Andricus lignicola (Hartig, 1840)

Mutun, Serap, Received, Hülya Karagözoğlu & Online, Published, 2015, Investigation of genetic variation among Turkish populations of Andricus lignicola using mitochondrial cytochrome b gene sequence data, Turkish Journal of Zoology 39 (5), pp. 721-733 : 729-730

4.1. Genetic variation of Andricus lignicola populations The base pair composition in A. lignicola cyt b haplotypes indicates that the collected data are genuine mitochondrial DNA, since anti-G bias is a characteristic of mitochondrial DNA genes ( Zhang and Hewitt, 1996). Neither pseudogenes nor heteroplasmy were detected when A. lignicola haplotypes were compared to the corresponding cyt b region of other insect species. In insect mitochondrial protein-coding gene segments, transitions are observed more often than transversions because of poor or deficient mtDNA repair mechanisms and tautomeric base pairing ( Brown et al., 1979). In the sequenced region of A. lignicola , the rate of substitution was well within the range of transition and transversion ratios of other insect species ( Jermin and Crozier, 1994).

freedom).

a) One group: all populations are accepted as a distinct group.

A total of 18 haplotypes were detected out of the 117 individuals of A. lignicola . Haplotype 1 (H1) is the most common haplotype and is found in 36 individuals representing eight populations (Afyon, Denizli, Düzce, Eskişehir, İstanbul, Kahramanmaraş, Kütahya, and Manisa). The frequency and wide distribution area across populations of H1 may imply that this haplotype is older than other haplotypes detected in this species. Indeed, phylogeographic studies revealed that more common haplotypes might have had a longer time to disperse through the distribution area compared to more geographically restricted haplotypes that may be derived ( Crandall and Templeton, 1993). Moreover, sharing haplotypes between/among populations may entail that A. lignicola had a widespread natural distribution in the region, as proposed by the distribution data of the species.

In A. lignicola average nucleotide diversity was 0.8%. In other gall wasp species genetic diversity shows wide variation. In A. coriarius nucleotide diversity was 0.5% in Iranian populations and 0.6% in Lebanese populations; however, in Turkey as it was placed in the main clade, nucleotide diversity was 1.5% ( Challis et al., 2007). RFLPbased haplotype and nucleotide diversity were 0.4631 and 0.3204 for A. caputmedusae ( Mutun, 2010) , 0.8089 and 0.115542 for A. lucidus ( Mutun, 2011) , and 0.45 and 0.054 for A. quercustozae ( Dinç and Mutun, 2011) . Overall assessment of the findings confirms high genetic diversity in the Anatolian populations of A. lignicola that is higher than in studied European populations of oak gall wasp species, and well within the range of other gall wasp taxa examined so far from Turkey. On the other hand, with respect to haplotype diversity, the Kütahya population had the highest variation followed by the Uşak population ( Table 3). Likewise, the highest nucleotide diversity was estimated for the Uşak population, followed by the Konya and Kütahya populations. However, in six populations (Afyon, Antalya, Denizli, Düzce, Eskişehir, and Kırıkkale) both haplotype and nucleotide diversity could not be observed (both values are 0) due to the detection of a single haplotype. Possible explanations for the low variation may be related to insufficient or different sampling sizes among localities. Haplotype richness is thought to be correlated with sampling size (Kalinowski, 2004); however, in our case there was no significant correlation between number of individuals sampled per population. This was further supported by the Chao-1 estimators in which only five populations displayed positive values. Alternately, speciesspecific demographic factors might have influenced these A. lignicola populations. However, observing only a single haplotype with balanced sampling across the distribution range may well be correlated with parasitoid attacks ( Hayward and Stone, 2006), which are common in A. lignicola populations and it may be possible that some of the lineages have been swapped out of these localities. Galls of oak gall wasps can be parasitized by inquilines and parasitoids, leading some of the individuals to fail to develop inside the gall ( Bailey et al., 2009). Synophrus politus , for example, infects galls of some cynipid species ( Washburn and Cornell, 1981). Infected galls are quite similar in coloration pattern and other phenotypic features when compared with the noninfected galls of A. lignicola . Similar effects have been reported for Andricus burgundus gall wasps ( Pujade-Villar et al., 2001); however, there is no report for A. lignicola infected by S. politus (except personal observations). In addition to inquiline and parasitoid attacks, there have been several reports of certain alphaproteobacteria, such as Wolbachia spp. , and fungi infecting and causing high mortality in gall wasps during the developmental processes of larvae ( Rokas et al., 2002). Parasitic attacks may even skew the sex ratio in oak gall wasps ( Atkinson et al., 2003).

Pairwise comparisons among eighteen A. lignicola haplotypes revealed that the highest number of base differences in mere counting was between haplotype 5 (Çanakkale population) and haplotype 18 (Uşak population) with 46 nucleotide differences. Morphologically indistinguishable species, or cryptic species, may lie within taxonomically defined species. One way to detect cryptic species is to use DNA barcoding; a higher level of sequence difference (barcoding gap) is observed between species and a lower level of genetic distance is observed within species ( Leasi and Norenburg, 2014). Cryptic species may be more common than was once thought ( Williams et al., 2012); a new cryptic oak gall wasp species from Turkey and Iran has been described recently that was previously classified under A. coriarius ( Challis et al., 2007) . The presence of a high sequence difference in our case may imply a cryptic species complex. Our preliminary ABGD analysis provided supporting results with an emphasis on the presence of four hidden lineages within A. lignicola . Since speciation is not always accompanied by morphological changes—in our case both gall wasp characters and overall gall characteristics do not show distinguishable differences—large genetic distances within this traditionally recognized species might be handled carefully and more deeply. Therefore, further research is necessary to identify the presence of a cryptic species complex within A. lignicola . Our largerscale studies to test this hypothesis are ongoing.