Table Of Content

HYM. RES. J. Vol. 17(1), 2008, pp. 57-63 Load Carriage during Foraging in Two Species of Solitary Wasps Joseph R. Coelho, Jon M. Hastings, Charles W. Holliday, and Angela Mendell (JRC) Institute for Franciscan Environmental Studies, Biology Program, Quincy University, 1800 College Ave, Quincy, IL 62301, U.S.A. (JMH, AM) Department of Biological Sciences, Northern Kentucky University, Highland Heights, KY 41099, U.S.A. (CWH) Department of Biology, Lafayette College, Easton, PA 18042, U.S.A. — Abstract. Foraging strategies of two species of solitary digger wasps (Hymenoptera: Crabronidae) were examined. Both species capture insects and return with them in flight to a burrowwhere they areused as foodfor larvae. The actualloading ofthewaspswas compared to a theoretical ideal, the m.aximum that they should be able to carry in flight. Bembix troglodytes Handlirschis arelativelylargewasp thatcarriesmanysmallfliestoitsburrows. Theflies aremuch smaller than the ideal size, but the choice of small prey appears to be adaptive in that it should reduce the rate of stealing of flies by conspecific females, which was a common event. Tachytes chrysopyga Cresson is a relatively small wasp that carries relatively large prey, grasshoppers and crickets, to its burrow. The average size of prey appears to be ideal; however, the distribution of prey size is so greatthatmanywasps were underloaded, while others were overloaded. Prey theft wasnotobserved in T. chrysopyga, and flexible flightbehaviors (e.g. short, hoppingflights) allow it to carry a broad range of prey sizes. These two wasp species may represent near extremes of a continuum ofbehavior among predaceous wasps. Although much is now known about the maximum load mass, the ratio of flight foraging and nest provisioning behaviors musclemass tobodymass (orflightmuscle of solitary wasps (O'Neill 2001, Evans and ratio, FMR) is 0.179 (Marden 1987). O'Neill2007), theirforaging strategies have Use of flight muscle ratio as a metric for not been extensively studied using the flight capability has several advantages, conceptual framework offered by optimal primarily its independence of the size of foraging theory. One approach to examin- thewasp. Forwasps thatcarrytheirpreyin ing foraging decisions in such species is to flight, the maximum force produced must develop a prediction of the optimal load equal orexceed the combined weightofthe size of the prey, and to test this prediction wasp and prey. Assuming it is optimal for against the size of actual loads. Fortunate- a wasp to carry the largest prey possible, ly, the maximal load size is simple to the FMR (now flight muscle mass divided calculate for species that carry the prey in by combined body and prey mass) should flight. Marden (1987) measured the maxi- approach but not exceed 0.179. We have mum lift force of a variety of flying tested this prediction in yellowjackets animals, including 33 bees and wasps, by {Vespula spp, Coelho and Hoagland 1995), progressively loading individuals with the eastern cicada killer (Sphecius speciosiis weights until they could not take off. The Drury, Coelho 1997), the great golden force production of flying animals is digger wasp (Sphex ichneumoneus L., primarily dependent on their flight muscle Coelho and LaDage 1999), and the carpen- mass. In Hymenoptera, this relationship is ter wasp (Monobia quadridens L., Edgar and quite strong (r^ = 0.99), and at the Coelho 2000). In no case yet examined has 58 Journalof Hymenoptera Research the FMR of foraging wasps fallen precisely flower-visiting flies, though most are rela- at the predicted value; however, investiga- tively small in size. Exceptionally small tion of the causes for the deviation from flies are provided to young larvae, while "optimality" has always revealed interest- older larvae are provisioned with larger ing insights into foraging behavior. In this flies. Digging and provisioning a single study we apply this method to examine the nest requires only about six days and from foraging strategies of two rather different 21 to 26 flies are required to fullyprovision crabronid wasps facing different selection a larva (Evans 1957). pressures. Tachytes Bembix Tachytes is a genus of digger wasps that The general biology of Bembix and stocks its underground burrows with specific details of B. troglodytes Handlirsch, Orthoptera carried in flight from foraging the subject of this study, are described by areas. Tachytes digs complex burrows with Evans (1957; 1963). Onlyrelevantaspects of multiple cells and packs each cell with up their behavior are summarized here. Bem- to 10 prey items. The egg is not laid until bix is a genus of crabronid wasps that hunt the cell is fully provisioned (Evans and flies (Diptera) and carry them to their Kurczewski 1966, Elliotand Salbert 1981), a burrow in flight. The female digs a single case of mass provisioning. Tachytes nest burrow in the ground and excavates a nest aggregations are sometimes associated cell. In progressive provisioners, an egg is with those of Sphecius, as they were in this laid and attached to the firstprey in the cell study. In one species, T. distinctus F. Smith, or to the substrate in the center of the cell. males establish perches near the nest In B. troglodytes, oviposition occurs in an entrance of a female and chase any insect empty cell, and the first fly is captured, that flies near, including brood parasites paralyzed and placed in the cell later such as Zanysson texanus (Cresson) (Lin (Evans 1957, Evans and O'Neill 2007). The and Michener, 1972). female continues to capture flies and feed MATERD^LS AND METHODS them to its developing larva. As the larva increases in size, the rate of provisioning We conducted our observations and increases (Tengo et al. 1996). A final flurry measurements on a large nesting aggrega- of foraging provides the larva with all the tion of Bembix troglodytes Handlirsch at the flies it will require and the female then HotSprings area ofBigBend National Park W seals the burrow and digs another. Bembix (N 29^ 10' 39.57" latitude, 102^ 59' burrows often occur in high-density aggre- 52.73" longitude. Brewer Co., Texas) from gations, perhaps as a response to parasite 22 to 23 May 2006. We noted other and predator pressure, as the relative aggregations, apparently of the same spe- incidence of parasites per nest decreases cies, at Santa Elena Canyon and Boquillas with increasing nest density in at least one Canyon, all in sandy areas immediately species (Larsson 1986). Bembix troglodytes adjacent to the Rio Grande River. These females close theburrow whenleavingitto aggregations were in approximately the hunt, but only when the larva is young. same locations as those we observed in The burrow is left open during the inten- previous years. Female wasps carrying sive foraging phase and closed only at prey were netted and weighed on an night. Bembix exploit the most abundant Ohaus Adventurer-Pro electronic balance flies available, apparently learning the to the nearestmg. The head, abdomen, legs richest sources of flies and repeatedly and wings of the wasps were removed exploiting them (Evans 1957, 1963). Bembix with scissors and the thorax mass deter- troglodytes preys upon a large variety of mined. Flight muscle mass was estimated Volume 17, Number 1, 2008 59 as 95% of thorax mass (Marden 1987). The mass in B. troglodytes. However, small mass of the prey fly and its thorax mass wasps were restricted to the smallest flies, were similarly determined. Several of the while larger wasps carried a greater range wasps were collected as voucher speci- of sizes of flies (Fig. 2A). mens and deposited in the Lafayette There were usually many wasps flying College Insect Collection and the Quincy about the nest aggregation, some of which University Life Sciences Museum. were doubtlessly males engaged in the sun We discovered a large colony of Tachytes dance (Evans 1957). Many, however, were chrysopyga obscurus Cresson nesting on an females. As successful hunters returned to earthen berm within an even larger aggre- the aggregation, they were nearly always gation of the eastern cicada killer {Sphecius pounced upon by conspecifics before they speciosus Drury). This berm was located had the opportunity toland. Oftenthe prey within a large chemical production facility was dropped, and then picked up by the (Flint Hills Resources) in Will County, IL, same or another female. Prey-laden fe- at N 41° 26' 39.31" latitude, W 88° 10' males were fast, maneuverable, surprising- 22.16" longitude. From 27 to 28 July, 2006, ly difficult to distinguish from unladen female wasps were collected as they females, and difficult to catch. The fly was returned to their burrows with prey. Body, tucked under the body and held tightly A thorax and prey mass were determined as with all legs. slightly larger profile described above. Several wasps were col- normally provided the investigator with lected as voucher specimens and deposited the cue that a wasp was carrying prey. with the California Academy of Sciences Bembix troglodytes with prey generally entomology collection. All prey were de- landed and entered their burrows very posited in the Quincy University Life rapidly if they escaped attempts at prey Sciences Museum. All data are reported theft. as mean ± standard error unless otherwise Digging activity was frequent. The this- indicated. Descriptive statistics were calcu- tledown velvet ant (Dasymutilla gloriosa lated using Microsoft® Excel 2003, while T (Sauss.)), a brood parasite, was common tests were performed using VassarStats in the area, and often observed digging in (Lowry 2007). the sand as well as entering openburrows. Brood parasitic satellite flies were fre- RESULTS quently observed perched atnestentrances and occasionally entering burrows, some- Bembix troglodytes times closely following prey-laden B. trog- Bembix troglodytes females averaged 99 ± lodytes down their burrows. 2(N - 48) mg in body mass and 37 ± Prey flies collected for this study were 0.003(N = 48) mg in thorax mass, resulting dismembered to determine their FMR, and inanunladenFMRof0.36 ± 0.004(N = 48). couldnotbe subsequentlyidentified to any FlypreyofB. troglodytes averaged 45 ± 3(N great degree. = 33) mg in body mass and 16 ± 1.1(N = 33) mginthorax mass, resulting in anFMR Tachytes chrysopyga of 0.33 ± 0.01(N = 32). Carrying flies Tachytes chrysopyga females averaged 52 resulted in a loaded FMR of 0.29 ± 0.006(N ± 2.6(N = 31) mg in body mass and 18 ± = 27) for female B. troglodytes and loaded 1.0(N = 31) mg in thorax mass, yielding an FMR ranged from 0.23 to 0.34. Hence, all unladen FMRof0.33 ± 0.007(N = 30). Prey fly-carrying females had FMRs well above items of T. chrysopyga averaged 50 ± the marginal FMR of 0.179 (Fig. lA). 4.8(31) mg, producing a mean loaded There was no significant relationship FMR of 0.18 ± 0.007(N = 30) in the wasps. between wasp body mass and prey body The latter value is very close to the 60 Journalof Hymenoptera Research A. Bembixtroglodytes I 1 12 14 16 18 0.2 22 24 26 28 0.3 32 34 0.36 Tachytes chrysopyga 0.1 12 14 16 18 0.2 22 24 26 28 0.3 0.32 0.34 0.36 Loaded FMR Fig. 1. The distribution of flight muscle ratios (FMR) in wasps carrying prey. Arrows indicate the marginal flight muscle ratio, below which take-off is not possible, a. Bembix troglodi/tes from south Texas, b. Tachytes chnjsopyga obsciirus from northern Illinois. /u - |eo- A. ..••''"""' .-••''• (/> 50 - ..-'*""" ™£ 4.0o - .....--••"..-•'••••. • « • ^ 30- ........^-' • • no ..-*'" . • > 20 - • • • 2 10 - • *• Bembixtroglodytes - 20 30 40 50 60 70 80 SO 100 110 120 130 / R E 120 - / >>^ ,- W 100 - C/} ,-' 1 30- > 60- S 40- ' O 20 - • ; .• Tachytes chrysopyga ^ oJ 1 1 1 1 1 1 20 30 40 50 60 70 80 90 100 110 120 130 Wasp body mass (mg) Fig. 2. The effect of wasp size on prey size in two solitary wasps, a. Bembix troglodytes from south Texas, b. TachyteschrysopygaobsciirusfromnorthernIllinois.Thelineswerefittedbyeyetoeachgraphtodemonstratethe maximum prey size for wasps ofa given body mass. Volume 17, Xl-mber 1, 2008 61 marginal FMR of 0.179, ^vhich lies ^vithin mass in 7. chrysopyga. But, as in B, the 95% confidence inten'al of loaded FMR troglodytes, small wasps were restricted to (0.16 to 0.19). the smallest prev, while larger wasps Ho^vever, the range of pre}^ mass (11 to carried a greater range of sizes of orthop- m 132 mg) resulted a very broad distribu- terans (Fig. 2B). tion of EMRs, ^vith many mdi\iduals far B. troglodytes ^vas significantly larger above (maximum = 0.26) and many far than r. chrysopyga m both bod\- mass (P belo^v (minimum = 0.099) the predicted < 0.0001, I = -10.85, df = 77) and thorax value (Fig. IB). At times, prey-loaded T. mass (P < 0.0001, T = -12, df = 76). chrysopyga were fast, maneu\'erable and Unladen FMR was significanth' higher in difficult to capture. However, some indi- B. troglodytes (P < 0.0001, I = -10.25, df = viduals performed short, hopping flights 76), as was laden FMR (P < 0.0001, T = alongthe ground and ^vere easily caught. It —10.25, df = 55). HoAvever, prev mass was ^vas \'ery common for the females with significantly higher in T. chrysopyga than B. prey to land, perhaps even releasing their troglodytes (P = 0.0025, T =^3.15, df = 62). prev for a fe^v seconds, en route to their burro^vs. We never obserA^ed prey stealing, DISCUSSIOX nor did we ever see conspecific females Bemhix troglodytes lurking around the burrows of others. In fact, we formd small numbers of aban- At first glance, B. troglodytes appears to doned prey in the area of the nest be a suboptimal forager, at least in the aggregation. Although satellite tlies were context of our FMR-based model. The often obser\^ed closely trailing prey-laden wasps could in theor\' decrease the time S. speciosus, perching near and entering spent hunting and the number of himting their burro^vs, we ne\'er sa^v evidence of trips by simply capturing and carrying such fUes similarly harassing female T. larger flies. This strategv ^vould doubtless chrysopyga in the same area. increase the number of offspringthat could During much of the day, many T. be reared during the wasps' short life chn/sopyga ^vere flving about lo^v to the spans, but there may be mitigating factors ground, though some ^vere certainlymales. involved. Larger flies maybe less available The burrows ^vere small and relatively or more difficult, energetically expensive inconspicuous, with entrance^vays some- and time-consuming to capture. The flies times near that of an eastern cicada killer. that were captured had lower FMR than There was no ob\Tous tumulus near the T. unladen B, troglodytes. FMR is a useful chiysopyga entrancesvay, though the smaU indicator of maneuverability (see Marden amount of dirt could easily have been 1989, Marden and Chai 1991). Hence, B. displaced bv weather or other disturbanc- troglodytes should have been able to cap- es. Entrance^vays appeared to be open, ture the tTies on the wing. Prey records for although they could have been sealed Bemhix include fast-flving tabanids (Evans deeper in the burrowv. 1957). Prey of I. chrysopyga were diverse It is possible that the foraging strategy of Orthoptera, including Gn-Tlidae (16 Enop- B. troglodytes rninirnizes prey stealing. Few terinae, 7 Oecanthinae), Tettigoniidae (1 wasps returning with prey made it to their Conocephalinae, 3 Copiphorinae) and Ac- burro^vs imchallenged, and though some rididae (8 Cvrtacamthacridinae). The sole of the attacking conspecifics could have adult among the prey ^vas a meado^v been males attempting to mate, theft by grasshopper (Conocephalinae). females was frequent. Evans (1957) did not There ^vas no significant relationship report prey stealing in B. troglodytes, but between ^vasp body mass and prey body the aggregationshe studied ^vere relativeh^ 62 JournalofHymenoptera Research diffuse. Ours was rather dense, which prey to the burrow. Prey need not be increases the likelihood of thievery or exceedingly large either, as Tachytes provi- harassment by males. Evans (1957) did sions with a variable number of prey report prey stealing in five often species of (Evans and Kurczewski 1966), and greater Beiubix examined. In the related wasp numbers could compensate for smaller Stictia hews (Fabr.), in which prey theft is size. common, the probability of being attacked In spite of being ground-nesting digger is directly related to the size of prey wasps with manybehaviors in common, B. (Villalobos and Shelly 1996). These authors troglodytes and T. chrysopyga provide inter- suggest that female Stictia carrying large esting contrasts in foraging strategy. As the prey were more vulnerablebecause oftheir vast majority of its prey are flightless, T. lower flight speed and maneuverability. chrysopyga does not require great maneu- Therefore, if B. troglodytes females took verability (bestowedbyhighunladenFMR) larger flies, theywould likely suffer greater to capture them, as does B. troglodytes. rates of conspecific attack. Small prey help Bembix troglodytes is a large wasp that takes them maintain a high le\'el of maneuver- many, smaU prey, while T. chrysopyga is a ability, as we observed, and probably small wasp that takes fewer, larger prey improve the likelihood of successful trans- (Fig. 2). Both revealed their maximum prey port of the prey all the way to the burrow. size, as demonstrated by thenearly straight Beiubix females can compensate for small line that can be drawn through the highest prey sizebyincreasing thenumber ofprey. points (at a give body size) in Figs 2 A and B. For jB. troglodytes, the maximum is Tachytes chrysopyga somewhat below the wasp's own body On first appraisal, T. chrysopyga appears mass,whileforT. chrysopygaitiswellabove to be an optimal forager (using our simple the wasp's body mass. In each species of FMR-based model), its loaded FMR being wasp, the same predictions were applied indistinguishable from the predicted value and essentially the same methods used. that would maximize load carriage. On Neither species conformed to these predic- average, T. chrysopyga takes prey thatmake tions;however,muchwaslearnedaboutthe full use of its load-lifting capacity. Howev- biology and behavior of each wasp. er, upon closer inspection, T. chrysopyga is, ACKNOWLEDGMENTS in fact, highly variable in the size of prey it takes, and consequently, the magnitude of We thank Wojciech Pulawski and Allan Hook for loaded FMR it experiences. These results species identifications. We are indebted to the are comparable to those of Elliot and National Park Serv-ice for permission to conduct Salbert (1981), who found that T. tricinctus researchinBigBendNationalParkunderNPSpennit (F.) prey varied from 36.3 to 214 mg, #BIBE-2005-SCI-003,andtoFlintHillsResources,Inc. forpermissionto conductresearchon theirpremises, averaging 93.8 mg. These prey are about and especially to the Wildlife Committee for provid- twice the size of those of T. chrysopyga, ingthe required supervision. Thisworkwassupport- which is not surprising, as T. tricinctus ed in part by grants from the Quincy University females,averaging 126.5 mg, areovertwice Faculty Development and Welfare Committee QRC), the size of T. chrysopyga. However, the theNorthern KentuckyUniversityCenter for Integra- tiveNaturalSciencesandMathematics(JMH),andthe approximately tenfold range in prey mass Lafayette College Academic Research Committee in both species may reflect a similarly (CWH). opportunistic foraging behavior. Prey need LITERATURE CITED notbe exceedingly small, as prey stealingis not apparent in these species. Overloading is compensated by behavior - short, hop- Coelhboe,haJv.iRo.r19i9n7.cSiecxaudaalksiilzleerdsim{oSprhpechiiisism aspnedciofsluisg)h.t ping flights being adequate to return some Oikos 79: 371-375. Volume 17, Number 1, 2008 63 — and J. Hoagland. 1995. Load-lifting capacity and S. Brace. 2004. Pre-social benefits of and foraging of three species of yellowjackets extended parental care. Nature428: 650-652. (Vespida)onhoney-beecorpses.FunctionalEcology Larsson, F. K. 1986. Increased nest density of the 9: 171-174. digger wasp Bembix rostrata as a response to and L. D. LaDage. 1999. Foraging capacity of parasites and predators (Hymenoptera: Spheci- thegreatgoldendiggerwasp,Sphexichneumoneus dae). Entomologia Generalis 12: 71-75. L. Ecological Entomology24: 480-483. Lin,N.andC.D.Michener.1972.Evolutionofsociality Edgar, P. K. and J. R. Coelho. 2000. Load-lifting ininsects. QuarterlyReviewofBiology47: 131-159. constraints on provisioning and nestbuilding in Marden, J. H. 1987. Maximum lift production during the carpenter wasp, Monobia quadridens L. (Hy- takeoff in flying animals. Journal ofExperimental menoptera: Eumenidae). Journal of Hymenoptera Biology 130: 235-248. Research 9: 370-376. . 1989, Bodybuilding dragonflies: costs and Elliott,N. B. andP. Salbert. 1981. Observations onthe benefitsofmaximizingflightm.uscle.Physiological nesting behavior of Tachytes tricinctus (F.) (Hy- Zoology 62: 505-521. menoptera: Sphecidae, Larrinae). Journal of the and P. Chai. 1991. Aerial predation and New York Entomological Society 88: 170-173. butterfly design: how palatability, mimicry and Evans,H. E. 1957. Studies ofthecomparativeethologyof the need for evasive flight constrain mass digger wasps ofthe genus Bembix. Cornell Univer- allocation. American Naturalist 138: 15-36. sity, NewYork. O'Neill, K. M. 1985. Egg size, prey size, and sexual . 1962, The evolution of prey-carrying mecha- sizedimorphismindiggerwasps (Hymenoptera: nisms inwasps. Evolution 16: 468-483. Sphecidae). Canadian Journal ofZoology 63: 2187- and F. E. Kurczewski. 1966. Observations on 2193. the nesting behavior of some species of Tachytes Tengo,J.,H. Schone,W. D. Kiihme,H. Schone,andL. (Hymenoptera:Sphecidae,Larrinae).Journalofthe Kiihme. 1996. Nesting cycle and homing in the Kansas Entomological Society 39: 323-332. diggerwasp Bembixrostrata (Hymenoptera Sphe- andK.M.O'Neill.2007. Thesandwasps:natural cidae). Ethology, Ecologyand Evolution 8: 207-211. history and behavior. Harvard University, Cam- VUlalobos,E.M. andT. Shelly. 1996.Intraspecificnest bridge. parasitism in the sand wasp Stictia heros (Fabr.) Field,J. 2005. The evolutionofprogressiveprovision- (Hymenoptera: Sphecidae). Journal of Insect Be- ing. Behavioural Ecology 16: 770-778. havior 9: 105-119.