The biology of Diadasina distincta (Holmberg, 1903) (Hymenoptera: Anthophoridae)

8 Pages·1994·2.7 MB·

by  R P Martins

PROr. ENTOMOL. SOC. WASH.
96(3), 1994, pp. 553-560
THE BIOLOGY OF DIADASINA DISTINCTA (HOLMBERG, 1903)
(HYMENOPTERA: ANTHOPHORIDAE)
RoGERio Parentoni Martins and Yasmine Antonini
Laboratorio de Ecologia e Comportamento de Insetos Depto. Biologia Geral, ICB-
UFMG. Cx. Postal 2486. 30161-900, Belo Horizonte, MG, Brazil.
Abstract.—Diadasina distincta Holmberg is a solitary, multivoltine. and philopatric
anthophoridbeethatnestedinaggregationsonavegetation-freedirtroadindirectsunlight.
Nests were provisioned by each female, and there were no burrow connections. Despite
that 63 different pollen types had been found in the nests, D. distincta is oligolectic
preferring Ludwigia suffruticosa L. (Onagraceae). Mating was promiscuous with males
strongly competing for mates. Adults appeared sensitive to low light intensity, lowering
nesting and mating activities during cloudy days. Females remained in their burrows at
night and males probably in the surrounding vegetation. Nests were heavily parasitized
by an Anthrax sp. (Bombyliidae) and by Leucospis genalis Boucek (Leucospidae). Both
bee and parasite prepupae ofthe previous generation remained dormant in the cells from
October to February.
Key Words: Diadasinadistincta, Anthophoridae, nestingbehavior, life-cycle, parasitism,
Anthrax. Bombyliidae
The biology and behavior ofneotropical Material and Methods
solitaryand presocial beesarepoorlyknown. Observations were made at the Esta(;ao
In Brazil, there have been some studies on EcologicadoCampusdaUniversidade Fed-
biocoenotic structure(Sakagami et al. 1967, eral de Minas Gerais (UFMG). Belo Hori-
Laroca et al. 1982), some aspects ofnesting zonte, Minas Gerais, Brazil, from January
biologyofseveral species(von Ihering 1905), 1991 toOctober 1992. Theindividual'sdai-
habitat selection by gregarious nestcrs of ly and monthly activities at the aggregation
earthen banks (Michcncr et al. 1958) and were ascertained by observation two days
theirnestingbehavior(Michcncrand Langc per week from March to August 1991, to-
1957, 1958 a, b, c, d, e: Michcncr and Sea- taling 180 h. In the following months the
bra 1959; Sakagami and Moure 1967; Sak- observationsweredoneopportunisticallyin
agami et al. 1967). Except for the study on the higheractivityperiod between 1 1:00and
nest spatial distribution patterns (Martins 15:00 h, for a total of 150 h.
and Figueira 1992) and the survey record Theaveragemonthlytemperatureandto-
in the state ofRio Grande do Sul (Wittman tal monthly rainfall records were from the
and Hoffman 1990), there arc no studies on Esta(;ao Climatologica do Ministerio da
the biology and nesting behavior ofD. dis- Agricultura in Belo Horizonte.
tincta. We present here data on the biology Twentynestswereexcavatedtoverifythe
and associate organisms ofD. distincta. existenceofcell orburrowconnections. Sol-
PROCEEDINGS OFTHE ENTOMOLOGICAL SOCIETY OF WASHINGTON
554
o^ofe
Fig. 1. Turret morphology ofD. distincta.
ubility degree for water was determined
submerging two cells in water for 10 min cm
I
and two others for 24 h. The nest depths Fig. 2. Nest structure of£>. distincla showing ver-
and turret sizes were measured with a cal- tical cell and nest entrance closures, (natural size, no
iper. We excavated three 20 cm- samples turret).
from which we obtained 186 cells that were
measured and kept in vials at 25°C and 12
hours photoperiod until bee or parasite excavated with the mandibles and the soil
emergence. Averages and standard devia- pushed out by hind legs and abdominal
tions were calculated for all the measure- movements. The soil pellets were then
ments. pushed out by the hind legs and remained
In September, November, and January scattered around the nests entrances (Fig.
we excavated 10 nests each month to as- 1). Each bee built also a curved turret at the
certainthedevelopmentalstagesofbothbees nest entrance (Fig. 1). These turrets were
and parasites. We opened 100 cells in the frequently rebuilt when destroyed by rain-
laboratorytoanalyzetheircontents,andob- fall or when bees were excavating or pro-
served the post-emergence behavior of 10 visioning the burrow. Complete nests (n =
males and 20 females of D. distincta. 15) were 4.0 ± 0.3 cm vertical one-celled
Voucher specimens of D. distincta are de- burrows with a 0.9 ± 0.2 cm in diameter
posited in the Snow Hall Museum of En- (Fig. 2). Cells (n = 64) were 0.79 ± 0.07 cm
tomology of the University of Kansas and in diameter and 1.05 ± 0.07 cm in length.
in the Laboratorio de Ecologia e Compor- We estimated that there were 12,400 nests
tamento de Insetos do Depto. de Biologia in the 4000 cm- aggregation, based on the
Geral da Universidade Federal de Minas 186 cellswe found in oursample of60 cm-.
Gerais (UFMG), in which are also depos- This aggregation has persisted for at least 5
ited voucher specimens of associated in- successive years at the same site. H. R. Pi-
sects. menta (pers. comm.) first recorded the ag-
gregation in 1988, 1989. 1990. Burrow and
Results
cells were molded by the female pygidium
Nesting behavior which tamped the soil. Thus both burrow
The nesting site was in a vegetation-free and cells walls become very smooth. The
dirt road in direct sunlight. Females exca- larvae lined the cell walls with a layer of
vated their isolated nests in dry and com- fecal material overlapped by a thin cello-
pactsandysoil, softenedwithwaterbrought phane-like layer ofnetlike silk threads that
to the nests in their crops. The nests were waterproofed the cells. Both cells sub-
VOLUME NUMBER
96, 3 555
mersed in water for 10 min and 24 hours days). They were very sensitive to low light
showed no water infiltration in the larval intensity. The patrolling flights quickly de-
chamber. Although the nests were located clined and stopped on cloudy days, orwhen
side by side and distributed in a regular the shadow of the observer fell on the ag-
fashion (Martins and Figueira 1992). there gregation. Then the males flew to the veg-
were mistakes by bees when they returned etation, where they waited for bright sun.
from provisioningtrips, i.e. beesentered the Mating in D. distiucta seems to be pro-
wrong nests. However, shortly afterward, miscuous. Several males strongly compete
such bees returned to their own nests. De- for females. Newly emerged femalesand fe-
spite nest proximity there were no intemest males provisioning nests were equally pur-
connections. sued. Males mounted the female's back,
Nests were individually provisioned by gripped her neck with his forelegs and her
females with pollen of 63 plant species. body with his middle and hind legs. Cop-
Nevertheless, only 5 types had frequencies ulation lasted 38.06 ± 2.97 seconds (n =
higherthan 5%inthesamples.Amongthese 30).
types, Ludwigia suffrulicosa L. (Onagra- Most ofthe copulations occurred at nest
ceae) had frequencies from 60% to 99%. entrances, and once a female was gripped
Therefore, D. distiucta is an oligolectic bee by a male, two to five males may try to
preferring L. suffrulicosa (Martins and Bor- dislodge the successful one. The group may
ges, manuscript in preparation). roll on the ground like a ball. In the labo-
The pollen, carried on rigid hairs of the ratory, newly emerged males and females
scopa ofhind legs, was molded into a ball opened their own cell closures with their
in each cell, then an egg was laid beneath it mandibles, emerging head first. A minute
on the floor ofthe cell. Eggs measured 1,86 later they initiated copulation.
±0.14 mm (n = 3) in length. Bees filled
onlyonefourth oftheburrowbeforeclosing Nesting dynamics and life-cycle
nest entrances with moist soil (Fig. 2). The adult bees emerged at the end ofthe
Sometimes, when turrets had not been de- rainy season, and the nesting activities oc-
stroyed by winds or rainfall, females closed curredduringFebruarytoOctober,thedrier
the nests keeping the turret intact. Some and colder months ofthe year (Figs. 3, 4).
females were observed initiating another The reproductive peak was in June, and no
nest shortly after completion of previous adults could be found from Octoberto Feb-
nests. The prepupae ofZ). distiucta are flac- ruary, when thebeesandtheirsynchronized
cid, the larvae do not spin cocoons, and the parasite prepupae are dormant in the cells.
cells are very strong. This oligolectic bee's life-cycle seems to be
The egg-adult development time ob- related to pollen availability, because its re-
served in the field was 27,29 ± 2,24 days productive period is synchronized with the
(n = 25). and there were at least three over- flowering period ofits preferred host plant
lapping generations between February and L. suffniticosa (as suggested by pollen anal-
October in this philopatric species, ysis of cell contents). This plant is very
abundant in a swamp within 50 m of the
Male patrolling and mating
nesting site.
Femalessleptinnestsandmalesprobably
Aggregation dynamics
sleptonthesurroundingvegetationonwhich
they also rested between patrolling periods. The aggregation-growing dynamics are
The males patrolled the aggregation on sun- quite interesting. Because nests were quick-
nydays between 8:00 and 15:00 h. the peak ly built in close proximityand regularly dis-
of activity being around 11:00 h (n = 15 tributed (Martins and Figueira 1992), the
556 PROCEEDINGS OFTHE ENTOMOLOGICALSOCIETY OF WASHINGTON
m
m
o
f
r
a
I
n
JASO1 N1 D1 JrFM
A M J J A S O
Months of Years
1891 1992
Activity -+- Rainfall
Fig.3. D, distincta's activity level and
rainfall variation at the Ecological
Station of UFMG/BH/MG/BRAZIL.
Fig, 3. ActivitylevelofD, distmaaandrainfallvariationattheEcologicalStationofUFMGBH/MG/Brazil.
aggregation seemed to run in a "nesting gregations ofdifferent size, it was impossi-
wave,"changingthedirectionwhenthenests ble to relate the level of parasitism to ag-
were been constructed after rains or when gregations ofdifferent sizes. However, in a
it met vegetation. Around 400 m from the somewhat isolated part ofthe aggregation,
principal aggregation, there were 12 newly we observed that 25 out of 30 nests were
initiated sites, including from one to 200 parasitizedhyAnthrax Thisevidencefor
sx>-
nests, but these aggregations nevergrew like the high intensity of parasitism is circum-
the principal one. In the 600 ha of the stancial.
UFMG
Campus only the principal aggre- The parasitizing behavior ofAnthrax sp.
gation persisted forseveral years. Patrolling is very similar to that ofother bombyliids
males were less present in the incipient ag- that parasitizesolitarybees. Femalefliespa-
gregations, and the primary parasite. An- trolled the aggregation, hovered around
thrax sp., was frequently found flying or nests, and twisted their abdomens to ovi-
perching near them. posit into the nest entrances. We do not
know how the bombyliid larvae enter the
beecells. However, it seems like they would
Mortality factors and havetoreachthecellbeforethebeesecurely
associate insects
closes the nest.
Diadasina distincta was heavily parasit- Leucospisgenalis(Leucospidae) isanoth-
ized by an Anthrax sp. (Bombyliidae). The er important parasite, but it seems to have
rate of parasitism could be calculated be- a lower impact on the bee population than
cause the Bombyliidae puparia remains in the bombyliid. The behavior ofL. genalis
the burrow after the flies emerge. Because was different from that of the bombyliid.
we were unable to distinguish among ag- Females flewaround the nests, landingnear
VOLUME 96. NUMBER 3 557
FMAMJJASONDJFMAMJJAS
1991 Months o( Years 1892
Activity Temperature
Fig.4. D.. distincta's activity level and
temperature variation at the Ecological
Station of UFMG/BH/MG/BRAZIL.
Fig. 4. Activity level ofD. dislincla and temperature variation at the Ecological Station ofUFMG/BH/
Brazil.
the tuirets. Then they entered the nests gation, but we do not know the kind and
through turret entrances while bees were degree of interference they had on the ag-
away on provisioning trips. Sometimes a gregation dynamics. Occasionally the nests
beereturnedtothenestandmettheparasite (n = 4) ofanother anthophorid bee, Ptilo-
inside. Then it was aggressively pushed out thrix phimata Sm., were taken over by D.
by the bee. We observed two other bom- dislincla.
byliidspeciesrarelyvisitingtheaggregation; Tiypoxylum aiireoveslilum Taschenberg
however, we reared onlyAnthraxsp. and L. (Sphecidae) was commonly seen flying and
genalis in the laboratory. entering into the nests of D. dislincla. An
Aggregation-associated ant lion larvae individualofT. aureoveslilwnenteredmany
caused no detectable interference with the timesone provisioningnest, pickinguppol-
bees. Unusually heavy and persistent rains len with the mandibles scattering it around
during January 1992 (513 mm) eroded the the nest. We excavated anestwith entrance
nesting site and exposed cells. Consequent- closed with mud quitedifferent from D. dis-
ly, some D. dislincladormant prepupaeand tincta's closures. Inside we found two spi-
nest parasites were killed by desiccation or dersand an egg. Thustheevidenceis strong
by patrolling ants such as Solenopsis sp. that T. aiireovestiliim can use D. dislincla
An unidentified Cremalogastersp. invad- nests. However, we do not know its impact
ed some nests (n = 5) and probably killed on the nesting dynamics ofD. dislincla.
the bee or parasite larvae. A predator A leaf-cutting bee Megachile neoxan-
Apiomerus sp. (Reduviidae) was observed ihoplera Cockerell female built a cell inside
picking up a female at the nest entrance (H. a D. dislincla burrow. However, this is only
R. Pimenta, pers. comm.). Termites built a general association between them because
their subterranean nests below the aggre- this leaf-cutter bee also uses other Antho-
558 PROCEEDINGSOFTHE ENTOMOLOGICALSOCIETY OFWASHINGTON
phoridae and termite burrows, as well crev- cells with fecal material, the larvae do not
ices in the soil, for nesting. spin cocoons.
The rebuilding ofthe turret after damage
Discussion and Conclusions is variable among solitary bees (Stephen et
Diadasina Moure has been placed as a al. 1969). While building the burrow and
subgenus of Diadasia (Michener 1979). provisioningthenestD. distinctaalwaysre-
Specimens we sent to Michener were iden- builds the turret after any kind ofdamage.
tified as Diadasina distincta by A. Roig-Al- However, like otherauthors (see Stephen et
sina. Thus Roig-Alsina and Michener be- al. 1969), we do not know the turret's func-
lieve that Diadasina is a valid genus. tion. Theturret didnotprevent intensepar-
However, we found only reference to Dia- asitism by Anthrax sp. in D. distincta, nor
dasia in the Roubik's (1989) book about parasitismbyL.genalis(firstrecordforDia-
tropical bees. dasina) whose females enter the nests
Becausetherearenoavailabledataon the throughout turrets. Otherwise. Diadasia bi-
biology ofDiadasina species, except for the tubercidata (Cresson) is heavily parasitized
records of D. distincta, D. riparia (Ducke) by the bombyliid Villa sp. (Linsley 1958)
andDiadasinasp. in thestateofRioGrande and there is no record ofLeucospis parasit-
do Sul (Wittmann and Hoffman 1990) and izing species ofDiadasia (see Stephen et al.
on the spatial distribution ofnests (Martins 1969, Roubik 1989).
and Figueira 1992), it is useful to compare Apiomerus reduviids are well-known
the biology ofD. distincta with related spe- predators of solitary bees on flowers (Lin-
cies of Diadasia for which there are some sley 1958, Roubik 1989). Predation at nest
data. entrances has also been observed (Roubik
Nests in D. distincta, as in most Em- 1989). Linsley (1958) observed mutillids
phorini (C. D. Michener, pers. comm.) are parasitizing species of Diadasia. Although
usually one-celled, and the burrows are several specieswere presentatthe study site
shorterthan those ofDiadasia species. Spe- we did not observe parasitizing nests.
cies ofthe latter build linear series ofcells Becauseno incipientaggregation hadbeen
in burrows branching from the terminus of enlarged and males rarely patrolled them,
the main burrow (Stephen et al. 1969). we supposed that theaggregation formation
There is high variability in the shape of and persistence could be dependent on the
the turrets among species ofDiadasia (Ste- frequencyofpatrollingmales. In spitemales
phenetal. 1969). The turrets ofD. distincta do not patrol aggressively the aggregation,
differ in details from those ofDiadasia spe- the presence ofa numberofmales flying on
cies, which are vertically or horizontally the aggregation could prevent flies from ap-
straight in D. olivacea Cockerell and D. en- proaching the provisioning nest; thus re-
avata (Cresson), compared with smoothly ducing parasitism success (see Martins and
curved in D. distincta. Like D. distincta, Figueira 1992). From this hypothesis we
among the species ofDiadasia only D. con- would expect to find an inversely-density
sociata Timberlake does not build a tu- dependent parasitism. However, more data
mulus around the turrets (Stephen et al. are called for in our attempt to understand
1969). Like in other ground-nesting bees the parasite functional response.
(Stephen et al. 1969), a secreted thread-like Males of some bees wait to mate with
material is used to waterproof cells. Most femaleson individual flowers. Oneexample
Emphorini larvaeon maturitylinetheircells of this for Diadasia australis (Cresson) on
with alayeroffecal materialandspin avery cactus flowers was recorded by R. Brooks
thin cocoon (C. D. Michener, pers. comm.). (Eickwort and Ginsberg 1980).
Despite that D. distincta females lined their As in other bee species (Eickwort and
VOLUME 96. NUMBER 3 559
Ginsberg 1980), in D. distincta the lack of strained, probably by lack ofpreferred food
male aggressivity in patrolling the aggrega- (Martins and Borges, manuscript in prepa-
tions allow other males to do the same. As ration). Consideringthat both speciescould
in other bees, such as Diadasia hnconis be exposed to similarselectivepressuresand
Cockerell (Ordway 1987), when a receptive that the reproduction in D. distincta is sea-
femaleofD. distincta isencounteredby one sonal, it is unknown why Diadasia species
male, other males will also attempt to cop- evolved univoltinism and D. distincta mul-
ulate with her, resulting in a tumbling mass tivoltinism. Lowering ofthe egg-adult time
ofmales surrounding the female. However, ofdevelopment Diadasia spp. would select
copulation in D. rinconis lasts several min- formultivoltism. Usingthe samereasoning,
utes, whereas in D. distincta it lasts only a Matthews (1991) asked why among tem-
few seconds. perate-zone sphecid wasps the most wide-
The life history and nesting biology ofZ). spread condition isunivoltinism, sincethere
distincta is similar in some respects to Dia- is ample time for the development ofa sec-
dasia rinconis that nests in Arizona and ond and even a third generation.
Texas (Ordway 1987). Both species nest in The evidence for dormancy in the trop-
dense aggregations. The egg-adult period ical Hymenoptera is very scant (Denlinger
lasts about 20 days in D. rinconis and 27 1986). The only case cited by Denlinger
daysinD. distincta. Adultsemerge from the (1986) in the Aculeata is for the anthopho-
ground and mate from 09:00 h to noon in rid Epicharis zonata Lepetelier. We suspect
the same dry months ofthe year (see Figs. thatdormancy in the tropical Hymenoptera
3 and 4 in this study and fig. 2 in Ordway is more common than previously supposed
1987). But, ". . . it is unlikely that females because we haveevidence forprepupal dor-
ofD. rinconis could nest in hard dry soil in mancy in three sphecid and two other sol-
late MayandJunebecausenewlyconstruct- itary ground nesting bees (Martins 1993,
edcellswereconsistentlyfoundin moistsoil Martins and Almeida, manuscript in prep-
(Ordway 1987)." Conversely, females ofD. aration).
distincta nested actively in the hard dry soil
becausetheysofteneditwith watergathered
Acknowledgments
in their crops. Moreover, some D. rinconis
pupae and adults apparently remained dor- C. D. Michener and M. L. T. Buschini
mant in their cells until the next year, sim- criticized an earlier version of the manu-
ilar to the prepupae ofprevious generation script. We thank two anonymous reviewers
ofD. distincta. and in the univoltine Dia- for their useful comments. A. Roig-Alsina
dasia opuntia Cockerell (Ordway 1987). identified D. distincta, C. D. Michener P.
These species became dormant in different plumata, Z. Boucek L. genalis and N. Ev-
stages of development by using the same enhuis Anthrax sp.; S. T. P. Amarante T.
strategy to circumvent unfavorable weather aureovestitum and Pe. J. S. Moure M. ne-
or food scarcity period. Nevertheless, it is oxanthoptera. H. F. Leitao Filho identified
more likely that food scarcity isthe primary L. suffruticosa. We thank zoologist Myrian
factor preventing all year reproduction in M. Duartc for the drawings. H. R. Pimenta
D. distincta because it is an oligoletic spe- helped with field observations. This re-
cies. search was supported in part by a CPq-
UFMG
Except for Diadasia afflicta (Cresson), Grant 23072022065/91 to R. P.
which is bivoltine, D. rinconis and other Martins. We also thank the Brazilian CNPq
known species of Diadasia are univoltine. for the research scholarships. This paper is
Although D. distincta was found to be mul- a contribution ofthe Graduate Program in
tivoltine, itsreproduction isseasonally con- Ecology, Conservation and Wild Life Man-
560 PROCEEDINGS OFTHE ENTOMOLOGICAL SOCIETY OFWASHINGTON
agementoftheUniversidadeFederaldeMi- . 1958b. ObservationsonthebehaviorofBra-
nas Gerais, Belo Horizonte, Brazil. zilian halictid bees. II. Paraxystoglossa jocasta.
Journal ofthe Kansas Entomological Society 31:
129-138.
References . 1958c. ObservationsonthebehaviorofBra-
zilian halictid bees. III. UniversityofKansasSci-
Bortoli,C.deandS. Laroca. 1990. Estudobiocenoti- ence Bulletin 39: 473-505.
co em Apoidea (Hymenoptera) de uma area res- I958d. ObservationsonthebehaviorofBra-
.
tritaem Sao Jose dos Pinhais (PR, Sul do Brasil) zilian halictid bees. V. Chloraliclus. Insectes So-
com notas comparativas. Dusenia 15: 1-112. ciaux 5: 379-407.
Denlinger,O. L. 1986. Dormancyintropicalinsects. . 1958e. ObservationsontheethologyofNeo-
Annual Review ofEntomology 31: 239-264. tropical anthophorine bees (Hymenoptera: An-
Eickwort, G. C, and H. S. Ginsberg. 1980. Foraging thophoridae). University ofKansas Science Bul-
and matingbehaviorin Apoidea. Annual Review letin 39: 69-96.
ofEntomology 25: 421—446. Michener, C. D., R. B. Lange. J. J. Bigarella. and R.
Ihering. R. von. 1905. Biologiadasabelhassolitanas Salamuni. 1958. Factors influencing the distri-
do Brazil. Revistado Museu Paulista6: 239-264. butionofbeesnestsinearthbanks. Ecology39(2):
Laroca. S., J. R. Cure, and C. Bortoli. 1982. A as- 207-217.
sociacao de abelhas silvestres (Hymenoptera: Michener, C. D. and C. Seabra. 1959. Observations
Apoidea)deumaarearestritanointeriordacidade on the behavior of Brazilian halictids bees. VI.
deCuntiba(Brasil): Umaabordagembiocenotica. Tropical species. Journal ofthe Kansas Entomo-
Dusenia 13(3): 93-117. logical Society 32: 19-28.
Linsley, G. L. 1958. The ecology of solitary bees. Ordway. E. 1987. The life history of Diadasia rin-
Hilgardia 27: 453-599. conis Cockerell (Hymenoptera: Anthophoridae).
Matthews, R. W. 1991. Evolutionofsocial behavior JournaloftheKansasEntomologicalSociety60(1):
insphecidwasps,pp. 570-602,InRoss,K.G.and 15-24.
R.W.Matthews,eds..TheSocialBiologyofWasps. Roubik, D. W. 1989. The Ecologyand Natural His-
Comstock PublishingAssociates. Ithaca. toryofTropicalBees.CambridgeUniversityPress.
Martins,R.P. 1993. 'X\\eh\o\o%yoiEdithamagnifica 514 pp.
(Perty 1834)(Hymenoptera: Sphecidae). Tropical Sakagami. S. F. and J. S. Moure. 1967. Additional
Zoology 6(1): 109-123. observations on the nesting habits ofsome Bra-
Martins. R. P. and J. E. C. Figueira. 1992. Spatial zilian halictids bees. (Hymenoptera: Apoidea).
distributionsofnests in Diadasma dislincta (Hy- Muschi 40: 119-138.
menoptera:Anthophondae).JournalofInsectBe- Sakagami. S. F.. S. Laroca, and J. S. Moure. 1967.
havior 5(4): 527-529. Wild bee biocoenotics in Sao Jose dos Pinhais
Michener.C.D. 1979. Thebiogeographyofthebees. (PR). South Brazil. Preliminaryreport.Journalof
AnnalsoftheMissouriBotanicalGarden66:227- theFacultyofHokkaidoUniversity,Ser.VI,Zool.
347. 16: 253-271.
Michener,C.D.andR.B.Lange. 1957. Observations Stephen,W.P.,G.E.Bohart,andP.F.Torchio. 1969.
on the ethology of some Brazilian colletid bees Thebiologyandexternalmorphologyofbees.Ag-
(Hymenoptera: Apoidea). Journal ofthe Kansas ncultural Experimental Station. Oregon State
Entomological Society 30: 71-80. LIniversity. 140 pp.
1958a. ObservationsonthebehaviorofBra- Whittman. D. and M. Hofi'man. 1990. Bees ofthe
.
zilianhalictidsbees.I.Pseudagapostemon. Annals RioGrandedoSul. Southern Brazil (Insecta. Hy-
oftheEntomologicalSocietyof.America 51: 155- menoptera: Apoidea). Inheringia. Ser. Zoologica
164. 70: 17-43.