Biology and Mating Behavior of Xylocopa virginica L. (Hymenoptera, Anthophoridae)

Transcription

1 Behavioral Ecology and Sociobiology by Springer-Verlag 1978 Biology and Mating Behavior of Xylocopa virginica L. (Hymenoptera, Anthophoridae) Dan Gerling l and Henry R. Hermann 2 1 Department of Zoology, The George S. Wise Center for Life Sciences, Tel-Aviv University, Tel-Aviv, Israel 2 Department of Entomology, University of Georgia, Athens, Georgia, USA Summary. The biology and behavior of X ylocopa virginica L. were followed in the area of Athens, Georgia (USA) from August 1973 until September Several nests were X-rayed regularly during the developmental period of the bees, and information about the life cycle and development was obtained. The bees emerge in June and stay in their nests with their sibs throughout the summer, fall, and winter, leaving only occasionally on warm days to search for nectar as food. In March and April males defend nest-side territories and mate with most females. The males die in the late spring and the females dig nests, provision them, and rear up to 8 or 9 bees in each nest. Many females live 2 years, and often 2 to 3 females are present in each nest, but only 1 works and lays eggs. Emergence of adults from pupae is nearly synchronized, irrespective of age. The large carpenter bees of the species X ylocopa virginiccl L. have been the subject of several publications (Rau, 1933; Balduf, 1962; Sabrosky, 1962). However, most reports are descriptive or observational in nature, and very little experimental and analytical work has been done to date. A better knowledge of the bees' biology, especially with regard to the nesting of and cooperation between adults, will contribute to an improved understanding of the evolution of the Hymenoptera in general and of the large carpenter bees in particular. In our work we attempted to accumulate some basic knowledge about X. virginica and tried to explain some of the phenomena observed. Experiments were carried out both in the field and in the laboratory. Insects were dissected in water or saline for an examination of their genera] internal condition. In some cases, their reproductive systems were sectioned and stained with Delafield's hematoxylin and eosin Y.

2 Pinc logs in which some bees nested were X-raycd with a Picker X-ray machinc at 30 kv and 3 ma for 40 to 48 s at a distancc of 48 in. Throughout our study. bees were markcd individually by gluing small plastic numbered discs to their notum ('Nummerierte Obalith-Plaettchen' Fa. Chr. Graze K.G.. Germany). Most observations were conducted upon bees that nested in wooden structurcs on the grounds of the University of Georgia campus in Athens, Georgia. Abundant matcrial was also madc available through the destruction of numerous wooden buildings by local construction companies. Temperatures were recorded daily in our main observation al'ea by means of a maximumminimum thermometer. Additional measurements, both within the ncsts and outside, werc carried out with a Yellow Springs Telethermometcr. All temperatures were recorded in 0 C. The study lasted from August 1973 to September Some bces were followed aftcr having been marked with plastic numbers; others were collected throughout the year and notcd for the following characteristics: wing and mandibular wear, ovarial size, fat (its amount and location), the presence of spcrm in the spermatheca, the condition of Dufour's gland and of the yellow glands. Thc developmental history was studied by taking X-ray pictures of nests I to 3 times a week, from egg deposition time to adult emergence. Concurrent observations were conducted upon nests that had trap bags attachcd to thcir exit holes. ests also were opened and their contents reared out in glass vials. X. uirginica is some 2 cm long; however, live bees are hard to measure reliably their length varies with abdominal extension and contraction. Weight was therefore used as a criterion for size. Weight averages ±SD for 94 overwintering males were 570 ± mg ( ) and for 173 females, 603 ± ( ). Spring Oying bees (3 males and 39 females) showed no significant deviations from the above (426 ±60.96 and 520 ±93.05, respectively). Not all female bees perform similar activities. Consequently, not all show the same amount of mandibular and wing-tip wear. Thus, while several females which were marked by us in October 1973 had badly worn wings and mandibles in the spring of 1974, others had wings and mandibles that appeared unused. Moreover, we found that many females live for two years. These facts made age determination by external characters difficult. However, it was usually possible to estimate a female's age by examining her fat body and ovaries in addition to the external appearance. The entrance hole of the nests penetrates the wood perpendicular to the grain. It usually is round, with a diameter of to mm and a length of 13 to 15 mm. There is an average of 2.4 (1-4) tunnels per nest (Table I). Their walls are parallel when they are new and become segmentally barrel-shaped after nesting (Fig. 1). This is due to the scraping of the wall by females in order to obtain wood shavings for constructing the intercellular partitions. The tunnels run with the grain of the wood and are more or less parallel with each other, except when running into a knot. A cross-section of the tunnels may either be round and measure 10 to 14 ml11 across, or it may be oval. Measurements of the oval holes were: 18 x 16, 18 x 12, 16 xis, 16 x 12, 15 x 14, and 15 x m.Tunnel length averages 17.5 (47-3.9) cm (Table 1).

3 Biology of Xy{ocopa rirginica L. 101 Table I. Bees and nests of X. virginiccl during the winter Number Cells Cells Overwintering bees per nest Nesting Cell Tunnel Yearly of per per females length length tunnel tunnels active nest Total Females Males Males per nest (mm) (em) enlargeper tunnel bees per per per ment nest per nest nest nest (em) nest with males Sample size Total Average Range SD J 2.J Fig. 1. X-ray radiographs of2 nests of X. virginiccl taken at different times in I A January; I B, May 21; 1 C, June 14; 2A, May 15: 2B, June 5; 2C, June 14. Nest extension is seen from I A to I B. Females marked during the winter are distinguished by a dark spot on the thorax. 2 Band 2C show parasitism by Anthrax ligrin/ls DeGeer (arroil's). E. egg; L. larva; P. pupa; N. nest entrance

4 Q. ~ 15 I- r\ I, \ ; I {ll \ I ~ 1\, I I \ I,\ \ : ft :\'1/ ' I \ (\ I' I ' I I '. I \ " '\ I \ I I I I I I V \,I, ' I I ".~, I I \ I \,J I I I "I \, I I I I ' \, \, I~~ I I", \: ~, I V'I I' ~ 0, 'J ~ 1 I -5 I 1 I IV V Month VI Fig.2. Maximum and mimimum temperatures in the observation barn, March-May 1974 Nest tunnels are not utilized for nesting throughout their length; rather, a cell-free area is left between the proximal cell and the nest entrance (Fig. 1). When starting to nest, females may either construct a new nest or utilize an old one. Of the 32 old nest tunnels that we examined, only 14 were reused during the 1974 season. Before reusing them, the bees excavated and extended (Fig. 1) 10 of the tunnels. They also excavated 4 tunnels which later were abandoned without being nested in. An average of 3.8 (8-1) cells per active tunnel or 6.6 (14-2) cells per nest were found. The cells were 17.5 mm ( ) long (Table 1), and the partitions were thinnest (I mm) in the middle and thickest (4 mm) along the edges. Low winter temperatures in Athens reach a few degrees below freezing. During March, when the first bees began to ny, there were only a few nights with subfreezing temperatures; the coldest day reached a high of 12 C (Fig. 2). The first males were seen on March 5, after 5 days with temperature maxima above 20 0 C. The first females emerged from their hibernacula 5 to 6 days later, after a warm spell of about 10 days. A temperature drop followed, and no night was observed again until the weather became warmer on March 27. At that time, however, both sexes appeared as soon as the day warmed up. These obser-

5 Biology of X.locopa virgil/ica L. 103 Table 2. X-ray record of nests C9 and C Cell May June I L L L LC LC LNF LNF P P P P A 2 L L L LC LC LNF LNF LNF P P P A 3 E L L LC LC LC LNF LNF P P P A 4 E E L L LC LC LNF LNF P P P A 5 6 E E L L LC LC LNF LNF P P P A 7 E E E L L LC L F L F L F P P A 8 E E L L LC LNF L F L F P P A 9 L L? LC LC LNF LNF -p? P A 10 L L LC LC LNF L F P? P A 12 E L LC LC LNF LNF P P P A 13 E E L LC LNF LNF P P P A 14 E E E L L LC LNF LNF P P P A 15 E E L L LC LNF LNF P P P A 16 E E L L L LC LNF P P A Numbers 1-8 represent-nests C and numbers 9-16, C9 A, adult; E, egg; L.larva: LC, C-shaped larva; L F, larva with no food: P. pupa:?, details not clear in X-ray: -, progeny died vat ions correspond well with our laboratory experiments that showed that overwintering females need about one week above 23 C before they start to ny. Thereafter, they will ny readily whenever the temperatures favor night activities. Egg laying started about May 1, and most of the eggs were deposited before the 15th of that month (Table 2). However, occasional eggs were still layed until July I. The duration of development in the field lasted 45 to 50 days; accordingly, the first adults appeared in the field nest on July 15 with a peak of emergence occurring a few days later. Adult bees emerging from the same nest emerge within a maximum of 5 days of each other (Table 2). Once emerged, the adults thoroughly break the cell partitions apart and push out tunnel refuse, which consists of broken partitions, faecal pellets, and pupal exuviae. We started to X-ray the field nests on May 7, at which time the first eggs in nests C9 and C had already hatched (Table 2). However, since the larvae were very small, we assumed that the eggs were deposited about 5 days earlier, on May 2. Accordingly, oviposition lasted for 10 days in C9, and 7 days in C. Eight cells were built in each nest, including I cell (No.3 in C9 and NO.5 in C11) in which the immature bee had died. All 14 larvae had pupated by June 14

6 and were still in that stage on the 20th. Six days later, on the 26th, we found that all adults had emerged, i.e., pupae that were at least 22 (C 1l.1) days old, emerged simultaneously with ones that were at most 20 days old (C, 7, 8). The average temperatures during that time were relatively stable and could not account for the accelerated pupal development of the latter. Several X-ray records showed a decrease in the number of progeny during the developmental period. Some of these cases were associated with the removal of dead material, i.e., of bee bread and the eggs or larvae that failed to develop, and in one case, a larva of a wasp. However, the removal of apparently healthy material was also recorded. As shown in Figure 1, nest 0.2 had on 15 May I marked and 1 unmarked female in its 3 tunnels. It had 6 progeny in the upper tunnel, 5 in the lower left, and a pollen slant in the lower right. A few days later, the unmarked female disappeared, and by ] une 14, the upper branch remained intact but the lower had only 2 progeny left. Since the unmarked female was absent, we assume that it was her progeny that had been removed by the marked female. Yet she failed to remove 4 larvae, one in the lower tunnel and 3 in the upper, that had been parasitized by Anthrax tigrinus DeGeer. It is noteworthy that the marked female was present in the nest before the unmarked arrived. Newly emerging bees have a soft cuticle and milky-white wings which turn first brown and then bluish black. Capability to f1y is attained 3 to 4 days after emergence, but the bees do not usually leave their nest for the first 2 to 3 weeks of their life. No food is present in the nest during that time, except for occasional old bee bread in cells that have failed to produce bees. Moreover, experimentation with different foods showed that bees at this stage do not eat pollen or bee bread, but do accept honey. From the time of emergence through the rest of the summer and fall the bees are in the 'Juvenile' stage (Rau, 1933). They remain in their nests most of the time, each bee most commonly with its siblings. Occasional bees were seen and caught foraging on f10wers during August and September, but numerous bees were caught only in October. On October we were able to mark 20 bees (5 males and 15 females) near a f10wering Eleagnum shrub within one hour (: 15-12: 15). Only I of the 21 females collected on that date had worn wings. The same bees often were caught repeatedly throughout the early half of the day and on successive days. Ten of the collected bees, representing a sample from various localities in Athens and northern Georgia, were dissected. All had full spermathecae. On August 25, after all nesting activity had terminated, 13 nests were examin~d for their contents. They contained between 1 and 8 adult bees each,

7 both males and females. None of the bees was marked, and only one female per nest had worn wings. Additionally, one marked female with fresh wings was found in a plastic bag that was hanging under an additional nest. The contents of another nest, which contained 3 males and 4 females, was collected on August 5, and examined. All had fresh wings and mandibles, but only one female had a full spermatheca and a turgid Dufour's gland. The stomachs of all bees contained nectar and no pollen, and the seminal vesicles of all males were full with active spermatozoa. During the winter, both sexes remain in nest tunnels. These usually are the tunnels in which they have developed but also may be newly acquired ones. Of a total of 269 bees examined, 95 or 35 % were males. This population occupied 38 nests with an average of 7 bees (4.5 females and 2.5 males) per nest. The minimum number of bees per nest was 2 (6 nests) and the maximum 15 (I nest). Males occurred only in 28, or 73.6 % of the nests, where they averaged 3.4 males per nest (Table I). The bees do not undergo a physiological diapause. They do not require a cold period before being reactivated, and become active, begin to fly and feed readily upon honey, when the temperature is raised above C for about one week. Bees that fed solely on honey were kept alive for over one month. Overwintering bees are characterized by very small ovaries, a deplete Dufour's gland, and well developed fat bodies, but, as mentioned, their weight does not differ significantly from that of spring flying bees. Mated and unmated bees overwinter together. Since we were reluctant to kill large numbers of females to determine if they were mated, we have only a small sample. However, the data which we collected indicate that though 83 % of the nests contain at least I mated female, the majority of the overwintering females is unmated. Dormant females that were activated in the laboratory in February fed readily and flew about, but showed no sign of ovarial development, irrespective of their spermathecal condition. Therefore, we examined the effect of a protein diet (Anderson, 1974) on their ovaries, hoping that a diet which sufficed to retain a honey bee queen in full laying capacity for a prolonged time would activate the ovaries of the carpenter bees. However, no ovarial development occurred. Spring flying females have many characteristics of the overwintering individuals, namely, unworn wings and mandibles, much fat in the abdomen, small undeveloped ovaries and a digestive system which is nearly empty except for the rectum that is full with grayish-white material. Many of them are unmated, but, as mentioned, each nest will usually contain one mated individual. Females started to fly in Athens during March 1974, and they immediately visited various flowers. The first two weeks of pollen carrying by the females are not accompanied by nesting activity. Rather, the pollen ends up in their digestive system, probably serving as a protein supply for ovarial development. Pollen was also found in the digestive system of some bees that did not collect it by themselves,

8 as evidenced by their clean bodies. We assumed that they ate pollen which was brought into the nest by other field-going females. During the month of April, several changes occurred in the bees. As they flew about, some bees with worn wings were found. Their legs, coxal bases, mouthparts, stomach, and later also the gut became laden with pollen. The ovaries developed and attained a typical condition in which one ovary was longer than the other. Simultaneously, large opaque oocytes appeared. Dufour's gland began to lose its flat, empty appearance and became filled with an opaque material. After mating occurred, spermatozoa could be seen in the spermatheca. Two pairs of heretofore unnoticed glands became prominent during that time. These paired glands, which occur on the ventrolateral region of abdominal segments 5 and 7 (Fig. 3), drain into their anterior intersegmental membranes. The glands, which were white and compact at the time the female emerged, and now became first light and then dark yellow, were each composed of numerous monocell ular secreting cells, each with a long ductule. Three pairs of similar glands were observed by the author in X. pubescens Spinola and by Altenkirsch (1962) in Anthophora acervorum L., where their secreting units were characterized as 'A mpu liencellen'. In 1974, provisioning and egg laying activities started in the beginning of May, when we dissected and examined 33 females. Seventeen had worn wings and mandibles, well developed yellow glands, Dufour's gland and ovaries, and a full spermatheca. Nine females were unmated, with new wings and mandibles and with undeveloped ovaries and glands; the rest were mated bees with varying degrees of wear and gland development. All but one unmated bee were caught on the wing without pollen on their bodies. The remaining unmated bee was in a nest. All the mated bees were in nests or carrying pollen.

9 Males established territories near active nest entrances. Each male defined the territories by flying to and fro within its boundaries, which were almost linear and several meters long, when in the immediate proximity of the nest. Territories that were 10 to 20 m from the entrance had a square shape and covered about 4 to 8 m 2. Flight speed was usually uniform with occasional hovering in one place, especially by males that were close to the nest entrance. The same marked males were present in one territory for as long as 2 weeks. They acquired food and rest in pauses made during sunless hours when no territorial flight occurred, and on several occasions during the day. For example, a male marked at 14:00 left a few minutes later and reappeared at 14:40, left again about 14:55, showed up at 16:00, and continued to hover in front of the nest. During his absence, his territory, in front of a nest entrance, was occupied each time by other males, but he always reacquired it upon returning. Uninterrupted territorial flight lasted from a few minutes to over one hour. The behavior of males towards intruders into their territory was tested in a number of experiments in which patrolling males were confronted with dead or living males or females which were suspended from a thread into the midst of their territories. The introduced bee, whether dead or alive, was ignored as long as it hung motionless. As soon as it was allowed to fly to and fro or was swung about, it was pursued. These results conform well with our additional observations on the behavior of territorial males which don't seem to notice other individuals as long as they hover about, but pursue any fast-flying organism in their territory. Males apparently did not mark nests or other objects chemically, either as a warning signal for other males or as an attraction for females. We examined the mandibular glands, which are known to produce a marking substance in X. hirsutissima Maidl (Velthuis & DeCamargo, 1975), but found that contrary to the latter, X. virginica had small glands that were of equal size in both sexes and produced no special marking substance. The nest finding capacity of males was tested. For this purpose, 22 males, all of which nested adjacent to each other, were collected on April 3. They were then released in groups of 4 to 6 males, at various locations 400 m to 3 km from their original nests. The weather in the next few days followed release was very stormy; however, on A pril1f, we retrieved 2 males near their original nest site, 6 more were seen there on April 16, I on April 21, and I on April 22. The first males to be retrieved came from the farthest release points. All of the release points were represented by the 10 retrieved males. Most frequently, the males catch the females and mate with them when the latter leave or return to their nests. Most mating occurred in April, but some mating continued during May. Copulations were observed under natural conditions. In addition, females that

10 were tied to a 40-cm long string were introduced into males' territories. This allowed us both manipulation of some mating conditions, and closer, controlled observations. Males did not approach any object including a female, unless it was in motion. Suspended, motionless females were ignored. Once a female was approached and contacted, the male mounted on her back and tried to push his abdomen under hers. Copulation occurred immediately. Thereafter, the male either left the female or attempted to mate again. Whenever a female alighted on a substrate, be it the ground for a free female, or our hand in the string-tied one, the couple immediately became disengaged, and the male usually left the female hovering above her at a very close distance. As soon as she took to wing again, or was pulled off our hand, he mounted her and copulation attempts continued. On two occasions, once in the insect net and once on the ground, a male did not leave the female when she landed. Instead, he held on to her with his 6 legs and flapped his wings vigorously, apparently trying to pick her up. His body was raised a little off hers following his efforts, but he was unable to lift her off the substrate. After about two minutes the male left the female but when she took wing, he followed her in close pursuit, whereupon she landed again and the same sequence followed. In addition to the behavior described above, occasional couples fly off to a great height and distance. Occasional courting behavior was seen in which the male and female face each other and fly slowly for a few minutes following each other. This behavior was not observed to be followed by mating or by flight as reported for X. californica californica Cresson (Cruden, 1966), and may indicate that the female is unreceptive, as in X. jlavorufa DeGeer (Watmough, 1974). Provisioning and nesting activities were followed from May until July. Nine nests were followed and X-rayed regularly; in addition, each flight of the marked bees from 25 nests was observed and registered between May 14 and 20. The latter included 2 complete, uninterrupted days and a few shorter observations. In each case, the entering or leaving female was noted as to her identifying number, the time and the existence or absence of pollen on her body or legs. The X-ray radiographs showed that 8 nests harbored 2 or 3 females per nest, whereas the ninth was occupied by 1 female only. In each nest, only I particular female was active throughout the nesting period, whereas the other females flew only rarely. Only in one case was one of the latter seen with pollen. A total of 159 flights were observed; of these, 89 were nest entries and 70 were exits in which females went to the field. Of the entries, 54 were with pollen and 35 without. Of the exits to the field, 34 resulted in pollen collecting and 36 did not. Females that returned without pollen stayed in the nest for less time than ones who came with pollen. The time spent in the nests was usually shorter than the time spent in the field trips, which lasted in about 50 % of the cases for over an hour. Forty-seven active bee-days were observed, including 22 complete days (early morning to darkness) and 25 partial days (4 to 7 hours of prime activity). We found 6 complete days during which pollen was brought back on each trip, 7 days when both pollen and nectar were brought, and 9 days when no pollen was brought. About half of the bees (24 vs. 22) carried pollen on their morning trip. In their last evening trip, a preponderance of bees came without pollen (20 vs. 4). The largest number of trips per day was 6 and the smallest was 1. Daily activity

11 '" ~ 16 co ~ o I I I I, I I!I 24 ti. E 22 ~ Hour of day Fig.4. Temporal distribution of the first daily exit to the field and the last daily entrance to the nest started when the temperature in the shed rose above 15 C, with a peak between 09:00-1 LOO (15-25 C). Some final return trips were conducted as early as 15:00 and 16:00, with temperatures still above 30 C; other trips took place later, until darkness. It is apparent that the first trip may be triggered primarily by a temperature rise to 15 C and above (Fig. 4). Final return to the nest may be governed by a number of factors, including light and the particular activity of the female at that time. X. virginica belongs to the temperate and cold-zone carpenter bees; as such, it has only one generation each year. According to Rau (1933) the bees develop within about one month and spend the summer-fall and winter waiting for next spring. Mating occurs once a year, in the spring shortly prior to nesting. Our findings confirmed the above. However, females marked in August 1973 were still active in September Moreover, mated females were found throughout the year, and during the nesting time some of them had well developed glands; ovaries of other bees were underdeveloped. The former, which were also the

12 actively nesting females, died shortly after the end of the nesting period but the latter lived on for at least another year. These findings indicate that females live at least two years, and that mating takes place usually in the spring of the first year, whereas nesting occurs in the second. Most nests examined contained more than one female during nesting time. However, only the older female was performing the nesting activities, digging, excavating and lining the cells, collecting and preparing food, and ovipositing. The younger bees hardly left the nest and had unworn wings and mandibles. Cohabitation of more than one female in the same nest during nesting time was also observed in X. nogueirai Hurd and Moure (Hurd and Moure, 1960), X. ji-ontalis and X. hirsutissima (Sakagami and Laroca, 1971). In X. virginicc[ it is not obligatory, since progeny in one nest developed normally with only the mother present. However, it was a frequent occurrence and involved active and inactive females, each belonging to different generations. It should be noted that some of the females found during the nesting period were unmated. These might have been unable to mate, or simply escaped mating during the first year of their lives, and would mate in the second. Other females might have nested during the first year of their lives, a fact that we could easily have overlooked. Therefore, we assume that although most of the bees form 2 sympatric populations, each alternating its appearance, at least some gene now may occur between them. In tunnels of carpenter bees the first laid eggs are the farthest from the only exit hole. The mode and order of emergence of the bees has been the subject of discussion and speculation from the days of de Reaumur (1742) on (Malyshev, 1931 ; Rau, 1933; Hardouin, 1948; Skaife, 1952). Mature larvae and pupae, in contrast to eggs and young progeny, are not sensitive to being moved about or stepped on by other bees. Tunnel diameter suffices for 2 bees to stand side by side. Therefore, emerging bees can break the partitions and walk anywhere they desire without harming the yet immature progeny. Yet, it appears that a mechanism exists that acts to synchronize the emergence time of young and old bees in the same tunnel- by causing faster development of the former. No experimental explanation is offered for this mechanism, but it is possible that the'farther the bee develops down the tunnel, the less oxygen is available, and hence, the longer developmental duration is. It is possible to speculate on the merits of the cohabitation and mutual tolerance of bees from the same and from different generations within the nest of Xylocopa. A mated female and unmated ones spend the summer, fall, and winter together, whereas during the nesting period an actively nesting mother tolerates a younger, not working, female in her nest. The wood-nesting carpenter bees apparently have a shortage of nesting material, either because of lack of suitable boards, or because of the difficulty of digging a new nest. Hence, the frequent re-use of existing nests until these are no longer suitable, Occupancy of the nest by the younger, nonreproducing bee during the first year of her life may give her priority in the utilization of the same nest a year

13 later. Toleration of the younger female by the older is probably associated with guard duties that the former assumes when tbe latter is on her foraging trips. The fact that such guard duties were necessary was demonstrated by the ease with which removal of absent bees' progeny occurred. These assumptions indicate that the younger bee would benefit from the successful development of her elder's progeny, for they would serve as bel' nest guards next year. They also indicate tbat the emerging bees would benefit by removing their sisters and thus decreasing the number of future competitors for nesting sites. Such removal is made improbable through the synchronization of emergence, coupled with tbe fact that the newly emerging bees are phlegmatic and ill coordinated during tbe first days of their adult life. AcknolVledgemel1ls. We wish to thank Dr. H.O. Yates I of the Southeastern Forest Experiment Station, U.S.D.A. Forest Service, Athens, Georgia, SA for cnabling us to use the X-ray equipment and for his continued interest in our project. We wish to acknowledge, with thanks, the technical help given to us by Mrs. J. Kryspin, N. Gerling, and Mr. A. Hefetz. Thanks are also due to Mr. U. Motro for statistical analysis, Ms. R. Souzin and Mr. S. Shaefer for the illustrations and to Mr. A. Shoob for the photographic work. We are indebted to Dr. P.O. Hurd, Jr., of the Smithsonian Institution, Washington, D.C., Dr. C. D. Michener of The University of Kansas, and Dr. A. Zehavi of Tel-Aviv University for reading the manuscripts and commenting on many points of our work. Altenkirch. G.: Untersuchungen jiber die Morphologie der abdominalen Hautdriisen einheimischer Apiden (Insecta, Hymenoptera). Zool. Beitr. 7, (1962) Anderson, L.M.: Pyridoxine requirement of the honey bee Apis mellifera for brood rearing. M.Sc. Thesis, University of Georgia, Athens, Georgia (1974) Balduf, W.W.: Life of the carpenter bee Xylocopa virginicci (Linn.) (Xylocopidae: Hymenoptera). Ann. Entomol. Soc. A mer. 55, (1962) Hardouin, R.: La vie des abeilles solitaires. 2nd ed. Paris: Gallimard 1948 Hurd, P.O. Jr., Moure, J.S., C.M.F.: A Jew-world subgenus of bamboo-nesting carpenter bees belonging to the genus Xylocopa Latreille (Hymenoptera: Apoidea). Ann. Entomol. Soc. Amer. 53, (1960) Malyshev, Sol.: Lebensgesehiehle der Holzbienen, Xylocopa Latreille (Apoidea). Zeitschr. Morph. Oekol. Tiere 23, (1931) Rau, P.: The jungle bees and wasps of Barro Colorado Island with notes on other insects. Kirkwood, Mo.: 1933 Reaumur, M. de: Memoirs pour servir it I'histoire des insects. Paris: Impr. Roy. 61xxxii 1742 Sabrosky, C.W.: Mating in Xylocopa virgil1iccl (Hymenoptera, Apidae). Proc. Entomol. Soc. Wash. 63, 184 (1962) Sakagami, S.F., Laroca, S.: Observations on the bionomics of some neotropical Xylocopine bees, with comparative and biofaunistic notes (Hymenoptera, Anlhophoridae). J. Fac. Sci. Hokkaido Univ., Ser.IV, Zool. 18, (1971) Skaife, S.H.: The yellow-banded carpenter bee, Mesolrichia cajjra (Linn.) and its symbiotic mite, Dinogamassus. braunsi Vitzthum..I. Entomol. Soc. S. Afr. 15,61-76 (1952) Velthuis, H.H., DeCamargo,.I.M.F.: Observations on male territories in a carpenter bee. X)'locopa (Neoxylocopa) hirswissima Maidl (Hymenoptera, Anthophoridae). Z. Tierpsychol. 38, (1975) Watmough, R.H.: Biology and behaviour of carpenter bees in Southern Africa..I. Entomol. Soc. S. Afr. 37, (1974)