Gard W. Otis, Ronald J. McGinley, Lyn Garling, and Luis Malaret.
Biology and Systematics of the Bee Genus Crawfordapis (Colletidae, Diphaglossinae).
Psyche 89(3-4):279-296, 1982.

This article at Hindawi Publishing: https://doi.org/10.1155/1982/68685
CEC's scan of this article: http://psyche.entclub.org/pdf/89/89-279.pdf, 2276K
This landing page: http://psyche.entclub.org/89/89-279.html

The following unprocessed text is extracted automatically from the PDF file, and is likely to be both incomplete and full of errors. Please consult the PDF file for the complete article.

BIOLOGY AND SYSTEMATICS OF THE
BEE GENUS CRAWFORDAPZS
(COLLETIDAE, DIPHAGLOSSINAE)
BY GARD W. OTIS', RONALD J. MCGINLEY~, LYN GAR LING^, AND LUIS MALARET~
Crawfordapis luctuosa (Smith) is a robust, dusky-haired bee, pres- ently known from only a few localities in Mexico and Central Amer- ica. Individuals can be as long as 24 mm and superficially resemble the more familiar diphaglossine bees of the genera Ptiloglossa and Caupolicana to which they are related. All three genera are placed in the Caupolicanini which is characterized by the complete pre- episternal groove and very elongate first flagellar segment. While Crawfordapis is currently considered to be monotypic, Michener (1966) raised the possibility that the material from the more northern localities (Mexico and Guatemala) may represent a distinct species. Much more material from different localities is needed before that problem can be considered. The purpose of this paper is to present biological observations made on Crawfordapis by three of us in Costa Rica (L.G., L.M., G.O.). In addition, the larva of Crawfordapis is described and the systematic interrelationships of diphaglossine genera are reviewed (R.M.).
Description of the Site
Two nesting aggregations of Crawfordapis luctuosa were observed approximately 5 km east of Monteverde, Province of Puntarenas, Costa Rica (lo0 18'N, 84O47'W) on trails at 1540 m elevation. The surrounding vegetation is best described as elfin forest, with some characteristic plant species being Lycopsidium cernuum, Senecio megaphyllus, Cfibadium sp., Gunnera sp. and Myrica phanerodouta 'Environmental Biology Dept., University of Guelph, ON NIG 2W1, Canada. 2Museum of Comparative Zoology, Harvard University, Cambridge, MA 02138. Present address: Dept. of Entomology, N.H.B. 105, Smithsonian Inst., Wash- ington, D.C. 20560.
'Dept. of Zoology, University of Florida, Gainesville, FL 32641 Manuscript received by the editor July 7, 1982



================================================================================

280 Psyche [vo~. 89
(details in Lawton and Dryer, 1980). Frequent rains deposit approxi- mately 3000 mm of precipitation annually on this area. Heavy mists augment this precipitation. Mean annual temperature is about 16OC (max = 27OC, min = 10å¡C and during prolonged rainstorms, the temperature can remain at 15-1 6O C for 3-4 days. The wind generally blows 15 to 20 km/ hr, but ranges from nearly calm to winds in excess of 100 km/ hr during rainstorms (R. Lawton, pers. comm.). Nesting aggregation No. 1 was directly on the Continental Divide, on a narrow ridge known as "La Ventana". The aggregation was first noticed in August 1975 and was still active but reduced in size to only 5 nests in February 1981. W. Guindon (pers. comm.) indicated the site was active as early as 1966. Bees apparently maintained nests in this area throughout the year (R. Lawton, pers. comm.). Nests were built in the lee of a slope which partially protected them from mist and rain (Fig. 1). In July 1977, there were 97 nests with tumuli in the approximately 18 m aggregation. The majority of the nests were within an elliptical area of about 8 m. Of the nests 29 were completely exposed in the trail, 50 were on nearly level ground and partially obscured by grasses and other herbs, and 18 were on the face of the sheltering embankment.
Nesting aggregation No. 2, located at the head of the valley on the road to Penas Bancas, was exposed to high winds and unprotected from rain (Fig. 2). In February 1978, the aggregation consisted of not more than 100-150 nests, but had enlarged to at least four hundred active nests by August 1978. Nests were found both on the edge of the road and down the adjacent steep, bare slope. By the last visit to the site in February 1981, the number of active nests had declined to 230. The adjacent slope had become covered with dense vegetation and lacked nests.
Description of the Nest
Entrances to active nests had tumuli approximately 7.5 cm in diameter and 4-7 cm in height (Fig. 3). The frequent rains obliter- ated tumuli of all nests except those in which bees were actively digging. Nest entrances consistently measured 1.0 cm in diameter. In each of three nests excavated in horizontal ground, the tunnel began nearly vertically for 7-14 cm and then continued downward at an angle of approximately 75'. In the diagrammed nest (Fig. 4)



================================================================================

19821 Otis, McGinky, Garling & Malare1 - Crawfardapis 28 1 Figures 1-2. Crawfordapis luctuosa nesting sites. Fig, I. Nest aggregation No. 1.
Most nests were on level ground, either exposed or partially obscured by grasses. A smaller number of nests were in the nearly vertical embankment which sheltered the site from wind and mist. Fig. 2. Nest a~gregation Nn. 2.



================================================================================

282 Psyche [Vot. 89
Figure 3. Nest tumutua. Nests in which bees were actively digging were readily discerned by the presence of a dirt mound around the nest entrance. the tunnel changed directions again at at a depth of 24 cm, can- timed downward at a 75O angle another 8 cm, and diverged into two tunnels. One of these angled toward the embankment at an angle of 50' from vertical for another 12 cm, then at a depth of 41 crn continued slightly below horizontal for another 18 cm. A single, terminal cell was found at the end of that tunnel. The other tunnel continued downward another 16 cm from the branching point before becoming nearly horizontal at a depth of 47 cm. This tunnel could not be followed because the soil was too soft. Two additional nests were excavated on the nearly vertical embankment at aggregation No. 1. These nests differed from those in level ground in having very short (3-5 cm) vertical portions of the tunnel before becoming nearly horizontal. Nest I3 (Fig. 5) had a single horizontal tunnel that extended 22 cm into the embankment, Along the slight downward slope of this main tunnel were 6 nearly horizontal lateral tunnels 4-8 cm long. The two closest to the extc- rior contained pupae and the tunnels had been filled with soil. The next three contained larvae, and the distal cell was empty. Nest C



================================================================================

19821 Otis, Me Ginley, Garling & Malaret - Crawfordapis 283 Figures 4-6. Diagrammatic representation of three C. luctuosa nests. Fig. 4. Nest
at base of the embankment. Figs. 5-6. Nests excavated in the embankment. (Fig. 6) had one cell (contents unrecorded) only 4 cm from the vertical entrance tunnel. Another 8 cm further down, the main tun- nel diverged into two. One tunnel continued to slope downward and contained two terminal cells with pupae. The other tunnel sloped slightly upward and had two lateral cells with larvae and an empty terminal cell which was 34 cm from the face of the embankment. Each completed cell was lined with a shiny, cellophane-like mem- brane which is characteristic of Colletidae. It was not possible to lift the cell and contents out of the soil as described for Ptiloglossa guinnae Roberts (Roberts, 197 1). Cell contents were soupy; fermen- tation odors were not recorded. Cocoons were tough, nearly trans- parent membranes 17 mm in diameter by 35 mm in length. General Activity Pattern
Females of Crawfordapis luctuosa were active aboveground primarily between the hours of 0930 and 1400 hr, with few bees leaving the nests after 1300 hr on observation days February 19 and 20, 1978 (Fig. 7). A similar activity pattern was recorded on July 16, 1977. This sharply contrasts with the crepuscular activity pattern of Ptiloglossa guinnae which occurs in similar habitats (Roberts, 1971). In some instances, bees were seen returning to the nests in the morning before any bees left the nesting aggregation. It is possible



================================================================================

284 Psyche [vo~. 89
that these bees had spent the night away from their nests as occurs with Bombus species at high elevations (0. R. Tayor, pers. comm.). Before leaving the nest for the first time in the morning, females often remained just below the nest entrance for a few moments and then upon exiting, hovered nearby for a short time before flying off. Temperatures within the nest entrance remained at 12OC through- out the day on February 20 when ambient temperature was between 10-1 1å¡C
Males remained outside the nests at all times. During the activity period of the females, the males flew over the nesting aggregation and nearby at heights of 1-3 m. They rapidly approached any flying object, including female C. luctuosa, swallows, a hummingbird, a ctenuchid moth and a dragonfly. Males often seized females return- ing to the nests but it was not ascertained whether copulations occurred.
Nest Visitation Behavior
On February 19 and 20, 1978,46 nests within a 2.16 m subarea of aggregation No. 2 were mapped and numbered. Sixteen female bees were captured while leaving nests. Each bee and her corresponding nest of origin were given an identifying color combination. The bees were marked by paint spots on the thorax, while their nests of origin were indicated by a wooden chip about 1 cm in length placed near the entrance. All observed departures from and arrivals to mapped nests were recorded by noting time, markings (or lack thereof) of bees and nest number or color. A nest "visit" was defined as the disappearance of the bee beneath the ground surface for any length of time.
Of the 46 nests mapped, 40 (80%) were entered at least once by a bee. The number of observed visits per nest made by marked or unmarked bees to the 16 color-coded nests ranged from 0-1 5 over the two days (Table 1).
Of the 16 marked bees, four were not observed again. The remain- ing 12 marked bees visited nests a total of 78 times. Four of the marked bees (BG, GO, YBY, OB) concentrated their visits on a single nest, while others entered up to 12 different nests over the two days (Table 2).
The duration of visits of both marked and unmarked bees varied widely from less than 1 minute to a maximum of 151 minutes. The



================================================================================

19821 Otis, McGinley, Garling & Malaret - Crawfordapis 285



================================================================================

Psyche [vo~. 89
visits
Table 1.
Number of visits by all bees to color-coded nests ~~~~~
Nest Color-Code
YB OY YO GO BG YBY Y OBO 0 OB BY YG B GB OG BOY Table 2.
Number of visits made by marked bees to particular nests. Marked
bee
Nest Categories
1 lo 2O 3O 4O 5O 6' 7' 8' 9' 10å 11' 12O Total visits 1
*Asterisked entries indicate that the visits were made to the nest from which the bee was originally captured.
YB
OY
YO
GO
BG
Y
YBY
OB
OBO
0
BY*,YG
visits of 1 minute or less were likely to have been exploratory rather than "working" visits. Of a total of 69 timed visits, 30 (43%) were in the "exploratory" category. The average duration of the remaining 39 "working" visits was 18.1 minutes (S = 26.36) (Fig. 8). The duration of foraging trips of 4 of the marked bees was noted by timing of their absences from the nest area. Whether or not they returned with pollen was not noted. Absences ranged from 19-42 minutes, the average being 30.4 minutes (n = 11, s = 13.79). It seemed that bees returned from foraging trips in distinct pulses of several bees at a time. The foodplants are not known. This preliminary data on the movements of marked bees to differ- ent nests generates several hypotheses for further testing. 1) Multiple nest entering is part of searching behavior for a female's own nest.
3 2 2 2 2 2
6 * 4 2 1 1 1
2 * 2 2 1 1 1
5 1 1 1 1
5*
1 1 1
3*
3*
2
1 1
Each visited one nest once.




================================================================================

9821
Otis, McGinley, Garling & Malaret - Crawfordapis tt in
F
t
0
IT)
F
CO
(0
I 1
0
I
0 0
I
0 F
tt <å´
CM
TT (0
TT in
--0
in
_ (0
TT
C M
tt
CO
0
tt
0
_ 0 0
(0 CM
CM
CM
CO
<t
0
(0
CM
-:
F
VI
0




================================================================================

28 8 Psyche [VOI. 89
2) Individuals are locating abandoned nests to provision rather than starting completely new ones.
3) Individuals are "stealing" provisions from nests of other individuals.
4) Individuals are usurping nests which other individuals are actively excavating or provisioning.
5) Several individuals are contributing to excavation, provisioning or oviposition within a single nest.
The latter hypothesis is particularly interesting since no species of the family Colletidae is known to be parasocial (Michener, 1974). It is plausible that females not only construct their own nests but also usurp partially constructed or provisioned nests of other females as has been reported in some other Hymenoptera that nest in aggrega- tions (Brockmann and Dawkins, 1979; Brockmann et al., 1979; Eickwort, 1975; Eickwort, 198 1; Eickwort et al., 1977). Further stu- dies of Crawfordapis luctuosa are needed to better understand its social behavior and biology.
Description of Larva
The following description follows the format used for describing other colletid larvae (McGinley, 198 1). HEAD (Figs. 9-1 3): (2) Labrum nonspiculate; (3) epipharynx and (4) hypopharynx spiculate; (5) maxilla spiculate on inner surface. (7) Head size normal in comparison to body (head not relatively large as in Xeromelissinae); (8) head capsule somewhat elongate, slightly produced in lateral view; (10) frontal swellings above antennae absent; (lOa) median frontal swelling above antennae absent (pres- ent only in Ptiloglossa). (14) Anterior tentorial pit low in position (high in all other known diphaglossines); (15) posterior tentorial pit at junction of hypostomal ridge and posterior thickening of head capsule; (1 5a) tentorial development unknown (tentorium of speci- men examined was incomplete, probably due to nearness of speci- men to pupation). (16) Posterior thickening of head capsule moderately developed (17) straight medially, not curved forward; (19) median longitudinal thickening of head capsule absent; (20) hypostomal ridge well-developed; (25) epistomal ridge complete but thin, (26) arching dorsally to level of antennae. (27) Parietal bands distinct, broad and shallow. (28) Antenna1 prominence absent; (29)



================================================================================

19821 Otis, McGinley, Garling & Malaret- Crawfordapis 289 Figures 9-13. Mature larva of Crawfordapis luctuosa. Figs. 9, 10. Head capsule, frontal and lateral view. Figs. 11-13. Right mandible, dorsal, adoral and ventral view.
antenna1 papilla a moderate-sized convexity, (3 1) bearing three sen- silla. (32) Clypeus moderate in length; (34) labrum not projecting in lateral view; (35, 36, 37) labral tubercles very well-developed, nar- rowly rounded and strongly projecting (unlike those of other diphaglossines); (38) labral apex emarginate, (39) without sensilla- bearing swellings as in other diphaglossines. (41) Mandibles elon-



================================================================================

290 Psyche [vo~. 89
gate, (42) moderately slender in dorsal view, (43) broad basally in adoral view; (44) mandibular spiculation absent; (46) outer surface of mandible smooth, distinct tubercle and setae absent; (51) apical portion of mandible, in adoral view, attenuate; (52) cusp moderately well-defined; (53) cuspal projection absent; (54) cuspal region multi- dentate; (55, 55a) dorsal apical edge with distinct, moderately large teeth; (57) apical concavity weakly developed; (60, 60a) ventral api- cal edge smooth, teeth absent. (61) Labiomaxillary region produced; (62) labium and maxilla distinct, (63) subequal in length. (65) Inner apical surface of maxilla rounded, not produced mesiad; (66) unlike all other known bee larvae except those of Ptiloglossa, maxilla with a longitudinal groove on adoral surface; (67) cardo and stipes sclero- tized; (69, 70) maxillary palpus moderately elongate and slender, (71) apically positioned on maxillary apex in lateral view; (72) galea absent. (73) Labium divided into prementum and postmenturn; (75) palpus elongate and slightly decurved, (76) subequal to maxillary palpus in length. (77) Salivary lips well-developed; (78, 79, 80) sali- vary opening narrow, circular, at end of long spoutlike salivary lips, (83) which project from a well-defined platelike structure at apex of labium; (82) apical labial swellings absent. (84) Hypopharynx nor- mal in size, (85) bilobed, (86) exceeded by labium and maxilla; (87) hypopharyngeal groove distinct, sclerotized laterally. BODY (Fig. 14): (88) Integument spiculate, density of spicules greater on dorsum than on venter; (93) body moderate in length, (94) robust, (95) widest posteriorly in lateral view; (96) interseg- mental lines moderately incised; (97) intrasegmental lines indistinct; (98) dorsal tubercles weakly developed, most prominent on abdom- inal segments 5-7; (103) lateral tubercles absent (present on one specimen from Panama); (104) ventrolateral tubercles absent; (106) abdominal segment 10 moderate in length, (107) rounded, (108) dorsal in attachment to segment 9; (109) venter of segment 10 slightly produced (very weakly so in specimen from Panama), (1 10) without conspicuous, darkly pigmented spiculation; (1 1 1) dorsal surface of segment 10 smooth, without lines or ridges; (1 13) anus apical. (1 14) Spiracles large, (1 15) not on elevations; (1 17) atrium very broad and shallow, (1 18) not produced above body surface; (1 19) atrial wall faintly ridged, (120) with four to five broken rings of spicules; (121) atrial rim absent; (122) peritreme wide; (123) pri- mary tracheal collar absent; (126) subatrium apparently extremely



================================================================================

19821
Otis, Mc Ginley, Garling & Malaret - Cra wfordapis 29 1
Figure 14. Mature larva of Crawfordapis luctuosa. short. (The structure of diphaglossine larval spiracles remains poorly understood, especially with regard to the subatrium. The atrium is connected to the trachea by a long, nonringed tube. While this tube is characteristic of diphaglossines, its internal structure and homologies are not known.)
MATERIAL STUDIED: Two postdefecating larvae; 5 km east Monte- verde, Puntarenas Province, Costa Rica; July 16, 1977 (G. W. Otis); specimens in the larval bee collection of the American Museum of Natural History. Two postdefecating larvae; Bouquete, Chiriqui Province, Panama; April 25,198 1 (R. W. Brooks); specimens in the personal collection of R. W. Brooks.
Analysis of Larval Characters
McGinley (198 1) described the mature larvae of 30 colletid species including those of seven diphaglossines. Two cladograms for the diphaglossine genera appeared to be most strongly supported by lar- val characters. One of the cladograms, for reasons discussed in the above mentioned paper, appeared to be the preferable working hypothesis of diphaglossine phylogeny. This cladogram is presented in Figure 15, with Crawfordapis now included. The polarities of the characters listed in Table 3 were determined by out-group compari- son, i.e., consideration of character state distributions in all other bee larvae as well as in nonapoid larvae, especially those of specoid wasps.




================================================================================

Psyche
[Vol. 89
Figure 15. Cladogram of diphaglossine genera based on larval characters. Numbers refer to the characters listed in Table 3. Black rectangles represent presumed apo- morphic characters; white rectangles represent plesiomorphic characters. Character 109 appears to be variable in Crawfordapis. The larval cladogram corroborates the currently recognized diphaglossine classification based on adult characters (Michener, 1966). Recognition of the Caupolicanini (Ptiloglossa, Crawfordapis, Caupolicana) is supported by the presence of the unusual salivary plate (character 83, Fig. 9) and the rounded, nonprojecting inner maxillary surface (character 65).




================================================================================

19821 Otis, McGinley, Garling & Malaret - Crawfordapis 29 3 Table 3.
Larval Characters Used in Diphaglossine Cladogram Plesiomorphic
Apomorphic
5. Maxilla spiculate
10. Median frontal swelling absent
17. Median portion of posterior
thickening of head capsule
straight
25. Epistomal ridge complete
53. Cuspal region of mandible
without distinct projection
55. Teeth on dorsal apical edge of
mandible distinct basally
65. Inner apical surface of maxilla
produced mesiad
66. Inner surface of maxilla
smooth
75. Labial palpus moderately
' elongate, straight
78. Salivary lips transverse, not
spoutlike
83. Salivary plate absent
84. Hypopharynx normal in size
I 85. Hypopharynx bilobed
109. Venter of abdominal segment
I
10 not produced
Nonspiculate
Present
Curved forward
Incomplete
Cuspal projection present
Teeth fused basally, forming distinct
platelike wedge
a. Inner surface rounded
b. Inner surface strongly produced
forward
Inner surface of maxilla with
longitudinal groove
Extremely elongate, decurved
Elongate, spoutlike
Present
Conspicuously narrow
Rounded
Produced
The sister-group relationship of Ptiloglossa and Crawfordapis is
strongly supported by the presence of a longitudinal groove on the inner maxillary surface (character 66). Weaker support for this rela- tionship is indicated by character 109, the projection of the venter of abdominal segment 10 (this projection is conspicuous in some spec- imens of Crawfordapis but is only weakly developed in one speci- men from Panama).




================================================================================

294 Psyche [vo~. 89
Adult Characters
In a study of adult diphaglossines, Michener (1966) discussed the similarities of Crawfordapis to Ptiloglossa and Caupolicana. Four characters found in Crawfordapis were said to be more or less Caupolicana-like: (4) outer hind tibial spur of male normal, articu- lated at base like inner spur; (6) lateral extremities of terga of male without areas of short, dense, erect hair; (7) sixth tergum of male with posterior margin not thickened or sulcate; (9) eighth sternum of male with apical process rather heavily pigmented, not down- curved. The similarity based on character 4 is definitely symplesio- morphic as the fusion of the hind tibial spur and the tibia is found only in Ptiloglossa. The other three characters appear to be plesio- morphic as well in that they represent the absence of some rather unusual features.
Similarities between adult Crawfordapis and Ptiloglossa appear to be apomorphic for diphaglossines: (1) clypeus strongly elevated above level of adjacent parts of face; (2) marginal cell prolonged basally as a narrow sinus to apex of stigma; (3) expanded second and third hind tarsal segments of female considerably expanded above. This evidence also supports the Crawfordapis-Ptiloglossa sister-group relationship indicated by larval characters, but must be considered tentative until a comprehensive cladistic analysis of adult colletids has been performed.
Crawfordapis luctuosa, a large colletid bee, was studied at two nest aggregations in the mountains of Costa Rica. The aggregations were in exposed sites formed by landslides or clearing. Female bees slowly abandoned the aggregations as they became overgrown with vegetation. Several nests are described. In contrast to the crepuscu- lar habits of the closely related genus Ptiloglossa, Crawfordapis was active primarily between 0930 and 1400 hrs. Some individually marked females showed a high degree of constancy in nest visita- tion, while others visited several nests in succession. The exact explanation of this behavior is not yet known. The previously unknown larvae of Crawfordapis luctuosa are described. Informa- tion from these larvae supports the placement of the genus in the



================================================================================

19821 Otis, Mc Ginley, Garling & Malaret - Crawfordapis 295 tribe Caupolicanini that was suggested from the systematic study of adults, and indicates that Crawfordapis may be the sister-group of Pt iloglossa.
We wish to thank C. D. Michener and J. G. Rozen, Jr., for making helpful suggestions on the manuscript and R. W. Brooks for the loan of Crawfordapis larvae from Panama. The Tropical Science Center, San Jose, Costa Rica provided permission to study Crawfordapis within the Monteverde Cloud Forest Reserve. The Organization for Tropical Studies provided logistical support for a portion of the field work. Thanks are due also to the community of Monteverde for its friendly support of biologists and appreciation of their work.
BROCKMANN, H. J. AND R. DAWKINS.
1979
Joint nesting in a digger wasp as an evolutionary stable preadaptation to social life. Behavior 71:203-245.
BROCKMANN, H. J., A. GRAFEN, AND R. DAWKINS. 1979 Evolutionarily stable nesting strategy in a digger wasp. J. Theor. Biol. 77:473-496.
EICKWORT, G. C.
1975 Gregarious nesting of the mason bee Hoplitis anthocopoides and the evolution of parasitism and sociality among megachilid bees. Evolution 29: 142-1 50.
EICKWORT, G. C.
1981
Presocial Insects in Social Insects, Vol. 11, ed. Henry Hermann, Aca- demic Press, NY, pp. 199-280.
EICKWORT, G. C., K. R. EICKWORT, AND E. G. LINSLEY. 1977
Observations on nest aggregations of the bees Diadasia olivacea and D. diminuta (Hymenoptera: Anthophoridae). J. Kansas. Ent. Soc. 50:l-17. LAWTON, R. AND V. DRYER.
1980
The vegetation of the Monteverde Cloud Forest Reserve. Brenesia 18:lOl-116.
MCGINLEY, R. J.
1981
Systematics of the Colletidae based on mature larvae with phenetic anal- ysis of apoid larvae (Hymenoptera: Apoidea). Univ. Calif. Pub. Ent. 91: 1-307.




================================================================================

296 Psyche [VOI. 89
MICHENER, C. D.
1966 The classification of the Diphaglossinae and North American species of the genus Caupolicana (Hymenoptera, Colletidae). Univ. Kansas Sci. Bull. 46:7 1 7-75 1.
MICHENER, C. D.
1974
The Social Behavior of the Bees. Harvard University Press, Cambridge, Massachusetts.
ROBERTS, R. B.
1971 Biology of the crepuscular bee Ptiloglossa guinnae n. sp. with notes on associated bees, mites and yeasts. J. Kansas Ent. Soc. 44:283-294.



================================================================================