Honey bee (Apis mellifera L.) colonies with either European or Africanized queens mated to European or Africanized drones alone or in combination were tested for defensive behavior using a breath test. The most defensive colonies were those with European or Africanized queens mated to Africanized drones. In colonies where both European and Africanized patrilines existed, most of the workers participating in nest defense behavior for the first 30 s after a disturbance were of African patrilines. Nest defense behavior appears to be genetically dominant in honey bees.
KEY WORDS: colony defense; Africanized honey bees; behavioral genetics; patriline; reciprocal hybrids; subfamily.
The most evident behavioral difference between European honey bees (Apis mellifera L.) (EHB) and Africanized honey bees (A. mellifera scutellata) (AHB) is colony defensive behavior (Stort, 1975; Collins and Kubasek, 1982; Collins et al., 1982). Unlike EHB colonies, which often mount low or moderate responses to intruders, the response of AHB colonies to disturbance is usually extreme. It is not uncommon for hundreds or even thousands of worker bees to attack a perceived intruder when an AHB colony is disturbed.
immediate, particularly within I m of the colony entrance. If the disturbance is brief, such as when a colony is jarred or bumped, workers will rapidly exit the hive in search of the intruder. This activity continues for about 30 s and then declines unless the disturbance continues (Spangler and Sprenkle, 1977).
The worker population in a honey bee colony is genetically diverse because the queen mates with up to 17 drones (Adams et al., 1977). Hence, the worker population is comprised of numerous patfilines. This caused us to ask, When a colony is disturbed, is the population of responding bees indicative of the colony population with regard to patriline composition or does a particular patriline predominate? Genotypic differences in response thresholds among the workers in different patfilines might determine which patriline responds to a disturbance (Robinson and Page, 1989, 1995). If AHB patrilines have a lower response threshold for colony defense than EHB, AHB patrilines should comprise a majority of workers responding to a colony disturbance, particularly immediately after the colony is disturbed. The purpose of this study wsa to determine the proportion of AHB and EHB workers participating in colony defense in hives headed by either AHB or EHB queens with both AHIR and EHB subfamilies present.
Two AHB colonies captured in swarm traps (Schmidt and Thoenes, 1990) in Tucson, AZ, in 1994 were used as the AHB parent colonies (hereafter referred to as AHB-I and AHB-2). Honey bees in the United States are characterized as Africanized based upon morphometric analysis (Rinderer et al., 1993). The AHB-I and AHB-2 parent colonies used in this study each had a probability of Afficanization of 1.0. Mitochondrial DNA (mtDNA) analyses were conducted to determine the matfiline of the parent AHB colonies (Smith, 1988; Sheppard et al., 1991). Workers from AHB-1 and AHB-2 exhibited the EcoRl mtDNA haplotype common in most African races, including Apis mellifera scutellata (Smith, 1988).
The EHB parent colonies were two unrelated European colonies (EHB-1 and EHB-2) carrying the Cordovan (cd) gene for body color. The cd phenotype produces distinctive yellow-brown bees due to homozygous recessive alleles for cuticle color cuticle. We chose cd lines to represent EHB so that subfamilies could be distinguished by body color.
The EHB-1 colony had a cd queen instrumentally inseminated with cd drones, while the EHB-2 colony, which was used as a drone source, contained a naturally mated cd queen. Daughter queens were reared from the AHB-1 and EHB-1 colonies using methods described by Laidlaw (1977). When the queens were 10 days old, they were inseminated with AHB-2 and EHB-2 drone semen (Fig. 1). The semen was collected alternately from individual AHB-2 or EHB-2
drones. To ensure that only AHB-2 drones were used in the inserninations, the colony was placed in a walk-in cage when drones began to emerge. The cd drones could be identified by the color of their cuticle. The inseminated queens were marked on their thorax with paint, a wing was clipped to prevent flight, and they were introduced into five-frume nucleus colonies. The colonies are hereafter referred to as "AHB-1 x mix" and "EHB-1 x mix."
To ensure accuracy in subfamily identifications, we inseminated a group of AHB-I and EHB-I queens with semen from just AHB-2 or EHB-2 drones (single drone line inseminations). This produced colonies of the parental types or reciprocal hybrids. Samples of worker bees were taken from each colony using sealed brood that emerged in an incubator. The bees were preserved as voucher specimens for use in identification of patrilines. The colonies also were included in tests to determine nest defense behavior.
Our crosses with single drone lines produced EHB-2 patriline worker bees that were distinctively different from those with AHB-2. The AHB-2 patriline bees were either completely black or had a black thorax and a solid black stripe across the bottom of each abdominal tergite. Conversely, workers with an EHB-2 patriline were brown and did not have any black on the first three abdominal segments.
Two months after the inseminated queens were introduced into colonies, the worker populations consisted only of offspring from the inseminated queens. Just prior to measurements of defensive behavior, the colony populations were estimated by determining the area of all frames in each colony that were covered by adult bees. Based upon the size of our frames (1759 cm 2 for both sides of the frame) and the average size of the worker bees from all crosses, a frame completely covered with bees contained 2000 individuals (Burgett and Burikam, 1985). The five-frame nucleus colonies were used for all measurements of defensive behavior.
Defensive behavior was measured by the number of workers responding to a human breath stimulus (Boch and Rothenbuhler, 1974). The responding bees were captured in a paper cone (treatment cone) with a clear plastic bag attached to the end (Fig. 2). The paper cone fit snugly into the colony entrance, which was reduced to a hole 2.5 cm in diameter. Inside the treatment cone was a piece of black Velcro (pile side) dangling from a short elastic string. The black Velcro served as both a target and an additional moving stimulus for the responding bees (Lecomte, 1952; Free, 1961). Each trial consisted of first placing a cone without the Velcro path (control cone) in the colony entrance for 30 s (the control period). After the control period, the control cone was removed and a cork was placed in the end to prevent the bees from escaping. Next a clear plastic tube (140 cm long with a diameter of 2.1 cm) was placed in the colony entrance. A coinvestigator blew into the colony entrance three times through the plastic tube. The tube was removed and the treatment cone was immediately placed in the colony entrance for 30 s (treatment period). The treatment cone then was removed and corks were placed in the entrances of the cone and the colony. The bees captured in the cones during the control and treatment periods were immediately frozen. All 21 colonies, including AHB-I, AHB-2, EHB-I, and those headed by queens inseminated with single and mixed parental-type semen, were sampled within the same 40-min interval once a week for 6 weeks.
The number of honey bees from each patriline (AHB or EHB) captured in the cone traps was recorded for every colony. The patriline was determined by comparing the workers with voucher specimens. Paired t tests were conducted to determine if there were differences in treatment and control-period data from
colonies of sister queens inseminated with the same semen type (Ott, 1977). A repeated-measures analysis of variance (ANOVA) was conducted to determine if the parental colonies were significantly different in their defensive behavior (Sokal and Rohlf, 1995). A second repeated-measures ANOVA was conducted using data from colonies where queens were inseminated either with a single parental type or with the AHB-2 and EHB- 2 semen mix. Data were transformed [log (x + 1)] prior to the ANOVA, because the variance was not homogeneously
distributed among the means. A paired t test was conducted to determine if there were differences in the average proportion of workers from each patriline captured in the cones. The percentage of worker bees in each patriline in the colony population was detennined by placing cages over areas of sealed brood and recording and number of emerged bees (100 115 worker bees per colony) with each patriline. An ANOVA was conducted to determine if the "mix" colonies differed in their proportions of workers with either patrilines. This was followed by a chi-square test to deten- nine if the proportion of hive bees in each subfamily differed from that collected in the treatment cone.
The colonies used for nest defense measurements differed in population size by about 1200 worker bees. Colonies with the largest populations contained about 4000 workers. Because some of the bees immediately flew off the frame or ran to the sides of a frame and onto the walls of the colony when the hive was opened, population estimates were relative approximations. Counts of honey bees captured in the cone traps were adjusted based upon the relative population size of the colony using the equation, (4000/colony population size) * number of bees in the trap. All data analyses were conducted on adjusted counts.
Worker bees immediately responded to the human breath stimulus by rushing out of the colony and into the treatment cone. In all cases, significantly more bees were captured in the treatment cones compared with the controls. Of the three parent colonies, AHB-I and AHB-2 were the most defensive (Table 1). Workers from the EHB-1 colony captured in the treatment cone were comprised of 56.4% EHB-I workers; the remainder of the bees were drifters from other AHB colonies, as determined by their body color. Inspection of the colony indicated that EHB drifters comprised < I % of the colony population.
Regardless of the queen type (AHB or EHB), the most defensive colonies were those containing workers fathered by only AHB-2 drones, Colonies containing both AHB-2 and EHB-2 patfilines were equally defensive. Worker progeny of AHB-1 queens inseminated with only EHB-2 semen were the least defensive. In fact, all colonies containing EHB patrilines (AHB-1 X EHB-2, AHB-1 x mix, EHB-1 X mix) had reduced defensiveness compared with colonies containing only the AHB-2 patriline.
The AHB-2 patriline comprised the majority of bees captured in the treatment cones in colonies containing both patrilines (Table 11). The proportion of bees with either patriline in the "mix" colonies was not significantly different (F = 0.44, df = 1,9, p = 0.53; proportion of each patriline in the "mix" colonies-AHB = 0.68, EHB = 0.32). The proportion of AHB-2 patriline bees captured in the treatment cones was significantly greater than their proportion
in the hives (AHB- I X mix, X 2 = 37.0, df = 3, P < 0.05; EHB-1 x mix, X 2 = 29.4, df = 3, P < 0.05).
The responding workers with AHB-2 patrilines appeared not to stimulate their EHB-2 patriline half-sisters into exhibiting defensive behavior during this interval. If they had, the proportion of workers with EHB patrilines in our samples would have been comparable to that found in the colony population.
Our study measured the defensive response in colonies during the first 30 s after a disturbance to capture those responding workers with the lowest response thresholds. Had we continued to disturb the colony, it is possible that workers with higher response thresholds would also have been captured. In those colonies with the AHB-2 patriline, we captured an average of about 5% of the colony population in the first 30 s after the disturbance. If this proportion is representative, an average-sized colony of 30,000 bees could mount a defensive response of 1500 bees in 30 s.
The inheritance of defensive behavior in honey bees has been the subject of numerous studies (Collins et al., 1984, 1987, 1988; Guzman- Novoa and Page, 1993, 1994). Our findings were similar to these reports in which most of the workers responding to colony disturbances were AHB hybrids (GuzmanNovoa and Page, 1994). This occurred even when a small number of AHB workers drifted to another colony as in the case of EHB-1. The most defensive colonies in our study were those with AHB patrilines, yet the reciprocal cross created the least defensive colonies. These results support the hypothesis that queen genotype has little effect on colony defensive behavior (Guzman-Novoa and Page, 1993). Our data also indicate that defensive behavior is a genetically dominant trait that might also be influenced by a paternal factor (Guzman-Novoa and Page, 1994).
Findings of this study support the genotypic threshold model for honey bee behavior (Page and Robinson, 1991; Robinson and Page, 1995). Based upon the defensive behavior of the parental colonies, the A1113-2 patriline workers appeared to have lower thresholds for nest defense than the EHB-2. In all instances, most of the workers responding to the colony disturbance were the AHB-2 patriline. Our data indicate that in colonies with AHB and EHB patrilines, the initial line of defense is comprised primarily of the AHB patriline. Hence, the frequency of AHB patrilines in a colony might ultimately determine its overall defensive behavior.
The authors thank S. Machtley, N. Morales, and M. Templin for excellent technical assistance, W. S. Sheppard for mtDNA analyses, and John Harbo, Roger Hoopingarner, Stan Schneider, and two anonymous reviewers for their suggestions on and constructive criticisms of early versions of the manuscript. This research was funded by a USDA Pilot Test Program Grant to G.D.-H. and A.C.
Adams, J., Rothman, E. D., Kerr, W. E., and Paulino, Z. L. (1977). Estimation of the number of sex alleles and queen matings from diploid male frequencies in a population of Apis mellifera. Genetics 86: 583-596.
Boch, R., and Rothenbuhler, W. C. (1974). Defensive behavior and production of alarm pheromone in honeybees. J. Apic. Res. 13: 217-221.
Burgett, M., and Burikam, 1, (1985). Number of adult honey bees (Hymenoptera: Apidae) occupying a comb: A standard for estimating colony populations. J. Econ. Entomol. 78: 1154-1156.
Collins, A. M., and Kubasek, K. J. (1982). Field test of honey bee (Hymenoptera: Apidae) colony defensive behavior. Ann. Entomol. Soc. Am. 75: 383-387.
Collins, A. M., Rinderer, T. E., Harbo, J. R., and Bolten, A. B. (1982). Colony defense by Africanized and European honey bees. Science 218: 72-74.
Collins, A. M., Rinderer, T. E., Harbo, J. R., and Brown, M. A. (1984). Heritabilities and correlations for sever-al characters of the honey bee, Apis mellifera. J. Hered. 75: 135-140.
Collins, A. M., Rinderer, T. E., Tucker, K. W., and Pesante, D. G. (1987). Response to alarm pheromone by European and Africanized honey bees. J. Apic. Res. 26: 217-223.
Collins, A. M., Rinderer, T. E., and Tucker, K, W. (1988). Colony defense of two honey-bee types and their hybrids. 1. Natur-al matings. J. Apic. Res. 27: 137-140.
Guzman-Novoa, E., and Page, R. E., Jr. (1993). Backcrossing Africanized honey bee queens to European drones reduces colony defensive behavior. Ann. Entomol. Soc. Am. 86: 352-355.
Guzman-Novoa, E., and Page, R. E., Jr. (1994)- Genetic dominance and worker interactions affect honeybee colony defense. Behav. Ecol, 5; 91- 97.
Lecomte, J. (1952). Recherches sur le comportment agressif des ouvrieres d'Apis mellifera. Behaviour 4: 60-66.
Ott, L. (1977). An Introduction to Statistical Methods and Data Analysis, Duxbury Press, North Scituate, MA.
Page, R. E., and Robinson, G. E. (1991). The genetics of the division of labor in honey bee colonies. Adv. Insect Physiol. 23: 117-169.
Rinderer, T. E., Buco, S. M., Rubink, W. L, Daly, H. V., Stelzer, J. A., Riggio, R. M., Baptista, F. C. (1993). Morphometric identification of Africanized and European honey bees using large reference populations. Apidologia 24: 569-585.
Robinson, G. E., and Page, R. E., Jr. (1989). Genetic basis for division of labor in honey bee colonies. In Breed, M. D., and Page, R. E. (eds.), The Genetics ofSocial Evolution, Westview Press, Boulder, CO, pp. 61-80.
Robinson, G. E., and Page, R. E., Jr. (1995). Genotypic constraints on plasticity for corpse removal in honey bee colonies. Anim. Behav. 49: 867-876.
Sheppard, W. S., Rinderer, T. E., Mazzoli, J., Stelzer, J. A., and Schimanuki, H. (1991). Gene flow between African- and European-derived honey bee populations in Argentina. Nature (Lond.) 349: 782-784.
Smith, D. R. (1988). Mitochondrial DNA polymorphisms in five Old World subspecies of honey bees and in New World hybrids. In Needham, G. R., Page, R. E., Jr., Delfinado Baker, M_ and Bowman, C. E. (eds.), Africanized Honey Bees and Bee Mites, Ellis Horwood, Chichester, UK,
Spangler, H. G., and Sprenkle, D. J. (1997). An instrument for quantifying honey bee defensiveness. Appl. Acoust. 50: 325-332.
Stort, A. C. (1975). Genetic study of aggressiveness of two subspecies of Apis mellifera in Brazil. 2. Time at which the first sting reached a leather ball. J. Apic. Res. 14: 171-175.
'Carl Hayden Bee Research Center, USDA-ARS, 2000 East Allen Road, Tucson, Arizona 85719. 2 Honey Bee Laboratory, USDA-ARS, BARC-East, Building 476, Beltsville, Maryland 20705.