MICROSATELLITE DNA POLYMORPHISM IN SELECTIVELY CONTROLLED APIS MELLIFERA CARNICA AND APIS MELLIFERA CAUCASICA POPULATIONS FROM POLAND

Genetic polymorphism in selectively controlled honeybee populations of A. m. carnica and A. m. caucasica in Poland, was characterized by microsatellite DNA analysis. All honeybee samples were analyzed for nine microsatellite loci: Ac011; A024; A043; A088; Ap226; Ap238; Ap243; Ap249 and Ap256, which were found to be polymorphic in both populations. The mean number of alleles per locus was 6.222 for A. m. carnica and 4.556 for A. m. caucasica. Average observed and expected heterozygosity values were calculated as 0.976 and 0.734 in A. m. carnica and as 0.933 and 0.603 in A. m. caucasica, respectively. For the nine microsatellite loci, a total of 76 alleles were found in both populations. Thirty-five private alleles were observed in A. m. carnica and 20 in A. m. caucasica. Information about allele frequencies, FST values and genotypic differentiation is given. Nei’s genetic distance between studied populations of A. m. carnica and A. m. caucasica was calculated as 0.384.


INTRODUCTION
European honeybee populations show considerable differences in morphological, behavioral and biological characters across their natural range in the world.Many of these geographically and biologically distinct populations have been recognized as subspecies by Ruttner (1992) that have been clustered into four main lineages: C (Carnica group); M (North and Western European group); A (African group); and the O group (Oriental group).Generally, about 27 subspecies and numerous ecotypes of Apis mellifera have been described (Ruttner, 1988;Sheppard et al., 1997;Sheppard and Meixner, 2003;Meixner et al., 2010;De La Rua et al., 2009;Meixner et al., 2013).A. m. carnica and A. m. caucasica subspecies that are currently the most widely bred in Poland belong to the C and O lineage groups.Initially these two subspecies were studied mainly by classical morphometry (Gromisz, 1978;1981), then by analyzing the venation of forewings (Tofilski, 2004;2008;Gerula et al., 2009) and by alloenzyme analysis (Ivanova et al., 2011;2012).There is no study about their genetic polymorphism based on microsatellite DNA analysis.
In this study, polymorphism in honeybee populations of A. m. carnica and A. m. caucasica selectively controlled in Poland was studied using 9 microsatellite loci.The objective of the study was to investigate and characterize genetic variability among both subspecies and to provide information about allele frequencies, number of alleles per locus, levels of polymorphism, observed and expected heterozygosity and genetic distance.

Biological material
Honeybee workers from A. m. carnica and A. m. caucasica colonies in the Apiculture Division of Research Institute of Horticulture in Pulawy, Poland, with instrumentally inseminated queens, were used for this study.Five colonies per population and 6 to 10 individuals per colony were tested.Collected worker bees were transported to the laboratory alive, kept at -20 0 C and then moved to tubes with absolute alcohol until use for DNA extraction.

Statistical analyses
Population genetic statistics were computed using GENEPOP package software version 1.2 (Raymond and Rousset, 1995).F ST values were calculated according to Weir and Cockerham (1984).The exact test for Hardy-Weinberg equilibrium and genotypic differentiation were performed using GENEPOP.Unbiased estimates and standard deviations of heterozygosity were calculated according to Nei (1987).

RESULTS
All nine microsatellite loci studied were polymorphic in both of the populations.The number of alleles observed and expected heterozygosity per locus for both populations are presented in Table 2.The mean number of alleles per locus was 6.222 for A. m. carnica and 4.556 for A. m. caucasica.The average observed heterozygosity values were calculated as 0.976 in A. m. carnica and 0.933 in A. m. caucasica.Additionally, the average expected heterozygosities for the studied populations were calculated as 0.734 in A. m. carnica and 0.603 in A. m. caucasica.Information about allele frequencies and private alleles for the populations is presented in Table 3. Thirty-five private alleles were observed in A. m. carnica and 20 in A. m. caucasica.There were no private alleles in the Ap249 locus for A. m. caucasica.The highest number of private alleles (6) was observed in A. m. carnica for A024 and A088 loci.
A test for the Hardy-Weinberg equilibrium was performed for both populations at nine microsatellite loci.We detected significant deviation (P<0.001) from Hardy-Weinberg equilibrium at 16 out of 76 population combinations of loci.Almost all deviations were in favor of homozygosites, except at Ap226 in A. m. carnica and Ac011 in A. m. caucasica.
The heterozygosity level within a subpopulation (F IS ), the heterozygosity level in total populations (F IT ) and the degree of genetic differentiation of subpopulations (F ST ) are presented in Table 4.All 9 loci illustrated an excess of heterozygosity in both populations.The mean F IT amounted to -0.330 (from -0.007 to 0.561).Mean F IS within populations (P<0.001) was -0.453 (from -0.085 to -0.612).The fixation coefficients of subpopulations for the loci studied within the total populations, measured as an F ST value, varied from 0.029 (Ap226) to 0.247 (Ap243), with a mean of 0.077 (Table 4).The mean gene flow (Nm) value was calculated as 4.42.It was calculated as 0.761 for Ap243 and as 8.476 for Ap226.Nei's genetic distance between the studied populations of A. m. carnica and A. m. caucasica was calculated as 0.384.

DISCUSSION
Gene heterozygosity is a suitable parameter for investigating genetic variation.Ott (2001) gave a definition that a polymorphic locus must have a heterozygosity of at least 0.10.In this aspect, all 9 microsatellite loci studied in our investigation had high polymorphism with a mean expected heterozygosity of 0.734 for A. m. carnica and 0.603 for A. m. caucasica, showing a high degree of genetic diversity and relative high selection potential.
Microsatellite studies on honeybee populations have been generally carried out for European and African honeybee subspecies (Frank et al., 1998(Frank et al., , 2001)), whereas recent studies have been published for island populations and Mediterranean honeybee populations (Dall'Olio et al., 2007;Franck et al., 2001;Bodur et al., 2007).According to these studies, the expected heterozygosity levels were highest among African honeybee populations -between 0.76 and 0.90 (Franck et al., 2001) and lowest among western Mediterranean honeybees -between 0.26 and 0.68 (Garnery et al., 1998;Franck et al., 2001).Addition-ally, the gene diversity for North Mediterranean honeybees varied from 0.528 and 0.637 (Dall'Olio et al., 2007).Lebanon honeybees including Middle Eastern honeybee populations were studied using microsatellite loci and the gene diversity for these populations was estimated to be 0.65 (Franck et al., 2000).Bodur et al. (2007) found gene diversity values between 0.54 and 0.68, which was very similar to Middle Eastern populations.
In our study, locus-based gene diversity values ranged between 0.180 (Ap243) for A. m. caucasica and 0.848 (A088) for A. m. carnica.Ap243 revealed the smaller value of 0.180 for gene diversity in  (2003) estimated gene diversities of 0.29 and 0.61 for M and C lineages, respectively.The gene diversity for Ap256 was detected as 0.616 and 0.705 for A. m. carnica and A. m. caucasica bees, respectively, in our study.For the same locus, the gene diversity value was estimated at 0.61 for M lineage, 0.79 for C and 0.75 for A lineages (Solignac et al., 2003).The gene diversity estimations for the Ac011 locus were 0.735 and 0.555 in our study and reported as 0.80 for M lineage, 0.43 for C and 0.80 for A lineages by Solignac et al. (2003) and as 0.816 by Chaline et al. (2002) in UK honeybee populations.
Additionally, F ST values have been determined for lineages and different populations by many studies (Frank et al., 2000;2001;Garnery et al., 1998;Dall' Olio et al., 2007;Bodur et al. 2007).According to Hartl and Clark (2007), F ST levels between 0 and 0.05 indicate slight genetic differentiation; levels between 0.05 and 0.15 indicate moderate genetic differentiation; levels between 0.15 and 0.25 indicate high genetic differentiation and levels larger than 0.25 designate highly significant genetic differentiation.According to this information, the F ST values for most of the loci in our study (0.029-0.088) demonstrated slight and moderate levels of heterozygosity, except for the Ap243 (0.247) locus where high genetic differentiation in both populations was established.Franck et al. (2000) revealed that the lineage pairwise F ST values for A and M lineages were smaller than 0.1, whereas C lineage pairwise values were higher than 0.1 levels.Furthermore, Franck et al. (2001) illustrated that pairwise F ST values for A lineage were between 0.01 and 0.12, for M lineage smaller than 0.1 and for C lineage between 0.17 and 0.024.According to Garnery et al. (1998), among western European populations (Portugal, France, Spain, Sweden and Belgium), pairwise F ST was calculated between 0.002 and 0.185 (Garnery et al., 1998).Dall' Olio et al. (2007) estimated the pairwise F ST value as 0.0009 and 0.0439 for A. m. ligustica clustered in C lineage and as 0.207 and 0.322 for the "mellifera" group.Bodur et al. (2007) estimated pairwise F ST values between 0.0 and 0.183 for Turkish honeybee populations using nine different microsatellite loci.
The gene flow (Nm) values provide information about the genetic divergence or genetic similarity of subpopulations due to gene flow.If the Nm value is smaller than 2, there is still considerable genetic differentiation among subpopulations.In our study, Nm was greater than 2 for most of the loci studied.This indicated small genetic differentiations among the populations.
In this study, private alleles for both populations are described.As their frequencies were more than 5%, all 35 private alleles observed in A. m. carnica and 20 private alleles observed in A. m. caucasica, together with the most frequent alleles in the populations (Table 3), could be successfully used as suitable genetic markers.
The results of this research provide new information concerning the genetic variability in A. m. carnica and A. m. caucasica selectively reared in Poland.Comparative data about genetic polymorphism in "carnica" and "caucasica" honeybees from Poland based on microsatellite DNA analysis are reported and discussed here for the first time.The results could be useful for selection and conservation purposes.

Table 1 .
Characteristics of microsatellite loci used for STR analyses of A. m. carnica and A. m. caucasica populations

Table 2 .
Number of alleles (Na), heterozygosity observed (Ho) and expected (He) per locus for A. m. carnica and A. m. caucasica populations

Table 3 .
Allele frequencies at the microsatellite loci studied.Private allele frequencies are marked with ( * ).

Table 4 .
The heterozygosity level of individual loci within subpopulation (F IS ), in total populations (F IT ), the degree of genetic differentiation of subpopulations (F ST ) and gene flow value (Nm).