Genetic variability is essential for adaptation and could be acquired via hybridization with a closely related lineage. We use ants to investigate thermal adaptation and the link between temperature and genetic variation arising from hybridization. We test for differences in cold and heat tolerance between Finnish Formica polyctena and Formica aquilonia wood ants and their naturally occurring hybrids. Using workers, we find the parental individuals differ in both cold and heat tolerances and express thermal limits which reflect their global distributions. Hybrids however cannot combine thermal tolerance of parental species as they are equally heat-tolerant to F. polyctena , but not equally cold-tolerant to F. aquilonia . We then focus on a single hybrid population to investigate the relationship between temperature variation and genetic variation across 16 years using reproductive individuals. Based on the thermal tolerance results, we expected the frequency of putative F. polyctena alleles to increase in warm years and F. aquilonia alleles to increase in cold years. We find support for this in hybrid males but not in hybrid females. These results contribute to understanding the outcomes of hybridization, that may be sex-specific or depend on the environment. Furthermore, genetic variability resulting from hybridization could help hybrid wood ants cope with changing thermal conditions.

The purification of total DNA was performed with DNeasy Tissue kits (Qiagen) following the manufacturer’s spin-column protocol for insects. Samples were genotyped based on previous genetic studies using nine microsatellite markers (Fe7, Fe17, Fy3, Fe19, Fe13, Fy15, Fy12, Fl29, and Fy13) and PCR conditions determined in earlier studies (Kulmuni et al., 2010; Beresford et al., 2017). Total DNA was amplified with polymerase chain reactions (PCR) in Veriti 96-well Thermal Cycler (Applied Biosystems) with fluorescent labelling. Genotypes were resolved by capillary electrophoresis with a 3730 DNA Analyzer (Applied Biosystems) using 500 ROX size standard. Lastly, genotypes were scored with GENEMAPPER version 4.0 (Applied Biosystems). As genotypes were compared to samples genotyped in an earlier study on the same system (Beresford et al., 2017) we used three reference samples of known genotypes in every PCR reaction to control for variation potentially introduced by different PCR and DNA Analyzer machines.