The Giant Nests of the African Stink Paltothyreus tarsatus (Formicidae, ) Author(s): Ulrich Braun, Christian Peeters and Bert Holldobler Source: Biotropica, Vol. 26, No. 3 (Sep., 1994), pp. 308-311 Published by: The Association for Tropical Biology and Conservation Stable URL: http://www.jstor.org/stable/2388852 . Accessed: 28/11/2013 01:48

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp

. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

.

The Association for Tropical Biology and Conservation is collaborating with JSTOR to digitize, preserve and extend access to Biotropica.

http://www.jstor.org

This content downloaded from 134.157.146.58 on Thu, 28 Nov 2013 01:48:10 AM All use subject to JSTOR Terms and Conditions BIOTROPICA 26(3): 308-311 1994

The Giant Nests of the AfricanStink Ant Paltothyreustarsatus (Formicidae, Ponerinae)1

UlrichBraun, Christian Peeters2, and Bert Holidobler Theodor-Boveri-Institut,Lehrstuhl Verhaltensphysiologie und Soziobiologie der Universitat,Am Hubland, D-97074 Wurzburg,Germany

ABSTRACT Fourteennests of the ponerineant Paltothyreustarsatus were excavated in Kenya and in the IvoryCoast. All colonies containedonly one mated queen. The workerpopulation varied markedlybetween colonies, reaching 2444 adults in one case. Nests consistedof manychambers located 30 cm to morethan 150 cm belowthe surface. The construction and arrangementof the chamberswere not particularlyelaborate, but the associatedhorizontal tunnel system (5-15 cm below the surface)was mostremarkable. In one nest,over 130 m of shallowsubterranean tunnels were uncovered. An area of 1200 m2 was coveredby the tunnelsystem of anothernest. Through multiple tunnel exits the are able to reachalmost any spot in theirforaging area. In some nestswe founda secondhorizontal tunnel system, 50- 80 cm below the surface,which connected nest chambers several meters apart.

Keywords: ants;foraging; monogyny; nests; Paltothyreus; Ponerinae; polydomy; queen; reproduction.

PALTOTHYREUS TARSATUS, a memberof the ant sub- galleriesby gentledigging with a trowel.A consid- familyPonerinae, is widelydistributed throughout erableamount of clearing was oftenrequired in order Africasouth of the Sahara (Wheeler 1922). It is a to proceed;branches and rootswere cut and ground hunterand a scavenger.Although the workers for- litterwas removed.Once large concentrationsof age individually,they can recruitnestmates with workerswere found, vertical digging proceeded. The chemicalsignals when retrieving large or more abun- excavatedtunnels were leftuncovered, and over- dant preyitems (H6lldobler 1984). The nestsof nightdigging activity by the ants led to thedetection P. tarsatusare built in thesoil. As notedin previous of many side tunnels.The lengthsof the tunnels fieldstudies, they have multipleentrances, and ma- weremeasured between intersections. The number turecolonies occupy rather large surface areas (Le- and locationof exitholes were not recordedin this vieux 1965; Kalule-Sabiiti& Banage 1977; H611- colony.Another ten colonieswere collectedin the dobler 1980, 1984; Dejean et al. 1993). In the National Park of the Comoe in the Ivory Coast courseof continuing ecological investigations of Pal- duringAugust 1992 and February-May1993. tothyreuswe have found that theirnests can be substantiallybigger in areaand morepopulous than previouslythought. RESULTS Paltothyreustarsatus colonies nest primarily in for- MATERIALS AND METHODS estedareas, but also in the transitionzone between forestand savanna.Our mostdetailed data on nest We excavateda total of 14 nests. Four colonies structurewere obtained from two colonies in Kenya. werecollected in the Shimba Hills Reserve(Kwale In one colony (August 1978), we especiallyex- district,south of Mombasa, Kenya); threeduring aminedthe tunnel system in the "core area" where July-August1978 and one duringMay 1991. The most of the brood and the queen were collected latterin particularwas verythoroughly dug out by (Figs. 1 and 2). There were about 25 exit holes two personsover 4 days (>40 man-hours)and leadingeither to tunnels5-15 cm belowthe surface, unexploredtunnels still remained. We began at an or directlyto the inhabitedchambers. The latter activeentrance, and followedthe shallowforaging werealso connectedtogether by subterraneangal- leries(Fig. 2). The core area could be recognized by the crater-shapednest entrancesand the con- I Received 19 March 1993, revisionaccepted 4 October 1993. spicuousmounds of soil and refusepiles (containing preyand cocoonremains); this was not the case in 2 Presentaddress: CNRS URA 667, Laboratoired'Etho- logie Experimentaleet Comparee,Universite Paris Nord, all colonies.The majorityof thenest chambers were F-93430 Villetaneuse,France. located 30-150 cm below the surface.The larger 308

This content downloaded from 134.157.146.58 on Thu, 28 Nov 2013 01:48:10 AM All use subject to JSTOR Terms and Conditions The Giant Nests of the AfricanStink Ant 309

Im

FIGURE 2. Verticalsection through the central cham- bersand connecting galleries of a P. tarsatusnest (Kenya; August1978). This drawing is basedon numerousfield

1m sketchesand photographs taken during the excavation.

FIGURE 1. Corearea of a P. tarsatus nest(Kenya; This mayexplain why so fewbrood were collected, August1978), showing the funnel-shaped entrances sur- roundedby mounds of soil and refusepiles. The under- especiallycocoons (although one cocoonwas carried groundtunnels leading to theforaging grounds are also bya workeralong a foraginggallery, indicating that indicated,together with exit holes. theywere present at thistime of the year). Two additionalnests excavated in July 1978 also yieldedonly one inseminatedqueen foreach chambers,measuring approximately 25 x 5 x 5 colony;the numbers of workerswere 392 and 446, cm, were foundcloser to the surface(ca 30-100 respectively.In both cases, however,we were not cm depth),the smallerchambers (4 x 3 x 1 cm) yetfamiliar with the complexity of the architecture weremostly located at a depthof 100-150 cm. In ofPaltothyreus nests, and it is thereforepossible that thisparticular nest we identifiedmore than 40 in- we did not uncoverother major parts of the nests. habitedchambers. We collectedfrom those cham- Our excavationsof ten P. tarsatusnests in the bers,from adjacent galleries and on the surface,a IvoryCoast (1992 and 1993) revealeda similar totalof 1 dealate queen, 916 workers,82 larvae, organizationas foundin Kenya.The sizeof thenest 45 pupae, 4 males, and 1 alate queen. We con- populationsvaried greatly (Table 1), suggestingthat firmedby dissectionthat the dealate queen was some colonieswere younger, while others may have inseminated.Despite our intensiveefforts we prob- been collectedincompletely. One colonyexcavated ably did not uncoverall nest chambers,missing especiallymore alate queens and males. As previ- ouslynoted (H6lldobler 1980, 1984) an extensive systemof shallowtunnels led fromthe nestcham- .~~~~ bersto the surroundingforaging grounds (Fig. 1). In a second colonyfrom Shimba Hills (May 1991), we aimed to revealthe entirenetwork of shallowtunnels (Fig. 3). The excavationstarted in an open, grassyarea, but the galleriessoon led to twodifferent forested areas. Over 130 m offoraging gallerieswere uncovered.This excavationrevealed thatone colonycan be distributedinto several nest- ing sites,labelled A, E, and F (Fig. 3). In A, where Sm chambersreached deeper than elsewhere,we col- lected518 workers,20 larvae,6 eggs,and I dealate queen (also inseminated);there were 208 workers FIGURE 3. The subterraneantunnel system of a P. tarsatus tunnels and 2 larvaein and 77 workersin F. In addition nest(Kenya; May 1991). Onlyshallow E, are indicated(5-10 cm belowthe surface). The arrow we caught 102 workerswalking along the uncov- indicatesthe spot where the excavation began. A, F, and eredforaging galleries. Thus a totalof 905 workers E arethe locations of deep chambers where workers and werecollected throughout the rangeof thiscolony. broodwere collected. Many workers were also found above Howeverwe wereunable to followall the foraging groundaround B and C. Stippledareas indicate forest and thicketvegetation; clear area represents grassland. galleries,or to excavatedeep chamberscompletely.

This content downloaded from 134.157.146.58 on Thu, 28 Nov 2013 01:48:10 AM All use subject to JSTOR Terms and Conditions 310 Braun, Peeters, and Holidobler

TABLE 1. Demographicdata from10 nestsof Paltothyreustarsatus excavated in theIvory Coast.

Dealate Alate Nest queen queens Males Workers Cocoons Larvae Eggs Aug. 1992 1 0 0 385 499 458 218 Aug. 1992 1 0 0 157 120 7 35 Aug. 1992 1 59 0 247 68 91 19 Aug. 1992 1 73 5 688 764 241 0 Aug. 1992 1 36 0 377 30 164 0 Feb. 1993 1 0 0 1552 94 1376 208 Feb. 1993 1 466 696 1191 444 129 945 Feb. 1993 1 408 131 1522 166 2016 139 April 1993 0 0 0 2444 1118 2789 177 May 1993 0 0 0 692 580 915 163 in April1993 wasconsiderably larger (numbers of specialcase of polydomy.Based on thesenew find- adultworkers, cocoons and larvae).In all colonies, ings,we have to considerthe possibilitythat some however,we foundonly one dealate queen. In four of the nests(10 to 20 m apart) whichH6lldobler nestswe found a secondsystem of deeper horizontal (1984) identifiedas belongingto separatecolonies, tunnels,50-80 cmbelow the surface, which led to may in factbe partsof one and the same colony. anothernetwork of chambers,3 to 5 m fromthe Generally,colony size maybe underestimatedsince, currentlyoccupied core area. Whether these cham- given the availabilityof numerousunderground berswere entirely abandoned or had just been evac- connectionsbetween different parts of the colony, uatedcould not be determined.In one nestwe brood and workerscan be evacuatedas the ants measuredthe subterraneanforaging tunnels cov- detect the vibrationscaused by human digging. eringan areaof up to 1200 M2. H6lldobler(1984) observedthat tandemrunning is used to recruitnestmates when parts of a nestare DISCUSSION accidentallydestroyed (as opposed to trail laying duringforaging); this specializationsuggests that Antsof the subfamily Ponerinae usually build rel- relocationwithin an existingnest complex may occur ativelysimple nests in thesoil or in rottingwood. frequently. A fewspecies exhibit more complex nest architec- The home rangeof P. tarsatusforagers above ture,e.g., Harpegnathos saltator (Peeters et al. 1994). ground is generallynot greaterthan 5 m and is Althoughthe design and construction ofthe cham- usually considerablysmaller (H6lldobler 1984). bersof P. tarsatusnests are not particularly elab- However,the underground trunk routes with mul- orate,the huge scale of theirsystem of chambers tipleexit holes enable the foragers to travelsecurely and tunnelsis remarkable.Our data indicatethat to manypoints in the colony'sterritory (up to 40 coloniescan be considerablylarger than previously m away fromthe main nest),before commencing assumed.In theIvory Coast Levieux (1967) ex- to forageabove ground.Mark and recapturetests cavated5 coloniesand foundan averagenumber indicatethat individualants show substantialfi- of 350 workersper colony,with a maximumof delityin theuse of specificexits (H6lldobler 1984). 450 workers.In Cameroun,Dejean et al. (1993) Such decentralizedterritories allow harvestingben- collectedtwo colonies with 260 and 311 workers. efitsfrom throughout the interiorof the largeter- Thelarger worker populations in our study may be ritorywhile reducingpredation on foragers(H611- relatedto the completeness ofour excavations. Col- dobler & Lumsden 1980, H6lldobler & Wilson oniesof P. tarsatusare large relative to otherpo- 1990). It also enables preciseorientation in the nerineants with a singlequeen (see Peeters 1993), restrictivelighting conditions of forests(H6lldobler althoughcolonies of Leptogenysdistinguenda can 1980). yieldmore than 30,000 workers(Maschwitz et al. We foundonly one dealatequeen in all colonies 1989). excavated. This confirmsthe findingof Levieux Our studyalso revealedthat large P. tarsatus (1965) that P. tarsatuscolonies are monogynous. coloniescan occupymore than one nestingsite, They are establishedby semi-claustralfoundation severalmeters distant from each other,but con- (C. Peeters,pers. obs. in South Africa;Villet et al. nectedby subterraneantunnels. This represents a 1989). Given the highlydispersed character of the

This content downloaded from 134.157.146.58 on Thu, 28 Nov 2013 01:48:10 AM All use subject to JSTOR Terms and Conditions The Giant Nests of the AfricanStink Ant 311 nestcomplexes, it is interestingto speculate about How thisis achievedis currentlybeing investigated themechanism of queen control in thisspecies. In in our laboratory. thelaboratory we isolated a largegroup of over 500 workerswithout their queen. No eggswere laid ACKNOWLEDGMENTS untilafter 11 weeks;after 15 weeksmore than 10 eggswere found. These observations suggest that Thiswork was supportedby grantsfrom the Deutsche theworkers were previously inhibited by the queen. Forschungsgemeinschaft(Leibniz Preis to BH and SFB 215, project18). We wouldlike to thankJ. Mugambi Sincethe colonies are large, we assumethat repro- forfield assistance in ShimbaHills, R. Bagine(National ductiveinhibition ispheromonal, and that the queen Museumsof Kenya) and E. Linsenmair(Field Station of signalis distributedtoworkers throughout the nest. theComoe National Park, Ivory Coast) for strategic help.

LITERATURECITED DEJEAN,A., G. BEUGNON, AND J.-P. LACHAUD. 1993. Spatial componentsof foragingbehavior in an African ponerineant, Paltothyreustarsatus. J. Behav. 6: 271-285. H6LLDOBLER, B. 1980. Canopy orientation:a new kind of orientationin ants. Science210: 86-88. 1984. Communicationduring foraging and nest-relocationin the AfricanStink Ant, Paltothyreus tarsatus Fabr. (HymenopteraFormicidae, Ponerinae). Z. Tierpsychol.65: 40-52. , AND C. J.LUMSDEN. 1980. Territorialstrategies in ants. Science210: 732-739. , AND E. 0. WILSON. 1990. The ants.Belknap Press of HarvardUniversity Press, Cambridge, Massachusetts and Springer-Verlag,Heidelberg-Berlin-London. KALuLE-SABiITI, M., AND B. BANAGE. 1977. Some aspectsof theecology of thestink ant Paltothyreustarsatus (Fabr) in Uganda. Geo. Eco. Trop. 1: 119-138. LEVIEUX,J. 1965. Descriptionde quelques nids de fourmisde C6te d'Ivoire (Hym.) . Bull. Entomol.Soc. France 70: 259-266. 1967. Recherchesecologiques dans la savane de Lamto (C6te d'Ivoire): donnees preliminairessur le peuplementen fourmisterricoles. La Terreet la Vie 3: 278-296. MASCHWITZ,U., S. STEGHAUS-KOVAC, R. GAUBE, AND H. HXNEL. 1989. A South East Asian ponerineant of the genus Leptogenys(Hym., Form.) with armyant lifehabits. Behav. Ecol. Sociobiol. 24: 305-316. PEETERS, C. 1993. Monogynyand polygynyin ponerineants with or withoutqueens. In L. Keller (Ed.). Queen numberand socialityin ,pp. 234-261. OxfordUniversity Press, Cambridge, England. , B. H6LLDOBLER, M. MOFFETT, AND T. M. MUSTHAK Au. 1994. "Wall-papering" and elaboratenest architecturein the ponerineant Harpegnathossaltator. Insectes. Soc. 41: 211-2 18. ViuLET, M., R. CREWE, AND H. ROBERTSON. 1989. Mating behaviorand dispersalin Paltothyreustarsatus Fabr. (:Formicidae). J. InsectBehav. 2: 413-417. WHEELER, W. M. 1922. Ants of the AmericanMuseum, Congo Expedition.A contributionto the myrmecology of Africa.Bull. Am. Mus. Nat. Hist. 45: 1-269.

This content downloaded from 134.157.146.58 on Thu, 28 Nov 2013 01:48:10 AM All use subject to JSTOR Terms and Conditions