Utah State University DigitalCommons@USU
Mi Bee Lab
12-8-1978
Pollen Manipulation and Related Activities and Structures in Bees of the Family Apidae
Charles D. Michener University of Kansas
Mark L. Winston University of Kansas
Rudolf Jander University of Kansas
Follow this and additional works at: https://digitalcommons.usu.edu/bee_lab_mi
Part of the Entomology Commons
Recommended Citation Michener, Charles D.; Winston, Mark L.; and Jander, Rudolf, "Pollen Manipulation and Related Activities and Structures in Bees of the Family Apidae" (1978). Mi. Paper 101. https://digitalcommons.usu.edu/bee_lab_mi/101
This Article is brought to you for free and open access by the Bee Lab at DigitalCommons@USU. It has been accepted for inclusion in Mi by an authorized administrator of DigitalCommons@USU. For more information, please contact [email protected]. ..:.-~.:.:.: .:.:.:.:.:•:.:•:•:•:•: .-:•:•:•:•:•:•:•;•:•:.:.:.:.: ...•.•-·····~---~-~-~--·~-~.~-~-~-=-~-~-~-~-~.~-~-~-~-~.~-~.~.~ ..-·~·-·~·: ::·..~ .. :.- ..- : :: :: , THE UNIVERSITY OF KANSAS
SCIENCE BULLETINMU==> . COMP . ZOOL.
LI -J ~ARY DEC1 81978
t--tARVAr POLLEN MANIPULATION AND RELATED ACTIVITIES AND STRUCTURES IN BEES OF THE FAMILY APIDAE By Charles D. Michener , Mark L. ,vinston , and Rudolf Jander Vol. 51, No. 19, pp. 575-601 Decemb er 8, 1978 ANNOUNCEMENT The University of Kansas Science Bulletin ( continuation of the Kansas Uni versity Quarterly) is an outlet for scholarly scientific investigations carried out at the University of Kansas or by University facu!ty and students. Since its incep tion, volumes of the Bulletin have been variously issued as single bound volumes, as two or three multi-paper parts or as series of individual papers. Issuance is at irregular intervals, with each volume prior to volume 50 approximately 1000 pages in length. Effective with volume 50, page size has been enlarged, reducing the length of each volume to about 750 pages. The supply of all volumes of the Kansas University Quarterly is now ex hausted. However, most volumes of the University of Kansas Science Bulletin are still available and are offered, in exchange for similar publications, to learned societies, colleges and universities and other institutions, or may be purchased at $20.00 per volume. Where some of these volumes were issued in parts, individual parts are priced at the rate of 2 cents per page. Current policy, initiated with volume 46, is to issue individual papers as published. Such separata may be purchased individually at the rate of 3 cents per page, with a minimum charge of $1.00 per separate. Subscriptions for forthcoming volumes may be entered at the rate of $20.00 per volume. All communications regarding exchanges, sales and subscriptions should be addressed to the ExcI-IANGELIBRARIAN, UNIVERSITY OF KANSASLIBRARIES, LAWRENCE, KANSAS 66045. Reprints of individual papers for personal use by investigators are available gratis for most recent and many older issues of the Bulletin. Such requests should be directed to the author. The International Standard Serial Number of this publication 1s US ISSN 0022-8850. Editor Eugene C. Bovee Editorial Board William L. Bloom Philip W. Hedrick Rudolf Jander Harvey Lillywhite Charles D. Michener Norman A. Slade Henry D. Stone George W. Byers, Chairman THE U N IVERSITY OF KANSAS SCIENCE BULLET IN Vol. 51, No. 19, pp. 575-601 Decem her 8, l 9i8 Pollen Manipulation and Related Activities and Structures in Bees of the Family Apidac1 3 CHARLES D . Mr cHENER:!, • 1, MARK L. \Vr:--; sTo:--;:!, A:'\D RuooLF JANDER~' 3 CONTENTS ABSTRACT 575 J:S:TRODUCTIO:S: 576 :\LH ERI:\LS 577 ;\{ETHODS 57S TRA:S:SPORT OF NEsTI'.\'.G .\IATERI:\LS 579 MA,IPCLATIO:'\ OF ScE:'\Ts B Y :\hLE Ec GLoss1,1 5~0 APID PoLLE'.\'.J-. L,sIPULATIO'.\'...... Sb! The Forelegs ---·----·--·-· ...... Sb! Pollen Packing ····---·------·••· 5il2 Typ e I Pollen Packing --·-----·········· . ·····-·•·-----·--· 5S2 Typ e II Poll en Packing ·-··------·- ·······-··············•-·· 583 Ip ~ilateral Type II Poll en Packin g ...... ·········-· 5113 Contralateral Pollen Packing in A pis and Bom b11s --··----···· 587 Contralateral Pollen Packing in l\leliponinae ········-·······-····· 58S The Corbicular Pollen Loa ABSTRACT Poll en m anipul ation is descri bed for all major groups of Api 2 Department of Entomo logy, 3 Department of 1 Contribution number I 659 from the Department Systematics an INTRODUCTION also associated with the transport of ma terials used in nest construction. The purpose of this paper is to describe movements whereby pollen and other ma The family Apidae is divided into four terials are collected and placed for trans distinct groups, the subfamily Meliponinae port in the apid "pollen baskets" or cor ( the sister group to all the rest according biculae, the derivation of these movements to \Vinston and Michener, 1977), the tribes from self-grooming behavior, and the Euglossini and Bombini ( currently united meaning of these movements and related in the subfamily Bombinae), and the sub structures for apid evolution. Morpho family Apinae. In the past, pollen-collect logical and behavioral features for collect ing behavior has been well described only ing and transporting materials play a cru for the Apinae; progressively less was cial role in the evolution and adaptive known about this behavior in the Bombini, radiatio n of the bees ( superfamily Apoi Meliponinae and Euglossini (Maurizio, dea). These features are therefore im 1968). portant both for bee taxonomy and for While some female bees (Euryglossi evaluating homologies and convergence nae, Hylaeinae) transport pollen to the (Michener, 1944, 1974; Jander, 1976; Win nest exclusive ly in the crop, most carry at ston and Michener, 1977). While the least part of their pollen harvest with the adaptations for transport of nonliquid ma help of hairs which are located differently terials in most families of bees relate pri in the various taxonomic groups. Such marily to pollen, in the Apidae they are hairs form the scopa, a term applied to POLLEN i\L\:-.:IPL'LATIO I'-: BY BEES ( APIDAE) 577 poll en-carrying h:1irs whether they are on tion). The transfer of dirt (in grooming) the outer sides of th e hind tihi :1e and basi is consistently fron1 anterior to more pos tarsi, the und er sides of th e ba sal segments terior legs and it is commonly discarded of the hind legs, th e sides of th e propo from the hi n angle~. C.D.:'-.1. and :'-.f. D. Breed observed pollen terial in the corbiculae. Bassindale (1955) m anipulation by Trigona as well as leg movements of male euglossinc bees in sou thw c~tern Colombia; gi\'es an ~,ccount of prnpolis packing by Dr. Breed made moving pictures w hich were later T. braunsi and Sakag~1mi and Camargo analyzed hy C.T)) ,1., R.J. m:iclc moving pictures of ( 1964) illustrate and give an account of resin collecting by ,·/ pis m cllifaa and of pollen col lecting by Rombw pt·11n.,yfra11ims. The films were the similar packing of cerumen for trans analyzed with the invaluable help of :i Super 8 port by T. postica. Th e latter authors re Lafayette Analyzer Projector. port that the mandibles cut out particles Another source of information is pollen on the legs of bees killed while collecting pollen. The of cerumen which are manipulated and anatomical details and th e location of pollen ac p ressed together to form a lum p by the cumula tion help in Jctcrmining the packing behavio r. mandibles an d forelegs (probably basitarsi, T hese Jata, collectcJ by C.D.:'-.1. anJ .:'-.!.L.W. and recordcJ largely as sketches anJ notes maJc at C.D ..M.). "Wh en the lump atta ins an Kourou, French Guiana, usually sub stantiate the approp riate size, one of the middle legs behavioral data. reaches forward, and using bristles on its Vari ous bcl1:l\·ioral matt ers that arc not directly related to filling the corbiculac arc mcntioncJ in underside (b~1sitarsu s? C.D.M.), the bee passing. To save space, referen ces in such cases arc transfers the lump very rapidly to the not u sually included, but can be founJ in l\1ichcncr corbi cular surface of th e hind leg of the (1974). The terminology uscJ for the m ovement s in sam e side, which is synchron ous ly moved rnh·eJ in pollen manipulation is th at of Jandcr a little forward." By repetition of these (1 976). In rubbing, tw o parts mo\'e back and forth, mo\'ements, the accumulation on the cor one against the other, wit hou t losing contact through out the action. In scraping , ~trokes in one direction bicula grows. Often the middle leg is ex involve contact, but the parts are separ ated for the tended back, an d gently presses the grow return strokes. For hairy structures such as many ing ipsilateral ( same side) corbicular bees have, the se terms may not seem iJcal; w ords like brushing or combing are more Je scriptivc. deposit. On e or both of the hind legs may How ever, the y J o not inJicate th e di stinction Jrawn be rais ed in a pernliar way above the between rubbing and scraping anJ are a\'oidcd ex wings; th e function of this mo\'ement, if cept when our observati ons are not good enough to make that Jistinction . any, is unknown. Oricntati on al terms for the legs, especially for On e of us (R. J.) has observed collecting movement s rclati\'e to the hind tibi a, can be con fusing . In the tibia's usu.ii posi tion, upward might and transport of mud by ,\J elip ona fasciata. be regarded as towarJ the base. Became it is move Biting \\'ith th e mandibles and scratching able, that usage is arniJed, anJ we me instead basad, with the forelegs, the bee loosens a bit of towarJ the base or femoral artic ulati on, anJ apicad, toward the apex . Th ese are simply Jircctional terms; moist mud, which is then taken up by a basad mo\-cmcn t can occur near th e apex of the the m andib les. With a backward motion, tibia. Because the tibia is often extendeJ po~ter iorly , one foreleg takes the bit of mud from the 11plt'ard is taken to be towarJ the upper mar gin, i.e. , at right angles to th e long axis of the tibia. Th e mandib les. The ipsila teral middle leg same directi on relative to the tibia can be calleJ scrap es the piece of mud from th e foreleg posttirior if the long axis of the tibia is cons idered by clasping the fore leg from the outside . to be vertical. This mo\· ement appears identical to the cleaning of th e foreleg by the middle leg TRANSPORT OF dming a normal cleanin g bout (Jander, NEST ING MATERIALS 1976). The middle leg then passes the Species of A/ elipona are regularly seen mud back\\' ard and presses it from the out collecting mud and carrying it in th e cor side into the corbicula. This is usually biculae, an d M elipon a and Trigona both follo\\' ed by patting mu\'ements of the carry cerum en from other nests, as well middle leg on to the mud in the corbirnla. as gums and resins, in a similar way. Some Only unilat eral ( ont side at a tim e) trans species of Tri go 11a also carry vertebrat e fer of mud was obser\'cd ; mud may be fecal mat erial, mud, or chewed plant ma - passed back\\'~1rd on one side severa l times 580 Tm : UN IVER SITY oF KAN SAS Sr.IENCE BuLLETIN hcf ore the bee uses the oth er side. After above for Trigona postica except that the both corbirnlae have been filled, but before hind legs are not raised (Sladen, 19126; taking flight, the bee t:1kes a larg er lump Rilsch, 1927; Meyer, 1953, 1954, 1956; Saka of mud betwee n th e mandibl es, first hold g:1mi and C1margo, 1964; see IBRA, 1976). ing it with both forelegs and then pressing One of us (R.J.) made additional observa it to th e m andible s. tions of such behavior, summarized as Eugloss a and Euplwia G1rry resin in follows: The mandibles gnaw the surface the corbi cub and Eu lacma carries verte of the resin while at the same time the brate feca l mat erial as well as resin. Eu forelegs and occasionally the middle legs plusia som etim es carries small pieces of scrape :rnd scratch the surface. Loosened bark stuck to th e resin, if museum speci pieces are released by the mandibles to mens with such bark arc meaningful. One both forelegs. In a very quick movement, fem :ile of th e Euglossa cordata group was one foreleg swings backward and the ipsi observed taking cerumen from fragments Literal middle leg grasps the foreleg from of an abandoned Trigo11a nest. Few ob the outside and scrapes off the propolis. servations were possible, so that details The femorotibial joint of the middle leg are not available, but the bee clearly de is sh:uply bent in this operation (as when tached pieces of cerumen with its mandi the middle leg cleans the foreleg in groom bles, th en hovered and while in the air ing, Jander, 1976) and the piece of propolis transf erred the cerumen to the corbiculae. sticks to the inner posterior side of the mid The middle legs clearly were seen to syn basitarsus, presumably held by the sharp chronously carry pieces back to the ipsi bristles in that area. Immediately the mid bt eral hind legs which were brought for dle leg swings backward and the piece of ward, the middle legs then patting the propolis is pressed by the mid basitarsus cerumen into the corbicube. The bee onto the corbicula of the ipsilateral hind then alit to obtain more cerumen. The leg. Then the middle leg is pulled forward cerumen masses on the corbiculae became while still in contact with the hind leg; larg e and irr egular. Becaus ·e of behavior the resultant scr:1ping movement leaves in other groups as well as in euglossine the resin in the corbicula. The pressing males described below, we suspect that :111dscraping may then be repeated a sec euglossines may not always hover to trans ond time. All these movements were per fer sticky materials to the corbiculae syn funned while the bees were on the ground. chronously with both middle legs. Non flying M eliponinae and Apinae have been MANIPULATION OF SCENTS seen to transfer sticky materials to the BY MALE EUGLOSSINI corbicula e asynchronously, with one mid Male euglossine bees collect certain dle leg at a time. scented substances from orchid flowers Bombus constructs its nests using a mix and other sources (Vogel, 1966) and can ture of wax and pollen. \Vax is secreted be attracted to scents provided artificially by the bees. So far as known, pollen is (Dodson et al., 1969). Interestingly, the not tr ansported differently for construction movements involved are similar to those of than for food ; in fact young larvae may pollen collecting by females, although other eat some of th e \Vax-poll en mixture. Trans male bees, so far as known, use similar port of pollen is described below. movements only for self-grooming. Several Apis mellif era, both in Europe and the species were observed; the following ob Africanized bees in Brazil, collects resin serv :uions apply equally to Euglossa, Eu (propo lis) or ceru mcn exactly as described plusia, and Eulaema, and largely support PoLLE :--t ~lANIPL'LATION BY B1-.Es ( ,\PJDAE) 581 th e detailed :iccounts ;md illustrations by ,\PID POLLE;\ ~L\~IPULATIO:\ Vog el (1966) and Evoy and Jones (19il). T11E FoREt.E<,s. ,\s in most other bees, (Th e middle tibia cited by th e latter au the front legs (basitarsi) and the proboscis thor s appears to have been i11 rea lit y the remove pollen from the anthers of flowers; middl e hasitarsus.) pollen 011 the proboscis :1.nd the he:1.d is Th e front tarsi are rubb ed on the ma subsequently scr:1ped ofI hy the for\\'ard terial contai11i11g the attractant; sometimes ~ind downw;1rd movements of the front they also scrape the hea d, especially the legs. Thus for most species at most kinds eyes, downward and forward as in groom or' flowers, the front legs are particularly ing 1110\'ernents. The probosci s is 1~ot ex important, hcing the primary pollen gath ert ed; the tips of the antennae are dir ecte d ering structures. Pollen is generally trans down to or almost to the scent source. ferred hackw:1rd directlv from the front After suc h rubbing, the bee usu ally hovers legs to the miJJic legs ,(basitarsi), whic h and while in th e air the rniclclle legs move al;o clean pollen from both dorsal and forward svnchronouslv and :1ppar ently \'entr;tl surfaces of the thorax. However, scrape the 'fore tarsi \\:ith the under sur these movements in the mcliponine genus faces of the ipsibteral mid basit~1rsi. Th e Tri gona are unusu:1.L as described below. middl e leg is prohably flexed as in normal In flowers whose anthers are readily cleaning of the foreleg. Th e middle legs access ible, Trigona species common ly bit e now mov e back, and :1t the same time th e an ~1nther repeated ly, loosening pollen, the hind legs are synchronously flexed and ante nn ae being bent down, their tips con rot ate d forward. Now the und er surfaces tacting th e an th er or nearly so. If there is of the mid basitarsi scrape sync hronou sly alre~1dv loose pollen. biting is unnecessary. upward, i.e., at right angles to the long At lea~t T. capitata, pallida, tlzoracica, and axes, across the outer su rfa ces of th e en nigerrima, and presumably all species, ex larg ed hind tibi:1e. The two structures are tend the probo scis, contacting the anthers nearl y parallel to one another, an d only at repeatedly ( Fig. 1). The pollen !s pr~ the end of each stroke may the mid basi sum ablv made sticky with nect:1.r m this tarsus contact the hai rs of the dorsal m eta wav, a~ is th e case with .,..Jpis. As these tibial groove. The whole sequence is re acti\'ities continue, bees brush the anthers peat ed ... seve ral tim es whil e the bee is and especi;11ly th e pro boscis with the front ho vering , before it alights :1gain at th e ursi, presumably accum ulating pollen_ o_n source of the scent, or departs. The gr eate r the h:1irs of the hasitarsi. The proboscis 1s part of the conuct of the middle tarsus scrap ed downward, toward its apex, with with the hind tibia is with the simp le, bot h forelegs (basiursi?) synchronously. convex, short-haired, outer, tibial surface The pollen is th en transferred by the front and not with the groo\'e which is supposed tarsi to :111 area of back\\':1rcl-directed, stiff to absorb the :1ttractant substan ces. k1ir s on th e ventral surface of th e m ese Rar ely, instead of ho,· ering, a bee gr asps pisternu m in front of :rnd between th e middle coxac , and to sim ilar h:1.irs on th e the ed b(Te of a leaf with its m:rndibl es after rubbing an attractant with its front t:1.rsi, middle and hind cox:1e. These movements and th en, hanging hy the m:1.ndibles, it :ire s\'nchrnnous, left an d right forelegs movi;1g simu lt:111cously, the bee being sup baoes throu ba h the leab movem ents descrih ed above. It thus frees the middle and hind ported by the middle and hind legs (Fig. legs for the m ovements usually perform ed 2) . ~foYing pictures show that the fo re while hovering. tarsi arc scr;1pecl forward ;1cross the coxal 582 T1rn UNIVERSITY oF KANSAS Sc1ENCE B u LL ETIN ~,re forag ing at flowe rs with abundant loose po llen th at adh eres to th e body, it is no do ubt bru shed oft of th e diff erent part s of the bo dy by th e inn er surfac es of th e b:isitarsi of all th e legs . T. pectoralis and T . ( Paratrigona) sp. collectin g poll en on Baccharis ( ?) flowers h ad pollen den sely caked on th e inn er sides of th e hind basi ta rsi. Su ch pollen would never be trans ferred to th e thor acic vent er, but must pass dir ectly to th e corbicula. Thi s is in contr ast to our observations on T. pallida and nigerrima which h ad littl e pollen on the hind basit arsi. Po LLEN P AC KI NG. W e recogniz e two basic types of pollen packin g in Apida e. Typ e I resem bl es th e m anip ulati on of nest ma terials, as describ ed above, in that sticky m asses arc pla ced dir ectly ont o the corbicu lae by th e middl e legs. In T ype II, pollen is placed near th e distal end of the cor bicula rath er th an dir ectly on th e corbicular 2 sur foce, ;ind is th en pu shed b:1.s:1.dinto the corb icul a. Special m orpholog ical features of th e di stal end of th e hind tibia and base of th e basitarsus are necessary for Type II pollen m anipulati on; th ese fea tur es diff er am ong th e group s of Apidae, :1s do th e po llen-packin g m ove m ents . T here exists, th erefo re, various subt ypes Fies. I, 2. Tri go na nigcrrima taking pollen from of Ty pe II. H cdyc/1i11m. In Fig. 1, in an int erval between biting , th e tip of the glossa is in cont act with the pollen P ACK I 7"G. source and th e forelegs arc about to scrape the glossa TYPE I PoLLEN It ma y be that and floral surfa ce. In Fig. 2 the forelegs arc trans all fem ale apid s reta in, as part of their ferring pollen to the thor:i.cic vent er and mid and behavioral repert oire, pollen packing in hind coxae. Th ese dr awin gs :ire based on rnoYing picture fr:imes, but mu ch deta il has been :iddcd since whi ch the middl e legs pla ce pollen dir ectl y th ey showed mainl y silh ouett es. int o th e corbicula e, ::ts noted bel ow for Trigona ( Tri gonisca) buyssoni ;ind for T. and mesepistcrn~d vestit ure to transfe r pol amalth ea on Cucurbita. F or th e form er len to the latter. spec ies thi s m ay be th e prin cipal m eth od. The behavioral reperto ire of Tri gona It was observe d on th e sam e flowers wh ere species must include somew hat diff erent T . nigerrim a w as tr J nsferring po llen in moveme nts for flowers with loose po llen the w:1.y m ore comm on for th e Apid ae that does not need to be freed wit h th e (Ty pe II). Most species pro babl y use Typ e mandib les and that may be picke d up by I packin g onl y for large masses of sticky parts of the body ~rnd ap pendages other mJter ial such as resin, mud , or th e larg e than the fron t tar si. When Trigona species po llen m asses of Curnr bita. PoLLEl': MANIPCLATio:--r BY BLEs ( APIDAE) 583 Trigona (Trigonisca) buyssoni bites the An old report (l lofTer, 1SS2) says that anthers of Hedyclzium, moistens the loos pollen is "pressed with the middle legs ened pollen with nectar (?) from the into the corhicub of the hind leg" by glossa, and places it on the mesepisternum Bombw. This may indicate that Type I ,,,ith the fore tarsi, as described above behavior is part of the repertoire of Bom ( under The Forelegs) in greater detail for bw, as it is of other groups. other Trigona species. Then, while still Even in Apis mcllifera, when the cor standing on the anther or petal, supported bicular pollen loads become large, the bee by front and hind legs, the middle legs may pat them many times with the ipsi move pollen from the thoracic venter to later:d middle legs, probably to smooth the corbicube. (The front legs could and compact them. Some pollen may be theoretically play a role in removing pollen added directly to the corbicular pollen from the thoracic venter, but if so, it must masses in this way, although the quantity be one front leg at a time rather than appears to be small (Parker, 1926). synchronously since at least one front leg TYPE II PoLLE:\ P.\CKI:\G. This type of must support the body. Nothing of the pollen packing inrnlves ( a) placement of sort was seen.) :Movement of the middle sticky pollen in the region of the hind legs is synchronous, and the pollen is tibiotarsal joint by movements which de scraped or patted onto the ipsilateral cor pend upon the kind of hee and the kind of biculae by the middle basitarsi. flower, and (b), most characteristically, The abundant, coarse, sticky pollen of pushing the pollen basad up the outer sur Cucurbita probably presents special prob face of the tibia from the region of the lems or opportunities for bees collecting it. tibiotarsal joint. This movement of the Our observations were made on Trigona pollen places it into the smooth and nearly amaltlzea, a form very close to and perhaps hairless corbicub. This is unlike the filling conspecific with T. silvestriana whose Type of the scopa of other fomilies of bees, where II handling of Cassia pollen has been ob the dense scopal hairs prevent such a served by Wille (1963). But on Curnrbita, process, and it is unlike the Type I apid after several transfers of pollen (like those process in which pollen is placed directly described under The Forelegs) from the into the corbicula by the middle leg. As anthers or from thick accumulations on described below, in ,--ipis, Bombw, and the corolla beneath the anthers to the probably the Euglossini, pollen is pushed mesepistern um and adjacent coxae, the bee, basally into the tibial corbicub by the while still in the flower and now supported auricle, i.e., the broadened base of the by front and hind legs, uses the middle basitarsus. ,\s emphasized by Buttel basitarsi to remove pollen masses from the Reepen (1915), ;\faidl (1934), and Win mesepisternum and pat them onto the ipsi ston and Michener ( 1977), ~frliponinae lateral corbiculae. The result, after several hase no auricle. ,\ majm objective of this such movements, is large but loose and ir study, therefore, was to learn how ?\1eli regular pollen masses on the corbiculae. poninae fill their corhiculae. The movem ents are similar to those of lpsilateral Type II Pollen Packing. T. postica described by Sakagami and Corbicular loa / \ 's_- ~ ·:,:,:-:::::::... ;ff'/{_--~ % \~-~ 1;r~ ~ /{; \\~ ~1\ *-1\~~t f I ' \ ·~ ~ ;1,/"' '\~ 1-\ t1/'. \ ""\ -~ ~ g 'Ill ) ,J\ 10 /{ \\ \ :\ \\ -\ j\ f}~~d~\\~ ~r~1- 1t /,/. ~ U,'°\ I ,1'"" fr\~ 12 F1Gs. 8-13. Hind leg of worke r of T,i gona pallida. Fig. S. T1b1a and ba sitarsus . Fig. 9, Same with large pollen mas s. Fig. 10, Basitar sus showing ba sal stru cture. Fig. 11. Tibia and ba sitarsus in anteri or \'iew. Fig . 12, ,\ pex of tibia, outer \·iew. Fig. 13, Distal view of apex of tibia. p, peni cillum; r, rastellum; a, auricular area; c, corbicula. seemingly pr essed against the pollen ma ss and th en curve distad at th eir tips. Th e in the corbicub. Thi s could be either ad longest bristles are the oute rm ost while ju stm ent of the pollen mass or a Type I those nearer the corbi cula r sur face arc addition of pollen, but the arrangement of progressively shorter ( not true of all spe poll en from different sour ces suggests th e cies). The hind b~1sitarsus except at irs former ( see section on Th e Corbicular tib ial articulatio n is offset mesally (Fig. Pollen Load in l\I eliponinae). 11) , so that th ere is a gap between its outer Th e pertin ent hind tibial and basitarsal surface and the cur\'ed apiccs of the bristles structur es of Tri gona pallida arc shown of the pcnicil lum. The structure suggests in Figure s S to 13. The penicillum is a row that through this space the ipsilateral mid of stiff bristles arising on the lower ( or dle basitarsus is drawn at right angles to ante rior) di stal angle of th e tibi a. The y the long axis of the hind tibia in ord er to s\\'eep upw ard ( or posteriorly) parallel to transfer pollen onto the latter. The apices the tibial apex, but \\'ell separated from it, of the pcnicillar bristles would scrape the 586 T 1rn UNIVE RSIT Y oF KA NSAS S c 1ENCE BULL ETIN pollen out of the k1irs on th e out er surface Examin ation of freshly kill ed pollen col of the mid hasitarsus, and the penicillar lector s of A/ clipo11aand Trigona frequently curvature, during poste rior mov em ent of show poll en on the tibial apex in the space the hind tibia l apex reLitive to the mid betw een the inner side of the penicillum hasitarsus, wo uld force such poll en basad and th e out er sides of the basitarsus and onto the outer sur face of th e tibia. The rast ellum. In T. pallicla a small brush progressive short enin g of th e penicillar ( Figs. 10 and 11) arising along a curved bris tles from the out er to the inner ones line on the basal part of the auricular area, makes the comb obliqu e, tending to push behind the penicillum, can, with backward the po llen aga inst the ap ex of the corbicula. flexion of the basitarsus, push such pollen T hu s the m ovem ent of hind leg relative to basad onto the corbicular surface. It con the middl e could ~done be responsible for sists of weak hairs hardly able to move the pushin g som e pollen basad into the cor whole pollen mass, but its effectiveness bicula. with small amounts of pollen was shown Th e posterior basal area of the outer by drawing a detached middle tarsus with surfac e of the hind basitarsus is provided pollen on it through the gap described with hairs which are directed posteroba above (between the penicillum and the sally (Fig. 10), not apically like most other hind basitarsus) on the hind leg of a hairs. We speak of this part of the basitar freshly killed worker of T. pallida. Part of sus as the auricular area because of its the poll en was combed off as expected by location, comparable to that of the auricle the penicillum, but the quantity was not of oth er subfamilies of Apidae. The hairs enough for the curv~1ture of the penicillum of the auricular area often have some pol to push much pollen basad onto the cor len on them and presumably serve to bicula. The basitarsus was then flexed scrape pollen off of the inner hairs of backward, with the result that the brush the ipsilat eral middle basitarsus. Laterad moved the pollen onto the distal end of the movement of the hind basitarsus would corbicula where it remained, held by its pr ess th e mid tarsus between the penicil stickiness. In M elipona, as in most Tri lum and the posterior basitarsus itself (in go11a,there is no defined brush, but rather cluding the auricular area) as the hind a hairy auricular area (Fig. 14) with hairs tibial apex is moved backward relative to directed posterobasally, as described above. the mid leg. Posterior flexion of the hind The frequent presence of small amounts basitarsus would then help to push pollen of pollen on this area in Melipo11a and that com es off onto the auricular area up Trigona suggests its importance in push ont o th e corbicula, thanks to the direction ing pollen upward. Such movement is of the hair s in that area. possible because of the highly flexible tibio As repeated passages of the mid basi tarsal joint, activated at least in Apis ta rsus add pollen to the distal end of the (Snodgrass, 1956) by three muscles. ipsilate ral outer tibial surface ( supple A casual observer might suspect from mente d by the pollen from the inner sur the curvature of the penicillar bristles that face of th e contralat eral hind basitarsus, the penicillum must somehow function as see below ), th e add ed pollen must pile up a scoop, accumulating material on its con and be pu shed basally. Th e motive forces GlVe surface. Actually if our interpretation are pres uma bly th e scraping mov ement of is correct, the penicillum scrapes pollen the hind leg ~1long th e mid leg as already onto its convex surface. Like the rastellar described, supp leme nted by back and front bristl es in Apis, the penicillar bristles comb (up an d dow n) flexion s of the basitarsus. in the direction of the apices of the hairs PoLLEN MANJPULAT1ox BY BEES ( APIDAE) 587 from which they are removing pollen. cumulates 011 the posterior basal projection Such scraping movements can remove in (auricle) of the basitarsus or between the definite :1mounts of pollen because new :1 uricle and the rastellum. Then, by pos poll en easily pushes away that already terior ( upw:1rd) flexion of the basitarsus, pr esent. T his would not be the case for the ;1uriclc forces pollen basad onto the a scoop. corbicula. Repetition adds more and more Contralateral Pol/e17 Pacl(illg in 1\pis pollen at the distal end of the corbicula, and Bombus. Although it largely occurs forcing the first pollen collected toward while the bees are in flight, the process the base of the tibia and ultimately filling of pollen manipulation and packing onto the entire corbicula with pollen, which the corbic ula e for transpor t has been stud may also he p:1ttecl from the outside by ied repeatedly for _,,..J_pis( for references, see the ipsibteral mi While the bee is h:rnging from the pre ssed :1gainst one another as during Aowcr, the hind legs (probably under sides grooming; contact is in th e vicinity of th e of hasiursi hut possib ly the rastelb) sweep tibiours:il joint :incl we believe that the back :111d fort h ( scraping or rubbing, we rastell um of each tibia combs pollen from cannot say wh ich) , tr ansversely, sync h ro the underside of the opposite basitarsus . nouslv, removing pollen from th e abdom This would lead to deposition of pollen inal ~tern a. In each suc h movement , as on the auricle, which would push it into they approach th e midventral lin e of the the distal end of the corbicula, just as abdomen, the distal parts of the tarsi meet in Ap is. ;111di nterfcre with the process. The result Co11tralateralPollen Packing in Meli is :1 longitudinal, midventral line of pollen, po11i11ae.Although Meliponinae lack auri not swept up by the basitarsi or tibiae. cles, most of them have rastella; it was While still hanging from the flower, th erefore natural to look for contrabteral the bee performs at least five movements corbicular loading similar to that known with the middle legs. They were observed for Apis. In some species of Trigo11a, to occasiona lly groom a foreleg by flexing contrabtcra l pollen packing is of little so that th e foreleg is scraped simultane importance, at least at the flowers where ously by mid femur and basitarsus. (This we made our observa tions. Thus the hind familiar movement is pr esumably function basitarsi of Trigona 11igerrima and pallida less for pollen collecting in the context of are brought rather close to one another Cassia Rowers.) Midlegs more often scrape beneath the body during poll en loading, anteriorly on the dorsum of the thorax so th:1t the middle legs can be drawn across and posteriorly on the venter between fore the out er surfaces of th e hind legs, as de and mid coxae. The latter movement scribed in the section on Ip silat era l Type II should sweep Cassia pollen from the tho Pollen Packing. The hind legs may touch racic venter. The middle leg s also, one at one another and make some basad-distad a tim e, extend backward, are· appressed alternate pumping movements. Moving between the hind basitarsi, then pull for pictures of hovering T. nigerrima taken ward as the hind legs are straightened from th e rear show occasions when the back ward. This scrapes pollen from the inner apex of a hind tibia, bearing the middl e basit :.1rsus onto the under sides of comb or rastellum, combs downward over the hind basitarsi. ( B::isitarsi of pollen the inner surface of the opposite or contra collecting individuals of Bombus caye11- lateral basitarsus, followed by the same 11ensis had abundant pollen on their under rnovement of the op posite leg. This alter surfoc es.-C.D. M.) Finally, midlegs oc nating or pumping movement is probably casionally pat the corbicu lar pollen loads, not important for pollen manipulation in usually starting at the hasal part of a pol the c:1ses most intensively studied, Trigona len rn;1ss and working toward the distal pallida on Ipomoea and T. 11igerrima on part. Hedycl1iu111,for in these cases the pollen Bombus in flight were photographed was picked up as described above, exclu while cleaning the forelegs wit h the mid sive ly by the front legs; little or no pollen leg s and the midlegs with the hind legs got onto the abdomen or hind tarsi. More movements also seen while bees were hang over, pollen was rarely found on the inner ing from Rowers. In addition, film analysis sid es of the hind basitarsi in the pollen revealed pumping mov em ent s of the hind collecting bees, and when present, there legs like those describ ed for A pis. It was was but littl e. The pumping mov ements verified that the hind basitarsi are not in these instances were probably cleaning P oLLEN .MANIPULATION BY BEES ( APIDAE) 589 or self-g room ing activit y or stereotyped ac hasitarsi appe1red to he outside the hind tivity th at m ay ha\' e import ance in po llen tibia, which ( together with the presence manipulation at oth er kind s of f1owc rs. of a strong penicillum) suggests that ipsi Th ese species, however, lik e all oth ers in lateral packing also occurs. T. ( Paratri th e suhge nu s Tri gona and cert ain species gona) impunctata is too small for detailed of th e sub ge nu s T etragona, h ave a large observations while hoYering. hut it very sericeous area , not covered w ith br istles, rarely seems to transfer pollen while rest on th e inn er side of th e hind bas itars us. ing on a Hower. One mid tarsus at a time Su ch a basita rsus mu st be in efficient, com is extended posteriorly and pulled forward p:u ed to that of oth er Api cb e. in bru shing het ween the two hind basitarsi which are pollen fro m th e middl e tarsus and th ere held with their inner surfaces apposed. fo re in contralat eral pollen pa cking. T his movement was seen performed by Th e oth er subg enera of Tri go na, lik e two pollen-co llecting indivicl 11als. T he on ly other M eliponina e, hav e th e und ersid e of prob lem in its interpretation is that this th e hind basitarsus full y bristl ed and it is is a typ ic d :1poid cleaning movement. The among such form s th at contr abt eral po l observe r (C.D.M.) believed that the move len packin g is m ost lik ely to occur. A s m ent was po llen transferral, but recog noted above, much po llen was found on nized that it could have been merely clean th e inn er sides of th e hind basitarsi of ing of the midd le tarsus. It was followed T. ( Scapt otrigo na) pectoralis and T. ( Para by hovering :rnd pumping movements of trigona) species collecting pollen on Bac the hi nd legs, suggesting pollen packing. clwris (?). Th e species of Scaura di scussed Asi de from the bristles on the inner in th e nest section mu st depend largel y on side of th e h ind basitars us, an essentia l contralat eral pollen packing. Mor e sig stru ctur e for contralatera l po llen packing nificantl y, th ere was much pollen on th e is th e raste llum. T h is is a ro w of bristles inn er sid es of the hind basitarsi of T. (Fig. 13) along the inner m argin of the ( Ceplwlotri go na) capitata and T. ( Para apex of the posterior tibia . I n Apis and trigo na) impun ctata collectin g pollen fro m Bombus the raste llum functions to comb Stylosanth es, even th ough thi s is a sm all po llen off of the inner surface of the con flowered leg um e wh ose poJlen was being tralate ral hind basiursus, and it extends rem ove d fr om the flowers by th e fro nt ta rsi more or less the fu ll width of the tibi al onl y. Tbu s, unlik e Scaura and th e bees apex. In M eliponinae the row of bri stles on Bacclwri s, tho se on St ylosa11tl1es were is common ly sho rter, being large ly poste not getting pollen on th e hody. Pollen rior to th e basitarsa l articulation and pen i must have been activ ely tr ;msferr ed to th e cillum, but probab ly has the sam e functio n ; inn er sid es of th e hind basitarsi. the pumping motio n of the hi nd legs in volves com bing of hi nd basita rsi by the Behavi oral ob serYations, whil e no t de rastella. The result is :iccumubtion of po l cisive, ind icated th e sam e conclu sion. Po l len from the contralater:d hasitarsus on len collectors of T. capit ata, at least those the apex of the tibia lateral to the rastellum, with lar ge po llen loads, seem to place th e from which location hasitarsal moYements mid tarsi bet\\' een th e hind tarsi wh ile combined with the hairs of the au ricular ho verin g, after visits to one to seve ral are:1 can presum:1bly push pollen hasad flow ers. Th ereafte r di sta d-basad p umping onto the corbicula, as in the case of ipsi m ovem ent s of th e hind legs were con later:il pollen p:1cking . spicu ous, th e inn er apic es of the tibiae ap par ently scraping th e inn er sur faces of the Proh:1hly another significant function basitar si. Som etime s, howeve r, the mid of the rastell11m is as a fence to prevent 590 THE UNIVERS IT Y OF K ANSAS S CIENCE BULL ETI N the pollen that is tGmsfcrred to the outer surL1cc of the hind leg from being pushed onto the inner surbce in the course of the tarsa l mm·e ments. Even if, as in th e Trigona pallida and nigcrrim a which we studied. little or no pollen is comin g from the inner surfac es of the hind basitarsi, pollen arriving on the outer side of the tibiot:1rs:il area might leak th rough and be lost on the inn er surface of the hind tibi a in the absence of a rastellum . Tlz e Corbicu lar Loa d in Meliponina e. Since in both ipsibt eral and contrabter al pollen-packing movements, pollen is pbc ed at approxi mately the hind tib iotarsal joint , there m ust be a mechanism for forcing pollen basad from th at area int o th e cor bicub. T his mechanism has been described Fie. 1--1. Ap ex of tibia :iml b:i,e of b:isit:i.rsus of .Hclip o11a fa, ciat<1,wor ker , , howin g by shad ing posi above, and it is reassuring th at pollen load s tiom of di tkrcnt kind, of pollen. Abbrnia tions as collected from two or thr ee dift erent kind s (or Fi g,. K-13. N ute th e pollen on the :i.uri cubr :i.re:i. behin d the pcnici llum. of flowers by M elip ona fasciata, /I.I. favo sa, and T rigo na fimbri ata indi cate th e ex erect or rnrl ed hairs th at enclose a corbicu istence of such a mechani sm. Th e pollens br space, as in A.pis , Bombus, and A/ eli in all such loads are arr anged as though po na, but extends posteriorly from th e each new kind were added from the distal corbicubr surfa ce. Th e pollen ma ss moves end of the tibia (Fig . H). pushing pre partly out over these hair s (Fig. 9) and viously acquir ed pollen basacl over th e the stickin ess of th e pollen and th e liquid corbicular sur face. If successive kind s of i11corpor:1ted with it hold s th e pollen both pollen were patted onto th e surb ce of the tu the corbirnbr surface and to the pos load (T ype I) , th e clifterent kind s would terior fring e of hairs. Th e pollen mass constitute shells one over the oth er; thi s may be shap ed to some degree, or adju sted, is not the case. In other apid subfamili es by pattin g movements of the middl e legs. the auricle of th e b::isit::irsus plays :rn es Specialization s of Meliponine Poll en sent ial role in pushing pollen basad int o Collectin g . Th ere doubtl ess exist, within the corbicula but , as noted before, the the behavioral repertoir e of various meli Meliponinae have no ::rnricle. ponin e bees, man y modiGcation s of the As the pollen mass 0 11 th e meliponin e patterns described above. Will e (1963) corbirnb enbr ges, it is held :111cls upp orted describes how Tri gona f. fulviv entris, by the erect hairs of the lower ( or ante f uscipennis , and si/v estri,11111cut holes in rior) corbicubr margin , by the erect "outer the tubular ::inth crs of Ca.Hia bifiora and penicillum " or par apenicillu m of species extr act pollen from them with th e glossa. such as Tri gona pa/Iida, by the penicillum Laro ca (1970) indicat es th :u T. spinipes prope r, and by th e few curled hairs at th e and .mbn uda behave srnilarl y with similar distal end of the posterior margi n of th e tubular Tibou c/1ina anth ers, while T. fulvi tibi:1. T he lung upper (or poste rior) frin ge ventri s guiana e cuts th e tip s off of the of the curbicub in most Tri gona species, Tib ouchina anth ers and then exploit s them however, does not contain rnore or less in the same way. POLLE:-- ~L\:--.-IPl 'L:\TJO:--.- BY BEES ( :\PID.-\E) 591 Similar observations were made by described previously. C.D.M. at flowers of anot her mcbsto The subgenus Scctura of Trigo11a, how maccous shrub in Fr ench Guiana. The e\'er. appears to consist of specialized flowers :ire managed differently by differ gleaners ancl \'isitors to inllorcscences con ent meliponin e bees, as follows: ;,\frli pona sisting of rclati\'ely hro:i.cl surL1ces from pseu do cen tris curls the body o\·er the group \\'hich pollen can he S\\'cpt up. The se bees of :rnthers and buzz es, receiving th e pollen h:1\'e cxtr:iorclinarily large and hairy hind from th e tububr anthers on th e under side basitarsi (;ilso hairy middle b;1sitarsi), the of the body :1s described for other bees by principle subgeneric characteriqic. Laroca }.-fichener ( 1962) ;u1<.IWill e ( 1963). Tri :ll1d Lauer ( l 0iJ) clcscrihe pollen collccti ng gona impunctata chews the bas:d thick by T. (S.) latitarsis from the cylinclric:11 parts of the anthers open and extracts pol i nHoresccnces of Pip er ancl from the leaf len with the glossa ancl fore tarsi. T. surfaces beneath the flowers of A maran fulvivent/Jis gu ian ae chews off the attenu t/Jus. They describe the use of the hind ate distal part s ( one third to one half) of hasitarsi for S\\·eeping up pollen and the th e ant hers, thu s providing an ent rance rubbing of the hind legs against one an much larg er than the small apic:i.l pore, other. as described below for T. lo11gula. and re:i.ches in to extract pollen \\'i th the T. latitaniJ· is a minute bee. undoubtedly gloss:1. It th en scrap es pollen ofT of the dit1i.cult to observe. From its morphology, tongue wth sim ult aneous dist:i.cl move we assume that its hcha\'ior is similar to ments of th e front b:1sit;1rsi. A sing le bee th:1t of T. longula, a larger species for often cuts th e :1pices off of most or :ill of which we obtained fuller. although still the :i.nthers in :i. flower before going on to incomplete, information on pollen rnanipu another. btion. Trigona species sometimes are gleaners. Trigona longula \\ ':l.S \'isiti ng a brge picking up pollen from corolla surfaces Howerecl Cassia species. Only rarely do wh ere it falls following visits tu ;mthers these bees go to the anthers. \\'hen they by other insects. Alth ough the pollen on clo. they collect pollen :111clmanipubte it. the surfac es is usu:dly invisible. bees collect so far as we could sec. like other species larg e pollen load s on the corbicub e from of the genus. l'\carly all the pollen collect th ese sources. \ Ville ( 1963) described such ing was hy gleaning from petal. bud, or beh:1vior for T. jat y, nigra, and testL1cei leaf surfaces bclo\\' the anthers. (Flowers corm's on Cassia biflora. On :1nothcr Cas were \'isitcd by Xylocopa, EulL1cnza, Eu sia species C.D.M. and M.L.\V . observed glossa, and Centris, \\'hose buzzing rele:1ses that T. im pznzct ata visited on ly the ant hers, pollen. :i.sshown by \\ 'ille. 1963.) On such biting them to get pollen, and T. pallida surfaces. there was no noticeable deposit usually did the same thin g. The latter of pollen, hut i mli\'iduals of T. lo11gula species, however. was sometimes :i.lso :i. were ;1blc to fill their corbicube there. In gle:rner, going over th e corolla surface he orclin:i.ry w:ilking. T. longulct moves like low the anth ers with the :1ntcnn:1l tips any other Trigona species. On pollenifer clown to the surfac e, the glossa slightly ous surfaces. hcm-c\·er. the miclcllc and hind excertecl, and th e mandible s moving; th e legs arc spLiyccl out, the inner surfaces of fore tarsi swept up the pollen. especially the h;1sit;1rsiagainst the substrate. the hind by scr:iping clistacl on the gloss:i.. to \\'bich h;1sitarsi hcnt forward :ilmost :1t right it probably stuck beG1use of regurgitated ;111glesto the body even though the f cmora nect:ir. Th e rest of the rn;111ipulation for and tihi:1c ;uc clircctccl hack\\' :1rd. \\'ith both T. impzmctata ;me! T. pctllidct was ;1s the legs in this position. the hce shul11es 592 T1IE UN!\'ER:-- JTY OF KA NS. \ ~ S C I ENCE BL ' LLETIN along, dragging the dist:i.l part of th e abdo men. The tips of the anten nae are bent down to the surfocc: the fro nt leg s perform ordi11;uy walking moveme nt s. The b:i.si tarsi :i.nd :i.bdomen must pick up pollen from a rather broad sw:1th as the bee mo\'es along. 1\ collecti ng bee frequen tly stops, raises its :i.bdomen. ~111d while supported by fore and mid legs, scrap es backward over the abdomina l surface sync hron ously with the inner sur faces of th e hind b:i.sitarsi. Then, beneath th e abdo men, it rubs or scrapes the inner surfaces of these basitarsi against one another or th e apices of the tibiae in the course of pumping movements of the hind legs. Then th e hind legs are low ered for support :rnd the middle legs are brought back usmlly synchronously to th e Fie. 15. Hind tibia an d tars us of worker of Tri gona regions of the :1pices of the hind tibiae, (Scwira) longul,z (modified from Schwar z, 1948) . and the mid basita rsal regions are pulled of th e hind basitars us, both from the sub fonv:1rd across the out er surfaces of the strate and from th e abdomen, is combed apices of the ipsibteral hind tibiae. This off th;H surfac e by the contralateral rastel- seems to be :1 shorter mov ement than in 1um during the obsened pumping move T. pallida, perhaps beca use the hind legs ment s. Pollen should pile up outside of are not moveable, being used in support, the r:1stellum; freshly killed pollen col and is ob liqu e, not Jt right :rngles to the lectors had pollen betw een rastellar bristles tibia. ;mcl on the outer rast ellar surface. Then L ess often the bee hovers and transfers the concave aurirnlar area, functioning like po llen to the corbicula e by m e:1ns of leg the auricle in other groups of apids, pre movem ent s that :1ppear to be similar to sum :1bly push es poll en up into the cor those of other Tri gona spec ies. Details , birnh The unusual width of the hind how ever, could not be discerned. hasiLm i, their use and that of the abdo Microscopic examination showed that m en for pollen collecting, and their gen the und erside of the :1pic1l p:1rt of the er;il outer convexity so that ~1 concave abdo men ( metasomal sterna -+ and 5) has auricular ;1rea can be present, all suggest hairs which ;ue curved downward at their th;n contralateral pollen transfer from the apices :ind thus should re1 Gtted by Laroca and Lau er ( 1973) for Examination of vario us females o( T. latit ,1rsis. r\s this would involv e a for Euglo.Ha, E up lwia, ;111dEulae m a, ki lled ward mov ement of material, something whil e collecti ng pollen. provi ded more in not seen in grooming or pollen handling furm ;1tion. T he fwnt legs nften had a hch avio r of any other bees (lander, 1976, littl e pollen on the inner surfaces of the and in preparation), we suspect an obser basitarsi and less nn the tibiae, more on vational error. This would not be surpri s the posterior margins of the inner basitar ing considering the minuteness of T. sal sur Lices th;rn clsewhl'.fc. The midd le latitar si.f. legs had simibr ly distributed pollen, even Tli e Euglossini. Pollen collecting by more pr edomi na ntl y along the posterio r EugloHa near cordata ~rnd ign ita was ob margin s. Pr esum ably it is the hairs of the served on sm:dl tubular flowers ( Sabicea) posterior parts of the inner sur faces th at that served also as a nectar source. Hairs transf er much of th e pollen. On the hind on th e bas;il part of the proboscis pull legs, th ere is littl e or no pollen on th e in pollen out of the flower. Wh en a bee ner surfac es of the tibiae or tarsi, except rears back to withdr aw the long proboscis, th ~u when the pollen load is very large, a it places the front tarsi (basitarsi?) on thin and brok en Ltyer m ay be present on either side of th e proboscis, scraping distad the inn er surfac es. several times with the front legs, thus pre Wh enever the pollen lo;:id is o( moder suma bly remo ving pollen from the pro ate or small size. pollen on the tibiotarsal boscis. articular region is limit ed to th e ;:irea of Whil e thi s is going on the epipha rynx: th e rast ellum, and it is on th e out er side which is extr;10rdin arily long in Euglossini. of the row of bristles, not on the inn er side. is exerted and probably add s nectar to the On the corbirnla. any sm;dJ pollen load is pollen. The proboscis, used for this pur always imm ediately above th e base of th e pose in ot her api ds. is so long in Euglos basitarsus, as sho\\'n in Fi gure 16. It rnu st sini th:n it could scarcely h;:ive this func be pu shed up to this position by th e auricl e tion. at th e pusteriur base of th e basitarsus. Th e Th e bee th en takes wing and hovers, auri cle is pr esent, but unl ike th at of Apis or rarely grasps a leaf edge with its mandi and Bombu .f, it is close ttgainst th e tibial bles and hangs. In either case, th e middl e ~1pex. which is so shaped th at th e base of and hind legs are freed of their support th e b~1sitar sus rid es over th e con\'ex tibial fun ction, so as to ;:illow pollen manipula ;1pex ;1s th e tarsus is moved relativ e to th e tion, which is rapid an d difficult to observ e. tibia. \\ 'hen th e pollen hid is slightly Th e front legs come back sync hronous ly larg er. as in most specimens ex:1mined. th e and appa ren tly the tarsi are scraped by th e small fring e at th e anterior base of th e basitar si of the Aexed middl e legs. in th e hasitarsus seems to have played a role usua l way. The btter then moved back (Fi gs. 16 an d 17). Thi s fringe mu st push synchron ously to contact the hind legs. pollen :-irnong th e h;1irs of th e ante rior At least part of th e time the mid leg pat s corbicular fringe, for they arc sur round ed the corbirnl a and pollen load on the out by po llen. Thu s if the tarsus is flexed side of the ipsiL.iter:11hind tibia. Rubbing h:-ick ward. th e aur icle pu shes pollen b;1sad. of the inner sur faces of the hind legs one whil e if it is flexed forwa rd. th e anterior on th e other (pumping motion) was also bas;d fringe docs so. Th ere is ofte n some probably observed; certainly such motion s pollen on the outer side of th e basitarsus, were visible. but contact between th e two but thi s is discont inu ous with th at on th e hind legs could not be verified . corbicub except when the pollen load is T 111c UN J\'l·. lt~ nY o F K ANS. \ s Sc11. I\!c 1. lh ' LLLTI N carrying behavior, three successive basic pha ses c rn he recognized. If J;mde r ( 1C )7(i) is corre ct i 11 considering the crop :is the ;111cestctl pollcn -c:1rryi11gstruct ure, the pol len-collecting liel1;1\'ior of 1 l ylacus illus trat es the first ph:1se, utilizing st ructures ;111d lieh;1\'iors :ilready present in the Sphe cid:1e. / lyl,zclts carries pollen in the crop, collccti ng it hy brushing it off the :rnthers with the foreleg<; :md then scraping it off e;1ch foreleg with ;1 com b 011 the maxillary ga le:1 (Jand cr, ]97(J). There is :ilso th e pos sibilit y th:1t Hyla eus e;Hs pollen dir ectly. Moreover, Hylaeus c m scrap e the head with the forelegs and then c:it the pollen that w:1s lodged 011 the head. There arc no known f-Iylaeus movements for trans F1<,,. I(,, 17. Stru d urc~ of hind leg "f fe m ale of ferring pollen from th e thorax ur :1bdomen E11g/o,,·a ig11ita. Ii ,, .\ pc x of t1hia and ln ~c of ha 5i tu the mouth; such pollen is w:1sted or the l.lr~L1', ~h.,wing h>· ~hadi n_g loc.1tion of a ~mall pollen lo:1cl. 17, ILi t· o f h.1~it.1r~Lh " f ~ame , ~h"wing auricle smal l :1mounts that stick to the body sur at right :rnd · 111;1ll fri nge at up p: -r left that ma y pla >· Lices ;1re brushed off in a cell, if our knowl a ro k in mo 1 ing pollen. edge of the repertoire of mov ements is enormous. Th e mov ement of the :1Uricle complete (J:rncler, in pr epar:Hion). This over the hulb or convexity of the tibi:11 prototypic pollen collectin g of Hyla eus per a;~ex readily c1Uses this discontinuity. sists into the m;1jor evolutionar y lines of 1\ll this supports the observation that th e bees, even though most pollen trans the rullrn is :1pplied to the outer surfoce port in such bees is externa l. of th e hind leg ne;u the tibio-tarsal joint In the evol uti onary line th :1t led to the by th e middl e leg. Apparently the rastel bmilies 1\nthophoridae and Apiclac, the lum serves primarily as a fence to keep original ga le:11comb of the primitive bees th e pollen from "lcaki ng" through onto is repbced by ;1 stip ital comb (Schremmer, th e inner surfoce. In :rny event, the r:1stel J CU~S ID ERATIONS 011 the hind legs as in the majorit y of the E\'uu·nu:-.: oF PoLL Ei\ C o LLE CTI :-.:c .\ N il bees. In cutypical pollen g:nhering, pollen Tin >:Sl' valves sync hro 11ous movements of the legs workers mo\'e so rapidly th:1t details of of a g iven pair. Thu s the tw o middle legs thci r pollrn manipuLnion hJ.\'C eluded us. even in a Trigonisca resti ng on a petal ( Our ;Jttl'mpts at observation were mostly simult aneously pl:tce pollen on th e ipsi 111:1cle:It !lowers of 1\/imosa. where the bter :il corbicul:te. I 11 thi s respect :1picls M clipona rushes around through the sta difTer from at le:1st man y of th e non-apid mens, :di p:lrts of its hairy body being bees, whi ch tr a11s(er po llen hack to the dusted with pollen.) tibi al scopa hy movement of one leg at ;1 I 11the species o( Trigona th;1t we h;i\'e tim e. (Oh servations of 11011-apicls in f1ight studied most thoroughly, pollen is trans ;1re still needed; th eir leg movements may ferred to the corbicub hy drawing the th en be synchronous.) Movement of both pollen-hearing middle tarsus across the middl e and hi 11d legs appears to facilitate :tpex of the ipsilateral hind tihia ;ind base pollen tr ;rnsfcr to th e corbicub, m :1king it of the hasit:irsus in such a way that the e;isier to get th e mid b sit:u sus to the pollen appe;1rs to be picked up and pushed prop er position rebtiv e to th e hind leg. bas:1d onto the corbicula hy a structure Simult :ineous movement of both mid and found only in this subfamily, the penicil hind legs is di fficult, however, whil e the lum. This hehavior resembles eutypical bee uses at least one of these pa irs of legs movements in that pollen is tr;111sfcrred to for supp ort. Most :1pids solve thi s probl em the hind leg hy the ipsibteral middle leg. by hovering wh en empl oying derived but is derived in th;it it is not placed di (T ype II) pollen-p:1cking methods, so th at rectly into the corbicub. In the course of th e legs are freed for th e pollen-tr ansfer this movement pollen will :1lso he scraped ri ng movements. Such beh:ivior has seem off of the mid basitarsus by hairs of the ingly been tr ansmitt ed to males of the :1uricular :1rca of the ipsiLner;il hind basi Eugl ossini whi ch use similar movement s t;1rsus. for tr ansferring scent s to th e hind tibi ae Pollen also ;ippears to be transferred whil e hoverin g. Pr esum ably a bek1vioral from inner sides of hind h;1sitarsi to contr;i patt ern th at evolved among fem ales w;is btera l corbicube in mcliponines. They :1ctiv:ited in rn:1les, whi ch in any c:1se mu st scr;ipe pollrn from the ;1b F or m:111yo( th e commonest species of or bulla of th e tibial apex, so that pollen Trigo11L1,the impo rt :rnce of the inner side from th e contrala tera l basitarsus pr esum of the hinJ hasit;1rsus in transf erring pollen abl y could not be pu shed betwe en these seems reduced . In th e subgenus Tri gona leg seg m ent s onto th e auricl e, as in Apis s .. itr. wh ich includ es the spec ies that we ;rnd Bombw. M oreover, the rastellum do es ha ve stu di ed mo st carcf ull y an d in m any not proj ect m esally in such a way that it species of the subg enu s T et1"i1gona, th ere co uld eas ily comb the contr :1late ral basi is a b rgc, hairl ess, scri ceous area occ up ying tar sal hairs, as it does in th e oth er thr ee up to half of th e inner ba sitar sal sur face. :1picl groups. Poll en mu st th erefo re be 1\ ccumulations of po llen on this surfac e pb ced on the out side of the hind leg by haYe not been seen a nd are not comm on the middl e leg, pr esum ably in th e area of e\'en am ong the h:1irs on the rest of th e th e tibiotarsal joint; then it is probab ly surface. These gro up s are placed am ong push ed ba sad int o th e hug e corbicula by th e derived subgenera of Trigo11a. Th eir the auricl e and other ba sal marginal parts rather hairless abdome ns and perhap s th e of the ba sita rsus which slid e over the tibial nature of the flowers usua lly visited by bulb as th e ba sitar sus mov es. them may mak e brushing of th e ab dom en If Win ston and Mich ener (1977) have of littl e im porLrnce. Probabl y contralateral corr ectly pr esent ed the cladi stic relation po llen tran sfer is m ore important in oth er ship s among apid groups, the E uglo s subge nera in whic h th e und er surfac e of sina e mu st h ave lost th e contralatera l pol the hind basitars us is fully covered with len transfer and th e relativ ely large auricle bristles of stiff hairs. charact eristic of Bombw and .dpis. The F in ally for Meliponina e, the gleaning great enlarg em ent of th e tibia associated bees of th e subg enu s Scaura repr ese nt a with carr ying hug e loads of nesting ma noteworthy developm ent, with a concave te ri als and pollen m ay hav e h ad this effect auricular area or false auricl e in th e swol by mrrowing th e space betw een the auri cle len hind ba sitar sus for pu shing po llen up and th e swollen tibial apex. A lternatively, ward onto th e corbicub. The broad ba si th e Euglossinae might be, in features of tarsi as well as th e ab clomin ;1l h air s are poll en manipulati on ( as well as in solitary used to bru sh up poll en from flow er and or para social behavi or), primitiv e Apidae leaf surfac es. that , lik e certain African M eliponin ae, In th e remaini ng Apidae (Apinae :rnd hav e nev er evo lved contralatera l pollen the tribes Bombini and Euglossini of the transf er. Many mor e observations of pol Born bi na e) pollen manipulation is also of len-collect ing E ugl ossinae are needed to derived types, but partia lly diff erent fro m set tle this probl em. those of .1V1elipo nin ae. In th e Ap ina e and DERI\'ATIO'.\'" OF POLLEN" MANIPULATION Bombini, whose hind tibial and tar sal FRo.:--.1 SELF-Grwo.:--.11N"c M o,'DIE NTS. Pollen struc tur e is very similar, pollen fr om th e transfer fr om th e foreleg to th e middle leg inner surface of the hind basit arsus is is indi stingui shabl e from the correspond combed off by the contralat eral raste llum; ing cleaning mo\' em ent (Jand er, 1976) and it sticks to the tibial apex out side th e rast el is th erefore considered homologom to it. lum, and is then pushed upward onto th e Th e middl e leg , with th e poll en on its corbicub by the stro ngly deve loped auricle. und er and inn er side, swings backward in In th e Euglossini there is :1lso an auricle eutypic1l pollen transfer towards the ipsi alth ough it app ears sma ll beca use of the lateral hind leg. The poll en on the middle enormo usly expand ed hind tibia. The leg is th en pr essed into th e scopa, and auri cle is closely appressed to th e swelling simultan eo usly th e middle leg is pulled Po1.1.1.:-..-~L-\ NIPl'L:\TJOK BY BF.ES ( ,\PID.\E) 599 fonvard ~111db :1s~1drelati\' c to the hind 1\ft cr a bee has cleaned its middle legs tibia, so that the clistad-directcd hairs of hetwccn the: hind basitarsi, it regularly the scopa scrape the pollen oft the middle cleans the latter hy pressing them flat leg. This fin;d scopal loading movement ;1g;1i11st each other and then rubbing them is precisely the opposite oL and presumably \\'ith alternating longitudinal pumping dcri\'ccl from, normal self-cleaning hc mo\'Cmcnts. Duri 11g these movements the ha\'ior, during which bees typic1lly rcmo\'e t:-irsi :-ire in continuous contact. Dirt is dust from the outer side of the hind leg thereby pushed distac.1hcc1usc all bristles with a clistad scraping mo\'emen t of the :tncl h:-iirs on the inner sicks of the hind ipsilater ~d middl e leg. This cleaning move basitarsi point i 11 that direction. [This ment. howeve r, is used inside the nest for cleaning mo\·cment was observed in all of scraping the collected pollen out of the the 60 species of bees in sc\'en families scopa. (Our remarks abo\'e ignore the listed by J:rncler (1976) :rncl difTers from femora l, trochanteral, coxal, propocleal, and the corresponding cle:111ingmovement of sternal scopal areas of many colleticls, halic most other l-1ymenoptcra which clean the ticls, and and rcnids, partly for bck of data, hind b:1sit;1rsusby pulling it past the tibio but also because in the context of the Api tarsal joint of the contralateral hind leg dae it is the tibial scopa that is important.) (Farish, 1()72; Wagner, 1959). >:ormally Patting of the corbicular pollen load termin;d tibial spurs improv e the efficiency with the middl e basitarsus, presumably to of this movement and the so-c:-illedstrigil adjust and shape the pollen m::iss, is seen of the hind leg of sphccids is its morpho in most or a.II apids. This mo\' ement dif logic:1! concomitant.] In the pollen rnanip fers from normal cleaning of the outside ubtion of Meliponinae, Bombini and Api of the hind leg and probably from Type I nae an almost typical apoi middle legs, since hackw~ird scraping of d cr beinsammelnden Bienen. Jenaische the cox:le hy the mi Thi s p;tptT \\' ;JS m;ide pm ~ibk hy :\'SF grant 1249. DEB 73-0(,:-Sl'i. Field \\'ork in French Cui ana by EvoY. W. 1-I. AND B. P. Jo:-.:u;. 1971. Motor \I.L.\\ '. and C.D.\I. 1,·a, pm,ihlc thanks to faciliti es patte rn s of the male cuglossin e bees maintained 111·(; ard Ot1,. I >;11·idand (;rctchen Roubik. no keJ by floral fra g rances. Anim. and Penelope F. Ku kuk for work on the .-\ fricanized honeybee under the dirl'ction of 0. R. Ta ~·lor (L'nited BehaY .. 19:583-588. State, Department of .\ _1.;riculture. contract I 2-1-1- PMus11, D. J. 1972. Th e cYolutionary impli 7001-3(,3). Field work h)· C.l).\ I. in Colo mbi a wa, ca tion s of qualitatiY e variation in the possible tha nks t;i facilities and \Thi ck pro\'ided by grooming behaYior of the H yme nop the Smithsonian T ropict! Research ln ,titute J.nd tera. ,\nim. DehaY .. thanks to the courto 1· of it, reproentati\'t :s in Cali, 20:662-676. ColombiJ, Dr. Reinaldo ])i az and Dr. \I. j. \Ve, t I-l oFFLR , E. 1882. Di e I-fonrn1 eln Steier Eberhard. nr. \I. I>. Breed photographed polle n marks, 1. I-Lilfte. 31. Jahresh eri cht Jer collecting bee, in Cnlomhi :1, using equipmen t kind ly La nJ es-Oberrealschule in Graz. 92 pp., len t h~· Dr. \\'c,t EberhJ.rd. The work of Dr. Rreecl pis. A, L a nd 2. and C.D.\!. in Colombia J.nd of C. 1).\1. in French Guiana was facilitJ.ted by funclin-g- through th e Uni- IBRA . 1976. Bibliography N'o. 16. Anno 1·er~it)· of Kansa, Endowment .-\ ssociJ.tion. R.J.\ tat ed bibliography on propolis. Inter work in Cm.tJ. R1u WJ.S made J1'.>s,ihlc by th e Or national Bee Resea rch Association. ganization for Tr "pictl Stud ic,. and hi, \\'or k in JANDLR, R. 1976. Grooming anJ pollen \la!J. y, ia by gr.i nt B~IS 7-1-131/,'I from the i'\J.tionJ.I Science Foun l'Abeille, I{. Cknn·in Ed .. Tome 111. arc below the wet season water t.1ble. Par is: :\la sson. Biotropica . .\IEYLR, \\'. 19';_). lkohachtungcn un