INTRODUCTION
In recent years, many insect species have experienced pronounced changes in their distribution, including several documented cases of range expansions (González-Tokman et al., 2020; Sánchez-Guillén et al., 2016). A common signature of range expansion is decreasing genetic diversity at the range limit, due to frequent founder effects and genetic drift (Slatkin & Excoffier, 2012). However, such a loss of diversity can be counteracted by hybridization and introgression in situations where the expanding species comes in contact with a formerly allopatric sister taxa (Rieseberg et al., 2007). This process is of increasing interest as many regions are facing changes due to anthropogenic pressures and ongoing climate change, which can lead to novel interactions and hybridization among expanding species (Sánchez-Guillén et al., 2016). Evidence is accumulating that both the extent and pattern of hybridization have important conservation implications due to their direct impact on biodiversity as a whole, and because this process can sometimes lead to the replacement of one of the hybridizing taxa (Abbott et al., 2013; Seehausen, 2004; Todesco et al., 2016). Thus, understanding evolutionary outcomes of hybridization is of increasing interest for gaining general insights into species dynamics and conservation in response to environmental change (Abbott et al., 2013).
Hybridization outcomes are manifold and include transfer of genetic material between species (potentially facilitating their adaptive evolution), fusion of species, genetic swamping of one species by another, elicit reinforcement of reproductive isolation between incompletely isolated species, and the origin of new species (Seehausen, 2004). The likelihood of these outcomes is highly dependent on intrinsic factors, such as the extent of reproductive isolation, as well as extrinsic factors, because hybrid fitness may be dependent on the environment (Coyne, 2004). The study of multiple contact regions of the same species pair allows to test the consistency of hybridization outcomes (Harrison & Larson, 2016). Thus, when the range of two closely related species is large and overlapping, and if this overlap results in hybridization, then hybridization at several spatial scales can provide insights into parallel patterns of hybridization in different demographic and ecological contexts (Abbott et al., 2013). Several empirical studies, mainly from plants (Buerkle & Rieseberg, 2001; Haselhorst & Buerkle, 2013; Sweigart, Mason, & Willis, 2007) and fishes (Aboim, Mavárez, Bernatchez, & Coelho, 2010; Mandeville et al., 2017; Nolte, Gompert, & Buerkle, 2009), but also from mammals (Good, Handel, & Nachman, 2008), amphibians (Vines et al., 2003) and insects (Gompert et al., 2014), have shown that the outcomes of hybridization varies among hybrid regions of the same species pairs while only few studies have shown consistency in the pattern, e.g., in plants (Buerkle & Rieseberg, 2001) and insects (Larson, White, Ross, & Harrison, 2014). Detecting parallelism in the prevailing genomic signatures allows gaining insights into the process of hybridization itself, for example, if species have genomic regions that introgress more readily than others.
Odonata are a group that is heavily affected by increasing temperatures and many species are changing their distributions (Hassall & Thompson, 2008; Hassall, Thompson, & French, 2007; Hickling, Hill, & Thomas, 2005; Lancaster et al., 2016; Ott, 2010). Within odonates, the damselfly genus Ischnura is extremely species rich (around 70 species) (Dijkstra & Kalkman, 2012; Sánchez-Guillén et al., 2020) and includes many closely related and recently diverged species that co-occur in partial sympatry over parts of their range (Sánchez-Guillén, Muñoz, Rodríguez-Tapia, Arroyo, & Córdoba-Aguilar, 2013; Wellenreuther & Sánchez-Guillén, 2015). This genus has also other interesting properties such as high local abundances and wide-ranging distribution patterns, and it is highly enabled for field identification and sampling, making this a suitable group to explore adaptive introgression caused by environmental change. Indeed, there are several examples of extensive hybridization between species within the Ischnura genus, for example, between I. denticollis and the endangered I. gemina in the San Francisco Bay (Leong & Hafernik, 1992; Sánchez-Guillén et al., 2014), and in Europe between I. graellsiiand I. saharensis, and between I. genei and I. elegans (Sánchez-Guillén, Córdoba-Aguilar, Cordero-Rivera, & Wellenreuther, 2014).
However, the best documented and explored case of hybridization inIschnura is between I. elegans and I. graellsii in Spain (Monetti, Sánchez-Guillén, & Cordero Rivera, 2002; Sánchez-Guillén, van Gossum, & Cordero-Rivera, 2005; Sánchez-Guillén, Wellenreuther, & Cordero-Rivera, 2012; Wellenreuther et al., 2018).Ischnura elegans and I. graellsii are among the most frequently studied damselfly species within the order Odonata, and as such have become eco-evolutionary model species that have been studied extensively for hybridization and reproductive isolation (Monettiet al. , 2002; Sánchez-Guillén et al. , 2005, 2011, 2012, 2014b; Wellenreuther & Sánchez-Guillén, 2016; Wellenreuther et al. , 2018). Ischnura elegans and I. graellsii are closely related and share many morphological, genetic and phenotypic traits (Sánchez-Guillén et al., 2011), including preference traits for habitats, and they occupy similar ecological niche space in Spain (Wellenreuther et al., 2018). Ischnura elegans has dramatically expanded its distribution in Spain during last 40 years (Sánchez-Guillén et al., 2011), where its distributional range nowadays overlaps almost completely with I. graellsii (Fig. 1A-B), which is endemic to Spain and Maghreb in north-western Africa. The Mediterranean coast represents the oldest hybrid region with the firsts records of I. elegans in the late 19th century (Ocharan, 1987), with increasing records over the country from the early 1980s and onwards (Ocharan, 1987), and a fast expansion from 2000 to the present (Fig. 1B). A previous molecular study with microsatellites estimated that in the Spanish hybrid populations 55-60% of the I. elegans individuals were introgressed or backcrossed, and that from 10-30% were F1 or F2 hybrids (Sánchez-Guillén et al., 2011). In the Spanish hybrid region, I. elegans has replaced I. graellsii from the coast and modified its original environmental niche to become more similar to the I. graellsii niche (Wellenreuther et al., 2018).
In the present study, we focus on two hybrid regions in Spain, one in north-west Spain (Galicia) and one in north-central Spain (La Rioja, Navarra and Avila). We used Restriction site–Associated DNA (RAD) sequencing and a recent I. elegans reference genome assembly (Chauhan et al., 2021) to identify genome-wide SNPs in individuals from the two hybrid regions as well as from eight allopatric populations from both species in Spain and adjacent countries. First, we used species-specific SNPs to analyze the fine-scale individual ancestry and proportion of individuals in different hybrid classes (ranging from pure individuals to F1 and F2 hybrids). With this analysis we were particularly interested in confirming the occurrence of hybrids and backcrosses, and in evaluating the local and regional scales of hybridization dynamics. We used data on the strength of reproductive barriers and colonization history in an aim to explain the present-day hybridization patterns. Second, we used the full set of SNPs to compare the degree genetic variation and genetic differentiation in the Spanish hybrid regions and adjacent allopatric regions. With these analyses we were interested in evaluating whether hybridization and introgression have increased the level of genetic variation of individuals in the different hybrid classes. We discuss our results in light of parallelism in hybridization and introgression within and between regions, or the lack thereof, and in relation to previous data on biased mating success among I. elegans and I. graellsii . This work is a step towards understanding the potential role of hybridization and introgression in facilitating range expansions.