Discussion
Hybridization results in variation with new combinations of alleles through the same mode of sexual reproduction in species, which may provide a basis for adaptive evolution (Abbott et al. 2013). However, sexual reproduction can result only in nuclear gene recombination, whereas cytoplasmic genes are maternally inherited (Birky 2001; Greiner et al. 2015). Thus, asymmetric gene flow between cytoplasmic and nuclear loci is considered to provide evidence of interspecific hybridization (Setoguchi& Watanabe 2000). Our results showed that R. flaviceps and R. chinensis exhibited asymmetric genetic differentiation between their mitochondrial and nuclear genes. There was low gene flow and high genetic divergence of mitochondrial genes, while there was high gene flow and low genetic divergence of nuclear genes between the species. Additionally, the identifiedR. flaviceps/R. chinensis colonies possessed R. chinensis/R. flaviceps nuclear genes. These results suggested that interspecies hybridization and gene exchange occurred in the two termite species in nature.
What are the possible factors causing hybridization? In termites, the overlap of the distribution and swarming of related species is common, and the sex-pairing pheromone is similar in sibling species (Aldrich& Kambhampati 2009; Capblancq et al.2015; Hartke& Rosengaus 2011). More importantly, neither males nor females possess external sclerotized genitalia (Hartke& Baer 2011). Thus, the indicated prezygotic reproductive barrier may be largely absent in some sibling termite species. Second, the initial colony is usually established under monogamy (Korb& Hartfelder 2008; Thorne et al.1999). Many monogamous individuals may fail to mate with conspecific partners due to the sex ratio, predators and/or other factors (Matsuura et al. 2002). In this case, their mating with heterospecific partners and nest building favour improved fitness when they encounter each other. Additionally, in some subterranean termites, workers can develop into secondary reproductives (Haverty& Howard 1981; Su et al. 2014), and dealates that fail to mate can also live for a long time; they may find a homosexual partner and co-establish a stable nest (Mizumoto et al. 2016). Hence, it is possible that foraging workers fuse to form a heterospecific virgin colony and then mate with heterosexual dealates, which provides a possible mechanism for hybridization between dealates and secondary reproductives developing from workers.
Adaptive introgression of genetic material between incompletely isolated species may cause various impacts on genetic diversity. It can contribute to genetic diversity or facilitate genetic assimilation depending on the size and construction of populations (Seehausen 2004). For social termites, many colonies are established by founders from the same natal colony (DeHeer& Vargo 2005; Vargo& Husseneder 2009). One closed colony can be composed of kin, which may lead to inevitable inbreeding. In fact, high genetic diversity is still maintained in natural populations (Huang et al. 2013; Vargo& Husseneder 2009). Fortunately, hybridization and introgression between incompletely isolated species may explain why genetic diversity can be maintained in natural populations of termites with respect to inbreeding. We previously noted that R. flaviceps and R. chinensis possess similar behaviours and lack conspecific preferences in the process of reproduction (Wu et al . 2019). At the same time, we found that artificial hybrid colonies can produce living offspring under laboratory conditions. Combined with the results regarding genetic exchange between the species in nature in this study, we further confirmed the speculation that the maintenance of high genetic diversity in termites is related to adaptive introgression and hybridization.
In general, incomplete reproductive barriers occur in young species because they remain in the initial stage of differentiation. For example, a host shift led to prezygotic reproductive isolation between the original species and host races in Rhagoletis pomonella , but the postzygotic reproductive barriers were invalid due to the short time of differentiation (Dambroski& Feder 2007). However, hybridization between long-diverged species often negatively affects the viability and fertility of hybrid offspring, preventing gene flow between species (Anderson et al. 2010; Brideau et al. 2006; Brucker& Bordenstein 2013; Fitzpatrick 2008; Moyle& Nakazato 2009). Inconsistent with that prediction, the differentiation time between R. flavicepsand R. chinensis is 2-3 million years (Bourguignon et al. 2016). Our results indicated that they exhibit incomplete reproductive barriers and that there is gene flow between the two species. Similar patterns of gene exchange are also reported in the related tapeworm speciesSchistocephalus solidus and S. pungitii despite their differentiation 20-25 million years ago (Henrich& Kalbe 2016). These results suggested that interspecific gene exchange shows not relationship with the time of species differentiation.
During the process of evolution and speciation, differentiated species are often re-fused due to diffusion or environmental changes, which can result in adaptive introgression or speciation via hybridization. The phylogenetic relationship of this evolutionary process cannot be described through a binary bifurcation structure but by a network, namely, reticulate evolution. Therefore, the existing phylogenetic tree based on a binary bifurcation structure cannot truly describe the evolution of species (Pennisi 2016). Additionally, hybridization can reduce differences between taxa or even cause a genetic homogenizing effect (Bernalet al. 2017; Schierenbeck 2011). Alternatively, hybridization results in the introgression of genetic material and accelerates speciation (Abbottet al. 2013; Bendesky et al.2017; Payseur& Rieseberg 2016). However, hybridization and gene introgression existed in the termiteReticulitermes (Dedeine et al. 2016). Neither the above-described reductions in genetic differences between species nor speciation are observed in the termiteReticulitermes . If this were the case, there would clearly be strong selection against hybridization to maintain species integrity in nature. Further studies on the maintenance mechanisms of species will still be required in the future.