encoptolophus sordidus and Melanoplus femurrubrum prevailed in late July–August nests. In Kansas, Ageneotettix deorum, Boopedon gracile, Melanoplus differentialis, M. sanguinipes, M. spp., and Syrbula admirabilis were noteworthy prey in May–June cells. Arphia simplex and Chortophaga viridifasciata were prevalent in late summer–fall Kansas cells. A fall generation of T. terminatus in Kansas (TX-54, 1965) stocked a preponderance of these two species in 38 cells. Prey collected from the cells included 71 specimens of Arphia simplex, 30 Chortophaga viridifasciata, and 1 Melanoplus femurrubrum. A sample of nymphal grasshoppers of suitable prey size net-swept from vegetation where the wasps were hunting yielded: 80 specimens of Chortophaga viridifasciata, green and gray-streaked color forms, 33 Arphia simplex, and 1 Melanoplus femurrubrum. Only grasshoppers equivalent in weight to those stocked in the cells (1–50 mg) were counted, 104 being in the 20–50 mg range and 10 in the 1–19 mg range. The Table 4. Species of prey of Tachysphex terminatus, listed in alphabetical order. Species of prey (nymphs) No. specimens Ageneotettix deorum Scudder 14 Arphia simplex Scudder 119 Arphia xanthoptera (Burmeister) 1 Boopedon gracile Rehn 64 Chloealtis conspersa (Harris) 27 Chorthippus curtipennis (Harris) 255 Chortophaga viridifasciata (DeGeer) 82 Dichromorpha viridis (Scudder) 11 Dissosteira carolina (Linnaeus) 16 Encoptolophus sordidus (Burmeister) 92 Melanoplus bivittatus (Say) 895 Melanoplus differentialis (Thomas) 39 Melanoplus femurrubrum (DeGeer) 427 Melanoplus keeleri (Thomas) 139 Melanoplus s. sanguinipes (Fabricius) 78 Melanoplus spp. 75 Mermiria sp. 1 Orphulella speciosa (Scudder) 5 Pardalophora apiculata (Harris) 1 Phlibostroma quadrimaculatum (Thomas) 4 Schistocerca americana (Drury) 7 Syrbula admirabilis (Uhler) 15 38 Northeastern Naturalist Vol. 16, Monograph 3 vegetation was net-swept for one hour each day from 22 September through 18 October at a time when the wasps hunted. The number of grasshoppers stocked in fully provisioned cells at a single locality ranged from 1–14 (TX-1; Table 5). Wasps usually stocked three or more prey per fully provisioned cell. Cells that contained only one or two grasshoppers were often closed prematurely in response to oncoming bad weather or the onset of dusk. Cells of first-generation New York and Kansas aggregations usually contained more prey, often twice as many, as cells of second or later generations of wasps (Table 5). In multi-celled nests with fully provisioned cells, cell 1 contained slightly more prey (mean = 4.4 ± 2.4, Table 5. Number of prey and aggregate prey weight per cell in Tachysphex terminatus nests. Site No. prey per cell Aggregate prey wt. (mg) TX-year3 n Range Mean ± SD n Range Mean ± SD 12-1960 5 2–4 3.4 ± 1.73 11-1961 6 5–9 7.5 ± 1.38 12-1961 12 2–5 3.8 ± 1.22 11-1962 90 1–12 5.7 ± 2.37 12-1962 5 2–4 2.4 ± 0.89 11-1963 16 2–9 5.8 ± 2.17 16 38–128 81.8 ± 27.87 1-1964 27 3–14 6.1 ± 2.73 27 27–196 77.4 ± 37.85 11-1967 1 7 7.0 1 85 85.0 21-1962 4 4–8 6.5 ± 1.73 22-1962 7 1–6 3.3 ± 1.72 21-1963 7 4–11 6.6 ± 2.41 5 37–152 81.6 ± 38.93 22-1963 14 1–7 3.9 ± 2.22 12 24–120 70.4 ± 31.53 191 1 6 6.0 1 125 125.0 192 10 1–6 2.2 ± 1.48 10 25–68 42.9 ± 10.53 201-1963 18 1-7 4.6 ± 1.76 14 29-142 72.9 ± 30.25 202-1963 61 1-8 3.1 ± 1.36 61 26-131 61.0 ± 25.07 20-1964 32 1–6 3.4 ± 1.24 32 33–148 81.2 ± 29.65 241 1 2 2.0 261 4 4–7 5.3 ± 1.50 311 3 5–7 6.0 ± 1.00 3 25–113 80.7 ± 48.42 362-1963 21 1-5 2.8 ± 0.98 21 31-130 65.8 ± 32.38 36-1964 5 2-6 2.8 ± 1.51 5 51–110 76.8 ± 23.77 382 7 3–10 5.9 ± 2.12 412 3 2–5 3.7 ± 1.72 442 11 1–4 2.5 ± 0.93 481 4 4–7 6.3 ± 1.50 502 5 1–4 2.4 ± 1.14 5 32–70 46.0 ± 14.56 542-1964 2 3-6 4.5 ± 2.12 2 78–100 89.0 ± 15.56 541-1965 33 2-10 5.9 ± 2.16 27 50-140 97.1 ± 28.00 542-1965 43 1-4 2.8 ± 0.92 38 20-125 61.3 ± 26.23 562-1964 1 3 3.0 1 111 111.0 561-1965 3 6 6.0 ± 0.00 2 27-59 43.0 ± 22.62 601 5 4-7 5.4 ± 1.34 5 46-92 66.7 ± 20.49 692 2 2-3 2.5 ± 0.71 2 27-34 30.5 ± 4.95 991 3 6-7 6.7 ± 0.58 1Spring generation. 2Summer generation(s). 3Table 1 provides a key and brief description for TX- sites. 2009 F.E. Kurczewski 39 range = 1–14, n = 98) than cell 2 (mean = 3.9 ± 2.0, range = 1–9, n = 68), but not cell 3 (mean = 4.8 ± 2.7, range = 1–10, n = 16). Four and five prey were found in two fully provisioned fourth cells. Most of the grasshoppers that females stocked in their cells were the same size and weight or smaller and lighter in weight than the wasps. Prey individuals ranged from 3.5 mm (Melanoplus bivittatus) to 13.5 mm (Chorthippus curtipennis) in body length. In New York, grasshoppers collected from first-generation cells weighed 3–52 mg (mean = 14.0 ± 9.7, n = 224), while prey taken from second-generation cells weighed 5–58 mg (mean = 19.6 ± 9.9, n = 325) (P = 0.057). In Kansas, individual prey from spring cells weighed significantly less (mean = 13.2 ± 9.0 mg, range = 4–52, n = 184) than prey from summer–fall cells (mean = 21.0 ± 11.5 mg, range = 3–52, n = 173) (P < 0.05). A few of the larger prey individuals (>40 mg) from both New York and Kansas had one hindleg missing. First-generation prey from New York and Kansas, including those collected from wasps, weighed, on average, significantly less than prey from second and later generations: New York (1963–1964)—first-generation mean = 12.7 mg, range = 3–52, n = 346; second-generation mean = 20.6 mg, range = 4–58, n = 504 (P < 0.05); Kansas (1964–1965)—first-generation mean = 13.8 mg, range = 4–52, n = 233, second- and later-generation mean = 19.4 mg, range = 3–52, n = 230 (P < 0.05). In multi-celled nests, the mean aggregate prey weight for cells 1, 2, and 3 supported the mean number of prey per cell data: cell 1 mean = 76.4 ± 35.5 mg, range = 13–196, n = 83; cell 2 mean = 73.8 ± 29.1 mg, range = 18–148, n = 57; and cell 3 mean = 84.9 ± 36.4 mg, range = 20–128, n = 14. The aggregate prey weight of one cell 4 was 108 mg. Cells with the lowest aggregate prey weights were often closed prematurely at the onset of inclement weather. There was a substantial reduction in the prey biomass per fully provisioned cell from first to second and later generations in New York and Kansas as indicated by a decrease in the aggregate prey weight per cell: New York (1963–1964)—generation 1 mean = 84.0 mg, range = 28–196, n = 56; generation 2 mean = 65.3 mg, range = 24–148, n = 156 (P = 0.091); Kansas (1964–1965)—generation 1 mean = 94.4 mg, range = 27–140, n = 32; generation 2 mean = 58.9 mg, range = 20–125, n = 50 (P = 0.075) (Table 5). This reduction resulted in first-generation wasps being slightly smaller and lighter in weight than second or later generation wasps: New York—spring mean = 19.3 ± 4.1 mg, range = 11–27, n = 9; summer mean = 20.7 ± 2.3 mg, range 17–25, n = 7 (t = -0.5041, P = 0.6322, df = 8); Kansas—spring mean = 14.8 ± 3.6 mg, range = 10–21, n = 26, summer mean = 18.6 ± 2.9 mg, range = 14–24, n = 12 (t = -3.4931, P = 0.0023, df = 20). New York females were significantly heavier (mean = 19.9 ± 4.9 mg, range = 11–27, n = 16) than Kansas females (mean = 15.7 ± 4.0 mg, range = 10–24, n = 49) (t = 1.7247, P = 0.0254, df = 20). Six hundred fifteen of 730 (84.2%) prey were positioned head inward and ventral side upward or on their side in the cells. One hundred five (14.4%) 40 Northeastern Naturalist Vol. 16, Monograph 3 grasshoppers were placed head inward and dorsal side upward in the cells. Six prey were put head outward and ventral side upward, and four grasshoppers were placed head outward and dorsal side upward in the cells. Cells that contained the largest number of prey usually had more grasshoppers placed in atypical positions. The wasps, in maneuvering to oviposit after bringing in their last prey, may have accidentally moved some of the grasshoppers from their typical head inward and ventral side upward positions to the atypical positions in which some of them were found. Some prey were not placed in a cell. Four nests had several grasshoppers aligned almost single-file in the burrow. Six of nine prey in one nest were positioned that way, with one grasshopper lying halfway between the cell and entrance. A single-celled nest with 14 grasshoppers had four of them aligned in the burrow in front of the cell because they couldn’t all fit in the cell. Egg. The wasp’s egg was usually laid on a grasshopper at the bottom of the cell, near the back end, in cells with three or more prey. An egg was almost never placed on prey at the front of the cell because this grasshopper would be the first one attacked by enemies entering a nest. The prey to which the egg adhered was positioned head inward and ventral side upward or on its side in 478 of 497 (96.2%) cells. Twelve eggs were put on grasshoppers that were placed head inward and dorsal side upward, 4 on prey put head outward and dorsal side upward, and 3 on grasshoppers positioned head outward and ventral side upward. Some prey placed head outward occupied cells that were closed prematurely at the onset of inclement weather. The grasshopper to which the egg was attached was often the largest and heaviest prey in the cell. Two hundred eleven of 300 (70.3%) eggs were affixed to the heaviest grasshopper in the cell, 53 (17.7%) to the second heaviest, 16 to the third heaviest, 6 to the fourth heaviest, and 14 to the lightest prey in the cell. Eleven cells held only a single prey. Eggs were placed on grasshoppers that weighed 5–58 mg (mean = 25.8 ± 10.9, n = 290), while other prey in the cells had a mean weight of 15.4 ± 9.6 mg (n = 869) (P < 0.05). Excluding cells with only a single grasshopper, egg-bearing prey comprised the first (20 examples), second (32), third (16), fourth (2), fifth (4) or sixth (4) individuals taken into entrances. A wasp’s egg was attached to the soft intersegmental membrane surrounding the base of a grasshopper’s forecoxa by the less tapered, distal end. The proximal end extended free to the opposite side, the egg lying transversely across the pro- and mesosterna between the bases of the fore- and midcoxae (Fig. 22). Of 486 eggs attached to prey, 223 (45.9%) were affixed to the base of a left forecoxal intersegmental membrane and 263 (54.1%) to the base of a right forecoxal intersegmental membrane. Eighty-two of 114 (71.9%) eggs in two-celled nests were both attached to the same (left or right) side of the base of a prey’s forecoxal intersegmental membrane. Only twelve of 53 (22.6%) eggs in three-celled nests were all placed on the same side of the base of a grasshopper’s forecoxal intersegmental membrane. 2009 F.E. Kurczewski 41 Figure 22. Tachysphex terminatus egg (arrow) lying transversely across thorax of Melanoplus sp. between bases of fore- and midcoxae (TX-1). Egg is affixed to soft intersegmental membrane surrounding base of prey’s right forecoxa. 42 Northeastern Naturalist Vol. 16, Monograph 3 The wasp’s egg was slightly curved, sausage-shaped, white or creamcolored, and measured (live) 1.6–2.1 mm (mean = 1.88 ± 0.11, n = 122) long and 0.4–0.6 mm (mean = 0.45 ± 0.05, n = 122) wide. Development. Five eggs from Kansas (TX-54) and 10 eggs from New York (TX-1), all from first-generation females, were followed through larval development to adult emergence. A total of five of the 15 (33.0%) eggs did not complete development and emerge. Eclosion from egg to larva occurred in about 42–48 hours at 27–32 ºC air temperature. The small (2 mm long) larva remained affixed to its grasshopper, feeding at the site of the egg attachment for an additional day. Larvae, 3.5– 4.0 mm long, were seen feeding in the cervical region of the grasshopper the following day. The larvae rapidly grew to 4.5–5.0 mm, devoured much of their pedestal prey, and moved to feed on other grasshoppers nearby. The larvae were ca. 8–9 (male) to 9–11 (female) mm long when they began to construct a cocoon of silk, saliva, and the surrounding sand grains. Initiation of cocoon construction occurred 5–6 days after oviposition. The ovoid cocoons, averaging 9.5 mm long and 5.0 mm wide, were completed on the 6th or 7th day. Female cocoons were slightly larger than male cocoons. Female emergence occurred 30–34 days after oviposition; males took 29–33 days to complete development and emerge. One female emerged 56 days after being laid as an egg. Tachysphex similis Literature review Krombein and Evans (1955) reported a female of T. similis straddling an immature Radinotatum (= Achurum) sp., probably carinatum (Walker) (Gomphocerinae) on the sand at Marco, Collier County, fl. The wasp held the grasshopper with her mandibles by the bases of its antennae. Krombein (1964) observed a female of this species excavating a burrow in sandy soil at the Archbold Biological Station, Highlands County, fl. She removed sand from the burrow using her forelegs simultaneously, raised her abdomen synchronously as the sand was flung backward, and kept her wings flat on the dorsum. This wasp spent much time on the surface, leveling the sand removed from the burrow. Krombein (1964) noted a second female on the sand flats along the Peace River, just west of Arcadia, De Soto County, fl. She flew in with prey, released the grasshopper about a centimeter from the entrance, removed the temporary sand closure, entered, came out, grasped the prey, and pulled it into the burrow headfirst. The nest was incompletely provisioned. The entrance was 4 mm wide. The straight burrow, 7.5 cm long, ended in a cell, 6 cm beneath the surface. The cell held five Melanoplus nymphs, 5–6 mm long. Krombein (1964) observed a third female nesting in flat, bare sand near Lake Annie, Highlands County, fl. She entered her nest with prey as described above and then made a final closure. During the closure, she repeatedly backed down the burrow with sand obtained from the surface. 2009 F.E. Kurczewski 43 The burrow went downward at a shallow angle for 2.5 cm and then descended almost vertically to a cell at a depth of 3 cm below the surface. The fully provisioned cell contained three Aptenopedes sp. and one Melanoplus sp., 5–8 mm long. The wasp’s egg, 2 mm long, was attached to the right forecoxa of the largest prey in the cell and laid between the fore- and midcoxae. A second cell was found ca. 4 cm from the first cell. It was located 4.5 cm beneath the surface and held 10 nymphal Aptenopedes sp., 5–8 mm long. The wasp’s egg was affixed to a medium-sized prey as described above. Kurczewski (1967) outlined the nesting behavior of T. similis in a study on its cleptoparasite, Hedychridium fletcheri Bodenstein (Chrysididae). The female first excavated a slanting burrow and terminal cell in sandy soil. She temporarily closed the entrance with loose sand and then flew to vegetation where she searched for and stung a small nymphal grasshopper. The wasp returned in flight to her nest with the prey, landed near the filled entrance, removed the sand closure, entered, exited, pulled the grasshopper into the nest, came out, closed the entrance with sand, and flew off in search of another prey. She repeated this provisioning sequence until the full complement of prey was placed in the cell. A fully provisioned cell may contain more than a dozen small grasshoppers. After laying an egg on one of the prey, the wasp permanently filled the burrow leading to the cell with sand and excavated a new burrow elsewhere or she dug and provisioned up to four other cells in the same nest. Elliott and Kurczewski (1985) presented information on two nests of T. similis from Wamego, Pottawatomie County, KS. One nest was single-celled and contained three prey. The second nest had four cells with three, three, five, and three prey identified as Melanoplus sp. (8 examples), Mermiria sp. (1), Oplia obscurus (Thomas) (2), and Pseudopomala brachyptera (Scudder) (8). Although not reported in their study, females collected from this site weighed 9.8–13.2 mg (mean = 11.7, n = 5). Prey grasshoppers taken from the nests or wasps averaged 10.3 mg (range = 1.8–20.8, n = 19). Eggbearing prey weighed 8.7–17.1 mg (mean = 13.3, n = 4). Other prey in the cells averaged 9.5 mg (range = 1.8–20.8, n = 15). Aggregate prey weight per cell averaged 38.5 mg (range = 28.6–53.2, n = 4). Pulawski (1988) recorded a female of T. similis from Marco, flpinned with a juvenile Schistocerca sp. and another wasp from St. Catherine Island, Liberty County, GA pinned with a nymphal Melanoplus devastator Scudder. Elliott (1996) observed females of T. similis nesting at three locations in The Bahamas: San Salvador Island, Cat Island, and Great Abaco. Wasps nested throughout much of the year, being absent only from November to February. Nesting sites occupied “hard-packed sand” and loose beach sand. The aggregation on San Salvador Island contained 18 active nests with conspecific entrances as close together as 7 cm. Females began excavations from flat sandy surfaces removing the pebbles with their mandibles. Burrow excavation was as described above. Burrow excavation duration ranged from 44 Northeastern Naturalist Vol. 16, Monograph 3 23 to 49 min. Females made 27–32 trips into their burrows to remove sand during an excavation. Time spent inside the burrow increased as a dig went to completion. Wasps spent much time on the surface toward the end of an excavation, leveling the sand removed from the burrow. Following burrow excavation, females temporarily closed their entrances with sand and then made circling orientation flights above the area before flying off to hunt prey. Wasps returning with prey deposited the small nymphal grasshopper near the entrance, removed the closure, entered, exited, and pulled the prey inside by its antennae. This behavior was repeated several times per cell with one female bringing five grasshoppers to her nest in 41 min. Larger prey were carried in a series of short flights. Final closure involved the female pulling sand into the burrow from the surface, with one wasp spending 54 min for this behavior. Individual wasps remained with a single nest for up to five days. Elliott (1996) excavated eight nests of T. similis on San Salvador Island. The nests were one- and two-celled. Entrances were 4–5 mm in diameter. Burrows were 3.5–9.5 cm long. Cells varied from 1 to 5 cm (mean = 2.3 ± 0.77, n = 10) in depth below the surface. Cells excavated in June were deeper (mean = 3.4 ± 0.6 cm) than those dug in April (mean = 1.6 ± 0.1 cm). One cell was 12 mm long. Completed cells contained two to seven (mean = 4.5 ± 1.0) prey. Twelve of 15 grasshoppers in the cells were placed head inward and ventral or dorsal side upward or on the side. The egg was positioned between the bases of the fore- and middle legs of the prey. Wasps captured with prey were 7–9 mm (mean = 8.1 ± 0.4) long, and the grasshoppers were 7–10 mm (mean = 8.4 ± 0.5) in length. The nymphal Acrididae were not identified. Wasps were occasionally pursued to their nests by unidentified satellite-flies, and they responded with counter-cleptoparasitic behavior. Kurczewski (2000) reported on 6 two-celled and 2 three-celled nests of T. similis from Brunswick and Onslow Counties, NC. Both nesting sites were comprised of moist, level sand. Females began excavating burrows as early as 0934 (EDT) at a sand surface temperature of 36 ºC. They hunted at sand surface temperature as low as 31 ºC and provisioned nests at sand surface temperature as high as 55 ºC. Eighteen fully provisioned cells were unearthed at an average depth of 28.6 mm (range = 14–36) below the surface. Two of the cells were 6 mm high and 12–13 mm long. Cells held an average of 5.5 (range = 3–8, n = 18) nymphal acridids identified as Melanoplus femurrubrum (51 examples), Schistocerca sp. (1), Orphulella pelidna (19), Dichromorpha viridis (26), Dissosteira carolina (1), and Arphia sp. (1). Grasshoppers were put in the cells head inward and ventral or dorsal side upward or on the side. The mean prey weight was 15.8 mg (range = 4–27, n = 99). Aggregate prey weights of two cells were 44 and 82 mg. One wasp weighed 12 mg. Wasp eggs, <2 mm long, were attached to the left (7) or right (11) forecoxal coria of the prey. H.E. Evans’ field note Initially identified as T. terminatus (det. R.M. Bohart), field note CSL 478 from the “Little Gobi Desert,” Pottawatomie County, KS (July 1, 1953) 2009 F.E. Kurczewski 45 pertains to the red-tipped abdomen color form of T. similis. The female landed with prey, removed the sand fill, took in the grasshopper, came out 30 sec later, and re-made the entrance closure. She brought in five additional grasshoppers at intervals of 8–25 min, then remained inside the burrow for 15 min, evidently making a final closure. The nest was two-celled, with the cells only 1–2 inches beneath the surface. The burrows were 2–2½ inches long. The two cells contained six and seven “tiny” unidentified acridid nymphs. The cell with six prey had the wasp’s egg attached to one of the grasshoppers. Field studies Nest location. Tachysphex similis nested in bare or sparsely vegetated, dry to moist, loose to compact fine gravel, sand, or loamy fine sand (Table 6). Most sites occupied level ground, but some nesting areas sloped as much as 20–45° with the horizon. All sites were surrounded or bordered by vegetation. Some nesting areas were modified substantially by human activity (Table 7). Five rather large aggregations (TX-4, 7, 55, 67, 98) had 50–75 wasps nesting simultaneously (Table 7). Tachysphex similis and T. terminatus occurred together in mixed aggregations at 17 locations, but some of the sites were not assigned code numbers because of small wasp numbers and minimal information. Soils. Females nested in white, gray, tan, or brown fine gravel, coarse sand, medium sand, fine sand, and loamy fine sand based on an analysis of soils from all nesting sites (six of which are described in Table 6). Color designation, hue, value, and chroma were not as important as soil particle size in nesting-site soil characterization (Table 6). Activity conditions. Females in eastern Kansas nested from May 7 (1966) to October 13 (1965). In southern Florida, females nested from February through November (R.C. Miller, College of Environmental Science and Forestry, Syracuse, NY, 1973 pers. comm.). Wasps were most active on bright, sunny days when the air temperature in the shade was Table 6. Soil characteristics at some Tachysphex similis nesting sites. Site Soil particle size (%) Value/ No.1 Fg2 Cs3 Ms4 Fs5 Color Hue Chroma TX-4 0.9 33.0 43.2 22.9 Gray 10 YR 5/1 TX-7 17.0 53.5 24.4 5.1 Light brown 7.5 YR 6/4 TX-636 55.1 40.7 3.1 1.1 Reddish yellow 7.5 YR 6/6 TX-677 9.7 60.0 23.6 6.7 Brown - - TX-768 0.2 2.5 33.9 63.4 Pale brown 10 YR 6/3 TX-94 0.6 4.7 39.8 54.9 Brown 10 YR 4/3 1Table 7 provides a key and brief description for TX- sites. 2Fine gravel. 3Coarse sand. 4Medium sand. 5Fine sand + silt/clay. 6Same site as TX-54 (T. terminatus) listed in Table 1. 7Same site as TX-56 (T. terminatus) listed in Table 1. 8Same site as TX-20 (T. terminatus) listed in Table 1. 46 Northeastern Naturalist Vol. 16, Monograph 3 27–33 ºC. As soon as it clouded over, females stopped all nesting activities. In Florida, wasps reopened the previous day’s burrow or exited recently excavated burrows in which they spent the night as early as 0825–0835 (EST). At one locality (TX-41 [superscripts refer to generations: 1= spring, 2 = summer]), five males appeared on the surface at a sand surface temperature of 28 ºC, 12 min before the first female (29 ºC). The next day at the same locality, two females issued from their burrows at a sand surface Table 7. Tachysphex similis site code numbers, site locations, site descriptions, and inclusive dates of observation. Inclusive dates Site No. Site location Site description of collection TX-3 Lake Annie, flSand ridge 27–28 Jun 1962 TX-4 1 km W Arcadia, flSand flats 1–2 Jul 1962, 25 Mar-4 Apr 1963, 26–30 Mar 1965, 5–8 Apr 1966, 31 Mar, 27–30 Jun 1967, 26–29 Mar 1968, 12–14 Mar 1986 TX-6 Palmdale, flSandy RR embankment 5–6 Jul 1962 TX-7 Archbold Biological Sand pile, sandy fire trail 6–7 Jul 1962, Station, fl27–30 Mar 1963, 9 Apr 1966, 27 Mar-2 Apr, 25 Jun-4 Jul 1967 27 Jun 1968 TX-16 Lakeport, flSandy field 29 Mar–3 Apr 1963 TX-32 6 km N Geneva, NY Sandblow 7–9 Jul 1963 TX-40 Vincentown, NJ Sandy field 11–12 Aug 1963 TX-43 Colonie, NY Sand barren 17–18 Aug 1963, 20 Jun 1968 TX-55 8 km E Manhattan, KS Sandy draw 21 May–10 Jul 1965 TX-58 Blackjack Creek, KS Sandbank 31 May 1965 TX-61 Abilene, KS Sandy field 12 Jun 1965 TX-63 6 km SE Lecompton, KS Sandy RR embankment 28 Jun–8 Jul 1965 TX-66 Lakeview, KS Sandy lakeside 3 Jul–27 Aug 1965 TX-67 8 km NW Lawrence, KS Sand flats 28 Jul–19 Aug 1965, 6–31 May 1966 TX-76 2 km S Auburn, NY Sandpit 13–28 Aug 1968, 4 Aug 1969 TX-81 Presque Isle State Park, PA Sandy beach 19 Jul 1968, 9 Jul 1969 TX-82 6 km NW Chittenango, NY Sandy field 12 Jun 1969 TX-87 2 km SE Erie, PA Gravel pit 6 Aug 1967, 29 Aug 1983, 4–11 Jun 1986 TX-94 Selkirk Shores State Park, NY Sandpit 28-29 Jun 1972 TX-96 Allegan State Game Area, MI Sandy trail 10 Jun 1992, 11–13 Jun 1993 TX-98 Canadian Forces Base Sandy trail 27 Jun–27 Jul 1996, Borden, ON 13 Jul 1997 TX-100 Cape Henlopen State Sandy parking lot 10-12 Jun 2005 Park, DE 2009 F.E. Kurczewski 47 temperature of 28 ºC and a cell-depth temperature of 20 ºC. Both wasps sat in the sun “warming up” for 30 min before they started hunting. Females were observed hunting and bringing prey to nests as early as 0850 at a sand surface temperature of 32 ºC. Burrow excavation occurred as early as 0910 at a sand surface temperature of 32 ºC. Wasps in Florida brought prey to nests and made temporary closures at sand surface temperatures as high as 62ºC and 63 ºC, respectively. They suspended their provisioning activities at sand surface temperatures of 64–65°, staying inside their burrows until the surface temperature decreased to 60–61 ºC, when the wasps resumed provisioning. The females nested until 1820 (EST). Emergence. First-generation females at one site in Kansas (TX-67) began emerging on May 6, 1966. Recently emerged wasps were collected, frozen, and weighed throughout May. Red-tipped females weighed, on average, 9.7 ± 1.6 mg (range = 8–13, n = 10) while all-black females weighed, on average, 10.7 ± 1.9 mg (range = 8–14, n = 12). Females of both color variants combined weighed, on average, 10.2 ± 1.8 mg (range = 8–14, n = 22). There was slight correlation between date of emergence and size (weight) of first-generation wasps (r = 0.4145), skewed by the fact that the first four and seven of the first 10 emergent females at the site were red-tipped. Burrow excavation. Females searching for nesting sites started burrows as described under T. terminatus. They walked zigzag on the sand, tapped their antenna tips on the surface, sampled the soil with their mandibles, and used their forelegs in unison to remove the loose sand. If unsuccessful in finding a suitable place to dig, wasps made low rapid flights to new areas and resumed searching. Thirteen females each made several false starts 10–30 cm apart. They abandoned such excavations after 0.5–3.5 min, the burrows being only 3–6 mm long. When abandoning a dig, a female paused, and flew off (n = 7) or she turned and flung sand into the opening with her forelegs for 4–6 sec before leaving (n = 6). Not all females dug trial excavations before completing a burrow. Some wasps remained where they started digging and finished a burrow. Eight females stayed in one place and completed burrows in level sand. Some wasps beginning burrows made brief interspersed hovering flights up to 8 cm high above the site. They turned around once or twice in mid-air and landed facing the entrance before resuming excavation. A wasp loosened the sand surface with her mandibles. She then used her forelegs in unison to rake the loose soil backward beneath her synchronously lifting abdomen. Tachysphex similis moved their forelegs about twice as fast as T. terminatus and much faster than T. apicalis at the same sand temperature. Their mid- and hindlegs were used only for support and walking in the burrow and entrance. Their antennae were bowed downward, the tips constantly tapping the soil. Their wings were held flat on the dorsum. Females backed from their excavations at intervals to remove the loose sand that accumulated in the burrows and entrances. This action was accomplished by throwing the sand backward with their forelegs beneath 48 Northeastern Naturalist Vol. 16, Monograph 3 their synchronously lifting abdomen. The intervals at which the wasps removed sand from their burrows and entrances, as measured by the duration they were inside and out of sight, usually lengthened as they dug deeper (Fig. 23). The irregular pattern in mean values in Figures 23, 24, and 26 is related to fewer than 10 observations per sand-removal set, considerable variation in the number of burrow entries to obtain sand, and the uneven duration of burrow excavations. Females removing sand from burrows usually backed half the distance of the surface accumulation or farther when clearing it from in front of an entrance. Some wasps removed the loose sand to the edge of the accumulation, 45–65 mm from their entrances. Four females raking loose sand backward went beyond the accumulation, backing as far as 75–142 mm from the entrances. The distance to which the wasps removed sand from the entrances increased as the excavation proceeded (Fig. 24). Seventeen females began to extensively level the loose sand in front of an entrance following 19–62 sand removal episodes, 8–45 min after starting. This leveling consisted of throwing sand backward with their forelegs as described above. Most females moved laterally or diagonally while smoothing the sand accumulation in front of an entrance. The leveling paths were similar to those of T. terminatus (Fig. 25). The leveling paths increased in complexity and the duration spent smoothing the accumulation lengthened as an excavation proceeded to completion (Fig. 26). Some wasps spent from 1 to 6.5 min for an individual leveling session following a sand removal episode. One female Figure 23. Mean interval (in seconds) between successive trips backward to remove sand during burrow excavation plotted against sand removal number in Tachysphex similis (bar graph averages in sets of five with standard error bars; TX-4, 7, 55, 63, 66). 2009 F.E. Kurczewski 49 (TX-71) spent 35 min leveling the sand surface near her entrance and only 15 min excavating inside her burrow. Only one of 17 (5.9%) females made no attempt to level the sand accumulation in front of her entrance. She left a sand deposit, 20 mm wide x 36 mm long x 4 mm high, in front of the entrance. Some wasps interrupted leveling the sand and made periodic hovering flights up to 20–30 cm high during which they turned completely around in mid-air and landed facing the entrance. Such flights may have familiarized the females with their surroundings to facilitate finding the entrance upon Figure 25. Leveling path for single sand removal episode near end of burrow excavation in Tachysphex similis (TX-7). Dashed lines are short, low flights. Figure 24. Mean distance (mm) traveled to remove sand during burrow excavation plotted against sand removal number in Tachysphex similis (bar graph averages in sets of five with standard error bars; TX-4, 7, 55, 63, 66). 50 Northeastern Naturalist Vol. 16, Monograph 3 subsequent returns with prey. Wasps also flew to and rested on vegetation between leveling sessions. Females spent considerable time on the surface, throwing sand back toward the entrance near the end of an excavation. This sand was subsequently used in making the temporary closure. Wasps often stopped flinging sand backward, walked to the entrance, and examined the fill up to 22 times per excavation. Antenna cleaning following a sand removal episode occurred up to three times per excavation. The last or last few intervals spent inside the burrow were considerably longer in duration than prior intervals and apparently represented the shaping of a cell. This duration amounted to 3.3–8.5 min (mean = 5.3 ± 1.9, n = 13). Moist sand was pushed into the entrance from below, plugging the opening and signaling this event. After making the cell, a female appeared headfirst in her entrance and began her temporary closure. Keeping her wings flat on her dorsum, she flung sand backward into the opening using her forelegs in unison for 3–12 sec (mean = 5.4 ± 3.1, n = 18) at various sand surface temperatures. A few females used their mandibles and end of abdomen, along with their forelegs, in temporarily closing the entrance. Some wasps remained at the far edge of their depressed entrance with head bowed downward and did not move onto the surface during temporary closure. Other females proceeded as far as 10 mm from the opening while throwing sand into it. Some closures were filled flush with the surrounding sand. But, most closures remained 1–3 mm depressed after the wasps departed. Females made circular orientation flights Figure 26. Mean duration (in seconds) spent on surface distributing sand removed from burrow excavation plotted against sand removal number in Tachysphex similis (bar graph averages in sets of five with standard error bars; TX-4, 7, 55, 63, 66). 2009 F.E. Kurczewski 51 of increasingly larger radii above the area. The orientation flight of one wasp increased from 6 to 30 cm in diameter, interspersed with 11 landings to face the filled entrance. Wasps made 23–67 (mean = 48.9 ± 14.6, n = 14) entries to remove sand from the burrow and entrance during an excavation. The duration spent to dig the main burrow and first cell averaged 38.6 ± 9.7 min (range = 18–57, n = 18). Additional cells. Females often excavated a side burrow and second cell in a nest after stocking the first cell with prey and ovipositing on a grasshopper, followed by side burrows that led to third, fourth, and fifth cells. The additional burrows and cells were usually dug off the main burrow (Fig. 21). Most of the soil from a new dig was packed into the previous burrow. It was transferred from the new burrow to the main burrow below ground and out of sight. The wasp backed into her entrance only occasionally when making the new burrow and cell. Orientation flights following temporary closures of the entrances were rarely seen and, if made, were brief in duration. When making additional burrows and cells in a nest, females first backed into entrances 3–14 min (mean = 8.6 ± 3.8, n = 8) after taking in the last prey for the first cell. They removed sand from the new burrow and cell only 0–20 (mean = 6.8 ± 6.8, n = 10) times during an excavation. Six of eight wasps spent lengthy (44–98 sec) individual intervals below ground about midway or later through an excavation, probably moving loose sand from the new burrow to the main burrow. One female did not remove any sand onto the surface from the excavation of her second and third cells. The maximal distance a wasp backed from her entrance when removing the loose sand from the new excavation was only 12–27 mm (mean = 16.6 ± 5.0, n = 8). There was no extensive leveling of the sand accumulation on the surface in front of the entrances by females (n = 8) because there was so little sand removed from the burrow. The last interval between consecutive appearances in entrances to remove sand from the burrow was lengthy (mean = 3.1 ± 1.5 min, range = 1.0–5.3, n = 6), apparently being used for shaping the new cell at the end of the recently excavated side burrow. Females exited the opening flinging sand backward with their forelegs in unison beneath their synchronously lifting abdomen when making the temporary closure. This closure took only 2–3 sec at sand surface temperatures of 53–60 ºC (n = 6). Afterward, the entrance remained slightly depressed or partly open at the top. Two wasps closed their entrances from within with damp sand and remained inside following excavation of a new burrow and cell. One female made two examinations of the fill and a brief orientation flight following her temporary closure. Females spent 9–20 min (mean = 14.6 ± 3.3, n = 8) to excavate a new burrow and cell. Females in Kansas (TX-671) completed up to five fully provisioned cells per day in May, but only two per day in August (TX-632). Females in Florida (TX-41, 71) completed up to three fully provisioned cells per day in late March–April, but only two per day in late June–July (TX-42, 72). 52 Northeastern Naturalist Vol. 16, Monograph 3 Rain burrows. With an approaching rain, many females hurriedly excavated a short simple burrow in the sand using their mandibles and forelegs in unison. This burrow did not terminate in an ovoid cell. They closed themselves inside such a burrow in a head outward position, plugging the entrance with damp sand from below. One sand plug was 8 mm long. Two rain burrows took 4 and 5 min to dig. The burrows were abandoned following the rain, remaining wide open and fronted by a small mound of sand. Two rain burrows in Florida (TX-42) were 39 and 31 mm long. They entered the sand obliquely at about a 40° angle with the surface. Hunting. After making an orientation flight, females often flew to a nearby source of vegetation to hunt for prey. A few wasps searched for prey on vegetation only 60–90 cm from their nests. Most females were observed hunting 15–30 m from the bare sand in which they nested. They were able to hunt many meters from their nests because the prey they captured were relatively small in size and transported rather rapidly in flight. Prey comprised mainly early nymphal stages of Acrididae found commonly on grasses and decumbent and upright plants. Some females hunted low to the ground, running rapidly on plants interspersed with hovering flights. Other wasps made rapid flights, 6–12 cm long, from plant to plant, walked slowly on the stems and leaves with their wings held flat on their dorsum, paused in certain places, and tapped their antennae, especially at the petioles of plants with compound leaves. Some females stopped and rested with their antennae held rigidly upward. If unsuccessful in locating suitable prey, females interrupted hunting activities and returned periodically in flight to their nests. Such returns were made from a few minutes to 4.5 hours apart. Females landed in front of their entrances, sometimes examined the fill, turned, and flew off, or they removed the temporary closure, entered, exited a few seconds later, re-made the closure, and flew away. Wasps returned empty-handed up to five times between bringing consecutive prey items to their nests. Seven wasps captured grasshoppers at distances of 3–4 m from their nests. Prey capture was preceded by hovering flights that turned into lightning- fast forward flights, 6–12 cm long, once a grasshopper was located. One female caught a Melanoplus sp., 3 mm long, on vegetation, flew to the ground, stung it in the right side of the ventral thorax for 2 sec, dismounted, cleaned her antennae, legs, and abdomen, remounted the prey, and stung it in the other side for 1 sec. She repositioned the grasshopper ventral side upward, climbed atop it, grasped the bases of its antennae with her mandibles and its body with her legs, and took flight. A second wasp stung a different Melanoplus sp., 4 mm long, in the left side of the thorax for 3 sec, turned the prey over, kneaded the base of its left foreleg with her mandibles for 6 sec, and flew away holding it underneath. A third female stung her prey in the left side of the ventral thorax for 4 sec, then kneaded the base of the foreleg on that side for 9 sec, grasped it ventral side upward, and flew off. This wasp returned to the same plant in <2 min, hovered next to a leaf, pulled a prey off the plant in <1 sec, flew to the ground with it, stung it for 4 sec, cleaned 2009 F.E. Kurczewski 53 herself for 7 sec, rested, grasped it as described above, and flew away. One female that had been repeatedly capturing species of Gomphocerinae stung a slightly larger oedipodine grasshopper. However, she abandoned this prey on the sand surface and flew away. Prey transport. The mean ratio of weight of prey to wasp observed in flight transport was 0.6:1 (range = 0.22–2.01, n = 197). Only a single grasshopper weighing 30 mg (wasp 14 mg, ratio 2.14:1) was carried to a nest on the ground. Five other prey weighing 31–45 mg (wasps 15–20 mg, ratios 2.06–2.58:1) were transported in low flights, 15–30 cm long, interspersed with ground transport. During the tail-heavy flights, the grasshopper’s abdomen projected beyond that of the wasp and its hindlegs dangled downward. The wasp’s wings produced a clearly audible, low-pitched buzzing sound. Such females rested often on the sand surface and low plants. All but one grasshopper was carried headfirst and ventral side upward, being held around the body with the legs and by the bases of the antennae with the mandibles (Fig. 27). One prey was carried headfirst and dorsal side upward. Most flights were made directly to the nests. Some wasps interrupted flights to land and pause momentarily on soil or vegetation before resuming transport to their nests. During pauses, they released the grasshopper as described for T. terminatus and, rarely, kneaded the prey's forecoxal corium with their mandibles. In Kansas, first-generation (May–June) wasps (TX-55, 1965) spent significantly less time between consecutive prey items than second- and later-generation (July–August) wasps from another site (TX-66, 1965): Figure 27. Female Tachysphex similis resting on sand following lengthy flight transport (TX-7). Wasp grasps bases of grasshopper’s antennae with her mandibles and, in flight, holds prey’s body with her legs. 54 Northeastern Naturalist Vol. 16, Monograph 3 mean = 6.5 ± 5.9 min, range = 1–33, n = 154 vs. mean = 11.2 ± 9.0 min, range = 2–38, n = 33 (P = 0.0084). One wasp provisioning in the morning in Florida (TX-71, 1963) spent an average of 34.9 ± 34.7 min (range = 8–109, n = 7) between successive prey items. Nest entry. Provisioning females landed on the sand and placed the grasshopper ventral side upward or on its side with its head near the entrance. More than 100 prey were placed with the head 4–21 mm, usually 7–9 mm, from the opening. A grasshopper was nearly always put in front of the entrance in a straight line with the direction of the burrow. Prey were not placed above the entrance or beside it on the left or right at right angles to the direction of the burrow. After releasing the grasshopper on the sand, four of 979 (0.4%) observations in Florida and Kansas at sand surface temperatures of 58–61 ºC indicated a single, very brief, low hovering flight prior to nest entry. Prior to entry, wasps tapped their antenna tips alternately on or held them against the fill. Upon locating the closure, they removed the sand fill using their forelegs simultaneously (Fig. 28). The sand was flung backward with the forelegs in unison as the abdomen lifted synchronously. A few females, after releasing their prey on the sand surface, lost track of the temporary closure, dug in several places surrounding the entrance, but never found their nest. Such wasps abandoned the prey on the surface and began a new burrow elsewhere. The relatively loose and often depressed closure of this species may function in helping females find their nest entrances as well as provide rapid entry on the hot sand. After releasing the prey and removing the sand fill, a female entered her burrow headfirst, turned around inside (Fig. 29), walked up the burrow, and Figure 28. Female Tachysphex similis has released a paralyzed grasshopper on the sand surface and begins to remove the temporary closure using her forelegs in unison (TX-7). 2009 F.E. Kurczewski 55 appeared in the entrance headfirst in 2–9 sec (mean = 3.3 ± 1.7, n = 730). A wasp walked to her prey, grasped it by the end of an antenna and walked backward into the burrow pulling it behind (Fig. 30). If in the early or middle stages of provisioning, she reappeared headfirst, temporarily filled Figure 30. Female Tachysphex similis grasps a grasshopper by an antenna with her mandibles and begins dragging it backward into her burrow (TX-7). Figure 29. Female Tachysphex similis is inside her burrow. A paralyzed grasshopper lies on the sand surface (TX-7). 56 Northeastern Naturalist Vol. 16, Monograph 3 the entrance with loose sand (Fig. 31), and flew off to obtain additional prey. The removal of temporary closure, entry, exit, re-entry with prey, exit, and new temporary closure averaged about 18 sec including 3–16 sec (mean = 6.7 ± 3.4, n = 662) to deposit the grasshopper in the cell. Most wasps did not proceed far from the opening when closing their entrances. They bowed their head downward and flung sand backward with their forelegs, staying mostly in the depressed area of the entrance. The mainly filled area often remained depressed 1–3 mm after the female flew away. Some wasps provisioned continually from mid-morning to late afternoon regardless of sand surface temperature. Eight Kansas females (TX-67) consecutively brought 13–33 grasshoppers to nests in 1.5–5 hours at sand surface temperatures ranging from 38 to 61 ºC. Four Florida females (TX-4) consecutively arrived at nests with 5–18 prey in 3–6 hours at sand surface temperatures ranging from 35 to 63 ºC. The cumulative hours per female did not include duration spent excavating additional burrows and cells in nests. The removal of the temporary closure prior to nest entry and temporary closure were made increasingly faster in a straight-line relationship as sand surface temperature rose from 35 to 60 ºC (Figs. 32, 33). There was essentially no significant difference in temporary closure removal or temporary closure rates of activity or duration between Kansas and Florida wasps (Figs. 32, 33). Wasps left entrances incrementally more open as they spent less time making the closures at higher sand surface temperatures (58–63 ºC). Final closure. Before filling a burrow, some females exited their entrances, made a few short flights around the area, rested briefly on the Figure 31. Female Tachysphex similis makes a temporary closure, using her forelegs in unison, after placing a grasshopper in the nest (TX-7). 2009 F.E. Kurczewski 57 surface, re-entered their burrows headfirst, came back out, turned around, reentered again, and began closing. A wasp began filling the burrow by breaking down the top and sides with her mandibles, twisting her head from Figure 32. Mean duration of temporary closure removal (in seconds) plotted against sand surface temperature (in degrees C) in Tachysphex similis (TX-4, 7, 55, 61, 63, 66, 67). Figure 33. Mean duration of temporary closure (in seconds) plotted against sand surface temperature (in degrees C) in Tachysphex similis (TX-4, 7, 55, 61, 63, 66, 67). 58 Northeastern Naturalist Vol. 16, Monograph 3 left to right and vice versa. The loose soil was thrown backward beneath her synchronously lifting abdomen, using her forelegs in unison. Falling at the end of the burrow in front of the cell, the loose sand was tamped rapidly into place with the bent end of the abdomen. One wasp atypically entered her burrow headfirst during 13 of 19 exits to retrieve sand for the fill. As a female increasingly filled her burrow with sand, she eventually appeared in the entrance 2.5–24 min (mean = 8.9 ± 8.4, n = 13) after taking in the last prey and ovipositing. As she walked onto the surface, her forelegs continuously raked sand backward into the opening beneath her synchronously lifting abdomen. Her wings were held flat on her dorsum. Staying outside at first only a second or two, she backed quickly into the burrow with sand. She spent increasingly longer periods of time on the surface to obtain sand as the closure went to completion (Fig. 34). Females came out of their burrows increasingly farther to get loose sand as the supplies near the entrances were depleted (Fig. 35). As a burrow filled with sand, the intervals at which the wasp came out to get additional soil gradually decreased in duration (Fig. 36). One wasp nesting on a slope turned around and, standing on hindlegs, broke down the burrow walls and entrance using mainly her mandibles. Antenna cleaning was not observed during 13 final closures. One wasp made a hovering flight above the area, turned around in mid-air, landed facing the entrance, turned away, and backed into the burrow raking sand backward. After filling the burrow, females spent 1.0–3.5 min smoothing the sand surface Figure 34. Mean duration (in seconds) spent on surface raking sand backward into burrow for final closure plotted against sand retrieval number in Tachysphex similis (bar graph averages in sets of five with standard error bars; TX-4, 7, 55, 61, 63, 66). 2009 F.E. Kurczewski 59 in front of an entrance. They periodically interrupted this behavior and walked across the sand to examine the fill 0–28 (mean = 14.4 ± 14.4, n = 10) times Figure 36. Mean interval (in seconds) between successive exits to obtain sand for final closure plotted against sand retrieval number in Tachysphex similis (bar graph averages in sets of five with standard error bars; TX-4, 7, 55, 61, 63, 66). Figure 35. Mean distance (mm) to which wasps walked to get loose sand for final closure plotted against sand retrieval number in Tachysphex similis (bar graph averages in sets of five with standard error bars; TX-4, 7, 55, 61, 63, 66). 60 Northeastern Naturalist Vol. 16, Monograph 3 during closure. Wasps made 9–23 trips (mean = 16.6 ± 5.1, n = 10) onto the surface to obtain sand for the fill. The amount of time spent filling the burrow and distributing sand atop the entrance before departing was 6–115 min (mean = 17.0 ± 32.1, n = 11). Two wasps in Florida (TX-42, 72) spent 105 and 120 min, respectively, to excavate, provision, oviposit on a prey, and close 2 one-celled nests. Nest structure and dimensions. Females dug short shallow burrows that entered the sand obliquely at angles of 21–46° with the surface. Thirty of 50 (60.0%) burrows penetrated the soil at lateral curvatures of greater than 22.5° (Fig. 21). Some burrows were nearly circular in configuration. Nests were one- to five-celled with three-celled nests being prevalent (Fig. 21). Burrows leading to second cells left the main burrows at distances of 17–84 mm (mean = 40.3 ± 21.1, n = 17) from the entrances. Burrows leading to third cells left the main burrows at distances of 10–66 mm (mean = 31.3 ± 21.8, n = 6) from the entrances. Distances between first and second cells Table 8. Burrow length and cell depth in Tachysphex similis nests.3 Site Burrow length (mm) Cell depth (mm) TX-year4 N Range Mean ± SD N Range Mean ± SD 32 1 63 63.0 1 28 28.0 42-1962 3 50-62 57.3 ± 6.43 3 30-32 31.0 ± 1.00 41-1963 2 79-95 87.0 ± 11.31 2 45-49 47.0 ± 2.83 41-1965 9 97-128 107.3 ± 10.76 9 49-67 60.9 ± 6.21 41-1966 6 81-99 89.8 ± 6.71 6 49-58 53.0 ± 3.10 42-1967 1 114 114.0 1 59 59.0 41-1968 8 85-107 95.4 ± 7.52 8 39-54 46.9 ± 5.00 62 1 100 100.0 2 45-65 55.0 ± 14.14 72-1962 1 75 75.0 1 34 34.0 71-1963 5 49-79 67.6 ± 11.39 5 37-52 42.8 ± 6.26 72-1967 6 59-73 67.2 ± 5.81 6 34-53 40.5 ± 6.89 72-1968 1 71 71.0 1 36 36.0 161 5 63-88 74.6 ± 9.86 5 45-58 50.8 ± 5.36 402 5 48-71 63.4 ± 9.24 5 27-34 30.8 ± 3.11 432 3 53-62 58.0 ± 4.58 3 31-41 37.3 ± 5.51 551 15 44-83 64.1 ± 12.98 15 16-37 27.1 ± 5.51 581 3 90-112 100.0 ± 11.14 3 37-39 37.7 ± 1.15 611 1 42 42.0 1 21 21.0 632 5 44-83 63.0 ± 15.28 5 17-47 34.8 ± 15.37 662 1 39 39.0 1 15 15.0 672-1965 2 43-55 49.0 ± 8.49 2 18-21 19.5 ± 2.12 671-1966 8 57-84 69.5 ± 8.65 14 25-49 33.6 ± 6.42 812 1 60 60.0 1 30 30.0 821 2 75-83 79.0 ± 5.66 2 37-38 37.5 ± 0.71 941 2 48-55 51.5 ± 4.95 2 18-20 19.0 ± 1.41 961 3 41-61 52.0 ± 10.15 3 22-25 23.3 ± 1.53 981 3 45-55 51.7 ± 5.77 3 28-34 30.7 ± 3.06 1001 3 58-63 60.3 ± 2.52 3 26-29 27.7 ± 1.53 1Spring generation. 2Summer generation(s). 3Includes cell length and cell height, respectively. 4Table 7 provides a key and brief description for TX- sites. 2009 F.E. Kurczewski 61 were 9–65 mm (mean = 39.6 ± 24.2, n = 16) and between second and third cells were 8–56 mm (mean = 34.2 ± 12.9, n = 9). Burrow diameters were 3–6 mm, enlarging to 4–6 mm at the entrances. Each burrow ended in a somewhat declined cell or short spur. The cells were oval to elongate oval in shape. Cell height, width, and length were similar at all localities. Fifty-one of 55 (92.7%) cells measured 5–8 mm high, 6–9 mm wide, and 9–12 mm long. Longer and deeper spring nests and shorter and shallower summer nests were the norms at several Florida and Kansas sites (Table 8). Florida cells were significantly deeper than Kansas cells throughout the nesting season: Florida mean = 47.6 mm, Kansas mean = 29.8 mm; t = 8.5796, P < 0.0001, df = 57 (Fig. 37). There was no correlation between wasp size (weight) and cell depth in Kansas (r = 0.0312) or Florida (r = -0.0584) when the two samples were separated. However, when the two samples from the different regions were combined, the correlation was much stronger (r = 0.5365; Fig. 37). Prey. Prey consisted entirely of small nymphal Acrididae. Overall, the wasps were rather unselective of their prey, capturing 14 genera and 24 species based on 812 specimens (Table 9). Several hundred specimens remain unidentified, so that number is probably on the low side. The prey subfamilies consisted of Cyrtacanthacridinae + Melanoplinae, (581 specimens, 71.6%), Gomphocerinae (122, 15.0%), and Oedipodinae (109, 13.4%). As many as three subfamilies, genera, and species of Acrididae were found in a single cell. Prey species differences associated with different generations of wasps were not evident in Florida, as the females nested throughout much of the year Figure 37. Weight (mg) of female Tachysphex similis plotted against cell depth (mm) in Kansas (TX-55, 58, 61, 63, 66, 67) and Florida (TX-4, 7). 62 Northeastern Naturalist Vol. 16, Monograph 3 and had access to many of the same species of grasshoppers, more or less, year-round. Most prey from Kansas were from first-generation (May–June) nests, making a comparison with prey from later generations difficult to make. The number of prey stocked in fully provisioned cells was 1–15 (Table 10). Most cells held three or more grasshoppers. First-generation (spring) cells generally held more prey than later-generation (summer) cells (Table 10). Some cells with only one or two prey were closed prematurely at the onset of bad weather or dusk. In Kansas, first-generation all-black females stocked 5–14 (mean = 8.7 ± 2.8, n = 15) grasshoppers per fully provisioned cell while first-generation red-tipped abdomen females stocked 5–15 (mean = 8.9 ± 3.3, n = 16) prey per fully provisioned cell, an insignifi- cant difference. The contents of early summer, fully provisioned cells from one location in Florida (TX-72 [1967, 1968]) with only one or two prey may have resulted from scarcity of grasshoppers of suitable size and quasi-droughty conditions. Only 196 mm of rain fell between January 1 and June 1, 1967 compared to the normal average amount of 357 mm at that location. The site received only five days of rainfall totaling 67 mm in April and May 1967. Two wasps each spent nearly three days between successful provisioning trips. Several days of net-sweeping around the nesting area indicated a paucity of nymphal Acrididae of appropriate prey size. Table 9. Species of prey of Tachysphex similis, listed in alphabetical order. Species of prey (nymphs) No. specimens Achurum carinatum (Walker) 10 Amblytropidia mysteca (Saussure) 4 Aptenopedes sphenarioides Scudder 48 Arphia granulata Saussure 65 Arphia simplex Scudder 32 Arphia xanthoptera (Burmeister) 1 Chloealtis conspersa (Harris) 12 Chorthippus curtipennis (Harris) 6 Dissosteira carolina (Linnaeus) 2 Melanoplus bivittatus (Say) 2 Melanoplus differentialis (Thomas) 77 Melanoplus femurrubrum (DeGeer) 32 Melanoplus keeleri (Thomas) 30 Melanoplus puer group 43 Melanoplus s. sanguinipes (Fabricius) 64 Melanoplus sanguinipes vulturnus Gurney and Brooks 43 Melanoplus seminole Hubbell 18 Melanoplus viridipes Scudder 5 Melanoplus spp. 164 Melanoplinae, unknown genus 21 Orphulella pelidna (Burmeister) 58 Psinidia fenestralis (Serville) 17 Schistocerca americana (Drury) 35 Spharagemon marmorata picta (Scudder) 6 Syrbula admirabilis (Uhler) 17 2009 F.E. Kurczewski 63 In multi-celled nests with fully provisioned cells, cell 1 usually contained more prey (mean = 10.0 ± 3.3, range = 2–14, n = 33) than cell 2 (mean = 6.0 ± 2.7, range = 1–12, n = 27) or cell 3 (mean = 5.4 ± 2.6, range = 3–10, n = 8). One fully provisioned cell 4 held 15 grasshoppers. Most of the grasshoppers stocked in the cells were the same size and weight or smaller and lighter in weight than the wasps. Prey individuals ranged from 3.0 mm (Melanoplus sp.) to 11.5 mm (Aptenopedes sphenarioides, Achurum carinatum) in body length. Florida prey weighed significantly less in spring (March–April mean = 8.9 ± 5.6 mg, range = 2–31, n = 174) than in early summer (late June–July mean = 19.6 ± 11.8 mg, range = 6–45, n = 38) (P < 0.001). In Kansas, prey of the spring generation weighed, on average, only about half as much as prey from mid- to late summer (May–June mean = 6.4 ± 3.2 mg, range = 3–21, n = 250 vs. July–August mean = 11.6 ± 6.8 mg, range = 3–31, n = 31) (P < 0.05). Table 10. Number of prey and aggregate prey weight per cell in Tachysphex similis nests. Site No. prey per cell Aggregate prey wt. (mg) TX-year3 n Range Mean±SD n Range Mean±SD 32 1 3 3.0 42-1962 1 3 3.0 41-1965 7 2–8 5.3 ± 2.21 7 26–102 53.7 ± 28.16 41-1966 15 2–12 6.4 ± 2.85 10 31–71 49.3 ± 12.50 42-1967 8 3–10 7.0 ± 2.48 1 75 75.0 41-1968 6 6–11 8.0 ± 2.27 41-1986 5 2–7 4.2 ± 2.17 5 20–67 41.8 ± 20.78 62 1 9 9.0 72-1962 1 3 3.0 71-1963 4 5–13 9.2 ± 4.35 4 34–98 67.8 ± 33.98 71-1967 22 3–10 6.4 ± 2.34 72-1967 6 1–2 1.67 ± 0.52 6 31–80 51.8 ± 20.72 72-1968 1 1 1.0 1 42 42.0 161 4 3–9 5.0 ± 2.71 4 33–77 52.3 ± 21.31 402 4 2–5 3.0 ± 1.41 432 1 5 5.0 551 12 5–14 7.9 ± 2.86 12 29–74 52.8 ± 16.65 581 3 5–11 8.7 ± 3.21 3 46–85 64.0 ± 19.67 611 1 5 5.0 1 29 29.0 632 5 4–5 4.4 ± 0.55 5 28–82 48.0 ± 21.01 662 4 3–8 5.3 ± 2.22 1 41 41.0 672-1965 1 5 5.0 1 37 37.0 671-1966 15 5–15 9.1 ± 3.20 12 32–91 56.9 ± 19.76 812 1 3 3.0 1 29 29.0 821 2 8–9 8.5 ± 0.71 2 37–49 43.0 ± 8.48 871 4 6–8 7.0 ± 0.82 941 2 8 8.0 2 42–51 46.5 ± 6.36 961 2 4 4.0 2 53–62 57.5 ± 6.36 981 2 5 5.0 2 56–83 69.5 ± 19.09 1001 3 5–7 6.0 ± 1.00 3 45–65 58.0 ± 11.27 1Spring generation. 2Summer generation(s). 3Table 7 provides a key and brief description for TX- sites. 64 Northeastern Naturalist Vol. 16, Monograph 3 The aggregate prey weight of cells from the various sites in presented in Table 10. Cells with higher aggregate prey weight probably produced larger female wasps while those with lower aggregate prey weight probably produced smaller male wasps, although this was not proven during the study. The cell contents of first-generation all-black females in Kansas had an aggregate prey weight of 29–91 mg (mean = 57.7 ± 18.7, n = 15), while that of first-generation red-tipped abdomen females in Kansas had an aggregate prey weight of 29–85 mg (mean = 51.6 ± 18.2, n = 13). In multi-celled nests with fully provisioned cells, cell 1 had a higher aggregate prey weight (mean = 59.3 ± 19.5 mg, range = 28–102, n = 25) than cell 2 (mean = 40.1 ± 10.7 mg, range = 26–61, n = 19) or cell 3 (mean = 45.8 ± 11.5 mg, range = 33–66, n = 6). A decrease in the aggregate prey weight per fully provisioned cell from spring to mid-summer was evident at a few sites (Table 10). Nesting wasps weighed 8–20 mg (mean = 12.5 ± 3.5, n = 77). Florida females were significantly heavier (mean = 15.7 ± 3.3 mg, range = 10–20, n = 28) than Kansas females (mean = 10.3 ± 1.7 mg, range = 8–14, n = 38) (t = 7.8622, P < 0.0001, df = 38; Fig. 37). Kansas females included all-black individuals (mean = 10.7 ± 1.7 mg, range = 8–14, n = 22) from six locations and wasps with a red abdominal apex (mean = 10.3 ± 1.7 mg, range = 8–13, n = 16) from four sites (TX-55, 58, 61, 67). The slightly heavier weights of the all-black females corresponded with the slightly higher cell aggregate prey weights of this color variant (see above). Three hundred fifty-eight of 450 (79.6%) prey were put in the cells in a head inward and ventral side upward or on the side position. Eighty-three (18.4%) grasshoppers were placed in the cells in a head inward and dorsal side upward position. Five prey were put in the cells head outward and ventral side upward, and four grasshoppers were put in head outward and dorsal side upward. Grasshoppers in atypical positions may have been accidentally moved by wasps maneuvering to oviposit on another prey in the cell. Egg. A female usually laid her egg on a grasshopper placed toward the bottom of the cell near the back end, especially where there were multiple prey in the cell. The grasshopper to which the egg was affixed was positioned head inward and ventral side upward in 167 of 180 (92.8%) cells. Eleven eggs were attached to prey that were positioned head inward and dorsal side upward and 2 to prey placed head outward and ventral side upward. The grasshopper to which the egg was affixed was often the largest and heaviest prey in the cell. Fifty-four of 90 (60.0%) eggs were laid on the heaviest prey, 28 (31.1%) on the second heaviest, 4 on the third heaviest; 2 on the fourth heaviest, and 2 on the lightest grasshopper in the cell. Three cells contained only a single prey. Eggs were attached to prey that weighed 3–44 mg (mean = 14.5 ± 9.5, n = 73). Other prey in the cells weighed 2–45 mg (mean = 7.1 ± 4.7, n = 274), a significant difference (P < 0.001). Egg-bearing grasshoppers were the first (2 examples), second (2), or third (3) individuals taken into entrances in seven of nine cells. Eggs were laid on other sequential prey in two cells. 2009 F.E. Kurczewski 65 The wasp’s egg was affixed to the soft intersegmental membrane surrounding the base of a forecoxa by the less tapered, distal end. The proximal end extended free to the opposite side, the egg lying transversely across the pro- and mesosterna between the bases of the fore- and midcoxae. Forty-six (51.1%) eggs were attached to the intersegmental membrane surrounding the base of a left forecoxa, and 44 (48.9%) eggs to the intersegmental membrane surrounding the base of a right forecoxa. Fourteen of 24 (58.3%) eggs in two-celled nests were affixed to the same (left or right) side of the prey’s forecoxal intersegmental membrane, and nine of 18 (50.0%) eggs in threecelled nests were attached to the same side of the grasshopper’s forecoxal intersegmental membrane. Live eggs were slightly curved, sausage-shaped, white or cream-colored, and measured 1.5–1.9 mm (mean = 1.77 ± 0.15, n = 12) long and 0.4–0.5 mm (mean = 0.43 ± 0.05, n = 12) wide. Tachysphex apicalis apicalis Literature review Krombein (1964) located one nest of T. a. apicalis in a 60° sand slope in Highlands County, fl. The female was bringing prey to her nest in flight, entering the open burrow, and exiting a few seconds later. This nest was excavated after the female began her final closure. The burrow went in “almost horizontally, turned downward at a right angle,” and terminated in a cell 4 cm from the sloped surface. The cell held seven nymphal grasshoppers, 7–10 mm long, of two different instars of Melanoplus sp., possibly puer (Scudder). The sausage-shaped egg, 2.2 mm long, was affixed on the left side extending transversely across the sternum between the fore- and mid-coxae of a “medium-sized” prey. Pulawski (1988) reported this species nesting in vertical portions of land tortoise burrow entrances at the Archbold Biological Station, fl. Tachysphex a. apicalis was noted nesting in steep sand slopes and cliffs in Florida (Kurczewski and Snyder 1968). Burrow excavation was begun using the mandibles with assistance from the forelegs in unison following penetration of the surface. Females did not temporarily close the entrance with loose sand after completing a burrow and cell. They flew off and returned in flight, often several times, with small grasshoppers. Most wasps entered their burrows without pausing, holding the prey underneath with their mandibles and legs. Some females, with larger grasshoppers, released them just within the entrance and pulled them in from inside the burrow. After finishing a nest, the female broke down the area of the entrance with her mandibles and packed the loose soil in the burrow with the end of her abdomen. Nests often contained several cells, with most cells each containing several nymphal Acrididae. Kurczewski (2000) located a three-celled nest of T. a. apicalis near the top of a sandy incline in Bladen Lakes State Forest, Bladen County, NC. The cells were unearthed at distances of 56, 62, and 73 mm, including cell length, 66 Northeastern Naturalist Vol. 16, Monograph 3 from the surface. They contained four, five, and seven small acridid nymphs identified as Melanoplus spp. (9 examples), Orphulella pelidna (5), and an undetermined species of Oedipodinae (2). The grasshoppers were placed in the cells head inward and ventral or dorsal side upward or on the side. The wasp weighed 23 mg and the prey averaged 14.3 mg (range = 9–27, n = 16). Wasp eggs, ca. 2 mm long, adhered to the basal left (1) or right (2) forecoxal intersegmental membrane of the innermost grasshoppers in the cells. Field studies Nest location. Tachysphex a. apicalis nested in bare soil or soil with a minimum of vegetation, often the upper portions of sand cliffs along watercourses. The cliffs formed 80–90° angles with the horizon. The soil consisted of tan, gray, or white, dry to moist, compact layers of fine to coarse stratified sand. Some females nested in soil where the slope was 45–60°. The high density of individuals nesting in the vertical slopes may have forced these wasps to occupy less vertical areas elsewhere. A few females attempted to nest in loose, level sand, but their attempts were largely unsuccessful. An aggregation of T. a. apicalis was found underneath a recently constructed concrete bridge spanning the Peace River, 1 km west of Arcadia, fl(TX-5). More than 50 wasps nested in a 1.2-m-high, hard-packed sand cliff forming the western bank of the river during 3–19 July 1962. Wasps also nested at that site 25–26 March 1963, 28 March 1965, 6 April 1966, and 29 March 1968. The entire nesting area was devoid of direct sunlight, contained no vegetation, and never received rainfall. In July 1962, nest entrances dotted the upper third of the cliff for a distance of 15 m. I unearthed 95 nests overall and 19 nests at the center of the aggregation in an area 55 cm long. Many entrances were 2–4 cm apart with only 1.7 cm separating two openings. Nests on the periphery of the aggregation were 30–50 cm apart. This species also nested in the upper portion of a 2.8-m-high, man-made sand cliff at the Archbold Biological Station, Florida. Wasps were observed there on 14 July 1962, 28 March 1965, and 1 April 1970. This site was not given a site code number because nests were out of reach for excavation and study. Several females were studied and photographed at a 3-m-high sand cliff forming one side of a drainage ditch near Lakeport, Glades County, fl(TX- 17; Fig. 38). This was a less dense aggregation than the one near Arcadia with the closest entrances being 15–17 cm apart. Forty-one nests were excavated in the upper third of the cliff 29 March–3 April 1963 and 29 March 1965. Five females and their nests were observed and excavated in a sand cliff forming one side of a Native American burial mound near Venus, Highlands County, fl31 March–4 April 1963 (TX-18). Activity conditions. Females in southern Florida nested from February to December (R.C. Miller, 1973 pers. comm.). Nesting activity near Arcadia was at its height on sunny days at air (shade) temperatures of 27–35 ºC. As air temperature never fell below 27 ºC when making observations, it was 2009 F.E. Kurczewski 67 impossible to derive a minimum activity temperature. Nesting declined noticeably or ceased entirely during periods of inclement weather, especially rainfall, even though the aggregation near Arcadia was sheltered beneath the concrete bridge. The rainfall must have impeded hunting and prey capture. Females at that locality began nesting in numbers about 0900 (EST). The earliest times for wasps bringing prey to nests were 0859, 0901, and 0908. By 1800 (EST), most females in the aggregation had stopped nesting and retired to open burrows, some with males already inside. Burrow excavation. When beginning burrows, females walked slowly in a zigzag manner across the sand cliff or steep slope. Wasps paused momentarily every few centimeters and tapped their antennae slowly but incessantly on the soil. Females resisted the pull of gravity by grasping the substrate with their tarsal claws. Some wasps selected narrow ledges from which to start digging. Such horizontal surfaces afforded them a positional advantage over the smooth cliff face. Some females made up to eight attempts to excavate a burrow before remaining in one place and finishing. Other wasps stayed in one place and completed the burrow they started. Females used their mandibles exclusively when starting an excavation, producing an audible buzzing sound. One wasp, excavating with her mandibles, clutched the soil around the opening with her tarsal claws for leverage. Extensive use of the mandibles was required because of the compact stratified sand. Loosening the soil with the mandibles formed a cavity in the cliff or slope. Females occasionally stopped, backed to the surface, and walked in circles around the opening. Figure 38. Western side of man-made drainage ditch near Lakeport, Glades County, fl(TX-17). Tachysphex apicalis nested in the vertical sand cliff. Tachysphex similis nested in level, loose sand of the field atop the cliff (TX-16). 68 Northeastern Naturalist Vol. 16, Monograph 3 A wasp used her forelegs in unison to throw the loosened sand backward down the cliff or slope and out of reach. Her abdomen lifted up and down in synchrony as the sand was flung backward. The mid- and hindlegs were used only for walking and support. Females threw sand backward at the rate of about 4–6 scuffs per second in direct sunlight (TX-17) and 2–3 times per second in the shade (TX-5). Wasps attempting to dig in very loose sand in the shade atop the sand cliff at Arcadia never completed a burrow because the surface sand kept falling into the excavation and they couldn’t remove it fast enough. One female spent over an hour making 16 excavations in the loose sand, but none went to completion. As an excavation progressed, a female rather slowly backed out to remove the loose sand that accumulated in the burrow and entrance. This removal was accomplished by flinging the sand backward beneath her abdomen with her forelegs in unison. Her wings were held flat on her dorsum. The intervals at which one wasp backed out to remove the loose sand usually increased from beginning to end of an excavation—from an average of 15.3 sec for sand removal episodes 1–10 to 79.2 sec for sand removal episodes 57–66 (TX-5). The first few intervals generally consumed more time than the next few intervals because it was necessary to penetrate the sand crust by digging exclusively with the mandibles. Females spent up to 3.8 min for the last interval inside the burrow, presumably to shape a cell. Wasps nesting in sand cliffs never backed farther than 13–16 mm when clearing soil from their burrows, and even that distance was attainable only when a narrow ledge was present. Most females were limited to backing out only 6–9 mm from the cliff face. Four wasps nesting in 10–20° sloped, sunlit sand removed the soil backward to maximal distances of 13, 17, 20, and 32 mm. One female’s maximal distance averaged 9 (6–12) mm during sand removals 1–10 and 22 (17–30) mm for sand removals 56–65. Females always backed out removing the loose sand in a rather straight direction, being limited by the cliff face. They almost never turned sideways or moved laterally. Very little sand accumulated around the entrance as most fell down the cliff face or steep slope. Only one of four wasps nesting in nearly level sand atop the cliff at Arcadia moved to the side when removing loose sand from her burrow. The result of her excavation was a shallow mound of soil about 30 mm wide and 50 mm long below and to the right of the entrance. This female made no attempt to level or remove the tumulus. Three wasps took 16, 43, and 72 min to complete a burrow and first cell. The duration of excavation was correlated with burrow length; the 16-minute long dig produced a 49-mm long burrow and the 72-minute long excavation resulted in a burrow 130 mm in length. After making the cell, a female appeared in her entrance headfirst with her wings held flat on her dorsum. Her antennae tapped the floor of the entrance, and her abdomen pulsated rhythmically. Some wasps then turned around, entered the burrow, flinging sand backward with their forelegs, turned around 2009 F.E. Kurczewski 69 inside the burrow, and re-exited headfirst 5–8 sec later (Fig. 39). Females then made a series of circular orientation flights in front of the entrance, periodically landing on the cliff face to rest and clean the antennae and mouthparts with the forelegs before flying off in search of prey (Fig. 40). Figure 39. Female Tachysphex apicalis exiting an open burrow to fly in search of prey (TX-17). Note long digging spines on left foretarsus. Figure 40. Female Tachysphex apicalis on a sloped sandy cliff ledge, pausing to clean her mouthparts with her foretarsi (TX-5). 70 Northeastern Naturalist Vol. 16, Monograph 3 Additional cells. After provisioning the first cell and ovipositing on a prey, the female may excavate a second and, later, third cell at the ends of short side burrows off the main burrow. While making these burrows and cells, the wasp appeared inside her entrance infrequently to remove loose soil. She packed most of the sand from subsequent burrows into the burrows of cells made earlier, below the surface and out of sight. A wasp backed into an entrance flinging sand backward or she emerged headfirst and raked sand backward into the burrow 20–25 min after beginning a side burrow and second cell. Females digging a side burrow and third cell backed from the excavation 11–13 min after taking in the last prey for the second cell. Wasps backed out with loose sand only 4–7 and 3–6 times during excavation of second and third cells, respectively. The intervals at which they removed the loose sand were of longer duration than for the first burrow and cell, averaging about 2 min. One female spent 32 and 27 min for excavation of side burrows and second and third cells, respectively (TX-17). Wasps usually did not make an extensive orientation flight after excavating a second or third cell, having already familiarized themselves with their surroundings. Females in Florida completed up to three fully provisioned cells per day in April (TX-171) and July (TX-52). Hunting. Females searched for prey on plants as far as 50–60 m from their nests. They made slow, deliberate, winding and hovering flights, 10–15 cm high, facing the vegetation. One wasp (TX-17) hovered in flight near the leaves and attacked a grasshopper from above. Two females stopped flying and landed to pounce on a Chortophaga australior and Aptenopedes sphenarioides, respectively, stinging the grasshoppers in their ventral thoraces. One unidentified prey eluded a hunting wasp by jumping away. Hunting females walked up plant stems, paused, and tapped their antennae on the petiole, in some cases continuing up the stem to the next petiole, walking onto the leaves and searching further (Fig. 41). The relative slowness with which a wasp searched for prey may be connected with the capture of relatively small nymphal grasshoppers that cannot leap far. If unsuccessful in capturing prey, a female stopped hunting and returned to her nest. She landed near the entrance, walked on the cliff face or slope tapping her antennae, entered the burrow, and exited 6–8 sec later or took flight without entering and returned to the hunting area. Several wasps returned to their nests in flight, did not enter, but hovered in front of the entrance for a second or two, turned in mid-air, and flew away. Periodic returns to the nest were made at intervals of 4–23 min (n = 40). Prey transport. After stinging a prey in the underside of its thorax near a leg base, the female moved aside and cleaned herself. She then often straddled the grasshopper ventral side upward and pressed her mandibles against the soft intersegmental membrane surrounding the base of the prey’s forecoxa. She grasped the bases of the prey’s antennae with her mandibles and its body with her legs and took flight toward the nest (Fig. 42). Rarely, grasshoppers were carried dorsal side upward instead of ventral side upward. The relatively small prey was almost always carried directly to the nest in 2009 F.E. Kurczewski 71 a prolonged flight. However, a few wasps interrupted their flight to land on vegetation and rest. All 199 examples of prey transport in this species involved flight. The mean ratio of weight of prey to wasp was 0.8:1 (range = 0.23–2.70, n = 80). Figure 41. Female Tachysphex apicalis searching for prey on sensitive plant (TX-5). Her antennae continually tap leaflets and petiole. Figure 42. Female Tachysphex apicalis flying toward entrance with prey (TX-17). She grasps grasshopper’s antennae with her mandibles and its body with her legs. 72 Northeastern Naturalist Vol. 16, Monograph 3 Nest entry. The grasshopper was usually carried directly into the open burrow, the wasp holding the prey tightly underneath (Fig. 43). As viewed from behind, a provisioning wasp had her wings spread dorso-laterally as she approached the cliff, but upon entering the burrow she rapidly brought them downward and backward. After alighting inside the entrance, she continued to run forward holding the grasshopper underneath and headfirst but carried farther back. Prior to entering, some females with prey landed nearby on the cliff or slope. Wasps with larger prey released the grasshopper inside the opening (Fig. 44), turned around, probably in the cell, walked up the burrow, grasped the prey by an antenna and, with it, backed down the burrow. Females that entered directly with prey reappeared headfirst in their entrances 6–33 sec (mean = 16.8 ± 8.3, n = 38) after entering and flew away. Females spent 2–48 min (mean = 14.4 ± 10.7, n = 33) between consecutive entries with prey. One wasp, whose entrance had been disturbed, flew in front of the cliff face for over an hour before abandoning her grasshopper. Final closure. A female began her final closure inside the nest in a headoutward position. She pulled down soil from the top and sides of the main burrow with her mandibles, sometimes producing a buzzing sound. The loosened soil was thrown backward with her forelegs beneath her abdomen as it synchronously moved up and down. This soil was tamped down at the end of the main burrow by rapid repeated blows from the end of her abdomen. The female first appeared in her entrance to get additional soil for the fill 5–22 min (mean = 13.5 ± 12.0, n = 2) after entering with the last prey for the cell. Figure 43. Female Tachysphex apicalis flying directly into an open entrance holding small grasshopper underneath (TX-17). 2009 F.E. Kurczewski 73 There was usually no loose soil available for use as fill, forcing the wasp to scrape sand from the top and sides of the opening with her mandibles. She removed sand from around the entrance, raked the loosened soil down the burrow with her forelegs, and tamped it in place by repeated blows from the end of her abdomen. The entrance enlarged considerably as she continued to scrape sand from around it. The wasp stood on hindlegs and, hanging onto the sides of the enlarged cavity with her midlegs, pulled down additional sand with her mandibles and forelegs (Fig. 45). After several minutes of such behavior, she walked around the opening a few times and flew off leaving a 12–15 mm high, 14–19 mm wide, and 5–10 mm deep depression in the cliff face or slope. Closures (TX-5) of burrows 49 and 130 mm long took 19 and 42 min, respectively, from beginning to end. Nest structure and dimensions. Burrows penetrated the sand cliffs in nearly straight lines sloping slightly with the cliff faces (Fig. 46). Only 56 of 147 (38.1%) burrows exhibited lateral curvature of more than 22.5° (Fig. 21). A few burrows coursed inward sinuously, following veins of soft sand. Burrows were 5.0–6.5 mm and entrances were 6.0–8.0 mm in diameter. The relatively large entrances were wide enough to allow simultaneous entry of wasp and prey. Several finished nests were three-celled (Figs. 21, 46). Females prematurely closed some nests with only one or two cells at the onset of rainfall, extreme cloud cover, or dusk. Most nests had second and third cells at the ends of short side burrows leading from the main burrow. Side burrows from the main burrow to a second cell left the main burrow at distances of 29–41 mm (mean = 36.2 ± 4.6, n = 5) from an entrance. Figure 44. Large prey of Tachysphex apicalis released just inside entrance (TX-17). Wasp turned around inside nest and pulled in grasshopper seconds later. 74 Northeastern Naturalist Vol. 16, Monograph 3 Side burrows from the main burrow to a third cell left the main burrow at distances of 17–25 mm (mean = 20.3 ± 4.2, n = 3) from an entrance. Some females in attempting to excavate an additional burrow and cell encountered clay, stones, roots, pieces of wood, and other debris. Wasps Figure 46. Enlarged entrance, main burrow, and first cell of Tachysphex apicalis nest exposed in excavation of cliff face (TX-17). Paralyzed grasshoppers are positioned head inward and mainly ventral side upward in cell. Sand-filled second and third burrows furcate from main burrow toward entrance. Figure 45. Female Tachysphex apicalis standing upright inside entrance area to pull down sand with her mandibles for final closure (TX-17). 2009 F.E. Kurczewski 75 Table 11. Burrow length (mm)1 in Tachysphex a. apicalis nests. Site (TX- , year) n Range Mean ± SD Arcadia, fl(TX-5, 1962) 95 32–95 62.9 ± 13.62 Arcadia, fl(TX-5, 1963) 4 49–62 56.0 ± 5.72 Arcadia, fl(TX-5, 1965) 1 130 130.0 Lakeport, fl(TX-17, 1963) 39 45–79 63.0 ± 8.22 Lakeport, fl(TX-17, 1965) 2 107–122 114.5 ± 10.61 Venus, fl(TX-18, 1963) 5 46–66 54.8 ± 7.40 1Includes cell length. Table 12. Species of prey of Tachysphex a. apicalis, listed in alphabetical order. Species of prey (nymphs) No. specimens Achurum carinatum (Walker) 26 Amblytropidia mysteca (Saussure) 10 Aptenopedes sphenarioides Scudder 5 Arphia granulata Saussure 22 Chortophaga australior Rehn and Hebard 16 Dichromorpha viridis (Scudder) 1 Melanoplus keeleri (Thomas) 1 Melanoplus puer group 13 Melanoplus sanguinipes vulturnus Gurney and Brooks 9 Melanoplus seminole Hubbell 179 Melanoplus seminole Hubbell (adult male) 2 Orphulella pelidna (Burmeister) 71 Psinidia fenestralis (Serville) 3 Spharagemon marmorata picta (Scudder) 11 Trimerotropis maritima (Harris) 74 abandoned such excavations and dug elsewhere, leaving short incomplete burrows or spurs. Females, sometimes in tandem with one or more males, spent the night and periods of inclement weather in a head-outward position in such a spur. Second and third burrows and associated cells were dug with equal frequency off the left (27) or right (28) side of a main burrow. However, females always excavated a second and third burrow and adjoining cells off one side or the other of the main burrow. The distance from first to second cell was 17–60 mm (mean = 34.9 ± 11.0, n = 18) and from second to third cell, 5–30 mm (mean = 22.1 ± 10.3, n = 5). The cells were oval-elongate in shape, 6–11 mm (mean = 8.2 ± 1.1, n = 45) high, 6–11 mm (mean = 8.3 ± 1.1, n = 44) wide, and 12–22 mm (mean = 15.5 ± 2.7, n = 45) long. Burrow length was variable at three localities, probably due to site limitations (Table 11). Cell depth was variable because of the diverse nesting sites. Cell depth from the top of the cliff or slope to the bottom of the cell ranged from 2 (TX-5) to 55 cm (TX-17). Prey. Females preyed almost entirely on rather small nymphal Acrididae belonging to the subfamilies Gomphocerinae (108 specimens, 24.4%), Oedipodinae (126, 28.4%), and Melanoplinae (209, 47.2%) (Table 12). Two adult male Melanoplus seminole Hubbell were included among the 443 prey that 76 Northeastern Naturalist Vol. 16, Monograph 3 were identified. The wasps captured 11 genera and 14 species of prey Acrididae (Table 12). All three subfamilies and as many as four genera and species of Acrididae were found in a single cell (TX-5). Most of the prey were common species, and the wasps appeared to be capturing whatever was available in the nesting area (A.B. Gurney, US National Museum, Washington, DC, 1963 pers. comm.). It was impossible to discern generational differences in prey species as both the wasps and grasshoppers were active throughout much of the year. The number of grasshoppers stocked in fully provisioned cells was 2–12 (mean = 7.1 ± 2.6, n = 40). The number of prey per cell at one locality (TX-5) decreased from an average of 8.6 (range = 4–11) in cell 1 to 6.7 (range = 4–10) in cell 2 to 5.5 (range = 2–9) in cell 3 (n = 23 cells). Individual prey ranged from 4.0 mm (Amblytropidia mysteca, Melanoplus seminole) to 19.0 mm (Achurum carinatum) in body length and weighed 4–50 mg (mean = 16.2 ± 22.3, n = 148). The wasps weighed 17.0–30.5 mg (mean = 23.8 ± 5.7, n = 5). Aggregate prey weight per cell was 46–197 mg (mean = 114.3 ± 50.3, n = 21). Of 459 prey placed in the cells, 410 (89.3%) were positioned head inward and ventral side upward or on their side (Fig. 46), 42 (9.2%) head inward and dorsal side upward, 6 (1.3%) head outward and ventral side upward, and 1 (0.2%) head outward and dorsal side upward. Egg. The wasp’s egg was usually laid on a grasshopper at the bottom or farthest back in the cell after the full complement of prey was placed inside. One egg was laid on the uppermost grasshopper in the cell. Twenty-five of 28 (89.3%) eggs were placed on a grasshopper positioned head inward and ventral side upward or on the side; 3 (10.7%) eggs were put on a prey that was head inward and dorsal side upward. Eleven of 21 (52.4%) eggs were laid on the heaviest grasshopper in the cell, 4 (19.0%) on the second heaviest, 3 (14.3%) on the lightest, and 3 (14.3%) on other grasshoppers. The eggbearing individual weighed, on average, 30.0 ± 10.1 mg (range = 13–46, n = 21) or nearly twice as much as an average-sized prey (mean = 16.2 ± 22.3 mg, range = 4–50, n = 148) (P < 0.05). The egg was slightly curved, sausage-shaped, white or cream-colored, and affixed by the less tapered distal end to the soft intersegmental membrane surrounding the base of a procoxa, lying transversely between the front and midlegs of the prey. Live eggs were 1.7–2.2 mm (mean = 1.93 ± 0.13, n = 11) and 0.4–0.6 mm (mean = 0.46 ± 0.05, n = 11) in length and width, respectively. Ten of 28 (35.7%) eggs were attached to the intersegmental membrane surrounding the base of the left forecoxa, and 18 (64.3%) to the intersegmental membrane surrounding the base of the right forecoxa. In 4 two-celled nests, both eggs were affixed to the same (left or right) side. However, in 2 threecelled nests, eggs were attached to both the right and left sides. Tachysphex apicalis fusus Literature review Most studies on the nesting behavior and prey of this geographic variant of T. apicalis are reported under the name T. fusus. Williams (1914) noted a 2009 F.E. Kurczewski 77 female in the town of Pratt, Pratt County, KS “endeavoring to fly” with an immature Melanoplus larger than herself. He noted another female in Ness County, KS nesting in a “nearly vertical bank of earth.” The burrow was “two and one-half inches” long, single-celled, and contained two immature “Tryxalinae” (= Gomphocerinae?) in the cell. Williams (1932) recorded Oxya chinensis (Thunberg) (Oxyinae) as prey of this species in Hawaii. Rau and Rau (1918) noted a female of “T. fuscus” digging in loose mortar of an abandoned building foundation in Missouri. The wasp attempted several excavations in over an hour to no avail. A second female of this species dug a horizontal burrow halfway down an ant-lion pit. Elliott and Kurczewski (1985) reported a female of T. apicalis fusus at St. Anthony, ID excavating in soft sand of a prior excavation. Upon completion of the burrow, she left the entrance open, oriented in a hovering flight, and flew away. Upon returning with prey, she landed near the opening or up to 1.5 m away before resuming flight to the nest. All grasshoppers were taken directly into the open entrance without pausing. The female averaged 5.5 min between eight successive provisioning trips and spent 2 min inside between entry and exit. The unfinished nest consisted of a burrow, 7.5 cm long, with a single cell. The cell contained seven nymphal acridids. The female was flying with an eighth prey when collected. The grasshoppers were all identified as Melanoplus sanguinipes. They were positioned in the cell head inward and ventral side upward or on the side. Insect Museum Specimens A female from near the Anacostia River, Prince George’s County, MD (October 13, 1878; Collection T. Pergande) was collected with two immature grasshoppers, Melanoplus femurrubrum and Chortophaga viridifasciata. The label reads “in sand-bank. tunnel filled with young C. f. rubrum & Tr. viridifasciata.” This notation was the first hint that this species nests in strongly sloping sandy situations. A female from Milt’s Pit, 4 mi. SW Davidsonville, Anne Arundel County, MD (September 2, 2007; M. Buck) was collected “along a sandy path … at the bottom of a slope, formerly a sandpit. The wings of the specimen show almost no wear—only a tiny piece missing from the forewing” (M. Buck, 2008 pers. comm.). Another specimen from Windsor, Essex County, ON, Canada (July 31, 2002; M. Buck) was captured “on a sand patch … amidst a small circular track eroded out by dirt bikes” (Buck 2004; M. Buck, 2008 pers. comm.). Field studies Nest location. Two females of T. apicalis fusus (TX-52) were observed nesting in a hard-packed, 2-m-high sand cliff along State Route 25 about 9 km south of Lakin, Kearny County, KS on 27–28 August 1964. One wasp was followed through five unfinished excavations and two fully provisioned one-celled nests. This small aggregation was probably in its late-season stages as inferred from three single-celled nests with unidentifiable 78 Northeastern Naturalist Vol. 16, Monograph 3 grasshopper fragments and mature wasp larvae ready to make cocoons. The five cells were located 85–156 cm from the top of the cliff. Two other females (TX-71) nested in a steeply sloped, 1.5-m-high, coarse sand–fine gravel bed of the Santa Fe Railroad, 6 km southeast of Lecompton, Douglas County, KS on 12–17 September 1965. One wasp was followed through six incomplete excavations and five successful singlecelled nests over six days. Nests were located about 70–80 cm from the top of the railroad bed. Activity conditions. Observations were made from 0950 to 1630 (CDT) at an average sand surface temperature of 44 ºC (TX-52) and from 1000 to 1630 (CDT) at air and sand surface temperatures of 25–29 °C and 45–49 ºC, respectively (TX-71). Burrow excavation. Females searched in shallow depressions or on narrow ledges, 1–5 cm wide, pausing frequently and sampling with their mandibles when looking for a place to begin an excavation. Wasps dug considerably with their mandibles before using their forelegs in unison to remove the loosened soil backward. An intermittent buzzing sound was produced when using the mandibles. Upon abandoning an excavation, some females turned away from the opening and threw sand into it using their forelegs in unison. One wasp spent 41 sec breaking down the top of the entrance with her mandibles before leaving. Wasps dug for 40 sec to 3.75 min before abandoning an excavation. Abandoned excavations started from the surface were only 2–5 mm deep, whereas one begun in a pre-existing hole was 28 mm long. Females abandoned from two to six excavations before remaining in one place and finishing a burrow. At intervals, a wasp backed from her excavation to remove the loosened soil that accumulated in the burrow. This action was accomplished by flinging the soil backward beneath her abdomen as it raised synchronously, using her forelegs in unison. One wasp (TX-52) made a hovering flight in front of the opening during 10 of her first 11 entries into the burrow to get sand. The intervals at which another female backed from her burrow to remove the loose sand usually gradually increased in duration as she dug deeper—from an average of 21 sec for the first 2 min to an average of 69 sec for the last 2 min. Females removed loose sand from the burrows and entrances 31–45 times (mean = 38.0 ± 6.1, n = 5) during an excavation. At the beginning, wasps stayed in the entrances only 3–5 sec before reentering the burrows, but near the end of an excavation, they remained in the entrances an average of 7 sec. The distance to which they backed from the entrance with the loosened soil averaged 13 mm throughout the excavation. This short distance was related to the limitation of the cliff or slope. Excavation of the main burrow and first cell took 18–43 min (mean = 30.4 ± 9.3, n = 6). After excavating two burrows and adjoining cells, one female (TX-52) appeared in her entrance headfirst each time moving her head from side to side, turned around, and reentered the burrow flinging sand backward with her forelegs. Before reentering, she hovered briefly in front of the entrance, 2009 F.E. Kurczewski 79 landed on the cliff face, and held her antennae straight out. Another wasp (TX-71), after excavating five burrows and cells, exited each time tapping her antennae on the entrance floor. She made a hovering flight about 15 cm in height, turning around once in mid-air in front of the entrance and reentered the burrow. Once, she made three such flights followed by reentry and re-exit. The different exit behavior of the two wasps may have been related to the different habitats—a vertical sand cliff (TX-52) and a sloped sandy-gravelly railroad embankment (TX-71). In both cases, exit behavior culminated in 22–29-second-long, 15–20-cm-high hovering orientation flights of increasingly larger radii followed by long, winding flights to the nearest vegetation. Females in Kansas (TX-522, 712) completed only one fully provisioned cell per day in late August–September. Hunting. One female (TX-52) was seen hunting on upright and decumbent vegetation about 20 m from her nest. She flew from plant to plant and walked up and down the stems and on the leaves tapping her antennae alternately. This wasp returned to her nest without prey at intervals of 8, 11, 14, and 46 min, landed near the entrance, and flew off or hovered in front of the opening for a second or two before flying away. Another female (TX-71) was seen hunting similarly. She caught a green Chortophaga viridifasciata on low-growing vegetation, stung it once in its ventral thorax near a leg base for 4 sec, and then again near a different leg base for 7 sec before grasping it and flying off. Prey transport. One female (TX-71) transported 17 grasshoppers consecutively in flight, regardless of their size. The wasp weighed 19 mg and her prey ranged in weight from 8 to 45 mg. The ratio of weight of the heaviest grasshopper to that of the female was 2.37:1. All grasshoppers were held ventral side upward by their antennae with the wasp’s mandibles and by the sides of their body with the wasp’s legs. Nest entry. The wasp took 12–67 min (mean = 45.6 ± 19.6, n = 8) between entering her nest with consecutive prey items over a three-day period (TX- 71). Her smallest grasshopper (weight = 24 mg) was taken directly into the burrow. The other seven individuals (weights = 26–45 mg) were released ventral side upward just inside the entrance with their abdomens protruding 1–4 mm beyond the opening. In each case, the female turned around, probably in the cell, walked up the burrow, and pulled the grasshopper down the burrow by an antenna 6 to 11 sec later. Twice, this wasp released a prey atop the railroad embankment and applied her mandibles to the base of its foreleg before resuming transport to the entrance. Final closure. One female made two final closures on separate days at 1430 and 1535 at sand surface temperatures of 49 and 45 ºC (TX-71). She appeared headfirst in the entrances 6 and 5.5 min, respectively, after taking in the last prey for the cell. With her head directed toward the entrance, she pulled down soil from the top and sides of the burrow with her mandibles and flung this soil backward beneath her abdomen with her forelegs in 80 Northeastern Naturalist Vol. 16, Monograph 3 unison. Her appearances in an entrance decreased from 23–62 sec at first to 7–10 sec later as she filled the burrow, except for one period where she remained out of sight for over 6 min. During the first closure, she stayed outside breaking down the entrance for only 3–6 sec, followed by a single hovering flight before backing down the burrow. For the second closure, she stayed outside breaking down the entrance for 4–13 sec before backing down the burrow, except for a 42-sec-long period of such behavior before disappearing inside for over 6 min. Following both closures, the wasp turned toward the entrance three times. Durations for the two closures were 11.5 and 18.0 min. Nest structure and dimensions. Females excavated short, single-celled nests that entered the cliff face almost perpendicularly (TX-52) or descended at an angle of 25° or less to the sloped embankment (TX-71). Burrows dug into the cliff face were straight or, at most, slightly curved, while those in the sloped embankment exhibited much curvature. Entrance diameter was 5–6 mm and burrow diameter, 4–5 mm regardless of soil type. Burrows in the hard-packed sand cliff were longer (mean = 62.6 ± 7.5 mm, range = 54–72, n = 5), including cell length, than burrows in the more loose soil of the sloped embankment (mean = 48.4 ± 7.1 mm, range = 41–60, n = 5). Cell size (length x height x width) averaged about the same at the two locations: 13.7 x 7.0 x 7.7 mm (TX-52: range = 13–14 x 7 x 7–8, n = 3), and 14.3 x 7.0 x 7.8 mm (TX-71: range = 12–18 x 6–8 x 7–8, n = 4). The cells were oval to oval-elongate in shape. Prey. Females preyed on nymphal Oedipodinae and Melanoplinae (Acrididae) as follows: Arphia simplex Scudder (6 examples; TX-52, 71), Melanoplus sp. (4; TX-52), and Chortophaga viridifasciata (15; TX-71). Chortophaga viridifasiata included both gray-streaked and green color forms (A.B. Gurney, 1969 pers. comm.). The number of grasshoppers placed in the cells for larval food averaged 4.0 (range = 4, n = 2; TX-52) and 3.4 ±0.5 (range = 3–4, n = 5; TX-71). Individual prey weighed, on average, 16 (n = 1; TX-52) and 31.4 ± 8.5 mg (range = 8–45, n = 17; TX-71). Two wasps weighed 15 mg (TX-52) and 19 mg (TX-71). Four other grasshoppers (probably Melanoplus sp.) collected from provisioning females measured 5.0–8.5 mm (mean = 6.6, n = 4; TX-52) in body length. The mean aggregate weight of prey per cell was 106.8 ± 19.8 mg (range = 88–140, n = 5; TX-71). Nineteen of 21 (90.5%) prey in the cells were positioned head inward and ventral side upward or on their side. One grasshopper was placed head inward and dorsal side upward (TX-71), and one was atypically positioned head outward and ventral side upward (TX-52). Egg. The wasp’s egg was as described for T. a. apicalis in shape, size, and coloration. It was affixed by the less-tapered, distal end to the soft intersegmental membrane surrounding the base of the prey’s forecoxa. The proximal end extended free to the opposite side, the egg lying transversely between the bases of the fore- and midcoxae. Four of five eggs (TX-71) were attached 2009 F.E. Kurczewski 81 to the base of the right forecoxal intersegmental membrane of the prey. The grasshopper bearing the wasp’s egg was placed head inward and ventral side upward in all five cells (TX-71). It averaged 33.4 ± 6.1 mg (range = 24–41, n = 5; TX-71) in weight, or slightly heavier than the mean weight of other prey in the cells (31.4 mg). The egg was laid on the heaviest (2 examples), second heaviest (1), third heaviest (1), or lightest grasshopper (1) in the cell. The egg-bearing prey was the first or second individual taken into the nest. Discussion The nesting behavior of seven species in the Tachysphex terminatus group is documented in this study. Six of the species nest in level or slightly sloping friable soil (Table 13). Tachysphex apicalis nests mainly in sand cliffs and strongly sloped sandbanks (Kurczewski and Snyder 1968). Tachysphex antillarum, T. terminatus, T. similis, and T. apicalis often form dense aggregations. Females of these species, T. clarconis, and T. alpestris excavate multi-celled nests, with the result that emerging adults are immediately in close contact. The males resemble the females in size and coloration. Sexual differences or recognition characters are not exaggerated as in the more solitary species (Kurczewski 1966b). Similarity in appearance of the sexes may be subtly connected with the gregarious nature of the nesting aggregation. In dense monospecific aggregations, males and females probably spend less time locating one another. The external morphology and nesting behavior of T. apicalis reflect its atypical habitat. Tachysphex apicalis has a shiny frons, slightly less Table 13. Important ecological and behavioral characteristics of the Tachysphex terminatus species group. Abbreviations: Levels sand surface = levels sand surface during burrow excavation; Temp clos = temporary closure of nest entrance; + = present; 0 = absent; - = no record; No prey/cell = number of prey per fully provisioned cell; No. cells/nest = number of cells per completed nest. Levels sand Temp Transport Prey No prey/ No. cells/ Stage, Species Habitat surface clos method size cell nest type prey clarconis Level sand + + Flight Small 3–6 2–3 Nymphal Acrididae antillarum Level sand + + Flight Small 6–11 1–5 Nymphal Acrididae alpestris Level sand - + Flight Small 2–13 1–3 Nymphal Acrididae linsleyi Level sand - + Flight Small 7 1 Nymphal Acrididae terminatus Level sand, + + Flight Small 1–14 1–5 Nymphal sandbank Acrididae, Tettigoniidae similis Level sand + + Flight Small 1–15 1–5 Nymphal Acrididae apicalis Sand cliff, 0 0 Flight Small 2–12 1–3 Adult, sandbank nymphal Acrididae 82 Northeastern Naturalist Vol. 16, Monograph 3 setaceous integument, stout mandibles, narrow clypeus with medial projection, and more rectangular frons than other species in the group— all possible adaptations related to cliff nesting and absence of temporary closure. As many as three males in single file and two males and a female of T. apicalis in tandem may occupy the same open burrow at night and during inclement weather, a rarity in this genus. Males of this species lack long foretarsal digging spines and do not excavate their own burrows. Males of T. terminatus and T. similis have well-developed tarsal rakes on their forelegs with which they dig short resting and overnight burrows in the soil (Kurczewski 1966b). In initiating a burrow, T. apicalis uses its mandibles considerably in order to penetrate the hard-packed sandy cliff face. The stout mandibles and medial clypeal lip projection of this species may be adaptations to facilitate such behavior. With respect to the velocity of digging, T. apicalis uses the forelegs more slowly than T. terminatus, and T. terminatus uses them more slowly than T. similis. This difference may reflect the different sizes of the species, T. apicalis being the largest, T. similis the smallest, and T. terminatus intermediate in size. There are no noticeable differences between T. terminatus and T. similis in the number of entries to remove sand from the burrow and duration of burrow excavation. The practice of partly filling the entrance of an abortive burrow before abandoning the excavation is well developed in T. terminatus and T. similis, species that make a temporary closure of the nest entrance. Species that inhabit level sand such as T. clarconis, T. antillarum, T. alpestris, T. linsleyi, T. terminatus, and T. similis temporarily close the entrance with loose sand upon leaving the nest. All species except T. alpestris and T. linsleyi have been reported to elaborately level the sand surface near the entrance prior to making the temporary closure (T. alpestris and T. linsleyi probably have not been observed enough to establish this behavior as part of their repertoire). The complexity and duration of the leveling process increase as burrow excavation proceeds toward completion. Tachysphex apicalis is limited by its habitat since most sand from burrow excavation falls down the cliff face or steep sandbank and out of reach. This species leaves the nest entrance open when it goes in search of prey. Tachysphex terminatus spends about twice as much time as T. similis for removal of the temporary closure and temporary closure of the entrance at the same sand surface temperatures regardless of geographic location (P < 0.0001). The more rapid entry and exit of T. similis may be related to its smaller size, faster movements, and looser temporary closure. Both species display these behaviors increasingly more rapidly in a straight-line relationship as sand surface temperature rises from 35 to 60 ºC. During removal of the closure and closure, T. terminatus makes hovering flights into cooler strata of air to avoid the hot sand when surface temperature reaches 55 ºC. Such flights increase in number as the sand surface temperature reaches 59– 61 ºC and then decrease in number, concurrently giving way to an aborted 2009 F.E. Kurczewski 83 temporary closure. Tachysphex similis does not typically exhibit hovering flights at high surface temperatures, but instead, makes an increasingly looser temporary closure and leaves the entrance progressively more open above 57 ºC. The end result for both species at sand surface temperatures exceeding 60°C is a partly open entrance to allow for rapid entry with prey and avoidance of the hot sand. Species in the terminatus group capture mostly early nymphal stages of grasshoppers that frequent the leaves and stems of vegetation. The wasps fly to a nearby source of vegetation to hunt for prey after making a characteristic orientation flight. They are able to hunt for prey at distances as far as 30–60 m from their entrances because they rapidly transport relatively small grasshoppers to the nests in flight. Relatively small prey and rapid flight transport allow the various species to capture and stock several to many individuals per cell within a rather brief period and complete up to 3 to 5 fully provisioned cells per day. Species in the terminatus group often transport prey directly to the nest in a prolonged flight. The species possess longer wings with greater total surface area than some other species of Tachysphex that practice ground transport of larger grasshoppers. An increase in the size of the hindwing (jugal lobe) in some species groups of Tachysphex apparently increases their flight velocity, enabling them to fly more rapidly than other groups with smaller hindwings (Pulawski 2007). The relatively larger, more robust thorax of species in the terminatus group may house a more extensive flight musculature to facilitate flight transport. The relatively shorter legs of these species probably represent an adaptation for holding smaller prey. Evans (1962) believed that “shorter legs may be better adapted for holding the prey tightly beneath the body in flight.” Species in the terminatus group that nest in level soil and make a temporary closure release the grasshopper ventral side upward or on its side with the head near the sand-filled entrance. Such a placement facilitates the wasp pulling the prey into the nest by an antenna. The grasshopper is nearly always positioned in front of the entrance in a straight line or at a slight angle with the direction of the burrow. Tachysphex apicalis, on the other hand, is confined to carrying its prey directly into the open entrance because of the limitations of the sand cliff or steep sandbank. Smaller prey are carried directly down the burrow, while larger individuals are released just inside the entrance and pulled in from inside after the wasp turns around. Species that nest in level soil rake the sand backward with the tarsal comb of the forelegs when filling the burrow and closing the fully provisioned nest. The apical segments of the foretarsi are bent medially to expose more spines for catching the sand grains and flinging them backward. The soil flung backward is then tamped in place at the end of the burrow with the shiny, trowel-shaped pygidium and sub-terminal abdominal segments. Tachysphex similis females make fewer trips onto the sand surface to retrieve soil for the fill despite the burrows of this species and T. terminatus being equivalent in length at common localities. 84 Northeastern Naturalist Vol. 16, Monograph 3 When nesting in cliffs and steep slopes, T. apicalis, T. terminatus, and, rarely, T. similis close the nest by pulling down sand from around the entrance with the mandibles, as they are limited from obtaining loose sand by the vertical or strongly sloping sites. During the final stages of this type of closure, females stand upright on hindlegs inside the enlarged entrance, grasp the sides of the cavity with their midlegs, and loosen sand with their mandibles. Such completed nests are never filled flush, but remain readily visible as a concave depression on the rather smooth surface. The nests of species in the terminatus group enter the level sand obliquely and are relatively short and shallow. The nests of T. apicalis are slightly sloped in sand cliffs and steep sandbanks and are also relatively short and shallow. Under optimal conditions, the completed nests of all species in the group are two- to five-celled. Multi-celled nests are usually associated with dense aggregations and nesting in close proximity. Conserving space by making a single multi-celled nest rather than several uni-celled nests is advantageous where the amount of suitable soil is limited. A decrease in T. terminatus and T. similis mean burrow length and mean cell depth from spring to late summer at many locations is inexplicable. The two or more generations of wasps often nest in approximately the same soils, thus eliminating a difference in soil type or texture as an explanation. Longer and deeper nests in the first generations of the two species coincide with higher soil moisture content. However, the reverse is true in other fossorial wasp genera where deeper nests are excavated in drier soil (Alcock and Ryan 1973, Hager and Kurczewski 1986, Kurczewski 2002, Kurczewski and Wochadlo 1998). Generational differences in mean burrow length and mean cell depth in T. terminatus and T. similis may be connected with seasonal day length and resulting sand temperatures or, in the Northeast, the amount of snow cover. In New York, females of T. terminatus are slightly larger in the second generation, having been reared from more prey biomass in the first generation cells (Elliott and Kurczewski 1975; F.E. Kurczewski, pers. observ.). However, there is no correlation between wasp size (weight) and cell depth in T. terminatus. In T. similis, on the other hand, larger Florida females consistently dig deeper nests than smaller Kansas females. There is no difference in mean cell depth between the equivalent-sized all-black and red-tipped females of T. similis in Kansas. Species in the terminatus group stock their cells with small, mainly nymphal grasshoppers (Acrididae). A few females of T. terminatus stocked several cells with a single small nymphal bush katydid (Tettigoniidae) in addition to small nymphal grasshoppers during spring at a single locality (Kurczewski 1966a). Tachysphex apicalis, the largest species in the group, rarely stores adult male grasshoppers of smaller acridid species. Very large sample sizes for T. terminatus and T. similis reveal equivalent percentages of Gomphocerinae, Oedipodinae, and Cyrtacanthacridinae + Melanoplinae as prey. Tachysphex apicalis stocks its cells with these same subfamilies of 2009 F.E. Kurczewski 85 prey, but in different proportion. All three subfamilies of Acrididae and up to three or four genera and species of prey may be put in a single cell of these three wasp species. Females are evidently unselective of the species of prey and capture whatever nymphal grasshoppers of favorable size are available and abundant near the nesting site (A.B. Gurney, 1963 pers. comm.). Tachysphex terminatus and T. similis, except in Florida, stock different species of prey in spring and mid- to late summer nests. There is a gradual decrease in the number of prey species through the nesting season. Grasshoppers from first-generation (May–June) nests include species of suitable size that are available only at that time of year. By mid-summer, most of these species are in the adult stage, too large for use, and the wasps switch to other, albeit fewer, suitable-sized species of prey. Most second-generation (mid-July–August) nests contain prey that were in the egg stage and unavailable for use earlier in the year. Prey selection is thus connected with the seasonal availability of acridid nymphs of appropriate stage and size. It is difficult to discern prey differences for the different generations of T. similis and T. apicalis in Florida because both wasps and prey grasshoppers are available almost continually throughout much of the year. Species in the terminatus group usually stock several or many small prey in a fully provisioned cell. Cells with only one or a few prey are often closed prematurely with the onset of inclement weather or dusk. One or two prey per fully provisioned cell is also associated with scarcity of prey due to drought or other unpredictable environmental events. There is a trend in T. similis and T. terminatus to stock more, smaller grasshoppers per cell early in the nesting season and fewer, larger grasshoppers per cell later in the nesting season. The number of prey per cell is approximately halved from first to later generations in T. terminatus. Not only are fewer grasshoppers stocked in the cells later in the year, but there is also a reduction in the total biomass of prey per cell, resulting in slightly smaller first-generation wasps the following year. Most of the grasshoppers captured by species in the terminatus group are put in the cells head inward and ventral side upward. Cells with the largest number of prey usually have more grasshoppers placed in other positions. In such cases, the wasp, in maneuvering to oviposit after bringing in her last prey for the cell, probably accidentally moves some of the grasshoppers from their original head inward and ventral side upward positions to the atypical positions in which some of the prey are found. The egg is usually laid on a grasshopper at the bottom or one placed farthest back in the cell, but only after the full complement of prey has been put in the cell. An egg is not affixed to a grasshopper at the front end as this prey might be readily discovered and attacked by enemies entering the cell. The prey to which the wasp’s egg adheres is nearly always placed head inward and ventral side upward or tilted slightly on its side. This grasshopper is often one of the largest and heaviest prey in the cell as it is advantageous for the young larva to begin feeding on a sufficient amount of food. In some 86 Northeastern Naturalist Vol. 16, Monograph 3 multi-cellular nests, a female places her eggs on one side or the other of the grasshoppers in most or all cells. Other wasps may alternate between left and right-sided affixation sites. Acknowledgments James Carpenter, Keith Goldfarb, Michael Ohl, and Wojciech Pulawski reviewed the manuscript. Wojciech Pulawski kindly agreed to serve as Guest Editor. Richard Archbold, Leonard Brass, Thomas Eisner, Bill Huff, John Lerg, Rob Line, Charles Michener, Verne Pechuman, and Mary Rabe aided with equipment, funding, or facilities. Noel Snyder, Edmund Kurczewski, George Matuza, Ann Peck, Nancy Burdick, and Margery Spofford assisted with field studies. Noel Snyder added numerous research tips and took the many photographs. Howard Evans initially proposed a comparative nesting behavior study of Tachysphex terminatus and T. tarsatus (Say) for my M.S. thesis and permitted me to use his field notes. Matthias Buck sent me his prey records and observations on members of this group. Richard Bohart, Karl Krombein, and Wojciech Pulawski confirmed my identifications of the wasp species. Ashley Gurney and Dave Nickle determined the prey Acrididae and Tettigoniidae. Dan Dindal and Charles Schirmer analyzed the soil samples. Diane Kiernan interpreted and analyzed innumerable statistical data. She introduced alternative ways to present some of the quantitative information. Don Artz formatted Figures 1, 16, and 17. Joe Stoll retouched and image-scanned the photographs. Funds for this study were provided by Sigma Xi-RESA Grants-in-Aid of Research, Cornell University Faculty Research Fund, National Science Foundation Postdoctoral Fellowship and Undergraduate Research Participation Programs, National Institutes of Health Postdoctoral Fellowship, State University of New York-RESA Grants-in-Aid of Research, Michigan Chaper of The Nature Conservancy, and New York State United University Professions. Literature Cited Alcock, J., and A.F. Ryan. 1973. The behavior of Microbembex nigrifrons (Hymenoptera: Sphecidae). Pan-Pacific Entomologist 49:144–148. Ashmead, W.H. 1894. The habits of the aculeate Hymenoptera. Psyche 7:19–26, 39–46, 59–66, 75–79. Bohart, R.M. 1962. New species of black Tachysphex from North America (Hymenoptera: Sphecidae). Proceedings of the Biological Society of Washington 75:33–40. Bohart, R.M., and A.S. Menke. 1976. Sphecid Wasps of the World. A Generic Revision. University of California Press, Berkeley, CA. 695 pp. Buck, M. 2004. An annotated checklist of the spheciform wasps of Ontario (Hymenoptera: Ampulicidae, Sphecidae and Crabronidae). Journal of the Entomological Society of Ontario 134:19–84. Elliott, N.B. 1996. The biology of a widely distributed North American wasp (Tachysphex similis Rohwer) in the Bahamas. Bahamas Journal of Science 4:24–28. Elliott, N.B., and F.E. Kurczewski. 1975. Seasonal variation in Tachysphex terminatus (Smith) (Hymenoptera: Sphecidae, Larrinae). Journal of the New York Entomological Society 82:268–270. Elliott, N.B., and F.E. Kurczewski. 1985. Nesting and predatory behavior of some Tachysphex from the western United States (Hymenoptera: Sphecidae). Great Basin Naturalist 45:293–298. 2009 F.E. Kurczewski 87 Evans, H.E. 1962. The evolution of prey-carrying mechanisms in wasps. Evolution 16:468–483. Evans, H.E. 1970. Ecological-behavioral studies on the wasps of Jackson Hole, Wyoming. Bulletin of the Museum of Comparative Zoology 140:451–511. Finnamore, A.T. 1982. The Sphecoidea of southern Quebec (Hymenoptera). Lyman Entomological Museum Research Laboratory Memoir 11:1–348. Fox, W.J. 1894. The North American Larridae. Proceedings of the Academy of Natural Sciences of Philadephia 45:467–551. Genaro, J.A., and C.S. Sanchez. 1992. Observations on the nesting behavior of Tachysphex antillarum (Hymenoptera: Sphecidae). Caribbean Journal of Science 28:218–220. Hager, B.J., and F.E. Kurczewski. 1986. Nesting behavior of Ammophila harti (Fernald) (Hymenoptera: Sphecidae). American Midland Naturalist 116:7–24. Krombein, K.V. 1964. Results of the Archbold Expeditions. No. 87. Biological notes on some Floridian wasps (Hymenoptera, Aculeata). American Museum Novitates 2201:1–27. Krombein, K.V. 1967. Superfamily Sphecoidea. Pp. 386–421, In K.V. Krombein and B.D. Burks (Eds.). Hymenoptera of America North of Mexico. Synoptic Catalogue (Agriculture Monograph No. 2). Second Supplement. United States Department of Agriculture, Washington, DC. Krombein, K.V. 1979. Superfamily Sphecoidea. Pp. 1573–1740, In K.V. Krombein, P.D. Hurd, Jr., D.R. Smith, and B.D. Burks (Eds.). Catalog of Hymenoptera in America North of Mexico. Volume 2. Smithsonian Institution Press, Washington, DC. Krombein, K.V., and H.E. Evans. 1955. An annotated list of wasps collected in Florida, March 20 to April 3, 1954 (Hymenoptera, Aculeata). Proceedings of the Entomological Society of Washington 57:223–235. Kurczewski, F.E. 1966a. Tachysphex terminatus preying on Tettigoniidae—an unusual record (Hymenoptera: Sphecidae, Larrinae). Journal of the Kansas Entomological Society 39:317–322. Kurczewski, F.E. 1966b. Comparative behavior of male digger wasps of the genus Tachysphex (Hymenoptera: Sphecidae, Larrinae). Journal of the Kansas Entomological Society 39:436–453. Kurczewski, F.E. 1967. Hedychridium fletcheri (Hymenoptera: Chrysididae, Elampinae), a probable parasite of Tachysphex similis (Hymenoptera: Sphecidae, Larrinae). Journal of the Kansas Entomological Society 40:278–284. Kurczewski, F.E. 2000. Distribution, behavior, and ecology of acridid-hunting Tachysphex from the Carolinas (Hymenoptera: Sphecidae). Journal of the Elisha Mitchell Scientific Society 116:113–123. Kurczewski, F.E. 2002. Seasonal variation in nesting behavior of Oxybelus bipunctatus (Hymenoptera: Sphecidae). Northeastern Naturalist 9:419–432. Kurczewski, F.E., and R.E. Acciavatti. 1990. Late summer–fall solitary wasp fauna of central New York (Hymenoptera: Tiphiidae, Pompilidae, Sphecidae). Great Lakes Entomologist 23:57–64. Kurczewski, F.E., and H.E. Evans. 1986. Correct names for species of Tachysphex observed by Evans (1970) at Jackson Hole, Wyoming, with new information on T. alpestris and T. semirufus (Hymenoptera: Sphecidae). Proceedings of the Entomological Society of Washington 88:720–721. Kurczewski, F.E., and B.J. Harris. 1968. The relative abundance of two digger wasps, Oxybelus bipunctatus and Tachysphex terminatus, and their associates, in a sand pit in central New York. Journal of the New York Entomological Society 76:81–83. 88 Northeastern Naturalist Vol. 16, Monograph 3 Kurczewski, F.E., and N.F.R. Snyder. 1968. Evolution of cliff-nesting in digger wasps. The Conservationist 23:28–31. Kurczewski, F.E., and D.L. Wochadlo. 1998. Relationship of cell depth and soil moisture in Oxybelus bipunctatus (Hymenoptera: Sphecidae). Entomological News 109:7–14. Miller, R.C., and F.E. Kurczewski. 1973. Ecology of digger wasps (Sphecidae). Pp. 204–217, In D.L..Dindal (Ed.). Proceedings of the First Soil Microcommunities Conference. US Atomic Energy Commission, Office of Information Services, CONF-711076, Syracuse, NY, October 18–20, 1971. National Technical Information Service, US Department of Commerce, Springfield, VA. Pulawski, W.J. 1988. Revision of North American Tachysphex wasps including Central American and Caribbean species (Hymenoptera: Sphecidae). Memoirs of the California Academy of Sciences 10:1–211. Pulawski, W.J. 2007. The wasp genus Tachysphex Kohl, 1883, of Sahara, Sub- Saharan Africa, the Arabian Peninsula, and Madagascar (Hymenoptera: Apoidea: Crabronidae). Proceedings of the California Academy of Sciences Fourth Series 58:1–698. Pulawski, W.J. 2008. Catalog of Sphecidae sensu lato (= Apoidea excluding Apidae). Available online at http://www.calacademy.org/research/entomology/ Entomology”Resources/Hymenoptera/sphecidae/Genera”and”species”PDF/introduction. htm. Accessed April 10, 2008. Rau, P. 1926. The ecology of a sheltered clay bank: A study in insect sociology. Transactions of the Academy of Science of St. Louis 25:157–260. Rau, P. 1946. Notes on the behavior of a few solitary wasps. Bulletin of the Brooklyn Entomological Society 41:10–11. Rau, P., and N. Rau. 1918. Wasp Studies Afield. Princeton University Press, Princeton, NJ. 372 pp. Spofford, M.G., and F.E. Kurczewski. 1990. Comparative larvipositional behaviours and cleptoparasitic frequencies of Nearctic species of Miltogrammini (Diptera: Sarcophagidae). Journal of Natural History 24:731–755. Spofford, M.G., and F.E. Kurczewski. 1992. Counter-cleptoparasitic behaviours of species of Sphecidae (Hymenoptera) in response to Miltogrammini larviposition (Diptera: Sarcophagidae). Journal of Natural History 26:993–1012. Spofford, M.G., F.E. Kurczewski, and D.J. Peckham. 1986. Cleptoparasitism of Tachysphex terminatus (Hymenoptera: Sphecidae) by three species of Miltogrammini (Diptera: Sarcophagidae). Annals of the Entomological Society of America 79:350–358. Strandtmann, R.W. 1953. Notes on the nesting habits of some digger wasps. Journal of the Kansas Entomological Society 26:45–52. Williams, F.X. 1914. The Larridae of Kansas. Kansas University Science Bulletin 8:121–213. Williams, F.X. 1932. Notes and exhibitions. Tachysphex fusus Fox. Hawaiian Entomological Society Proceedings 8:17.