vol,. 13 NO. 4 THE SOUTHWESTERN ENTOMOLOGIST DEC. 1988

NATURAL ENEI,IIES OF : WASPS IN EAST TEXASI/

D. A. Dean, M. Nyffeler and W. L. Sterling Department of Entomology, Texas A&M University College Statj.on, IX 77843

ABSTRACT prey analysis of sphecid wasps was conducted in an east Täxas färrnland area. Two spider farnilies, Araneidae and , comprised > 9OA of the vtaspsr diet. The rnost abundant prey of these hlasps were the orb-weavers, Gea heptaqon (Hentz) (57.L2 of total), and Acanthepeira stellata (ltalcken- aer) (L7.72). The southern black widow, mactans (F.i is.st), !üas also an important component of the diet of these ntasps. Renaining farnilies represented < LoA. These sphecid wasps are fairty selective hunters on two-three fänilies of spiders. Individual wasps occasionally rnay choose predominantly froro a single species of spider. Data of other workers are compared to our studY.

TNTRODUCTION

Fernale mud dauber lrasps provide spiders as food for thei-r immatures. The wasps construct rnud nests containing several tubuLar brood cells. After being stung and paralyzed, the spiders are transported to the nestsi and each brood cell is provisioned with sorne spiders. The ltasp then lays an egg on one of these spiders. After hatching, the ntasp larva consumes the well-preserved spiders in the brood ceII, pupates, and later the adult emerges through a hole chewed in the brood ce1l. Two species of sphecid wasps that occur in the United States often exist syrnpatrically. They are the yellow-legqed mud dauber, Sceliphron caementarium (Drury), and the steel- , SgIyb:LeD californicum (saussure) , an obligate hyperparasite. Sceliphron caementarium builds its own nests; g.- cJlifornicurn breaks into the nests of S. caementarium, empties the brood cells, filts then with its own prey' lays an egg and ttren reseals the nest (Rau 1935b, Muma and Jeffers fSas1. The predatory behavior of the two wasp species are described in detail by Eberhard (1970) and Coville (L976, L987). we analyzed the diet of mud dauber $tasps in.east Texas to determine tüeir ecological role as predators in a rural area.

I'IATERIALS AND METHODS The studies hlere conducted during the sumner of l-985 on

l/ulrrnenoptera:

283 the L. N. Brown farm, located 5.6 kn west of Austonio, near Crockett, Houston Co., east Texas. The farm buildings are surrounded by extensive grassland that serves as a horse pasture and j-s occasionally rnown. The area is dorni_nated by cotton fields and grazing land. wasp nests hrere removed frorn inside and outside of farm buildings in June and Ju1y. Dorris (l-969) noted that wasps ernerged frorn nests in Arkänsas from June to early September, and June and July were the best months for collecting. A total of 69 nests were collected; 23 nests contained spiders and 2O of these nests contained identifiable spiders. Wasps from the remaining 46 nests had consumed the prey and emerged. The prey were rernoved frorn the brood cÄf:_s and preserved in 708 ethyl alcohoI. They were l_ater identified under a microscope to the lowest taxon possible. Voucher speci-rnens were deposited in the corlection of the Departnent of Entomology at Texas A&M University. Besides the actual prey of the wasps, we al_so assessed their potential prey (spiders occurrinq in ihe waspsr environ- nent) by sampling spiders from the cotton pl_ants and grassland (composed of various grasses and low growing annual Dicotyledonae) near the farm. Twenty-five- O-Vad samples (Dietrick 1"961-) consisting of l- m of row äach were taken weäk1y in cotton for l-3 weeks irom June to early september (totai spiders = 923). Ten D-Vac samples were take-n eväry three weeks in grassland (330 spiders) on four sanpling dates from July to 9af1y September, and an undetermined number of samples were taken by sweeping in grassland (L.252 spidersl on nine sarnpling dates from June to early septernber. They were returned tö ttrÄ laboratory and rater identified and counted under the micro- scope. Taxonomic literature of spiders is scattered amongr numerous publicati.ons, but spider Genera of North agise ly Roth (L985) contains references to rerrGi-ons or ttre variou3 groups. Vogel (L970) presents a list of spider species recorded from Texas along with references.

RESULTS AND DISCUSSTON - Tl" prey-were collected frorn Sceliphron caernentariurn or glarybion californicurn nests lraurä-rJ.- sincE-w" could not identify.the wasp species by ids nest and since adurt wasps do not remain with the nest, we present the prey of both wasp species grouped together (Table l-). The waspst prey consisted exclusively of spiders. Most $tasp nests contained one to three species from one or two !anili99. Spiders of the farnilies araieidae (7s.LZ of total), Theridiidae (t_3.08), oxyopidae (6.9?), sai-ticidae (j_.6ti-i-n ; Lycosidae (0.2t), and Thomisidae @.äZ) were found tire vaspst nests. lhe Araneidae and Theridiidae comprised > 9Og of the hraspsr diet. Spiders often captured by the wasps included the orb-weav- ers_ qea heptaqon (Hentz) (57.i.t) and Acanthepeira stellata (Walckenaer) (L7.72) which made up 752 of the waspslrey. The next most abundant components in the waspst diet-wele the cornL'-footed spiders Tidarren haernorrhoidalä (Bertkau) (7.f_?) and.I=,atrodectus nactans (F.) (5.9S) and the lynx spider Oxyopes Eafticrls Hentz (6.58). However, atJ_ !. ha-emor;hoidalE-;äre found in one nest, and most o. sarticul were rouna in another nest. of the spiders cofiecTEil rrequently by wasps, G. heotagon, A. stellata and o. salticus were äfundant on the

284 \Ni\o ONoOON.$\tcOaONi \OtsN N-t naoN ;til g :l corao tI NJ6 N

coNn rl N tco d

{t !l .f rl odo \o

Öd Nd *l -l N

.") c) _, =l n6 0$ NO 9d N N

s il N r!l ! .dEqq'iEHd.dq*Eqdq dqi a) 0 (! E o tqi ; q) a 6l! il q ü5^^? n ! _y ik 0) 0 u (!(u (ü > H äl Gil(!trlf !0.i I N E 0 r 0l i= o! o ! O *l r! il ho i N c @ i616le4($!O! F ,!t#i 3 il Clqt :lol iläl uli ! iE!3E >. ;l F =.1@^l gl Hl !nl ül fl 3l tlrr i Et 'lt q) q, €l ?", lr E6l q) El El :l $l#l äl il il il ! ! ol .El ä. (! 6Xl äl IlHl R4 äl Tl rl 11, ;l E ! .'t ,el 0, El EtH El ;i;i "J Et !;i *t! ri* äl# * -el gl I kr ol A El äH €l älll ä äl EH €H ill I't 6 - d:l< Frl g IlgllE älä il! Fvl !l rll #l Jl !l ä81 I dl

285 foliage of mixed ltrasses, annual Dicotyledonae and cotton plants of that farrn area (Table 2). Lairodectus mactans is also common in that area (underrepresented in Table 2 based on our sampling rnethods)-. Tidarren haernorrhoidale was probably the most abundant spider in the barn where many wasps buill their nests. Thus, the wasps captured prinarilli spiders that were abundant in that area. However, a comparison of actual Pfey (Table 1) and potential prey (Tab1e 2) indicates that those wasps are selective predators, since the species diversi- ty of the potential prey is higher than that of the actual prey.

TABLE-2. Percent species composition of spiders (potential prey of Sphecid I{asps) from an east Texas Farmland Area in thä Vicinity of a Mixed Colonv of Sphecid l{aso Nests fsummer t9a5l -

Sweep D-Vac D-Vac Uloboridae crlonosus (Walckenaer) 0.9 Dictynidae.Ulobgrus o.2 Dictvna secrregata certsch & Mulaik O.L Theridiidae o.3 3.3 Achaearanea qlobosa (Hent.z) L.2 o.2 Argyrodes fictiLiurn (Hentz) o.3 (F.) 0.1 0.9 Theridion australe Banks o.2 o. 3 3,.7 Theridion crispulum Simon 0. l_ Theridion murarium (Ernerton) 0. l_ 0.3 Thrnnoites so- 0.3 Tidarren so- 0. t- Mysrnenidae incredula (Gertsch & Davis) 0.3 Linyphiidae_.Mvs$?lq o.L Ceraticelus sp. o.2 0.9 o.L Epgriqone eschatologica Crosby 0.3 0.8 Eriqone autumnalis Emerton 0.1 o.2 Grammonota texana (Banks) o.4 0.9 0. l_ Other Linyphiidae 0.3 o.7 Araneidae Acanthepeifa stel_lata (Walckenaer) r-.8 0. 6 3.5 Aroiope trifasciata (Forskat) 0.3 0.6 Cvclosa turbinata (Walckenaer) o.2 0.6 0.5 Eustala anastera (Walckenaer) 0.9 o.2 Gea heptagon (Hentz) 0. 6 26.4 l_. 1 Glenoqnatha foxi (Mccook) o.2 2.4 1.0 llansora gibberosa (Hentz) 0.9 0.9 o.2 Neoscona arabesca (Walckenaer) 4. L 0.9 o.7 Tetraqnatha laboriosa Hentz 0.5 0.6 2.2 Other Araneidae 0.1 Mimetidae Mimetus hesperus Charnberlin o.1 0.L Pisauridae Dolomedes sp. 0.L Pisaurina mira (Watckenaer) o.1

286 TABLE 2. cont- Lycosidae Lvcosa rabida (Walckenaer) o.2 o.6 o.L Pardosa sp. o.2 1.2 l_.5 Schizocosa avida (Walckenaer) 0. 1_ o.6 Other Lycosidae 0.3 Oxyopidae Oxyopes apollo Brady 0.1 o.3 o.3 Oxyopes salticus Hentz 32 .5 32.4 67.2 Peucetia viridans (Hentz) l_. 3 0.3 0.3 Clubionidae Cheiracanthiun inclusum (Hentz) 0. t- 1.3 Clubi-ona catahrba Gertsch 0. 1_ Anyphaenidae Avsha gracilis (Hentz) o.2 2.O Vlulfila saltabunda (Hentz) 0.6 o.2 Thomisidae Misumenoides formosipes (Walckenaer) o.2 Misumenops asperatus (Hentz) 0. l_ Misurnenops celer (Hentz) 0.9 0.3 Misunenops dubius (Keyserling) o.2 Misumenops sp. (irnrnatures) 30. 0 10. 3 2.8 Xysticus sp. o.3 Philodromi-dae Philodromus sp. 4.2 5.5 t_. 3 Tibellus duttoni (Hentz) 4.3 2.4 Salticidae Eris sp. 0.1 Habronattus coecatus (Hentz) O.l_ 0.6 Hentzia pahnarum (Hentz) O.l_ 0.3 Metaphidippus qal-athea (Irfalckenaer) 7.7 2.4 2.r Phidippus audax (Hentz) 5.i- 2.4 l_. I Phidippus pius Scheffer O.l_ Sarinda hentzi (Banks) O.L 0. l_ Thiodina puerpera (Hentz) O.2 o. t_ Zyqoballus rufipes (G. & E. peckharn) 0.3 o.3 o. l_

Unfortunately, in this study we were not able to determine which of the spiders listed in Table l_ were kill_ed by C. cali-fornicurn or S. caernentarium. However, other authors were able to determine these two wasp speciest diets separately, as represented in Table 3. The follording pattern emerges: C. californicum captured predominantly spiders of the iarnilies Araneidae and Theridiidae (> 9OZ of the waspst di-et). In Oklahona, Latrodectus mactans was found to be an rrindex speciesrr prey of C. californicum (Horner and Klein l-g7gr. Sceliphron caementarium attacks prirnarily spiders of the families Araneidae, Thomisidae, and Salticidae. Spiders of the genera Oxvopes and Misumenops are typical prey of S. caementarium (Horner and Klein 1,979). The sirnilarj-ty of the diets of C. californicum and S. caementarium was calculated by the niche overlap formula of Horn (1966) using prey data from the l_iterature (Muma and Jeffers 1945, Horner and Kl-ein L979). Overlap values range fron zero to l-.0, from entirely different to identical diets, respectively. The calculation was carried out at farnily level of the identified prey items, but these values rnay be overest-

287 irnated when compared with calculation at species level. From the data of Muma and Jeffers (1945) we calculated an overlap value of 0.36, and a value of 0.39 from the data of Horner and Klein (L979). These values j-ndicate that the diets of both wasp species in general are distinct, even if some spiders (orb-weavers) are conmon to the diets of both.

TABLE 3. Proportion of Five Major Food Components in Diets of Sceliphron cäementarium and Chal-vbiori californicum.

sceliphron caementariUm4/ ch.l-ybior, califorriicurnS/

MD MO MS OK MD Mr! .)I' ET. 8g zz z4 zz Theridiidae 0.5 8.1 67.A 85.9 29.5 25.2 Araneidae 55.7 A7.9 36.7 29 .5 26.9 t-3 .9 70. o 72.7 Oxyopidae 9.5 2.2 3.O 5.9 1.6 - 0.4 Thonisidae l_6. l_ 4.9 l-9. o 63. 1 l_.8 0. L - L.l_ Salticidae r-0. L 4.o 20.0 1. O 1.6 0.1 - 0.1 Others 8.1_ 1.0 13 .2 0.5 o.3 0.5 0.5

B/Uo: ttaryland - Muma and Jeffers (l-945). MO: Missouri (S.c.) - Rau (1935a), (C.c.) - Landes et a1. (l_e87). MS: Mississippi - Dorris (l_97O). OK: Oklahoma - Horner and Klein (t-979). FL: Florida - Landes et al. (L987).

The _species-specific prey preferences may be explained by the different hunting behavior of the two wasp Species. Rau (1935b) and Muma and Jeffers (1945) found that S.-caementarium hunts ämong foliage and flowers, while g. califoinicurn searches for spiders near the ground and around and in buildings. - A comparison of our data (Table t-) with thoSe fron the literature reveals that the prey of mosi wasp nests collected in thi.s study fit the foräqing pattern ot c. californicun (exceptions: nests no. 2, 4, LO, f:, fO1. Evidence for this pattern is indicated by the following observations: L) foliage-hunting spiders (Oxyopidae, Thomisidae, Salticidae), apparently typicat prey of S. caernentariurn, are rnissing-spiders in tLä prey of most hrasp np,stsi 2) in BOt of the wasp nests, were found that build their webs close to the ground (G. heptagon, irnrnature A. ste1lata, L. rnactans). f,atrodectüs magfans, described in literature as an ilindex. prey" of S. 9g]+fAXqiSgU (Irving and Hinman 1935, Rau 1935a, Härner and Kl,ein L979), were found in four wasp nestst 3) a large number of the barn-dwelling spider T. haenorrhoj-dale was found in wasp nest no. 15; and 4) urost trasps that hatched from nud daubei nests brought into the laboratory hrere g. californicun. On the farm C. calj-fornicurn rras more common tfr.n g. Sphecid wasps could be useful in ttreir"egrngnt.riun. coftecUion of

288 poisonous spiders of the genus Latrodectus (Irving and Hinnan 1"935, Rau 1-935a). fn Maryland and Oklahoma ca. 2SZ of the diet of C. californicum consisted of Latrodectus (Muna and Jeffers 1945, Horner and Klein L979). At the same tine these hrasps are natural enemies of spiders which feed on cotton ,s (Nyffeler et al. 1988). These spj_der-hunting s/asps are themselves occasionally killed by web-building spiders (Obin r_e82 ) .

ACKNOWLEDGI.{ENT We thank L. N. Brovn, farmer in Austonio. for pernission to conduct this study on his property and T. L. payne for the use of his trailer as a field laboratory. We also Lhank S. ,f. Merritt for identification of the wasps and R. G. Breene and W. R. ltartin for their review of this päper.-froject This research was nade possible by the Expanded Researci H-2591--2LOO of the Texas Agricultural Experinent Station and by a postdoctoral fellowship of the Swiss National Science Foundätion granted to one o.f us (M.N.). Approved for publication as Ta 23966 by Director, Texas Agriculturat Experirnent Station.

LITERATURE CTTED Covil1e, R. E. L976. predatory behävior of the , californicum (: Sphecidae) . -pan-pacific Entomol. 52. 229-233. Coville, R. E. 1987. Spider-hunting sphecid wasps. Ecophysiol- ogy of Spiders. Ed. W. Nentwig. Springer-Veilag, Näw york, 436 pp. Dietrick, E. J. l-96L. An inrproved backpack rnotor f,an for suction sanpling of insect populationJ. ,f. Econ. Entomol. 54. 394-395. Dorris, P. R. L969. Splders collected from mud-dauber nests in clark County, Arkansas. Ark. Acad. scj-. proc. 23: g8-90. Dorris. P. R. l-97O. Spiders collected from mud-dauber nests in Mississippi. J. Kans. Entornol . soc.43:10-l_l-. Eberhard, ll. L97O., The predatory behavior of two rdasps. Agenoi4eus hurnilis (pornpilidae) and Sceliphron caementarium (Sphecidae), on the orb weavinq spiaär Argnefts cornutus (Araneidae). psyche 7?. 243-25j,. Horn, H. s. l_966. Measurenent of overlap in comparative ecological studies. An. Nat. LOO. 4Lg-424. Horner, N. V., and J. H. Klein, Jr. j_929. Spider prey of two mud dauber wasp species in Cornanche County, Oklahorna (Hymenoptera: Sphecidae) . Environ. Entomol. 8: -30-31_. Irving, W. c., and E. H. Hinnan. l-935. The blue mud-dauber as a predator of the black widow spider. Seience AZ. 395-396. Landes, D. 4., M. S. Obin, A. B. Cady, and J. H. Hunt. 1987. Seasonal and latitudinal vari.ation in spider prey of the rnud dauber Chalvbion californicum (Hyrnenoptera, SphLcidae) . J. Arachnol. 15. 249-256. Muma, M. H., and W. F. Jeffers. ].g4S. Studies of the spider prey of several mud-dauber wasps. Ann. Entomol. Soc. Am. 38: 245-255. Nyffeler,-M., D. A. Dean, and I{. L. Sterling. L98g. prey selection and predatory irnportance of orb-wäaving spiderä (Araneae: Araneidae and Uloboridae) in Texas cotton. Environ. Entornol. (In press). obin' I{. s. r-982. spiderJ living at wasp nesting sites: what

289 constrains by mud-daubers? Psyche 89: 321-335' Rau, P. 1-935a. The wasp, Chalvbion cyaneum Fab. preys upon the black widow spider, Latrodectus nactans' Fab. (Hymen', Araneae). Entomol. Nehls 461 259-260. Rau, P. fg:Sf . The spider prey of the mud l^Iasp' Sceliphron caementarium (Araneae, Hymen': sphecidae)' Entomol' Ne\^Is 46: 267 -27 0 . Roth, V. D. 1-985. Spider Genera of . Privately published by the author. Available from Am. Arachnol' soc' Vogäl, B. R. igzo. Bibliography of Texas spiders. Arnadillo Papers 2: L-36.

290