Discussion:
Documenting the establishment and formation of new hybrid zones in real time is critical for understanding the spatial and temporal nature of these regions of genetic interchange (Abbott et al., 2013; Mallet, 2005). In addition, understanding the dynamics of hybridization between native and non-native species may be particularly important for understanding how invasive species become established and spread, because reproducing with a native species could alleviate Allee effects that limit the establishment of small populations due to stochastic disturbances and mate-finding (Ellstrand & Schierenbeck, 2000; Espeland, 2013; Mesgaran et al., 2016; Pfennig, Kelly, & Pierce, 2016; Yamaguchi, Yamanaka, & Liebhold, 2019). Here, we document the real time formation of a clinal (sensu Taylor et al., 2015) hybrid zone, following the introduction of the invasive winter moth to the northeastern United States. Our analyses suggest that the location of the center of this hybrid zone might not be regulated primarily by environmental variables, but appears to be behaving as a tension hybrid zone. Tension hybrid zones are characterized by their independence from environmental variables, narrow geographic width, low frequencies of hybridization, and with the geographic location determined by a balance between dispersal (dependent on population density) and selection against hybrids (Barton & Hewitt, 1985; Key, 1968; Smith, Hale, Kearney, Austin, & Melville, 2013). As shown by examining our two transects, the hybrid zone is narrow, with a mean of ~ 40 km across all years in both transects, and hybridization rate is low (~ 6 % in both transects). The location of the hybrid zone also appears to be dependent on population size, which would influence dispersal rate because the center of the hybrid zone is near the region where the population size of winter moth drops to where it is similar to the endemic Bruce spanworm populations (Figures 7 and 8). The final feature of a tension zone, low hybrid fitness, has also been demonstrated in this system. Laboratory rearing of winter moth and Bruce spanworm produced 93.5 and 94.1% viable eggs, respectively, while crosses between winter moth females and Bruce spanworm males produced 60.8% viable eggs and crosses between Bruce spanworm females and winter moth males produced just 22.1% viable eggs (Havill et al., 2017). The near complete lack of Bruce spanworm backcrosses also indicates low hybrid fitness in this system (Havill et al., 2017; Andersen et al., 2019; this study). Interestingly, these two species appear to have few pre-zygotic barriers to hybridization since they share the same sex pheromone (Elkinton et al., 2011) and have overlapping mating flight periods (Andersen, unpublished data). The barriers to hybridization in this system, therefore, appear to be almost entirely made up of post-zygotic incompatibilities resulting from > 500,000 years of allopatric divergence (based on an averaged observed mitochondrial percent divergence between these two species of 7.5% documented in Gwiazdowski, Elkinton, DeWaard, & Sremac, 2013; and the newly calibrated mitochondrial mutation rate of approximately 14.5% per million years for insects presented in Key, Frederick, & Schul, 2018).
Separating environmental factors (e.g., climate, land use, etc.) from population factors (e.g., dispersal, abundance, hybrid fitness, etc.), may not always be entirely feasible, and could, in part, explain why there is a paucity of documented examples of this type of hybrid zone between an introduced and a native species. However, by comparing our two transects that differed in extreme minimum temperatures (Massachusetts from -20°C to -24°C, and Connecticut ~ -17°C), our results indicate that extreme minimum winter temperatures are not constraining the geographic location or width of the winter moth x Bruce spanworm hybrid zone. That said, it should be noted that researchers in Europe have observed that populations of winter moth can rapidly adapt to changes in environmental conditions (van Asch, Salis, Holleman, van Lith, & Visser, 2013), and as such the winter moth x Bruce spanworm hybrid zone presents an exciting system to study the combined roles of local adaptation and hybridization in the establishment an invasive species under changing climate regimes.
In contrast to direct environmental constraints on the location of the hybrid zone, we believe that population factors are more important for explaining the differences in relative population densities of these two species and therefore the stability and dynamics of the hybrid zone. One such constraint might be top-down pressure by natural enemies of both species. The biological control of winter moth in North America is one of the best-known examples of the successful use of importation biological control (Van Driesche et al., 2010) to reduce the ecological and economic impacts of a non-native forest defoliator with a broad host range (Elkinton et al., 2015; Embree, 1966; Kimberling et al., 1986; Roland & Embree, 1995). Recently, Elkinton et al. (2021) showed that the introduction of a single specialist natural enemy to the Northeast was able to convert winter moth to non-pest status. These introduced natural enemies have been incredibly effective at reducing the abundance of winter moth in high density locations, but at low densities, numerous authors have found that native pupal parasitoids play an important role in regulating winter moth population sizes (Frank, 1967a, 1967b; Horgan, 2005; Horgan & Myers, 2004; Latto & Hassell, 1987; Raymond et al., 2002; Roland, 1994; Roland & Embree, 1995, Broadley 2018). For example, in the Northeast, Broadley (2018) found that mortality caused by native generalist pupal parasitoids was lowest in the eastern coastal regions and increased as she sampled locations into the western interior portions of this region. Pupal parasitism could therefore play an important role in limiting the population sizes of both species, and as a result providing the necessary balance for a tension hybrid zone to exist in this system (see Taylor et al., 2015). It will be interesting to observe whether the location of the hybrid zone shifts east as the population density of winter moth continues to decrease due to the impacts of biological control. Indeed, the eastward retreat of the hybrid zone in 2018, the last year of our study (Figures 7 and 8), may indicate that this has begun.
It is commonly acknowledged that during the invasion process, the probability of establishment of non-native species can be influenced by native predators, parasitoids, competitors, and/or microbial communities through a process known as biotic resistance (Alpert, 2006; Dawkins & Esiobu, 2016; Kimbro, Cheng, & Grosholz, 2013; Levine, Adler, & Yelenik, 2004). For several decades there has been considerable concern expressed in the literature about the risk of hybridization between native and introduced species resulting in the “hybridization to extinction” of the native species (Allendorf et al., 2001; Ayres, Zaremba, & Strong, 2004; Hinton, 1975; Levin, 2002; Levin, Francisco-Ortega, & Jansen, 1996; Prentis, White, Radford, Lowe, & Clarke, 2007; Rhymer & Simberloff, 1996; Todesco et al., 2016; Wolf, Takebayashi, & Rieseberg, 2001). Under a tension hybrid zone model, however, the continued exchange of genetic material and the resulting production of low-fitness hybrids, could result in a reduction in the rate of spread of the introduced species by stabilizing the geographic center of the hybrid zone, creating what we believe is an underappreciated form of biotic resistance to invasion (sensuLevine et al., 2004). As such, we encourage additional research into the possible role of hybridization for limiting the establishment and spread of non-native species.