This is a first award application requesting funding to study the genes and gene action involved in defensive behavior in honey bees. The project will focus on the two identified QTLs recognized to affect defensive behavior in Africanized honey bee colonies. Both single gene effects as well as social interactions with those genes will be addressed by the development of behavioral assays that permit characterization of defense behavior and its components at both the whole colony and individual level. The PI lists a number of hypotheses to be tested with respect to defensive behavior. These hypotheses are: (1) that stinging behavior on individual bees is affected by the same QTLs that influence behavior at the colony level; (2) that defensive behavior per se can be partitioned into number of components through the development of specific behavioral assays and markers linked to targeted QTLs; (3) that the alleles at QTLs show genetic dominance; (4) that defensive behavioral QTLs can be precisely mapped by correlating individual level behavior with the segregation of linked markers. The PI reports a number of relevant preliminary studies and observations. Using RAPD markers, a high density map of one marker every 100 to 200 kb has been developed for the honey bee, and the PI had the major role in this development. This map has already been used to map the sex-determination gene. Preliminary studies of 172 backcross colonies derived from hybrids of Africanized and European bees have identified two QTLs (designated sting-1 and sting-2) that are associated with variation in standard behavioral assays for stinging behavior, such as the number of stings in a defined period. The proposed investigation will extend and better define this association at both the individual and colony level. This proposal first calls for the development of behavioral assays that will maximize the defensive behavioral differences that exist between Africanized and European honey bees. Emphasis will be placed on aspects of behavior that may be quantified at the individual level. Most prior studies have emphasized colony-level stinging response, and this has tended to ignore intracolony interactions in the expression of both individual and overall colony response. The defensive components to be examine are likelihood to sting first, likelihood to sting, time to respond, and total numbers of stings. One approach that will be used is the marking of individuals in synthetic colonies of mixed genetic origin. This will allow an assessment of the interaction between colony genotype composition and genotype behavior, as well as the effect of colony size on genotypic behavior. For example, if Africanized genotypes are the first to sting in defense of a colony does this increase the tendency to sting of bees of European genotype. In addition, the effects of age will be addressed as well as extrinsic features of colony environment, such as the proximity of nectar sources. In the end, those conditions and components of behavior that maximize differences between the two races of bees will be used to address questions concerning colony and individual differences, degree of dominance, and to further map putative QTLs. The proposal is very specific with respect to the origin of parental stocks, and details of all inter race crosses and backcrosses. In general, the mapping of effects at the colony and individual levels emphasizes different designs with respect to F1 and F2 backcrosses, recognizing the haplo-diploid form of inheritance in hymenoptera. Initially, the sources of the Africanized lines will be 20 colonies of bees verified to be or African origin by mitochondrial DNA analysis and morphological features. These bees will come from established and characterized feral bee swarms collected in the state of Guerrero Mexico and selected using standard behavior assays for colonies showing the strongest defensive behavior, assuming that these lines have the greatest likelihood of not segregating for defensive alleles. One European queen will be used as the source of all virgin queens for F1 crosses with recombinant drones of hybrid origin. A foundation stock of European colonies will be established from crosses between individuals from colonies with the lowest defensive scores. Basically, colony specific behaviors are examined by taking drones (males) produced from F1 hybrid females and crossing them back to identical females of either origin and thus producing individual colonies where female workers are all identical within colonies with respect to their fathers recombinant genome. Differences between colonies reflect different meiotic products derived from the hybrid females. In the study of individual differences, Within colony variation in (female) workers arises in crosses where the mother is heterozygous with respect to origin and daughters differ with respect to mei osis in the mothers germ line. The paternal genome is constant in this case. Within the colony, the association of individual components of behavior with candidate QTL markers is followed. Individual molecular marker genotypes are assessed by either sacrificing whole individual bees following behavior characterization, or in some cases by sampling stingers left after attack on a leather patch. Finally, it is proposed to map at higher resolution the two putative QTLs identified in the preliminary studies. This will be accomplished by increased saturation of the genome with four different types of markers and examining their segregation with respect to individual behaviors (the original mapping was carried out at the colony level and only involved 172 meioses). By characterizing 1000 individuals in backcrosses, it should be possible to refine the map position to within 0.3 cM with 95% confidence. It is proposed to use STS primers designed from the RAPD markers, SSCP polymorphisms and further development of another set of RAPD markers from 1000 10-mer primers, screened in bulk.
