Introduction
Population genetics within a host-parasite association are complex systems dependent on many ecological features of both counterparts (Criscione, Poulin, & Blouin, 2005; Barrett, Thrall, Burdon, & Linde, 2008; Sweet & Johnson, 2018). Within hybrid zones (HZ) and secondary contact zones (SCZ) this picture is likely to become even more complicated, possibly giving rise to new unexpected patterns. Unfortunately, very few studies have been devoted to this aspect of host-parasite interactions. From the most general point of view, it is assumed that since parasites are dependent on their host, their genetic structure will tend to mirror the host. In this respect, two assumptions are frequently expressed. First, the degree of congruence with the host is dependent on traits connected to the parasitic life-style, such as the degree of host-specificity, transmission mode, presence of dispersal stages, etc. (Maze-Guilmo, Blanchet, McCoy, & Loot, 2016). Generally, the more intimate the association, the higher the degree of congruence. However, this general view may be twisted by many specific traits of the particular host-parasite association. For example, the population structure of heteroxenous parasites (parasites with more than one host in their life cycles) is likely to reflect the least structured host, since any potential structure is erased by the more motile host (Jarne & Theron, 2001; Louhi, Karvonen, Rellstab, & Jokela, 2010). Similarly, with longer free living stage(s), the genetic structures of the host and the parasite become more incongruent (Jarne & Theron, 2001). The second assumption is about the speed of diversification: since the parasites have a shorter generation time, they undergo faster genetic diversification, which may eventually lead to the parasite’s duplication (Page, Lee, Becher, Griffiths, & Clayton, 1998; scenario a in Figure 1). The assumption about the higher mutation rate in parasites was demonstrated in several studies (Nieberding, Morand, Libois, & Michaux, 2004; for ectoparasites: McCoy et al., 2005; Whiteman, Kimball, & Parker, 2007; but see Gómez-Díaz, González-Solís, Peinado, & Page, 2007; Jones & Britten, 2010 for the opposite results).
Although many studies have been devoted to comparing phylogenies and population structures of host-parasite associations, only a few analyzed these processes in connection to SCZ and HZ of the hosts (reviewed by Theodosopoulos, Hund, & Taylor, 2019), and only recently, de Bellocq et al. (2018) focused on detecting HZ in parasite populations. Using two parasites of the house mouse Mus musculus , the nematodeSyphacia obvelata and the fungus Pneumocystis murina , they found that within the host’s HZ both parasites create their own HZ. They also demonstrated that the parasites (reaching higher genetic divergence) created significantly narrower HZs than the host (scenariob in the Figure 1).
From a theoretical point of view, the assumptions and empirical evidence discussed lead to a third possible scenario: during secondary contact the host does not create a HZ but rather re-establishes a panmictic population, while the parasite accumulates a degree of genetic differences which prevent re-establishment of panmixia but does not lead to speciation (scenario c in the Figure 1). A paradoxical result of such an event would be establishment of a parasite’s HZ within the host’s panmictic population, which on a microevolutionary scale would function as a “host-cleansing filter”: passing through this filter from area A to area B (Figure 1d), the hosts rid themselves of area A parasites and acquire the area B parasites. To our knowledge, this “filter” has never been observed in nature. In fact, the presence of a parasite’s HZ in the scenario described here is difficult to guessa priori , as it is not indicated by the host’s HZ. However, in our previous work (Martinů, Hypša, & Štefka, 2018) we presented the genetic structure of postglacial Europe recolonization by the mice of the genus Apodemus and their ectoparasite, the lousePolyplax serrata , which corresponds to such scenario (Figure 2; see below for details).
Similar to all sucking lice, P. serrata is a permanent homoxenous ectoparasite with strict host specificity, which is transmitted almost exclusively during physical contact of its hosts. As such it falls into the category of highly intimate parasites displaying a high degree of congruence with their hosts. In Figure 2, we summarize the main features of the population genetic pattern obtained by analysis of 379 bp mitochondrial haplotypes (Martinů et al., 2018). It shows that P. serrata is composed of several genetic lineages (Figure 2d) with different host-specificities and geographic distributions. This indicates that even such traits as the degree of host specificity may be very flexible and change rapidly at a shallow phylogenetic level. For example, the so-called specific (S) and non-specific (N) lineages, although closely related (sister lineages) and living in sympatry, differ in degree of their specificities, one being exclusive to Apodemus flavicollis, while the other can also live onA. sylvaticus . However, the most intriguing part of the pattern was detected within the S lineage. On the mtDNA based phylogenetic trees, the host (A. flavicollis ) and the parasite (S-lineage ofP. serrata ), display the same basic structure. Their samples collected across all of Europe form two genetically distant clusters, suggesting recolonization from two different refugia (the taxa designated by red and blue colours in the Figure 2; see Martinů et al., 2018 for discussion). However, while the two host’s clusters have already spread across the entirety of Europe, their lice did not follow the same process. Instead, their two sub-lineages, designated asspecific east (SE) and specific west (SW), ceased their dispersion after reaching the SCZ in the middle of Europe (Figure 2). This disparity is surprising since the high intimacy of lice should predetermine them to mirror genetic structure of the host (e.g. Harper, Spradling, Demastes, & Calhoun, 2015).
To obtain a more complex picture of secondary contact in P. serrata , in this study we analyze three patterns derived from metagenomic data of 13 louse specimens collected across the SCZ: nuclear SNPs, complete mitochondrial genomes, and complete genomes of the symbiotic bacterium Legionella polyplacis . We use these analyses to retrieve two kinds of information. First, we compare nuclear and maternally inherited markers (mitochondrial genomes, symbiotic genomes) to demonstrate a narrow HZ between the SW and SE lineages of the lice. Second, we address two possible causes of the SE/SW incompatibility suggested in our previous work (Martinů et al., 2018). P. serratacarries the intracellular obligate symbiont Legionella polyplacis(Říhová, Nováková, Husník, & Hypša, 2017) which could be incompatible with the non-native genetic background. Similarly, since thePolyplax louse mitochondria are fragmented into 11 minichromosomes (Dong, Song, Jin, Guo, & Shao, 2014) a rearrangement of their genetic composition could theoretically lead to the SE/SW incompatibility. We therefore compare complete mitochondrial and symbiotic genomes to assess the degree of their divergence.