EDP Sciences logo
Search
Menu
All issues Volume 40 / No 6 (November-December 2009) Apidologie, 40 6 (2009) 617-626 Full HTML
Free Access
Issue
Apidologie
Volume 40, Number 6, November-December 2009
Page(s) 617 - 626
DOI https://doi.org/10.1051/apido/2009041
Published online 17 July 2009

© INRA/DIB-AGIB/EDP Sciences, 2009

1. INTRODUCTION

The natural distribution area of Apis mellifera L. include Africa, Europe and Western Asia. In these geographical areas exist races or subspecies, which are locally adapted to environmental conditions (Ruttner, 1988). Based on morphological and ecological data, 29 honeybee subspecies have been recognised (Engel, 1999; Sheppard and Meixner, 2003) and grouped in four morphological branches (Ruttner, 1988). The morphometric branch-C consisted of five honeybee subspecies: A. m. siciliana Grassi (later included in the African evolutionary A-lineage based on molecular analyses, Sinacori et al., 1998), A. m. ligustica Spinola, A. m. cecropia Kiensenwetter, A. m. macedonica Ruttner and A. m. carnica Pollman (Ruttner, 1988, named after Engel, 1999). The natural distribution of the Carniolan honeybee (A. m. carnica) covers the Balkan region from Eastern Austria, Slovenia, down to Croatia and to a lesser extent Hungary and Bulgaria (Ruttner, 1988). This subspecies is appreciated for its gentleness and easy managment for honey production and crop pollination and, therefore, has been introduced in other European countries, such as Germany and other continents, including Australia and the Americas (Moritz et al., 2005).

Table I

Details of samples location, number of analyzed colonies (n) and haplotypes in each location. Number in brackets referred to the locations showed in Figure 1.

Croatia contains a range of variable climates and habitats. Therefore, knowledge of its honeybee population is important to gain understanding of overall honeybee biogeography. In relation to climatic and geographical variability, three ecotypes of the Carniolan bee (Pannonian, Subalpin and Dalmatian) were morphologically described (Ruttner, 1988), although later the same author inferred the existence of only Alpine and Pannonian ecotypes and several regional variations such as Dalmatian (Ruttner, 1992). Recent genetic studies with both microsatellite and mtDNA markers (Sušnik et al., 2004) were unable to distinguish Croatian bees from Slovenian bees, probably because the beekeepers from both countries exchange honeybees. At the mtDNA level, a single haplotype (C2c) has been detected in populations from both countries.

Some mitochondrial haplotypes can be differentiated through variation in the mitochondrial tRNAleu-cox2 intergenic region (Garnery et al., 1993, 1995, 1998; De la Rúa et al., 1998, 1999, 2000; Franck et al., 1998, 2000a, 2000b; Cánovas et al., 2008). Subspecies belonging to the C evolutionary lineage show a short sequence and absence of length variability in the tRNAleu-cox2 intergenic region. Therefore, only five haplotypes have been reported to date, exclusively based on five single nucleotide polymorphisms (Franck et al., 2000a, 2000b; Sušnik et al., 2004). Consequently sequence data from this conservative mtDNA region are more useful than length polymorphism to estimate molecular variability within C lineage subspecies. Microsatellite variation is also suitable to detect recent population events, including introgression (Jensen et al., 2005) and admixture of local and introduced honeybee populations (De la Rúa et al., 2001).

The aim of the present study was to assess the genetic variability of Croatian honeybee coastal populations to increase the overall genetic knowledge of the subspecies A. m. carnica and to seek molecular evidence for the aforementioned ecotypes and regional variation. For appropriate comparisons within C-branch taxa, samples from Italy (A. m. ligustica) and Greece (A. m. macedonica) were also included in the analyses.

2. MATERIAL AND METHODS

2.1. Sampling and DNA extraction

Adult honeybee workers were sampled from 45 colonies in Croatia (N = 20), Italy (N = 20) and Greece (N = 5), during 2006–2008 at seventeen different locations (Tab. I, Fig. 1). Samples were preserved in absolute ethanol and kept at –20 °C until they were processed in the laboratory.

thumbnail Figure 1

Map of East Europe with the distribution of haplotypes in the Croatian subpopulations 1 and 2 and the reference Italian and Greek populations. Sampling sites names are showed in Table I.

A single honeybee worker per colony was used for mtDNA analysis (N = 45), while five honeybee workers per colony (N = 225, but see results) were used for microsatellite analysis. Total DNA was extracted from three right legs each honeybee worker using a 5% Chelex solution (Walsh et al., 1991).

2.2. Mitochondrial DNA analysis

The tRNAleu-cox2 intergenic region were PCR amplified with the primers E2 (5’-GGCAGAATAAGTGACATTG-3’) located at the 5’ end of the gene tRNAleu and H2 (5’-CAATATCATTGATGAACC-3’) located close to the 5’ end of the gene cox2 (Garnery et al., 1991) in a total volume of 25 μL, following the conditions described by Garnery et al. (1993). To determine the amplicon size, aliquots of 3 μL of each sample were run on a 1.5% agarose gel, stained with ethidium bromide and photographed over a UV light screen. Twenty-two μL of the PCR product were then digested with DraI and separated by 4% Nusieve gel electrophoresis to reveal RFLPs. Amplicons of each sample were purified with isopropanol and ammonium acetate and submitted to sequencing (Secugen S.L., Madrid, Spain) with the primer E2.

Each sequence was manually checked for base calling then a multiple sequence alignment was performed with the program MEGA version 4 (Tamura et al., 2007) including for comparison published sequences (Franck et al., 2000b; Sušnik et al., 2004).

2.3. Microsatellite analysis

Multiplex PCR reaction with five microsatellite loci A7, A113, Ap43, Ap55 and B124 (Estoup et al., 1995), was performed in 10 μL total volume containing 50 mM KCl, 10 mM Tris HCl (pH 9.0), 1.2 mM MgCl2, 0.3 μM of each dNTP, 0.8 μM of each primer, 1.5 units of Taq polymerase (Bioline) and 2 μL of DNA extract. Annealing temperature was set at 54 °C. PCR products were visualized by capillary electrophoresis and sized with an internal size-standard (Servei Central de Suport a la Investigació Experimental, University of Valencia, Spain). Alleles were subsequently scored using GeneMapper v3.7 software (Applied Biosystems).

2.4. Analysis of microsatellite data

Population genetic parameters were calculated with GenAlex (Peakall and Smouse, 2006). Genetic diversity within populations was evaluated by computing allele frequencies, observed (Ho) and expected (He) heterozygosity. Hardy-Weinberg equilibrium was tested with Genepop (Raymond and Rousset, 1995).

A Bayesian model-based clustering method for inferring population structure and assignment of individuals to populations probabilistically based on their multilocus genotypes, and thereby estimates of the posterior probability for a given number of genetic populations (K) were obtained with the software STRUCTURE v 2.2 (Pritchard et al., 2000). An admixture model assuming correlated allele frequencies was used. The results were based on simulations of 80 000 burn-in steps and 1 000 000 MCMC (Markov Chain Monte Carlo algorithm) iterations. Five runs for each K-value (K = 1 − 10) were used to estimate the most likely value of K. The number of populations was defined using the value of ΔK as described in Evanno et al. (2005).

Principal coordinates analysis (PCA) via covariance with standardization of the individual genetic distances was performed to find and plot the relationships between the individuals belonging to the different populations and inferred subpopulations.

3. RESULTS

3.1. Mitochondrial DNA

The DraI-test can discriminate between the major mitochondrial lineages of A. mellifera. However insertions, deletions or single point mutations not directly involved in the restriction sites, can only be detected by sequencing. The analysis of sequence data at the intergenic region produced five different haplotypes all of them ascribable to the East European C-lineage and characterized by the presence of a single Q sequence: four have been found in Croatia, three of them (C1, C2c and C2d) previously described (Franck et al., 2000b; Sušnik et al., 2004), and one (C2e) new. A fifth haplotype (C2i) was newly described in Greek samples (GenBank accession numbers FJ824582, FJ824583, FJ824584, FJ824585, FJ824586 and FJ824587).

A total of seven polymorphic sites were found among the sequences of 572–570 bp (Fig. 2) and named following Franck et al. (2000b). The polymorphic site 1, showed a single nucleotide deletion that provided diagnostic differentiation between the C1 and C2 haplotypes, while the remaining positions allowed us to distinguish among different types of C2 haplotype. The new haplotypes C2e and C2i were distinguishable from C2d because of a single nucleotide deletion and a G → A transition at the polymorphic sites 3 and 5, respectively.

thumbnail Figure 2

Sequence of the tRNAleu-cox2 intergenic regions of the C haplotypes (C1 sequence corresponds to the mtDNA fragment from positions 3363-3935 bp published by Crozier and Crozier, 1993). C1, C2a and C2b haplotypes were described by Franck et al. (2000b) and C2c and C2d haplotypes by Sušnik et al. (2004). C2i are newly described in this work. Numbers correspond to variable polymorphic sites (in bold) (GenBank accession numbers FJ824582, FJ824583, FJ824584, FJ824585, FJ824586 and FJ824587).

Haplotype distribution per locality is detailed in Table I. The four haplotypes detected in Croatia had the following overall frequencies: 0.45 (C2e), 0.35 (C1), 0.15 (C2c) and 0.05 (C2d). All Italian samples bore the same C1 haplotype, while two haplotypes were found in Greece, (C2d and C2i with a frequency of 0.80 and 0.20 respectively) (Fig. 1).

3.2. Population structure basedon microsatellite data

Only those samples that amplified three or more loci were included in the analyses (90% of Croatian, 67% of Italian and 76% of Greek samples).

The number of scored alleles at the five microsatellite loci varied from five (loci A113 in Italy and Ap55 in Greece) to 16 (locus A7 in Croatia). The average allele number within population varied between 9.6 (Italy) and 11.0 (Croatia). Gene diversity measured as expected heterozygosity (He) ranged from 0.645 (Italy) to 0.796 (Greece) on average (Tab. II). Honeybee populations from Greece and Croatia deviated significantly from the Hardy-Weinberg equilibrium (P < 0.05).

Table II

Microsatellite variation in Croatian, Italian and Greek A. mellifera populations. Sample size (N), number of detected alleles (n), observed (Ho) and expected (He) heterozygosity per locus and mean values ± SD by population.

Population genetic structure was inferred through the Bayesian clustering method: the highest posterior probability of the data set was detected when using a model that assumed four populations (Evanno et al., 2005). The Croatian population was split into two separate subpopulations: 1 spread at the north (Istra-Pidzan and Mali Losini) and 2 distributed at the south (Sutivanac, Brac and Korcula). These subpopulations showed different level of influence of the Italian population, while a small fraction of the Greek population showed influence of the subpopulation-1 of Croatian honeybees (Fig. 3).

thumbnail Figure 3

Results of STRUCTURE analysis using admixture and correlated allele frequencies models. Individuals are represented by vertical lines, grouped by inferred populations (K = 4). Division of individuals into coloured segment represents the assignment probability of that individual to each of the K groups.

A principal coordinate analysis (PCA) based on the first two principal coordinates was performed to investigate population patterns based on the genetic distance among individual samples (Fig. 4). Samples from Croatian subpopulation-1 were located in the quadrant 4 together with the samples from Greece, samples from Croatian subpopulation-2 were located in quadrant 3 with a few samples in the quadrant 2, Italian samples were grouped in a separate cluster located between the quadrants 1–2, being a few of them spread to the quadrant 3.

3.3. Introgression and hybridization

Assignment test (GeneAlex) allocated seven Italian individuals to the Croatian subpopulation-2 and four Greek individuals to the Croatian subpopulation-1, while one and two individuals of the Croatian subpopulations 1 and 2 respectively, were assigned to the Italian population.

These data were corroborated with the STRUCTURE software (Tab. III). When the distribution of the individual admixture proportions in pre-defined populations (Italy, Greece and Croatia) was analysed using the admixture and correlated allele frequency models and the four inferred populations (K = 4), most of the Croatian samples were shared in two subpopulations (Croatia-1 = 51.7% and Croatia-2 = 39.9%) while the remaining 6.3% of individuals were assigned to the Italian cluster; moreover, 19.7% of individuals of the Greek population and 14.7% of the Italian individuals were assigned to Croatian 2 and 1 inferred subpopulations, respectively.

thumbnail Figure 4

Distribution of A. mellifera individuals based on the genetic distance analysed with principal coordinate analysis (PCA).

Table III

Proportion of membership of each pre-defined population (Croatia, Italy and Greece) to each of the four inferred populations (K = 4). Pairwise analyses were performed with STRUCTURE exploring admixture and correlated allele frequency models. N = number of individuals.

4. DISCUSSION

Molecular genetic markers provide insights into the geographical distribution of the genetic diversity of A. mellifera subspecies. The assessment of present population structure is an essential knowledge for an effective management policy. In this work we analyzed mitochondrial and nuclear DNA markers to study the population structure of honeybees from coastal Croatia. Both marker sets showed that the situation of the Croatian honeybees was more complex than expected from previous published data.

In the intergenic tRNAleu-cox2 region, two new polymorphic sites were detected (one deletion and one transition), in addition to the five already observed in the A. mellifera subspecies of the C evolutionary lineage (Franck et al., 2000b; Sušnik et al., 2004), thus increasing the known molecular variation within this lineage. Four out of the seven C-haplotypes known to date have been found in Croatia, whereas Sušnik et al. (2004), in their study about the molecular variation of A. m. carnica in Slovenia that also included ten samples from Croatia, reported only one (C2c). Haplotype C2c was also detected in the present study but with a lower frequency, thus partially confirming the conclusion reached by Sušnik et al. (2004) about the close relationship between Slovenian and Croatian bees. These results are likely due, among other things, to the bi-directional commercial exchanges between beekeepers.

Ruttner (1992) described the presence of two morphologically distinct ecotypes of A. m. carnica according to zoogeographic zones named Alpine and Pannonian, and a regional variation called Dalmatian spread in the Adriatic seaside of Croatia. Although we lack samples from the Alpine (Austria and Slovenia) and the Pannonian regions (Hungary and Romania) with which to compare, our molecular analyses did demonstrate the presence of two different Croatian subpopulations within the Dalmatian variation. Based on the geographic origin of our samples, the subpopulation-1 corresponds to Northern (Istra Pidzan and Mali Losini) and the subpopulation-2 to Southern localities (Sutivanac, Brac and Korcula). Both subpopulations present different haplotype frequency, with C2e being more frequent at the Northern than at the Southern subpopulation.

Microsatellite analyses also supported this distinction and suggest that both subpopulations were (or still are) subject to introgression events from neighbouring A. mellifera subspecies. Introgression from Italian A. m. ligustica might have occurred mainly in the subpopulation-2 in congruence with the occurrence of the haplotype C1 characteristic of A. m. ligustica. These results suggest that introductions of honeybees from neighbouring countries occur, in contrast to the conclusion depicted by Sušnik et al. (2004). This introduction could be either due to natural migration, with Italian honeybees that have flown across the Adriatic Sea using the islands as stepping stones (the average distance from the Croatian coast to Italy is 160 km, but at some Northern points only 40–80 km separate both coasts, being a flying distance for a honeybee swarm with the help of summer west winds) or through human-mediated queen trade. This putative gene flow could happen in both directions as depicted from the microsatellite analyses. Although not a single C2 haplotype has been observed in Italy (Franck et al., 2000a, present data), genetic introgression from the Croatian subpopulation-2 into the Italian population was inferred from the Bayesian analysis of the microsatellite variation, and the assignment test that predicts that 14.7% of the Italian bees had a Croatian origin. We hypothesise that the introgression is produced by Croatian drones mating Italian queens. Evidence of A. m. carnica introgression into A. m. ligustica has been demonstrated by Dall’Olio et al. (2007), not only in the northern natural hybridization zone, but also further south. This introgression may result from either natural migration or transport by beekeepers or both.

The haplotype C2d, which is the most frequent in the Greek population, was observed in only one colony in the Croatian subpopulation-1. This fact and the gene flow inferred from the STRUCTURE and assignment test analyses, suggests the introduction of honeybees from the neighbouring A. m. macedonica Greek populations.

In conclusion, the analysis of Croatian honeybees with molecular markers provides additional evidence of the variation evident in the sequence of the intergenic tRNAleu-cox2 region and its usefulness to study genetic diversity among and within A. mellifera evolutionary lineages. Biodiversity found in this region enhances the need to conduct similar analyses (with both microsatellite markers and mitochondrial sequence data), in honeybee populations from neighbouring regions where major sources of the autochthonous Carnolian bee still remain.

Acknowledgments

Thanks are due to N. Kezic, F. Hatjina and M. Bouga for their help in providing samples. P. De la Rúa and I. Mu noz are supported by programs of the Spanish Ministry of Science and Innovation. The editor S. Fuchs and one reviewer made useful comments that improve the manuscript.

References

  • Cánovas F., De la Rúa P., Serrano J., and Galián J. (2008) Geographic patterns of mitochondrial DNA variation in Apis mellifera iberiensis (Hymenoptera: Apidae), J. Zool. Syst. Evol. Res. 46, 24–30 [Google Scholar]
  • Crozier R.H., and Crozier Y.H. (1993) The mitochondrial genome of the honey bee Apis mellifera: complete sequence and genome organisation, Genetics 133, 97–117 [PubMed] [Google Scholar]
  • Dall’Olio R., Marino A., Lodesani M., and Moritz R.F.A. (2007) Genetic characterization of Italian honeybees, Apis mellifera ligustica, based on microsatellite DNA polymorphisms, Apidologie 38, 207–217 [CrossRef] [EDP Sciences] [Google Scholar]
  • De la Rúa P., Galián J., and Serrano J. (1998) Mitochondrial variability of honeybees populations from the Canary Islands, Mol. Ecol. 7, 1543–1547 [CrossRef] [PubMed] [Google Scholar]
  • De la Rúa P., Galián J., and Serrano J. (1999) Varibilidad mitocondrial en poblaciones de abejas de la miel del Sureste Peninsular, Invest. Agr. Prod. Sanid. Anim. 14, 24–30 [Google Scholar]
  • De la Rúa P., Galián J., Serrano J., and Moritz R.F.A. (2001) Genetic structure and distinctness of Apis mellifera L. populations from the Canary Islands, Mol. Ecol. 10, 1733–1742 [CrossRef] [PubMed] [Google Scholar]
  • De la Rúa P., Simon U.E., Tilde A., Mortiz R.F.A., and Fuchs S. (2000) MtDNA variation in Apis cerana populations from the Philippines, Heredity 84, 124–130 [CrossRef] [PubMed] [Google Scholar]
  • Engel M.S. (1999) The taxonomy of recent and fossil honey bees (Hymenoptera: Apidae, Apis), J. Hymenop. Res. 8, 165–196 [Google Scholar]
  • Estoup A., Garnery L., Solignac M., and Cornuet J. (1995) Microsatellite variation in honey bee (Apis mellifera L.) populations: Hierarchical genetic structure and tests of infinite allele and stepwise mutation models, Genetics 140, 679–695 [PubMed] [Google Scholar]
  • Evanno G., Regnaut S., and Goudet J. (2005) Detecting the number of clusters of individuals using the software STRUCTURE: a simulation study, Mol. Ecol. 14, 2611–2620 [Google Scholar]
  • Franck P., Garnery L., Solignac M., Cornuet J.-M. (1998) The origin of West European subspecies of honeybees (Apis mellifera) new insights from microsatellite and mitochondrial data, Evolution 52, 1119–1134. [CrossRef] [PubMed] [Google Scholar]
  • Franck P., Garnery L., Celebrano G., Solignac M., Cornuet J.-M. (2000a) Hybrid origins of honeybees from Italy (Apis mellifera ligustica) and Sicily (A. m. sicula), Mol. Ecol. 9, 907–921. [CrossRef] [PubMed] [Google Scholar]
  • Franck P., Garnery L., Solignac M., Cornuet J.-M. (2000b) Molecular confirmation of a fourth lineage in honeybees from the Near East, Apidologie 31, 167–180. [CrossRef] [EDP Sciences] [Google Scholar]
  • Garnery L., Cornuet J.-M., and Solignac M. (1992) Evolutionary history of the honey bee (Apis mellifera L.) inferred from mitochondrial DNA analysis, Mol. Ecol. 1, 145–154 [CrossRef] [PubMed] [Google Scholar]
  • Garnery L., Franck P., Baudry E., Vautrin D., Cornuet J.-M., and Solignac M. (1998) Genetic biodiversity of the west European honeybee (Apis mellifera mellifera and A. m. iberica). I. Mitochondrial DNA, Genet. Sel. Evol. 30, 31–47 [Google Scholar]
  • Garnery L., Mosshine E.H., Oldroyd B.P., Cornuet J.-M. (1995) Mitochondrial DNA variation in Moroccan and Spanish honey bee populations, Mol. Ecol. 4, 465–471. [CrossRef] [Google Scholar]
  • Garnery L., Solignac M., Celebrano G., Cornuet J.-M. (1993) A simple test using restricted PCR-amplified mitochondrial DNA to study the genetic structure of Apis mellifera L, Experientia 49, 1016–1021. [Google Scholar]
  • Garnery L., Vautrin D., Cornuet J.-M., and Solignac M. (1991) Phylogenetic relationships in the genus Apis inferred from mitochondrial DNA sequence data, Apidologie 22, 87–92 [CrossRef] [EDP Sciences] [Google Scholar]
  • Jensen A.B., Palmer K.A., Boomsma J.J., and Pedersen B.V. (2005) Varying degrees of Apis mellifera ligustica introgression in protected populations of the black honeybee, Apis mellifera mellifera, in northwest Europe, Mol. Ecol. 14, 93–106 [Google Scholar]
  • Moritz R.FA., Härtel S., and Neumann P. (2005) Global invasions of the western honeybee (Apis mellifera) and the consequences for biodiversity, Ecoscience 12, 289–301 [CrossRef] [Google Scholar]
  • Peakall R., and Smouse P.E. (2006) GENALEX 6: genetic analysis in Excel. Population genetic software for teaching and research, Mol. Ecol. Notes 6, 288–295 [Google Scholar]
  • Pritchard J.K., Stephens M., and Donnelly P.J. (2000) Inference of population structure using multilocus genotype data, Genetics 155, 945–959 [PubMed] [Google Scholar]
  • Raymond M., and Rousset F. (1995) GENEPOP (version 1.2): population genetics software for exact tests and ecumenicism, J. Hered. 86, 248–249 [Google Scholar]
  • Ruttner F. (1988) Biogeography and Taxonomy of Honeybees, Springer Verlag, Berlin. [Google Scholar]
  • Ruttner F. (1992) Naturgeschichte der Honigbienen, Ehrenwirth Verlag, München. [Google Scholar]
  • Sheppard W.S., and Meixner M.D. (2003) Apis mellifera pomonella, a new honey bee subspecies from Central Asia, Apidologie 34, 367–375 [CrossRef] [EDP Sciences] [Google Scholar]
  • Sinacori A., Rinderer T.E., Lancaster V., and Sheppard W.S. (1998) A morphological assessment of Apis mellifera from Palermo, Italy, Apidologie 29, 481–490 [CrossRef] [EDP Sciences] [Google Scholar]
  • Sušnik S., Kozmus P., Poklukar J., and Meglič V. (2004) Molecular characterisation of indigenous Apis mellifera carnica in Slovenia, Apidologie 35, 623–636 [CrossRef] [EDP Sciences] [Google Scholar]
  • Tamura K., Dudley J., Nei M., and Kumar S. (2007) MEGA4: Molecular Evolutionary Genetics Analysis (MEGA) software version 4.0, Mol. Biol. Evol. 24, 1596–1599 [Google Scholar]
  • Walsh P.S., Metzqer D.A., and Higuchi R. (1991) Chelex 100 as a medium for simple extraction of DNA for PCR-based typing from forensic material, Biotechniques 10, 506–512 [PubMed] [Google Scholar]
  • Whitfield C.W., Behura S.K., Berlocher S.H., Clark A.G., Johnston J.S., Sheppard W.S., Smith D.R., Suarez A.V., Weaver D., and Tsutsui N.D. (2006) Thrice out of Africa: ancient and recent expansions of the honey bee, Apis mellifera, Science 314, 642–645 [CrossRef] [PubMed] [Google Scholar]

All Tables

Table I

Details of samples location, number of analyzed colonies (n) and haplotypes in each location. Number in brackets referred to the locations showed in Figure 1.

In the text
Table II

Microsatellite variation in Croatian, Italian and Greek A. mellifera populations. Sample size (N), number of detected alleles (n), observed (Ho) and expected (He) heterozygosity per locus and mean values ± SD by population.

In the text
Table III

Proportion of membership of each pre-defined population (Croatia, Italy and Greece) to each of the four inferred populations (K = 4). Pairwise analyses were performed with STRUCTURE exploring admixture and correlated allele frequency models. N = number of individuals.

In the text

All Figures

thumbnail Figure 1

Map of East Europe with the distribution of haplotypes in the Croatian subpopulations 1 and 2 and the reference Italian and Greek populations. Sampling sites names are showed in Table I.
In the text
thumbnail Figure 2

Sequence of the tRNAleu-cox2 intergenic regions of the C haplotypes (C1 sequence corresponds to the mtDNA fragment from positions 3363-3935 bp published by Crozier and Crozier, 1993). C1, C2a and C2b haplotypes were described by Franck et al. (2000b) and C2c and C2d haplotypes by Sušnik et al. (2004). C2i are newly described in this work. Numbers correspond to variable polymorphic sites (in bold) (GenBank accession numbers FJ824582, FJ824583, FJ824584, FJ824585, FJ824586 and FJ824587).
In the text
thumbnail Figure 3

Results of STRUCTURE analysis using admixture and correlated allele frequencies models. Individuals are represented by vertical lines, grouped by inferred populations (K = 4). Division of individuals into coloured segment represents the assignment probability of that individual to each of the K groups.
In the text
thumbnail Figure 4

Distribution of A. mellifera individuals based on the genetic distance analysed with principal coordinate analysis (PCA).
In the text