2008 NORTHEASTERN NATURALIST 15(Monograph 2):67–76
III. Ecological Observations on Tachysphex pechumani
Erin K. Moan1 and Elliot J. Tramer2,*
Abstract - This study addressed aspects of the ecology of Tachysphex pechumani
(antenna-waving wasp), a rare solitary wasp. Wasps nested in compacted sand, primarily
where vegetative cover was 5–30% and vegetation height was 0.5 to 33 cm.
Miltogrammine flies and predators presented challenges to wasp success at the study
sites. The timing and duration of T. pechumani’s breeding season conformed to an
apparent abundance peak of acridid grasshopper nymphs used by wasps, based on
sweep-net data. Heat increased wasp activity to an upper threshold of 56 °C at the
sand surface, at which temperature wasps ceased working. Weak flying abilities of
both sexes suggested that populations were effectively isolated and vulnerable to
habitat loss or fragmentation. Management for wasps should include maintenance of
open sites and prohibition of sand disturbance.
Tachysphex pechumani Krombein (antenna-waving wasp) is a rare insect
that inhabits partially exposed sandy soils characteristic of oak savanna
and pine barrens. It is distinguished from other Tachysphex wasps by its
brassy golden face and orange apical antennal segments, which are incessantly
waved up and down. The 8.3–12.1 mm female wasp is predominantly
black with two or three red basal abdominal segments; male wasps may be
6.0–9.7 mm and are entirely black (Kurczewski et al. 1970). Adults appear
for approximately five or six weeks surrounding the summer solstice. Males
are observed for only about a week, hovering near and mating with females
as they emerge. Females live three to four weeks; after mating, they hunt
and provision burrows. A female wasp paralyzes an immature grasshopper
(primarily of the genus Melanoplus), drags it back to a previously dug burrow,
lays a single egg on it, and fills in the burrow (Kurczewski and Elliot
1978). Because this animal requires expanses of barren sandy soil free of
overhanging vegetation, T. pechumani may be a good “indicator species” for
oak savanna ecosystems, which are rapidly disappearing due to land development
and fire suppression. As its required habitat disappears, the wasp’s
existence may become precarious. Although the antenna-waving wasp’s life
history and geographic distribution have been previously studied, there is
much to be discovered regarding this rare animal’s ecology.
Conservation biologists have specified several characteristics of a species
at high risk of extinction (Meffe et al. 1997), and these characteristics
12554 Township Road 444, Sugarcreek, OH 44681. 2University of Toledo, Department
of Environmental Sciences, MS 604, Toledo, OH 43606. *Corresponding author
68 Northeastern Naturalist Vol. 15, Monograph 2
fit the antenna-waving wasp. One characteristic is limited geographic range
or highly localized populations. Tachysphex pechumani is limited to areas of
southern Ontario, the lower peninsula of Michigan, the Oak Openings region
of northwest Ohio, New Jersey Pinelands National Reserve, and northwest
Indiana (Kurczewski 1998, 2000). Another characteristic of species at risk
is small population size. Small populations tend to lose genetic diversity
over time, leading to increased probability of extinction. Tachysphex pechumani
population sizes are generally unknown. However, at least in the Oak
Openings region, patches of sand suitable for burrowing by T. pechumani
are small and widely scattered. Furthermore, observations of nesting wasps
(Kurczewski 1998, Kurczewski and Elliot 1978) suggest that T. pechumani
may be a weak flyer. If so, poor dispersal ability would limit movement between
sites and restrict the wasps’ ability to find other suitable nesting sites
if their home site is destroyed or overgrown. In the Oak Openings region,
T. pechumani may be especially at risk due to human activities. Conversion
of Oak Openings habitats to residential and commercial development is
proceeding rapidly, and sandy openings in protected areas have experienced
closure by invading woody vegetation. Tachysphex pechumani has not been
reported in Ohio outside the Oak Openings and is considered “endangered”
in the state (Kurczewski 1998). Our study is aimed at improving understanding
of its ecology and suggesting proper habitat management for its benefit.
One goal of our study was to characterize T. pechumani burrow sites
in terms of vegetation structure and prey density. We also questioned why
wasps restrict their mating and nesting to such a short time period (early
June to mid-July). Hypothesizing that wasp activity might be limited by temperature,
we investigated the effect of heat on wasp activity. We also asked
whether the timing of adult wasp activity corresponded with seasonal prey
density, an idea suggested previously by Kurczewski (1998).
An important question is whether T. pechumani subpopulations are part
of a larger, interbreeding metapopulation, whose members disperse readily
and create gene flow. If so, small subpopulation sizes would be less of a
concern. Since oak savanna is being rapidly lost to land development and
fire suppression, it is also valuable to know whether wasps could disperse to
new habitat if an occupied site is destroyed or overgrown.
Finally, it was important to identify what management techniques might
provide and maintain suitable T. pechumani habitat. Since the wasp is
designated as endangered in Ohio, it is important that individuals and organizations
know how best to protect remaining T. pechumani populations.
Materials and Methods
Our study location was Oak Openings Preserve Metropark in Lucas
County, OH. From the park’s 12 known T. pechumani sites, we chose three
that appeared to contain the largest numbers of wasps, estimated from casual
observation in 2001. One site (“Monclova;” 41º33'27"N, 83º51'35"W) was a
utility road turnaround near an old sand quarry. Wasps nested in compacted
2008 E.K. Moan and E.J. Tramer 69
sand directly in the track. The second and third sites (“Tansel Lane;”
41º34'12"N, 83º51'35"W) were in a sandy opening adjacent to a Prunus
serotina Ehrh. (black cherry)/ Quercus velutina Lam. (black oak) woodland.
The opening contained two nesting sites, “Tansel East” and “Tansel West,”
separated by about 20 m of low, shrubby vegetation. The distance between
Monclova and the two Tansel sites was approximately 2.1 km.
We began recording temperature on June 7, 2002, utilizing a CR10X
datalogger (Campbell Scientific, Logan, UT) with a thermister placed at
a depth of 30 to 40 mm, which is Tachysphex pechumani’s burrow depth
(Kurczewski 2000). The datalogger recorded the temperature every 30 minutes
for the entire 2002 field season. In 2003, mercury laboratory thermometers
were placed in direct sun on the sand surface whenever researchers
were at wasp sites. This method may have given a temperature that differed
somewhat from actual “sand-surface temperature,” normally measured with
a shaded sensor, as the sun may have heated the mercury in the thermometer
more or less than it was heating the sand—thus the actual measurement taken
represents the net effect of heat flow by conduction between the sand and
the thermometer bulb and the heating of the thermometer bulb by solar radiation.
We adopted this method to attempt to approximate the temperatures
the wasps were actually experiencing, since their bodies were not shielded
from direct sunlight. For convenience, here we refer to these temperatures
as “sand-surface temperatures.” Every 15 minutes, temperatures were recorded
and notes were taken on wasp activity or absence. Temperature from
lab thermometers was also noted when a wasp appeared dragging prey, and
again when she stuffed the grasshopper into her burrow.
To characterize vegetation at T. pechumani burrow sites, vegetation
structure was surveyed over a time span of three days at the conclusion of the
first season (July 3–5, 2002). We used a meter square, centered on the burrow
entrance, to estimate percent vegetative cover. We also measured the tallest
individual of each species present. Twenty-three squares were sampled at
Monclova, 18 at Tansel west, and 6 at Tansel east.
Prey density and size were investigated May 30 through July 19 of 2002
via sweep netting. A 30-cm diameter insect net was swung along the same
transect through three areas used by the wasps for hunting. The vegetation
along the sweep routes was low (maximum height 30 cm), within the range
of heights used by wasps foraging for prey (Kurczewski 2008). Each site
was sampled six times (approximately every 14 days). Grasshoppers caught
were measured and the species recorded if known. Only acridid species
used by the wasps were tallied; “long-horned” grasshoppers and individuals
of the genera Spharagemon and Dissosteira were released. Kurczewski
(1998) speculated that the wasp’s lifespan corresponds with peak abundance
of potential prey grasshoppers. Sweep-net data were used to determine the
seasonal pattern of abundance of T. pechumani potential prey.
We compared grasshopper sweep-net data with grasshoppers caught and
then abandoned by antenna-waving wasps. Wasps frequently abandoned
70 Northeastern Naturalist Vol. 15, Monograph 2
paralyzed prey for a variety of reasons (see results). Thus, we were able to
collect grasshoppers taken by wasps without removing them from the wasps’
burrows, a procedure we wished to avoid because of T. pechumani ’s endangered
status. We assumed that abandoned prey represented a random sample
of grasshoppers being caught by the wasps, thus reflecting the size range of
captured prey. Based on our observations, we are confident that this assumption
To determine effect of temperature on wasp behavior, the number of
completed burrows each day was compared with the daily maximum temperature
(typically occurring between 1500 and 1700 hours) and also with
the temperature at 1100 hours, which appeared to be the peak of daily wasp
activity. Temperatures used were those recorded by the CR10X datalogger.
Wasp marking was performed in both field seasons. Females were netted,
placed in vials on ice within a cooler to render them temporarily inactive,
then marked between the bases of the forewings with nontoxic, permanent
Faskolor paint (Parma International, North Royalton, OH). Wasps were
returned to the cooler to allow the paint to dry, then released after a brief
warm-up period. Care was taken to avoid applying paint on wing bases,
where it might interfere with flight, or on the abdomen, where it could interfere
with breathing through the spiracles. Marked wasps behaved normally
and similarly to unmarked wasps. In 2002, wasps were marked as early in
the season as possible using a different color for each subpopulation to determine
if dispersal occurred among females of different subpopulations.
All statistical analyses were performed using SAS Version 8 (SAS Institute
Results and Discussion
At our study sites, wasps nested in flat areas where sand was moderately
compacted. Shifting dunes, slopes, and loose sand were not utilized by
antenna-waving wasps for burrow sites. Percent vegetative ground cover
for our sites ranged from 2% to 66%, with a mean of 19.8% and a median
of 16%. Three-fourths (34 out of 45) of the data points fell between 5% and
30% cover (Fig. 1A). The height of surrounding vegetation ranged from 0.5
cm to 68 cm, with a mean of 23.6 cm, a standard deviation of 15.4 cm, and a
median of 23.5 cm. Eighty percent (36 out of 45) of the values were between
0.5 cm and 33.0 cm (Fig. 1B). Plants occurring most frequently at study sites
were Rumex acetosella Linnaeus (sheep sorrel), Krigia virginica Linnaeus
(dwarf dandelion), Rubus flagellaris Willd. (dewberry), and many grasses,
including a number of Panicum species.
Adult wasps were first observed June 12 and June 14 in 2002 and 2003,
respectively. Males were observed for several days hovering near burrow
areas and mating with females. Ten to twelve days later, five to seven
females were often observed simultaneously at our study sites, and many
other females were hidden by surrounding vegetation as they hunted or returned
with prey. Most females were gone by the July 4 of both years, but
2008 E.K. Moan and E.J. Tramer 71
occasionally a single wasp was observable at the sites throughout the first
two weeks of July. After mid-July, no T. pechumani were seen.
Antenna-waving wasps contend with many obstacles to successful reproduction.
Antenna-waving wasps are subject to predation by Miltogrammine
flies known as satellite flies. The flies wait at a burrow site until a wasp
dragging prey appears, then attempt to dart in and lay their own larva on the
grasshopper while the wasp is reopening its burrow. The fly larva searches
Figure 1. A. Percent vegetative cover at Tachysphex pechumani (antenna-waving
wasp) burrow sites in 2002. B. Maximum vegetation height (cm) at T. pechumani
burrow sites in 2002.
72 Northeastern Naturalist Vol. 15, Monograph 2
for the wasp egg and destroys it before consuming the grasshopper (Spofford
et al 1986). At least two species of satellite flies were observed at our
study sites, and several of their pupal cases were discovered in soil from
wasp areas. It is also possible that velvet ants (Mutillidae) are parasitoids of
T. pechumani pupae; although no T. pechumani researcher has reported this,
velvet ants are known to parasitize the pupae of other wasps, chewing a hole
in the pupa case and then depositing their own egg inside (Evans and Eberhard
1970). Velvet ants were often seen at the study sites. Predators may also
destroy wasp burrows—on one occasion a skunk excavated a sizeable hole
next to a burrow marker at a site, possibly consuming the buried grasshopper
with wasp egg attached. Shrews and moles may also eat buried grasshoppers.
Finally, adult wasps may be preyed upon by Cicindela repanda Dejean
(tiger beetles). On several occasions we witnessed tiger beetles lunging at
female antenna-waving wasps. Wasps typically avoided the attacks by flying
a short distance, in some cases causing the wasps to leave prey momentarily
or abandon it completely. We also noted one case of direct mortality when a
tiger beetle caught and ate a male antenna-waving wasp.
We compared grasshopper sweep-net data with grasshoppers caught and
then abandoned by antenna-waving wasps (Table 1). If a wasp could not find
her burrow, failed to successfully reopen it, or was attacked by a tiger beetle,
she abandoned her prey. The abandoned grasshoppers were 16 to 21 mm
(mean = 17.6, n = 19) in length. Few sweep-netted grasshoppers exceeded
16 mm in length until June 16–18. By then, the sweep-netted grasshoppers
averaged 18.7 mm long, with a range of 10 to 25 mm, which encompasses the
entire abandoned grasshopper range. The sweep-net data from June 26 and 27
are also within the size range of the grasshoppers taken by antenna-waving
wasps. However, by July 6, the mean grasshopper length had increased to 20.3
mm, almost exceeding the abandoned grasshopper size range. There was still
a range of 17- to 22-mm grasshoppers present, but note that abundance had
decreased—only 9 grasshoppers were caught on that date.
Another interesting observation regarding prey was that immature shorthorned
grasshoppers of the genera Dissosteira and Spharagemon were not
taken by T. pechumani even though they were of ideal length. Nymphs of
these two genera have a stiff mid-dorsal crest on the thorax. We hypothesize
that T. pechumani is unable to straddle and sting these nymphs because of the
Table 1. Abundance and size of grasshoppers sweep netted in 2002, compared with size of prey
abandoned by Tachysphex pechumani (antenna-waving wasp) in both 2002 and 2003.
Date N Range (mm) Mean size (mm)
May 30 29 5–17 10.7
June 7 18 6–19 12.9
June 16–18 25 10–25 18.7
June 26–27 17 11–25 17.2
July 6 9 17–22 20.3
July 19 10 15–28 20.5
Abandoned 19 16–21 17.6
2008 E.K. Moan and E.J. Tramer 73
crest. It is interesting to note that adult Spharagemon and Dissosteira, which
are far too large to be preyed on by the wasp, lack this prominent crest.
Wasp provisioning activity peaked between 1100 and 1200 hours
(Fig. 2A). There was a smaller peak of activity in the late afternoon around
Figure 2. A. Time of day of Tachysphex pechumani (antenna-waving wasp) burrow
completion in 2003. B. Surface temperature (°C) at time of T. pechumani burrow
completion in 2003.
74 Northeastern Naturalist Vol. 15, Monograph 2
1600 hours, with a break inbetween. This bimodal daily activity pattern confirms Kurczewski’s (1998) observations.
There was a significant positive correlation between the number of burrows
completed in a day and that day’s high temperature at burrow depth
(r = 0.61, p = 0.01, n = 16). Warmer weather appears to induce T. pechumani
activity, at least up to a threshold. Our observations showed that T. pechumani
began burrowing and hunting when surface temperatures warmed
to between 25 and 30 °C. Wasps provisioned burrows most actively at 37
to 46 °C (Fig. 2B), but burrow provisioning continued until an apparent upper
threshold of 56 °C was reached, usually in the early afternoon on sunny
days. In some cases during hot afternoons, a wasp carrying prey would not
leave the shade, but halted at the sun-shade interface. Wasps exhibiting this
behavior made periodic dashes onto the sunny sand, but always returned
quickly to rest again in the shade. One wasp carrying prey spent 64 minutes
in this type of behavior, wandering in a 1.5-meter circle but unwilling to go
into the sun, where the surface temperature was 57 °C at 1242 hours. She
eventually did run onto the sunny sand, quickly opened her burrow, and then
spent over 3 minutes (much longer than average) completing the burrow,
since she frequently took time to hover over the sand, apparently cooling
herself. Another wasp exhibited the same aversion to the sunny sand surface
(surface temperature 56 °C), and abandoned her grasshopper after only 8
minutes at the sun-shade interface. Kurczewski (2008) noted that antennawaving
wasp females leave nesting sites when surface temperatures reach
55 °C, an observation very similar to our own. Many female solitary wasps
curtail their activities at temperatures in this range (O’Neill 2001).
In 2002, we detected no movement of marked females between subpopulations.
Even at the two Tansel sites, which were separated by only 20 m of
easily navigable vegetation, no color-coded wasps were found at the neighboring
site. We observed that antenna-waving wasps make mostly low, short
flights for 5 meters or less before settling again on the sand to run or walk;
Kurczewski (1998) and Kurczewski and Elliot (1978) also describe these
flight patterns. In 2003, however, a wasp marked at the Tansel east site was
sighted at the Tansel west site, 37.4 m from where it was marked. It exhibited
typical hunting behavior, running over and under dried leaves amidst vegetation
in an approximately half-meter area (Kurczewski and Elliot 1978). In two
years of observation, this wasp was the only one seen at such a distance from
its marking point. Based on our large number of observations of T. pechumani
flight patterns, we suggest that their ability to disperse may be limited.
If an antenna-waving wasp does possess limited dispersal ability, the
conservation implications are serious. In the Oak Openings region, suitable
wasp habitat occurs in small, isolated patches separated from one another
by several km of unsuitable terrain. Additionally, in this area, habitat destruction
due to development is an imminent threat to this wasp. If a site is
destroyed during months when wasps are underground, there is no chance of
survival. If this occurs during wasp nesting season, there would have to be
appropriate habitat very close by for successful colonization to occur.
2008 E.K. Moan and E.J. Tramer 75
In addition, low gene flow may put small populations at risk from
inbreeding even in appropriate habitat. It is unknown whether male T. pechumani
disperse after emergence; however, their flight pattern appears as weak
as that of females. If males did disperse, inbreeding would not be a concern,
but dispersal of males alone would not save a population threatened by habitat
destruction. More research is needed regarding antenna-waving wasps’
mobility and dispersal abilities.
Canopy closure due to natural ecological succession is also a potential
threat to wasp habitat, and would likely cause gradual dwindling of the population
as less exposed sand was available for nesting. In the Oak Openings,
active management by cutting and controlled burning is employed to prevent
the incursion of woody vegetation into open sand barrens.
Our observations suggest several conclusions about antenna-waving
wasps’ habitat requirements. One necessity is full sun on the site for most of
the day. Wasps were not active on overcast days or when the sand was wet.
If a site had sun early in the morning, the sand dried quickly and wasps could
begin digging. Shade did not appear to be completely detrimental, however,
because in the hottest hours of the afternoon, wasps dragging prey used shaded
areas to rest and cool themselves. Another habitat requirement is freedom
from encroaching vegetation. During the wet, cooler summer of 2003, grass
coverage expanded at two of the study sites, and wasps created fewer burrows
in those areas than in the hotter, drier year of 2002 when grass growth
was at a minimum. Most likely there is a natural cycle of expansion and
reduction of wasp habitat as the weather cycles through wet and dry years.
Light burning for vegetation removal should not be harmful to T. pechumani
as long as it is not performed during nesting season; Kurczewski (1998)
maintains that the wasp is adapted to light wildfires due to a life history as
a savanna-dwelling animal. A third requirement for T. pechumani sites is
prohibition of vehicle traffic, which is essential during nesting season (early
June through mid-July). Light traffic during antenna-waving wasps underground
phase is not detrimental, but any traffic which would cause sand to
be churned up or displaced (such as all-terrain vehicles) should be prohibited
year-round. In sum, a proactive management plan would include keeping an
antenna-waving wasp site free from excessive overhanging vegetation that
would shade the site, maintaining relatively open sand via controlled burning
or lopping, and prohibiting vehicle traffic as mentioned above. These
recommendations are similar to current land-management practice for the
preservation of oak savanna, and will maintain suitable habitat for other
savanna-dwelling plants and animals as well.
Financial support for this project was provided by Toledo Naturalists’ Association.
We thank Frank Kurczewski for sharing his wealth of wasp knowledge and giving
helpful advice, and Kevin O'Neill for acting as manuscript editor. We also thank
John Jaeger and Metroparks of Toledo for not only permitting, but encouraging,
76 Northeastern Naturalist Vol. 15, Monograph 2
our research at Oak Openings Metropark. Naturalist Bob Jacksy of Metroparks
contributed invaluable field observations, advice, and spectacular photographs. Dr.
Deb Neher and Dr. Michelle Grigore gave guidance and assistance throughout the
research process. We also thank field assistants Heather Iler, Brian Putman, Melanie
McLaughlin, Mary Croy, and Ben Proshek for countless hours of wasp watching.
Evans, H.E., and M.J.W. Eberhard. 1970. The Wasps. The University of Michigan
Press, Ann Arbor, MI. 265 pp.
Kurczewski, F.E. 1998. Distribution, status, evaluation, and recommendations for
the protection of Tachysphex pechumani Krombein, the antennal-waving wasp.
Natural Areas Journal 18:242–254.
Kurczewski, F.E. 2000. First record of Tachysphex pechumani (Hymenoptera: Sphecidae)
from Indiana. Great Lakes Entomologist 33:41–43.
Kurczewski, F. 2008. II. Nesting behavior of Tachysphex pechumani (Hymenoptera:
Crabronidae). Northeastern Naturalist 15(Monograph 2):33-66.
Kurczewski, F.E., and N.B. Elliott. 1978. Nesting behavior and ecology of
Tachysphex pechumani Krombein (Hymenoptera: Sphecidae). Journal of the
Kansas Entomological Society 51:765–780.
Kurczewski, F.E., N.B. Elliott, and C.E. Vasey. 1970. A redescription of the female
and description of the male of Tachysphex pechumani (Hymenoptera: Sphecidae,
Larrinae). Annals of the Entomological Society of America 63(6):1594–1597.
Meffe, G.K., C.R. Carroll, and contributors. 1997. Principles of Conservation Biology.
2nd Edition. Sinauer Associates, Inc, Sunderland, MA. 729 pp.
O’Neill, K.M. 2001. Solitary Wasps: Behavior and Natural History. Cornell University
Press, Ithaca, NY. 406 pp.
SAS Institute. 2000. SAS Software Version 8. SAS Institute, Cary, NC.
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