Introduction:
In light of the unprecedented number of introductions of nonnative species (Mack et al., 2000), one of the most pressing research needs for evolutionary biologists and ecologists is to identify the factors that influence the establishment of species that have negative ecological and economic impacts (Suarez & Tsutsui, 2008). Multiple introductions (Dlugosch & Parker, 2008), including cryptic ones (Roman, 2006), are thought to play an important role in providing the diversity required to overcome genetic bottlenecks associated with the establishment of populations in novel ecosystems (Darling, Bagley, Roman, Tepolt, & Geller, 2008; Facon, Pointier, Jarne, Sarda, & David, 2008). However, when multiple geographically disjunct populations of an invasive species become established, it is often unclear whether the species is a serial invader (i.e., each population was introduced independently) or whether the separate populations represent establishment from within the invasive regions under a “stepping stone” model (see Cerwenka, Alibert, Brandner, Geist, & Schliewen, 2014; Lombaert, Guillemaud, Cornuet, Malausa, Facon, & Estoup, 2010; Oficialdegui et al., 2019; Tonione, Reeder, & Moritz, 2011 for examples). Identifying which mode of introduction occurred (serial invader or stepping stone) is necessary for the study of genetic and ecological factors that drive invasion success as independent populations are necessary for robust hypothesis-based testing (Kang, Buckley, & Lowe, 2007), and this information is also crucial for focusing management efforts (Floerl, Inglis, Dey, & Smith, 2009).
When reconstructing regions of origins, and determining the numbers of introductions of a focal organism, ideally historical records should be observed and robust genetic analyses performed (e.g., Lynch & Saltonstall, 2002; Schwenk, Brede, & Streit, 2008). Unfortunately, historical records may not exist for all introduced species, and it is not uncommon for an introduced species to go unnoticed for long periods of time before becoming an invasive pest. Genetic analyses have the ability to independently reconstruct regions of origins, and in some instances provide estimates for the times of introductions (Auger-Rozenberg et al., 2012; Barker, Andonian, Swope, Luster, & Dlugosch, 2017; Javal et al., 2019; Lesieur et al., 2019; Lombaert et al., 2010; Oficialdegui et al., 2019; Zardus & Hadfield, 2005). However, the power of these analyses are constrained by numerous factors including the underlying genetic structure of the species, the number of generations since the introduction, the effective size of the founding population(s), the strength of the bottleneck the population(s) experienced, and/or the presence of differing selective pressures in the native and introduced regions. As such, one common finding is for introduced populations to be reconstructed as genetically “distinct” from all sampled source populations (e.g., Barker et al., 2017; Wu et al., 2015). This result could be due to the tendency for commonly implemented Bayesian genetic clustering algorithms to over-split populations (Frantz, Cellina, Krier, Schley, & Burke, 2009) or an artifact generated during the interpretation of results (Lawson, Van Dorp, & Falush, 2018).
Here we explore the invasion history of the economically damaging defoliator the winter moth, Operophtera brumata L. (Lepidoptera: Geometridae). In its native distribution across Europe, North Africa, and western Asia, winter moth defoliates a wide range of tree and shrub species (Ferguson, 1978). Populations of winter moth in Europe have been used as a model for the study of population ecology (Varley, Gradwell, & Hassell, 1973), and this species has been critically important for understanding the importance of spatial-synchrony (Jepsen et al., 2009) and synchrony of hatch with host tree bud-burst (Embree, 1965 ,Varley & Gradwell, 1960, Visser & Holleman, 2001) – a factor studied in other invasive defoliator populations as well (e.g. Hunter & Elkinton, 2000). The invasion history of winter moth in North America has been well documented, with populations of winter moth first reported in the 1930’s in Nova Scotia (Embree, 1967; MacPhee, 1967; MacPhee, Newton, & McRae, 1988), the 1950’s in Oregon (Kimberling, Miller, & Penrose, 1986), the 1970’s in British Columbia (Gillespie, Finlayson, Tonks, & Ross, 1978), and in the 1990’s in coastal regions of northeastern United States (Elkinton et al., 2010; Elkinton, Liebhold, Boettner, & Sremac, 2014). These populations are thought to have been introduced by the movement of infected nursery stocks (Ferguson, 1978). However, where in Eurasia these populations were introduced from, and whether these represent a single introduction that was then spread across North America, or multiple introductions (or some combination of these) is unclear. Previously, a genetic examination of the invasion history of winter moth in North America was conducted, but was unable to discern these patterns due to low levels of mitochondrial DNA diversity in both introduced and native samples (Gwiazdowski, Elkinton, DeWaard, & Sremac, 2013).
To overcome this limitation, we examine the invasion history of winter moth in North America using polymorphic microsatellite loci amplified from individuals collected across its native and introduced regions. We specifically examine how many times winter moth was introduced to North America, and when possible, determine the specific source location. Lastly, we comment on the role of hybridization between winter moth and a North American congener the Bruce spanworm (O. bruceata Hulst), and the effects of genetic bottlenecks on the establishment of invasive winter moth populations.