African honey bees (Apis mellifera scutellata) dramatically changed the South American beekeeping industry as they rapidly spread through the Americas following their introduction into Brazil. In the present study, we aimed to determine if the management of European-derived honey bees (A. mellifera sspp.) could reduce the relative abundance of African-matriline drones at regional mating sites known as drone congregation areas (DCAs). We collected 2,400 drones at six DCAs either 0.25 km or >2.8 km from managed European-derived honey bee apiaries. The maternal ancestry of each drone was determined by Bgl II enzyme digestion of an amplified portion of the mitochondrial Cytochrome b gene. Furthermore, sibship reconstruction via nuclear microsatellites was conducted for a subset of 1,200 drones to estimate the number of colonies contributing drones to each DCA. Results indicate that DCAs distant to managed European apiaries (>2.8 km) had significantly more African−matriline drones (34.33% of the collected drones had African mitochondrial DNA) than did DCAs close (0.25 km) to managed European apiaries (1.83% of the collected drones had African mitochondrial DNA). Furthermore, nuclear sibship reconstruction demonstrated that the reduction in the proportion of African matriline drones at DCAs near apiaries was not simply an increase in the number of European matriline drones at the DCAs but also the result of fewer African matriline colonies contributing drones to the DCAs. Our data demonstrate that the management of European honey bee colonies can dramatically influence the proportion of drones with African matrilines at nearby drone congregation areas, and would likely decreasing the probability that virgin European queens will mate with African drones at those drone congregation areas.
Citation: Mortensen AN, Ellis JD (2016) Managed European-Derived Honey Bee, Apis mellifera sspp, Colonies Reduce African-Matriline Honey Bee, A. m. scutellata, Drones at Regional Mating Congregations. PLoS ONE 11(8): e0161331. https://doi.org/10.1371/journal.pone.0161331
Editor: Raul Narciso Carvalho Guedes, Universidade Federal de Vicosa, BRAZIL
Received: January 11, 2016; Accepted: August 3, 2016; Published: August 12, 2016
Copyright: © 2016 Mortensen, Ellis. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the paper and its Supporting Information files.
Funding: Funding for this research was awarded to JDE and ANM (project number: 00113280) by the Florida Department of Agricultural and Consumer Services under the advisement of the Honey Bee Technical Counsel [http://www.freshfromflorida.com]. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
African honey bees, Apis mellifera scutellata Lepeletier (1836), are native to semi-arid areas in sub-Saharan Africa. African-derived honey bees rapidly spread through the Americas following their introduction into Brazil in the 1950s [1,2]. Despite the efforts of beekeepers and local agencies to conserve European−derived honey bees (including A. m. ligustica Spinola (1806), carnica Pollmann (1879), mellifera Linnaeus (1758), caucasica Pollmann (1889), and iberiensis Engel (1999) [3,4]), African-derived honey bees now predominate the American tropics [5,6]. African-derived honey bees in the Americas are genetically distinguishable from their ancestral population in Africa . However, for conciseness, we will refer to both African- and European-derived honey bees in the Americas as African and European honey bees, respectively.
In South and Central America, managed African honey bees quickly become the predominate subspecies upon entering new territory . African honey bees are better adapted to the tropical environment of South and Central America than are European honey bees and they outperform European honey bees in terms of survivability and honey production there . However, the heightened defensive behavior expressed by African honey bees has dramatically changed the South American beekeeping industry. For example, beekeepers relocated their apiaries into remote, isolated locations and hobbyist beekeeping has become nearly nonexistent .
In contrast, African bees are considered undesirable by beekeepers bees in temperate climates for a number of reasons. First, African honey bees are less likely to outperform temperate adapted European bee races in temperate climates . For example, European honey bees perform well in high population densities where food is the limiting resource , while African honey bees tend to abscond to areas with plentiful resources . The portion of the bee population with more European ancestry seems favored over that with an African ancestry when competition for limited resources arises . Moreover, the African honey bee’s elevated defensive behavior and high propensity to swarm/abscond make African honey bees undesirable to many beekeepers  and a potential safety concern for animals and humans. Therefore, beekeepers managing European bees in areas where African bees are present alter their management practices to limit the integration of African genetics into their managed bees.
One strategy used by beekeepers trying to limit the impact of African bees in the managed European bee population is saturating the environment with European colonies. Beekeepers believe this saturation can modify the regional reproductive population so that European queens are more likely to mate with European drones rather than hybridize with African drones, though this has not been tested directly.
Honey bee mating occurs at distinct locations called drone congregation areas (DCAs ). Large numbers of drones aggregate daily at DCAs between 10–30 m above the ground with the intent to mate with a queen . Drones typically do not travel far to a DCA. The majority of drones attend a DCA < 0.5 km from their colony and very few travel > 2.0 km to a DCA . “Africanization” of managed European honey bee colonies occurs through paternal introgression of African genetics (i.e. a European queen mates with an African drone ). Therefore, the ability to modify the proportion of African drones at a DCA could reduce the likelihood of Africanized colonies in managed apiaries.
In Chiapas, Mexico, Loper et al.  simultaneously employed three techniques to manipulate the reproductive population: 1) trapping and removal of African drones prior to introducing European drone-source colonies; 2) introduction of large numbers of European drone-source colonies to flood the regional population; and 3) strategic placement of those drone-source colonies. However, they were unable to ascertain which factor was most impactful due to the experimental design.
In this study, we aimed to determine if the management of European honey bees could decrease the proportion of African drones present at nearby drone congregations areas. We expected that the management of European colonies would result in proportionally more European drones, and correspondingly fewer African drones, at nearby DCAs.
Materials and Methods
The northern expansion of the African honey bee population in the Southeastern United States has been stable in central Florida since 2005 . Six DCAs were identified within the African honey bee range in Orange and Osceola counties, Florida per . Three of the six DCAs were located within 0.25 km of 96, ten-frame, commercial European colonies kept in Langstroth-style hives. The three remaining DCAs were located > 2.8 km from any managed colonies. Since drones typically travel < 2.0 km to a DCA , drones trapped>2.8 km from managed colonies were considered representative of the feral population.
Drones were collected through the duration of the flight time using a Williams  drone trap equipped with a queen lure constructed of a 2 cm piece braided dental roll died black with India ink. The queen lure was baited with 1 mg of synthetic 9−oda and suspended 20 cm below the trap opening [22–24]. The Williams trap was attached 5 m below a white 1.2 m chloroprene balloon. The height of the trap was defined as the distance from the bottom of the trap opening to the ground . A minimum of 400 drones was collected from each of the six DCAs. Within each DCA, drones were caught at two heights: a minimum of 200 drones from 10 m above the ground and a minimum of 200 drones from 30 m above the ground. Drones were preserved in 95% ethanol and transported to the laboratory where they were stored at −80°C until molecular processing.
Mitochindrial (mtDNA) Analysis
Four hundred drones (200 drones from 10 m above the ground and 200 drones from 30 m above the ground) were randomly selected from the total drones collected from each DCA for mtDNA analysis. Total DNA was extracted from a hind leg of each drone using 10% Chelex . Maternal ancestry was determined for each drone using PCR-RFLP for a diagnostic portion of the mitochondrial Cytochrome b gene . A 485−bp fragment of the cytochrome b gene was amplified by the PCR then digested by Bgl II restriction enzymes as described by Pinto et al. . European cytochrome b fragments contain the Bgl II restriction site and are cleaved into 291− and 194−bp fragments. African cytochrome b fragments lack the Bgl II restriction site and remain intact (485−bp) . The PCR−RFLP products were visualized on 2.0% agarose gels to identify cleaved (European) or intact (African) mtDNA fragments for each drone.
Subsets of 200 drones were selected from each DCA for nuclear DNA analysis (1200 drones total). Ten microsatellite loci (HB−THE−03, HB−THE−04, HB−C16−05, HB−SEX−01, HB−SEX−03, A7, A79, A88, A107 and B124 [12, 28, 29]) were amplified via PCR (Table 1) from the total DNA extracted for the mtDNA analysis. All forward primers were fluorescently labeled with FAM, VIC, PET, or NED, and all reverse primers were pigtailed (addition of GCTTCT ) to minimize stutter on the fragment length results. Individual 10 μl PCR reactions were run for one or two loci at a time. The decision to plex loci for PCR reactions was based on the fluorescent label of the forward primer for those loci and the previously published fragment length range of the loci. Only loci with the same fluorescent label and distinct fragment length ranges were plexed for PCR reactions. The PCR conditions used for all microsatellite reactions are described by Shaibi et al. . The PCR products from each reaction were then diluted by the ratios described in Table 1 and combined into the two analyses plexes. The final plexes were sent to the University of Florida’s Interdisciplinary Center for Biotechnology Research and fragment lengths were analyzed with a 3730xl DNA Analyzer (Applied Biosystems).
All PCR reaction were 10 μl total volume. Data are the loci that were grouped for PCR reactions (Reaction Plex); names of the microsatellite (Loci Name); the previously published fragment length of the amplified microsatellites (Fragement Length Ranges); Primer sequence information (including the fluorescent label of the forward primer and the pigtail sequence on the reverse primer (highlighted in bold), Primers); the molar concentration of each primer in each PCR reaction (Primer Concentration); the post reaction plexing groups for analysis; and the dilution ratio for each reaction plex prior to generating the analysis plex (Dilution for Analysis).
Microsatellite fragment lengths were scored using GeneMarker© 2.4.0 (SoftGenetics). Any individuals with failed amplification at more than two of the ten loci were excluded from statistical analysis. There were two linkage groups represented in the 10 microsatellites analyzed, HB−THE-−03 & HB−THE−04 and HB−SEX−01 & HB−SEX−03. Haplotypes were manually assigned to individual drones for each linkage group . Sibship reconstruction was run by COLONY© 184.108.40.206 (ZSL Institute of Zoology) to estimate the number of colonies that had produced the drones present at each DCA [31,32,33,34].
The proportion of both maternal ancestries (European and African) was compared for drones from each DCA by the proximity of the DCA to managed European honey bee colonies (near (0.25 km) or distant (> 2.8 km)) and height of the DCA from the ground (10 or 30 m) using Pearson’s χ2 test in JMP® 11.0.0 (SAS Institute Inc).
Counts of the number of colonies that had produced the drones present at each DCA were log transformed and compared via one-way ANOVAs based on the proximity of the DCA to managed European honey bee colonies (near (0.25 km) or distant (> 2.8 km)). Additionally, mtDNA results were combined with nuclear sibship data to determine the number of colonies that had African mtDNA and had produced the drones present at each DCA. The counts of colonies with African mtDNA were transformed by logging the count + 0.5. One−way ANOVA was run on the transformed African mtDNA colony data based on the proximity of the DCA to managed European honey bee colonies (near (0.25 km) or distant (> 2.8 km)). Lastly, the proportion of the total colonies detected that had African mtDNA was calculated and compared via one-way ANOVAs based on the proximity of the DCA to managed European honey bee colonies (near (0.25 km) or distant (> 2.8 km)). All drones used in the nuclear analysis came form the 10 m elevation sampling. Therefore, vertical height within the DCA was not included in the nuclear DNA analysis as it was in the mtDNA analysis.
All DCAs were located on private property, and permission for sampling was granted by the property owner. No formal permits were required for field collections or laboratory analysis because honey bee research is not regulated by animal use committees such as the Institutional Animal Care and Use Committee (IACUC).
Mitochondrial DNA analysis results indicate that DCAs distant to managed European honey bee apiaries (> 2.8 km) had significantly more (χ21, 2400 = 427.83, p <0.0001) African matriline drones (34.33% of the collected individuals) present than did DCAs located close to managed European honey bee apiaries (1.83% of the collected individuals; Table 2). There was no detectible difference in the vertical distribution of African or European drones within DCAs (χ21, 2400 = 0.11, p = 0.74).
Nuclear microsatellite analysis indicates that DCAs near to managed European honey bee apiaries (0.25 km) had significantly more (F = 45.17; df = 1,5; p = 0.0026) colonies contributing drones than did DCAs distant to managed European honey bee apiaries (> 2.8 km; Table 3). Additionally, significantly more (F = 45.2364; df = 1,5; p = 0.0025) of the colonies contributing drones to DCAs distant to managed European honey bee apiaries (> 2.8 km) had African mtDNA than did the colonies contributing drones to DCAs near to managed European honey bee apiaries (0.25 km). Furthermore, we saw a significant (F = 11.90; df = 1,5; p = 0.0261) reduction in the proportion of colonies with African mtDNA that were contributing drones to DCAs near to managed European honey bee apiaries.
Our data demonstrate that the management of European honey bee colonies can influence the proportion of drones with African matrilines at nearby DCAs dramatically. Furthermore, our nuclear analysis confirms that the observed decrease in African matrilines at DCAs near managed apiaries is not merely the result of a large number of managed European colonies producing drones that flood the proximal DCAs. Rather we see an increase in the total number of colonies contributing drones and a decrease in the total number of African matriline colonies contributing drone s to DCAs near managed apiaries. Cumulatively, these changes in population structure lead to a drastic reduction in the presence of African matriline drones at DCAs that are near by managed apiaries.
Our results suggest that the management of European honey bees could be a viable option for limiting the introgression of African bee genetics into European colonies. Managed European colonies become “Africanized” when virgin European queens mate with African drones . Increasing the proportion of European drones present at a DCA is expected to increase the likelihood that virgin queens at that DCA will mate with European, opposed to African, drones . By reducing the likelihood that European queens mate with African drones, the frequency and degree of Africanization in managed apiaries would be reduced.
Virgin queens fly to more distant DCAs than the proximal DCAs that drones from the same apiary attend. Drones typically travel less than 0.5 km to a DCA whereas queens are estimated to travel 2.0 km to a DCA [20, 35]. Therefore, altering the proportion of drones at DCAs nearby an apiary may not modify the proportion of African drones at the DCAs that virgin queens produced within that apiary may attend. Further research is needed to determine what concentration and distribution of managed European colonies are needed to modify the regional reproductive population to the extent that virgin queens produced in the apiaries providing drones would themselves attend a DCA with a high proportion of European drones. There likely is a balance between the number of European bee colonies needed in an area to saturate the regional DCAs and the resources available in the environment needed to sustain those colonies.
Past studies have shown a slight tendency towards like-subspecies matings where African and European queens mated with drones of their same subspecies 58% and 64% of the time, respectively . Several theories, including partial physiological barriers, temporal isolation, or spatial isolation, exist to explain these tendencies [36–39]. In the present study, drones were collected from both 10 and 30 m elevations at each DCA to determine if there was a vertical distribution of African and European subspecies within the DCA. However, no detectable difference in vertical flight behavior of African and European drones was observed. This suggests that spatial isolation is not likely to be a contributing factor to any assortative mating that may be occurring.
The central Florida honey bee population offers a unique opportunity for understanding the interactions of African and European honey bee populations and the Africanization process. Furthermore, experimentation in this region promises to offer valuable insight into possible population manipulation techniques to limit the introgression of undesirable African honey bee genetics into the managed European honey bee population.
S1 Table. Collection site, molecular profile, and maternal group assignment information for individual drones.
Data are the name of the DCA in which the drones were caught (DCA ID), the distance of that DCA from a managed European honey bee apiary (DCA Location), the ID number of each drone (Bee ID), the fragment lengths of the allele present at each microsatellite loci (HB−C16−05, A107, B124, A79, A88, A7, HB−SEX−03, HB−SEX−01, HB−THE−03, and HB−THE−04), the haplotypes assigned based on the alleles present within the linkage groups (HB−SEX Haplotype and HB−THE Haplotype), the mitochondrial PCR−RFLP results (mtDNA, E = European, A = African), and the maternal group assignment made by COLONY based on the unlinked microsatellite fragment length and the linkage groups (COLONY Mother ID (by DCA ID)). Note that: 1) "**" denotes missing allelic information in allele and haplotype data columns. 2) Loci HB−SEX−03, HB−SEX−01, HB−THE−03, and HB−THE−04 are shaded grey to indicate that these data are not directly inputted into COLONY for maternal analysis due to linkage. Rather the linkage pairs are utilized to generate the haplotypes that are then used in the maternal analysis with the other unlinked loci. 3) Haplotype assignments and mother ID's are determined by DCA. Therefore, HB−SEX Haplotype (HB−THE Haplotype, or Mother ID) 1 at AK is not equivalent to HB−SEX Haplotype (HB−THE Haplotype, or Mother ID) 1 at WD (OSC, SIED, LBV, or MK).
We graciously thank Arthur Mathisen for the coordination and use of experimental apiary sites and numerous other locations while searching for DCAs; Mario Jacob and the employees of D&J Apiary for providing and managing European-derived honey bee colonies; Niko and Gudrun Koeniger for their invaluable assistance with DCA identification and drone trapping techniques; Mark Dykes, Katy Evans, Stephanie Long, and Teresa Pitts for assistance identifying DCAs; and Cindy Tannahill for helping optimize the molecular protocols.
- Conceptualization: ANM JDE.
- Data curation: ANM.
- Formal analysis: ANM JDE.
- Funding acquisition: ANM JDE.
- Investigation: ANM.
- Methodology: ANM JDE.
- Project administration: ANM.
- Resources: JDE.
- Software: ANM JDE.
- Supervision: JDE.
- Validation: ANM JDE.
- Visualization: ANM JDE.
- Writing - original draft: ANM.
- Writing - review & editing: JDE.
- 1. Kerr WE. Introdução de abelhas africanas no Brasil. Arasil Apicola. 1957;3: 211–13.
- 2. Kerr WE. The history of the introduction of African bees to Brazil. The S. Africa Bee Journal. 1967;39: 3–5.
- 3. Sheppard WS. A history of the introduction of honey bee races into the United Sates: Part II. Amer Bee J. 1989;129: 664–667.
- 4. Sheppard WS. A history of the introduction of honey bee races into the United Sates: Part I. Amer Bee J. 1989;129: 617–619.
- 5. Hall HG, Muralidharan K. Evidence from mitochondrial DNA that African honey bees spread as continuous maternal lineages. Nature. 1989;339: 211–213. pmid:2566122
- 6. Diniz NM, Soares AEE, Sheppard WS, Del Lama MA. Genetic structure of honeybee populations from southern Brazil and Uruguay. Gen and Molec Bio. 2003;26: 47–52.
- 7. Schneider S, DeGrandi-Hoffman G, Smith DR. The African honey bee: Factors contributing to a successful biological invasion. Annu. Rev. Entomol. 2004;49: 351–376. pmid:14651468
- 8. Michener CD. The Brazilian bee problem. Ann Rev Entomol. 1975;20: 399–416.
- 9. Michener CD. Final Report Committee on the African Honey Bee. Washington, D.C.: National Academy of Sciences; 1972
- 10. Taylor OR. African bees: potential impact in the United States. Bull Entomol Soc Amer. 1985;31: 15–24.
- 11. Ratnieks FLW. Africanized Bees: Natural Selection For Colonizing Ability. In: Spivak M, Fletcher DJC, Breed MB editors. The "African" honey bee. Boulder: Westview Press, Inc; 1991. pp 119–135.
- 12. Pinto MA, Rubink WL, Patton JC, Coulson RN, Johnston JS. Africanization in the United States: replacement of feral European honeybees (Apis mellifera L.) by an African hybrid swarm. Genetics. 2005;170: 1653–1665. pmid:15937139
- 13. Taylor OR. The past and possible future spread of Africanized honeybees in the Americas. Bee World. 1977;58: 19–30.
- 14. Jean-Prost P. Queen mating. Apimondia. XVII Int. Beekeep. Congr. 1985; 404–408.
- 15. Ruttner F. Reproductive behavior in honeybees. In: Holldobler B, Lindauer M, editors. Experimental behavioral ecology and sociobiology. New York: Springer; 1985. pp. 225–236.
- 16. Tribe G. Drone mating assemblies. South African Bee J 1982;54: 99–100.
- 17. Loper G, Fierro M. Use of drone trapping and drone release to influence matings of European queens in an Africanized honey bee area (Hymenoptera, Apidae). J Apic Res. 1991;30: 119–124.
- 18. Agriculture Research Services. Map of the spread of Africanized honey bee by year. United States Department of Agriculture. 2011. Available: http://www.ars.usda.gov/research/docs.htm?docid=11059&page=6.
- 19. Mortensen AN, Ellis JD. Scientific note on a single-user method for identifying drone congregation areas. J Apic Res. 2014;53(4): 424–425.
- 20. Rowell GA, Taylor OR, Long-Rowell MA. Spatial Dynamics of the Honey Bee Mating System (Apis mellifera L.). J Kansas Entomol Soc. 1992;65: 218–222. Stable URL: http://www.jstor.org/stable/25085359
- 21. Williams JL.Wind-directed pheromone trap for drone honey bees (Hymenoptera: Apidae). J Econ Entomol. 1987;80: 532–536.
- 22. Gary NE. Chemical mating attractants in the queen honey bee. Science. 1962; 136:773–774. pmid:17752107
- 23. Gries M, Koeniger N. Straight forward to the queen: pursuing honeybee drones (Apis mellifera L.) adjust their body axis to the direction of the queen. J Comp Physiol A Neuroethol Sens Neural Behav Physiol. 1996;179: 539–544.
- 24. Scheiner R, Abramson CI, Brodschneider R, Crailsheim K, Farina W, Fuchs S, et al. Standard methods for behavioural studies of Apis mellifera. In V. Dietemann; J. D. Ellis; P Neumann (eds). The COLOSS Beebook, Volume I: Standard Methods for Apis mellifera Research. J Apic Res. 2013;52(4). doi: https://doi.org/10.3896/ibra.1.52.4.04
- 25. Walsh PS, Metzger DA, Higuchi R. Chelex 100 as a medium for simple extraction of DNA for PCR-based typing from forensic material. Biotechniques. 1991; 10:506. pmid:1867860
- 26. Crozier Y, Koulianos S, Crozier R. An improved test for Africanized honeybee mitochondrial DNA. Cell Mol Life Sci 1991;47: 968–969.
- 27. Pinto MA, Johnston J, Rubink WL, Coulson RN, Patton JC, Sheppard WS. Identification of Africanized honey bee (Hymenoptera: Apidae) mitochondrial DNA: validation of a rapid polymerase chain reaction-based assay. Annal Entomol Soc Amer. 2003;96: 679–684.
- 28. Estoup A, Garnery L, Solignac M, Cornuet JM. Microsatellite variation in honey bee (Apis mellifera L.) populations: hierarchical genetic structure and test of the infinite allele and stepwise mutation models. Genetics. 1995;140(2): 679–95. pmid:7498746
- 29. Shaibi T, Lattorff HM, Moritz RF. A microsatellite DNA toolkit for studying population structure in Apis mellifera. Molec Ecol Res. 2008;8(5): 1034–6.
- 30. Ascunce MS, Bouwma AM, Shoemaker D. Characterization of 24 microsatellite markers in 11 species of fire ants in the genus Solenopsis (Hymenoptera: Formicidae). Molec Ecol Res. 2009;9(6): 1475–9.
- 31. Human H, Brodschneider R, Dietemann V, Dively G, Ellis JD, Forsgren E, et al. Miscellaneous standard methods for Apis mellifera research In V. Dietemann; J. D. Ellis; P Neumann (eds). The COLOSS Beebook, Volume I: Standard Methods for Apis mellifera Research. J Apic Res. 2013;52(4). doi: https://doi.org/10.3896/IBRA.220.127.116.11
- 32. Kraus FB, Koeniger N, Tingek S, Moritz RFA. Using drones for estimating colony number by microsatellite DNA analysis of haploid males in Apis. Apidologie. 2005;36: 223–229.
- 33. Jones O, Wang J. COLONY: a program for parentage and sibship inference from multilocus genotype data. Molec Ecol Res. 2010;10: 551–555.
- 34. Wang J.Sibship reconstruction from genetic data with typing errors. Genetics. 2004;166: 1963–1979. pmid:15126412
- 35. Jensen AB, Palmer KA, Chaline N, Raine NE, Tofilski A, Martin SJ, et al. Quantifying honey bee mating range and isolation in semi-isolated valleys by DNA microsatellite paternity analysis. Conserv Gen. 2005;6: 527–537.
- 36. Kerr WE, Bueno D. Natural crossing between Apis mellifera adansonii and Apis mellifera ligustica. Evolution. 1970;24: 145–155. Stable URL: http://www.jstor.org/stable/2406721
- 37. Taylor O, Spivak M. Climatic limits of tropical African honeybees in the Americas. Bee World. 1984;65: 38–47.
- 38. Hall HG. DNA studies reveal processes involved in the spread of New World African honeybees. Florida Entomol: 1992;75: 51–59. Stable URL: http://www.jstor.org/stable/3495480
- 39. Quezada-Euan JJG, May-Itza WDJ. Partial seasonal isolation of African and European-derived Apis mellifera (Hymenoptera: Apidae) drones at congregation areas from subtropical Mexico. Annal Entomol Soc Amer. 2001;94: 540–544.