fc,* •, •• j^: -l".- GENOMIC DNA RESTRICTION FRAGMENT LENGTH POLYMORPHISMS AT A HIGHLY POLYMORPHIC LOCUS DISTINGUISH OLD AND NEW WORLD SUBSPECIES OF THE HONEY BEE, Apis mellifera L. : ;',iS'"-. ^••»tr . ff^ * ;^ ..' :i By -^v- 5 MARGARET ANNE MCMICHAEL «f' ^t A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 1994 ACKNOWLEDGMENTS am grateful for the efforts of so many people, and am bothered by the I I omission of many names from these two pages. My thanks go to all who have helped me. First and foremost, Glenn Hall took a chance on me and has been patient and generous in providing the opportunity, direction, and freedom have I enjoyed in the course of obtaining my doctorate in his lab. The members of my advisory committee provided tremendous assistance; Andy Cockburn, Jim Maruniak, and Buffy Bondy. Don Campton generously donated his time in the initial stages of my research, and Jan Conn reviewed my manuscripts and postdoctoral research proposal. Chip Taylor provided a memorable week in I Linnares, Mexico, and taught me the zen of smoker-lighting. Lois Lemmerman Myeroff, Case Western Reserve University, is responsible for teaching me all I knew about molecular biology when I was in the Mapstone/Goldthwait lab and as such was instrumental in my move to Florida. Alejandra Garcia and Raquel McTeirnan have been immeasurably helpful to me here, and my way would have been rough without their tireless and unselfish assistance. Marjorie Hoy's lab members-particularly Owain Edwards (who also reviewed my manuscripts), Jim Presnail, Dr. Jey, and Greg McDermott-have been instrumental in advancing my research efforts. Reg Coler, Scott Yocom, and Owain Edwards helped me make slides for the meetings and for my departmental seminar. ';,^ Dr. Strayer, Faith Oi, and Hugh Smith taught me a few things during one my of best experiences here; assisting in Principles of Entomology. Don Hall, Tom Dykstra, and Robin Goodson provided invaluable opportunities for me to lead their classes and tours of the Bee Lab. Scott Yocom recruited me for the Linnean Team, which was an exciting, not to mention a humbling, experience. Deserving additional, special mention are John Strayer and the great students he recruited for the department, with whom have had the honor and pleasure I to work. ";,i- '•; * 4 ." Kr ill " Ji irf TABLE OF CONTENTS ACKNOWLEDGMENTS ii ABSTRACT )'. v ! . . INTRODUCTION 1 IDENTIFICATION AND GEOGRAPHICAL DISTRIBUTION OF ALLELES AT LOCUS 178 5 Introduction 5 Materials and Methods 6 Results 8 Discussion 30 ALLELE FREQUENCIES AT LOCUS 178 REVEAL HYBRIDIZATION OF EUROPEAN AND AFRICAN BEES IN THE NEOTROPICS 36 Introduction 36 Materials and Methods 36 Results 40 Discussion 58 LOCALIZATION OF VARIATION AT LOCUS 1 78 IN Apis mellifera (L.) BY RESTRICTION MAPPING 65 Introduction 65 Materials and Methods 67 Results 69 Discussion 93 CONCLUDING REMARKS 99 LITERATURE CITED 103 BIOGRAPHICAL SKETCH Ill IV Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy GENOMIC DNA RESTRICTION FRAGMENT LENGTH POLYMORPHISMS AT A HIGHLY POLYMORPHIC LOCUS DISTINGUISH OLD AND NEW WORLD SUBSPECIES OF THE HONEY BEE, Apis mellifera L. By I Margaret Anne McMichael April 1994 Chairperson: H. G. Hall Major Department: Entomology and Nematology Honey bees (Apis mellifera L.) of African and European ancestry were distinguished by analysis of restriction fragment length polymorphisms (RFLPs) defined by two enzymes at a highly polymorphic locus corresponding to genomic probe pB178. Thirty-six Msp\ restriction fragment patterns, or variants, and thirty-three Dde\ variants were identified. Sixty-five pairwise combinations of the Msp\ and Dde\ variants, referred to as alleles, were found among the individual drones tested. Variants and alleles were discontinuously distributed in USA and South African drones: only one Dde\ variant and no Msp\ variant (hence no allele) were common to these two sample populations. I The diversity in the variants and alleles found in the South African drones was greaterthan in the USA drones. Msp\ variants were discontinuously distributed among workers bees of the subspecies -4. m. mellifera (west European), A. m. ligustica and A. m. caucasica (east European), and A. m. scutellata (South African). Ancestry in New World bees was inferred from variant and allele frequencies at this locus. In USA bees, variants characteristic of east and west European bees were found at frequencies consistent with previously identified nuclear and mitochondrial DNA markers. In neotropical drones, European and African origins were assumed for variants that were similar in fragment pattern to variants in the USA and South Africa, respectively. The alleles identified in ''^i' the majority of neotropical colonies were African. There was little evidence of hybridization of African and east European bees in neotropical bees: variants specific to South African bees were detected at the highest frequencies, while east European-specific variants were absent or detected at very low frequencies. A variant that may be specific for A. m. mellifera vjas found throughout the neotropics at frequencies that varied from 1 9% to 33% but did not increase clinally in a northward direction. v^ ; Determination of the allelic relationships between Msp\ fragments and between Dde\ fragments was initiated by mapping these restriction sites within the probe, and correlating fragments on Southern blots to the locations of restriction sites on the resulting map. " ' vl INTRODUCTION Descendants of ancestral honey bees migrated from the site of their ' J origin, presumed to be northeast Africa or the Middle East, and became widely distributed in the Old World (Garnery, Cornuet & Solignac 1992). Subsequent adaptation to a wide variety of ecological conditions resulted in the evolution of subspecies of the honey bee. Apis mellifera L. (Ruttner 1988). These subspecies are distinguished on the basis of physical, behavioral, and ecological criteria and are defined quantitatively by discriminant analysis of morphological characters (Daly & Balling 1978; Ruttner 1988; Ruttner, Tassencourt & Louveaux 1978). Honey bees were first introduced to the New World by 17th and 18th century European settlers. For over two hundred years, the genetic diversity found in New World honey bees resulted from the importation of primarily four European subspecies or races: west and north European A. m. mellifera (referred to here as west European); north Mediterraneany4. m. ligustica Spinola and A. m. carnica Pollmann (which in this report will be called east European; Ruttner 1 988); and the east European beeA. m. caucasica Gorbachev (referred to here as east European) (Kent 1988; Kerr, DeLeon & Dardo 1982). — -nK•'~'—-'~>'•'- — -''-' - A," Despite abundant nectar sources in the neotropics, self-sustaining feral populations of European bees did not become established (Michener 1975). The poor performance of European bees in the tropics has been attributed to I their failure to adapt to environmental conditions vastly different from those in which they evolved (Ruttner 1 988). To improve commercial honey production, queens of a central and south African race, A. m. scutellata Lepeletier, were imported into Brazil in 1956 for experimental breeding with European bees. Swarms of African bees escaped into the wild (Kerr 1967) and subsequently proliferated to form large feral populations that spread through tropical South and Central America. African bees entered Texas in 1 990 and Arizona in 1 993. I . The release and spread of African bees has been disruptive to the behavior and ecology of the melliferous flora and fauna in the neotropics (Cantwell 1974; Michener 1975; Roubik 1980, 1989; Spivak, Fletcher & Breed 1991; Taylor 1977; Winston, Taylor & Otis 1983). Explanations for the migration pressure resulting from the introduction of the bees from Africa and for the rapid colonization of New World territories previously unoccupied by honey bees have focused on the extent to which neotropical bees represent an :/, I admixture of African and European subspecies (Hall 1990, 1991, 1992a; Hall & Muralidharan 1989; Rinderer 1986; Rinderer ef a/. 1985, 1991; Sheppard ef al. 1 991 ; Smith, Taylor & Brown 1 989; Taylor 1 985). The ability to distinguish .V>' honey bee subspecies is essential to ascertain the extent of hybridization. In turn, the degree of hybridization will influence the northern limit of African bee -"!f?S? introgression in the USA, about which there is considerable debate (Dietz 1 986; Dietz, Krell & Pettis 1986; Rinderer 1986; Roublk 1986; Southwick, Roubik & Williams 1990; Taylor 1977; Taylor & Spivak 1984; Villa, Gentry & Taylor 1987; Villa, Rinderer & Collins 1993). The retention of African morphology and behavior in neotropical bees has stimulated the quest for the identification of their ancestry, which will aid in understanding the mechanisms contributing to their phenomenal success. Efforts to ascertain the extent of interaction between the extant European bees and the African bees in the neotropics have been hindered by a lack of genetic markers specific for each Old World subspecies introduced. Intermediate values of morphometric values and allozyme frequencies, taxonomic characters traditionally used to identify honey bee subspecies and subspecies groups, can only suggest that hybridization between races has occurred (Daly 1991; Del Lama eta/. 1988; Nunamaker ef a/. 1984; Spivak ef a/. 1988; Sylvester 1982, 1986). Thus, supplemental genetic characters subject to little environmental modification have thus been sought. ^ , Mitochondrial DNA (mtDNA) has been used effectively to identify honey bee subspecies within populations in biogeographic studies (Arias, Soares & Nobrega 1990; Cornuet & Garnery 1991; Garnery, Cornuet & Solignac 1992; Hall & Smith 1 991 ; Meixner, Sheppard & Poklukar 1 993; Smith & Brown 1 988, 1990; Smith 1991; Smith et al. 1991). Honey bee mtdna types are discontinuouslydistributed amongsubspecies, reflectingthesustainedtemporal separation and independent evolution of temperate and tropical lineages (Avise et al. 1987; Cornuet & Garnery 1991). African mtDNA has been found in almost all feral neotropical colonies. The small proportion of European mtDNA I that has persisted has remained largely confined to managed apiaries (Hall and Muralidharan 1989; Hall & Smith 1991; Smith, Taylor & Brown 1989). These findings supplied strong evidencethat unbroken matrilines spreading as swarms have been primarily responsible for the expansion of the African bee population .5 in the neotropics (Hall & Muralidharan 1 989; Hall & Smith 1 991 ; Smith, Taylor & Brown 1989). Yet the feaItures of mtDNA that make these molecules good genetic markers (i.e. uniparental inheritance, no recombination) limit the information that can be obtained from their analysis. Genomic RFLPs at several loci have been identified that distinguish east European from west European and African bees, but not west European from African bees (Hall 1 990, 1 992b). Frequencies for the markers indicated limited hybridization of African with European bees, but it was not possible to assess ,,- hybridization with west EuroIpean bees. Alleles at two loci, found at different frequencies in African and west European bees, were found at intermediate I frequencies in neotropical bees, which could have resulted from hybridization between these two groups of subspecies (Hall 1992b). The objective ofthe research reported here was to identify genomic DNA (molecular) markers or characters that distinguish among the east European, west European, and African groups of subspecies. , .