Cell Size
Mating Behaviour
  David A. Cushman logo  

Size related Mating Preferences in Honey Bee Drones

A paper by Joseph R. Coelho and Orley R. Taylor Jr.


We constructed models of honey bee queens as black, hollow tubes supplied with queen pheromone. Between models differing only in inside diameter (ID), drones "preferred" the models of larger ID, but drones mating with different sized models were not significantly different in size. Between models differing only in outside diameter (OD), drones mating with the smaller OD models were significantly smaller than those mating with larger OD models, but both models were mated equally frequently. In both experiments, copulating drones were significantly smaller than trapped drones. Assortative mating for OD may be a result of matching tactile cues. Among populations of honey bees, drone size should covary with queen size, resulting in stabilizing selection within each sex.


Mating in honey bees (Apis mellifera L.) occurs in the air, 10 - 40 m from the ground (Ruttner, 1966). Drones (males) take flight in the afternoon, passing along well-defined flyways, reorienting themselves at drone congregation areas (DCAs), without landing until returning to the hive (Ruttner, 1966; Loper et al., 1987). Because hundreds and sometimes thousands of drones can be in flight at the time of mating (Taylor, 1984a) and virgin queens are rare, the operational sex ratio is strongly skewed toward males. Drones are extremely variable in size, spanning nearly a two-fold range in thorax mass (Coelho, 1991b). Since mating occurs in flight under conditions which are potentially competitive for males, it is of interest to know if there is size-related differential success in mating among drones.

Because matings occur high above the ground and often involve high speed aerial chases by large numbers of drones, direct observations of the relative success of drones with different characteristics have not been possible. In addition, recovery of successfully mating drones is difficult because they become instantly paralyzed during copulation, and immediately fall to the ground (Ruttner, 1966). Therefore, drone behavior has been studied by observing drones attracted to suspended live queens, dead queens, queen models, and pheromone-impregnated wicks in drone-rich areas (Gary, 1963; Butler and Fairey, 1964; Ruttner and Ruttner, 1968; Gary and Marston, 1971; Gerig, 1971; Koeniger et al., 1979, 1989; Taylor, 1984a, 1984b; Lensky and Demter, 1985; Koeniger, 1986).

A drone attempting to copulate grasps the queen's abdomen with all six legs and forcibly everts the endophallus into the sting chamber of the queen (Koeniger et al., 1979). The size and shape of the queen may, therefore, have effects on the types of drones which mate her. To determine the effect of queen morphology on drone mating behavior, we examined the characteristics of drones copulating with queen models of differing dimensions.

Materials and Methods

Experimental Protocol

The study was conducted on the West Campus of the University of Kansas, Lawrence (Douglas County) during May, 1989. Twenty queen models were simultaneously suspended 10 m from the ground on a cord between two towers, and lowered every 5 min to remove post-copula drones, which remained stuck to the models (see Taylor, 1984b). Copulators were sealed in marked 1.5 ml centrifuge tubes and frozen for later measurements. Each model was fastened to the line by the "head" via an L-shaped wire holder (similar to that described by Taylor [1984b]) which prevented lateral movement and oriented the model downwind. The location of the towers was believed to be across a drone flyway, as opposed to a drone congregation area, and ˜70 m from an apiary (˜20 colonies) which was managed for high drone production by placing 1 frame of drone comb in each hive body.

For comparison, a sample of the drone population in the immediate vicinity of the towers was taken using a Taylor (1984a) drone trap suspended from a kite. Three such samples were taken during each afternoon while experiments were being performed, and at approximately the same height as the models. Drones removed from the trap were killed in a cyanide-charged insect-killing jar and frozen in airtight bags for later measurements.


Models of honey bee queens were constructed from the plastic tip covers of 1 ml syringes (Becton-Dickinson). A "petiole" was created by cutting out a wedge of plastic behind the "thorax", the enlarged portion of the tube which originally had fit around the tip of the syringe (Figure 1). A "head" was created by removing the small, black, rubber tip from the syringe plunger and inserting it into the anterior end of the "thorax". Each tube was painted black with a felt-tip marker. A key feature of these models was that the "abdomens" were flexed ventrally by bending them at the "petiole". Observations on drones attempting to mate with earlier prototypes indicated that drones often could not locate the "sting chamber", tending to probe beyond it to the ventral surface of the abdomen. This ventral flexion appeared to greatly facilitate mating, perhaps due to its mimicking the natural flexion of a real queen's abdomen.

A bit of cotton was forced down the tubes to the gap at the "petiole". The cotton balls were supplied with two drops of methylene chloride extract of queen honey bees. In a given trial, all models received queen extract from the same bottle to insure approximately equal olfactory attractiveness, and experiments were begun immediately after application of the extract. The total length of the model was 22.7 mm. The "thorax" was 6.8 mm in outside diameter and 4.7 mm in length, while the "abdomen" was 11 mm in length.

Model of honey bee queen used in mating experiments

Figure 1. Model of honey bee queen used in mating experiments.


To insure that the dimensions of the models fell within a reasonable range, six virgin queens were measured for abdominal width at the broadest point, length of abdomen, and width of sting chamber, using a hand-held micrometer. Virgin queens had a mean abdominal width of 4.8 ± 0.1 mm, abdominal length of 11.2 ± 0.4 mm, and sting-chamber width of 2.1 ± 0.1 mm. Thorax masses of drones were determined to the nearest mg on a Precisa 100M-300C balance. Wind speed was determined with a Davis Instruments (San Leandro, CA) Turbo Meter rotary anemometer to the nearest 0.1 m/s. Ambient temperature was measured to the nearest 0.1°C with a Bailey BAT-12 digital thermocouple thermometer. Data are reported as mean ± SE.

Inside Diameter Test

Twenty models were constructed, all with the same outside diameter (OD) of the "abdomen", 4.7 ± 0.02 mm. The "sting chamber" of the tube was bored to the desired inside diameter (ID) using an electric drill. Ten were bored to an ID of 2 mm, and ten to an ID of 3 mm. The two types of models were placed alternately along the line. During the experiment ambient temperature ranged from 23.5 to 27°C, and wind (from the southwest) was between 1 and 2 m/s in speed. Skies were partly to mostly cloudy.

Outside Diameter Test

Twenty models were constructed, and all were bored to the same ID (3 mm). The "abdomens" of the models were thickened to the desired OD by repeatedly dipping them into melted beeswax mixed with black fabric dye. Ten "small" models had OD of 4.92 ± 0.02 mm, and ten "large" models had OD of 6.34 ± 0.05 mm. Again, the two types of models were placed alternately along the line. During the experiment ambient temperature varied from 29 to 30°C, skies were clear, and a southeast wind was gusty, varying from 0.5 to 4 m/s. .



Drone bees approached from below and downwind of the models, presumably in response to queen pheromone and perhaps to the visual stimulus of the model highlighted against the sky. Groups of drones formed characteristic "comets" oriented toward the models (Gary, 1963). Drones were strongly attracted to these models (as opposed to larger prototypes). During the peak flight period from approximately 1500 to 1600 hrs, all of the models were "mated" in less than 2 min. Drones began flying at about 1400 hrs, but generally did not copulate until ˜30 min later. Drone flight generally ceased at ˜1700 hrs. Drones were strongly attracted to the Taylor drone trap, except under very cloudy conditions. They generally entered in large numbers, while none was seen to escape.

Inside Diameter Test

Drones mating with models differing only in ID were not significantly different in thorax mass (80.3 ± 1.7 vs. 79.8 ± 2.0 mg (P = 0.81, T-test). There was, however, a distinct preference for the larger ID models (18 small vs. 74 large, P < .001, binomial test; Figure 2). The thoraxes of both groups of copulating drones in this test weighed significantly less than those of a trapped sample of the population (87.9 ± 2.0 mg, n = 14, P < 0.01, T-test).

Size distribution of honey bee drones copulating with queen models of differing inside diameter

Figure 2. Size distribution of honey bee drones copulating with queen models of differing inside diameter.

Outside Diameter Test

Drones mating with the smaller OD models had significantly lower thorax mass than those mating with the larger OD models (78.4 ± 1.5 mg vs. 82.5 ± 1.3 mg respectively, P = 0.038, T-test); however, there was no overall preference for either model (49 small vs. 67 large, P > 0.05, binomial test; Figure 3). Thoraxes of both groups of copulating drones in this test were again smaller than those of a trapped sample of the population (86.3 ± 1.2 mg, n = 85, P < 0.01, T-test). Extreme variability in the size of drones was evident in this experiment in that copulating drones ranged in thorax mass from 48.0 to 103.0 mg. Similar levels of variation were found for other morphometric measurements, such as wing length and wing area.

Size distribution of honey bee drones copulating with queen models of differing outside diameter [at same scale as inside dia bar chart]

Figure 3. Size distribution of honey bee drones copulating with queen models of differing outside diameter.


Drones demonstrate a strong preference for queen models having a specific range of ID's of the simulated sting chamber. We tested models having two ID's, 2 and 3 mm, and drones "preferred" the 3 mm ID models, despite the fact that real queens are closer to 2 mm in ID of sting chamber. These results confirm the data of Gary and Marston (1971), who tested models ranging in ID from 1.6 to 4 mm, and found that drones "preferred" models with ID's of 3.2 and 4.0 mm. This apparent preference for sting chambers of larger than normal diameter may be due to the ease of detecting an open sting chamber if it is larger. The size of the sting chamber is probably assessed when the pubescent terminal segments of the drone's abdomen make contact with the queen model's abdomen, and if the chamber is open sufficiently, further stages in copulation are elicited (Gary and Marston, 1971). Hence, larger ID's would more readily result in the completion of copulation. In contrast, drones in general had no strong preference for a model of particular OD.

Drones did not assortatively mate on the basis of ID -- a "sting chamber" of acceptable size is apparently mateable by drones of all sizes. They did, however, mate assortatively on the basis of OD -- larger drones mated the large OD models, and smaller drones mated the small OD models.

It is possible that differential mechanical and tactile cues resulted in assortative mating by drones. The drone's first two pairs of legs grasp the model abdomen dorsally, while the last pair of legs grasp it laterally and ventrally (Gary and Marston, 1971). Small drones may have been unable to grasp the larger OD models properly in order to copulate successfully, or they may not have received the proper tactile stimuli for copulation to be elicited. The same might be said of large drones mating with small OD models. Size matching could occur through cues for mounting position. If drones use alignment cues from the anterior portion of the queen, small drones could be too far forward on the queen abdomen and the large drones too far back for copulation to be completed.

No matter what proximate factors account for assortative mating by drones, this mating pattern could have powerful ecological and evolutionary consequences. A limited degree of positive assortative mating has been observed between subspecies of Apis mellifera (Kerr and Bueno, 1970; Koeniger et al., 1989). Differences in the sizes of subspecies are well documented (Daly and Balling, 1978; Rinderer et al., 1985; Ruttner, 1987). It is conceivable that mating behavior of drones based on size is partly responsible for assortative mating and maintenance of races of honey bees.

Variation in queen size can occur as a result of age of larvae used for queen rearing, resource availability, and rearing conditions. We have shown that drones are extremely variable in size. In one study drones of sympatric African and European subspecies differed in body mass (Rinderer et al., 1985); therefore, it is likely that drone size has a heritable component. The mating tendencies we observed would, consequently, result in a generalized pattern of size matching between the sexes. The intensity of selection acting on drone size should be a function of the frequency distribution of queen size. Stabilizing selection for drone size would result if queen size is normally distributed. Among populations of honey bees drone size should, therefore, track mean queen size. Most matings would be between intermediate-sized drones and intermediate-sized queens. Hence, drones and queens would covary in size, resulting in stabilizing selection within each sex.

In a queenless colony, some workers lay eggs which develop into very small drones (Cale and Rothenbuhler, 1975). One might expect these drones to have little opportunity to mate because large drones have larger thoraxes and higher thorax temperatures, and, consequently, greater force production and faster flight (Coelho, 1991a; 1991b; Ross & Coelho, 1995). Due to the need for matching cues, however, small drones may be better able to mate small queens.

In this series of experiments copulating drones were smaller than a sample of the population collected by a trap. Our previous efforts have indicated that the size of copulating drones relative to the trapped sample varies from day to day. In fact, sometimes copulators are larger (O.R. Taylor, Jr. and J.R. Coelho, unpublished data). The size of copulating or noncopulating drones may be related to meteorological or other factors that may be identified in future studies.


The authors would like to thank J.A. Hunley and C. Pennuto for collecting preliminary data which led to this work. This study was supported in part by a University of Colorado Fellowship and a Graduate School Foundation Fund Award to JRC.

Literature Cited

Butler, C.G., and Fairey, E.M. 1964. Pheromones of the honeybee: biological studies of the mandibular gland secretion of the queen. Journal of Apicultural Research 3:65-76.

Cale, G.H., and Rothenbuhler, W.C. 1975. Genetics and breeding of the honey bee. In Dadant & Sons, (eds.) The hive and the honeybee. Dadant & Sons, Hamilton, IL, pp. 157-184.

Coelho, J.R. 1991a. The effect of thorax temperature on force production during tethered flight in honeybee (Apis mellifera) drones, workers, and queens. Physiological Zoology 64(3):823- 835.

Coelho, J.R. 1991b. Heat transfer and body temperature in honey bee (Hymenoptera: Apidae) drones and workers. Environmental Entomology 20(6):1627-1635.

Daly, H.V., and Balling, S.S. 1978. Identification of Africanized honeybees in the Western Hemisphere by discriminant analysis. Journal of the Kansas Entomological Society 51:857-869.

Gary, N.E. 1963. Observations of mating behaviour in the honeybee. Journal of Apicultural Research 2:3-13.

Gary, N.E., and Marston, J. 1971. Mating behaviour of drone honey bees with queen models (Apis mellifera L.). Animal Behavior 19:299-304.

Gerig, L. 1971. Wie Drohnen auf Kniginnenattrappen reagieren. Schweiz. Bienenztg. 94:558-562.

Kerr, W.E., and Bueno, D. 1970. Natural crossing between Apis mellifera adansonii and Apis mellifera ligustica. Evolution 24:145-148.

Koeniger, G. 1986. Mating sign and multiple mating in the honeybee. Bee World 67:141-150.

Koeniger, G., Koeniger, N., and Fabritius, M. 1979. Some detailed observations of mating in the honeybee. Bee World 60:53-57.

Koeniger, G., Koeniger, N., Pechhacker, H., Ruttner, F., and Berg, S. 1989. Assortative mating in a mixed population of European honeybees, Apis mellifera ligustica and Apis mellifera carnica. Insectes Sociaux 36:129-138.

Lensky Y., and Demter, M. 1985. Mating flights of the queen honeybee (Apis mellifera) in a subtropical climate. Comparative Biochemistry and Physiology 81A:229-241.

Loper, G.M., Wolf, W.W., and Taylor, O.R. Jr. 1987. Detection and monitoring of honeybee drone congregation areas by radar. Apidologie 18:163-172.

Rinderer, T.E., Collins, A.M., and Pesante, D. 1985. A comparison of Africanized and European drones: weights, mucus gland and seminal vesicle weights, and counts of spermatozoa. Apidologie 16:407-412.

Ross, A.J., and J.R. Coelho. 1995. The effect of thorax temperature on flight speed in honey bee drones (Apis mellifera). Pages 987-991 in Proceedings of the Undergraduate Research Conference. Volume III (Robert D. Yearout, ed.). University of North Carolina at Asheville, 1149 pp.

Ruttner, F. 1966. The life and flight activity of drones. Bee World 47:93-101.

Ruttner, F. 1987. Biogeography and taxonomy of the honeybees. Springer-Verlag, Berlin.

Ruttner, F., and Ruttner, H. 1968. Untersuchungen uber die Flugaktivitt und das Paarungsverhalten der Drohnen. 4. Zur Fernorientierung und Ortsstetigkeit der Drohnen auf ihren Paarungsflugen. Zeitschrift fur Bienenforschung 9:259-265.

Taylor, O.R., Jr. 1984a. An aerial trap for collecting drone honeybees in congregation areas. Journal of Apicultural Research 23:18-20.

Taylor, O.R., Jr. 1984b. A mating tube for studying attractiveness of queen honeybees and mating behaviour of drones. Journal of Apicultural Research 23:21-24.

Written... 18 September 2005,
This page has actually been validated by W3C Javascript Navigational elements removed as per W3C Link Checker version 4.1 (c) 1999-2004 Requirements