Reliability of an island mating apiary under routine management

Introduction Islands are widely used to achieve controlled matings of virgin queens because queens and drones are not expected to cross open water during mating flights. However, uncontrolled matings on island mating areas have repeatedly been reported for many islands in close proximity to mainland shorelines (c.f. Neumann et al., 1999b). With increasing distance between island and mainland one would expect increasing reliability of mating control. Moreover, uncontrolled matings are more likely when the number of drone producing colonies [DPCs] is low. Under routine mating apiary management conditions several strong DPCs are to be placed on the island in order to provide enough drones to saturate local drone congregation areas [DCAs] and prevent matings with undesired drones from the mainland (c.f. Neumann et al., 1999b). Here we test the reliability of a mating apiary on the island of Neuwerk under local routine conditions with an ample supply of drone producing colonies typical for managed island mating apiaries. Eight strong DPCs (22 frames) and 99 mating nuclei with virgin sister queens were placed on the North Sea island of Neuwerk. Six of the 80 nuclei with mated queens were used for the analysis (Table 1). Neuwerk is 9.3 km of open water away from the mainland with other drones. The virgin queens were allowed to mate freely for a three-week period during the local reproductive season (July). After the queens initiated oviposition, sealed worker brood was sampled from each mating nucleus colony and sealed drone brood from each DPC. DNA was extracted from flight muscle tissue of pupae and genotyped at nine polymorphic DNA microsatellite loci (B124, A7, A24, A35, A43, A76, A88, A107 and A113; Estoup et al., 1994) using routine protocols (Moritz et al., 2003). DNA was amplified with fluorescent dye labeled primers (FAM, HEX and TET). Allelic DNA fragment sizes were scored in two electrophoresis runs in a DNA sequencer (ABI 310) using Genotyper © software, an internal size standard (GeneScan 500) and the protocols of the supplier. The putative queen genotypes of the DPCs were derived from pooled drone samples (n=20 drone offspring for each DPC queen, Moritz et al., 2003). Likewise, the putative genotypes of the island-mated queens and of their drone mates (=number of observed matings) were inferred from the worker offspring using Mendelian rules (n=20 workers each queen). The number of estimated matings was calculated as in Neumann et al. (1999b). To detect uncontrolled …