Reproductive mode and mating system in the invasive wintercreeper vine (Euonymus fortunei) in southwestern Ohio: polyembryony, apomixis, selfing, and outcrossing

Abstract

Spread of invasive species can be impacted by their mode of reproduction (asexual vs. sexual) as well as the mating system (outcrossing vs. selfing). This is especially the case in the evergreen wintercreeper vine [Euonymus fortunei (Turcz.) Hand.-Maz.], which was originally brought to the United States for horticultural purposes and is now considered invasive across the Midwest. Wild wintercreeper populations consist primarily of a single polyploid genotype, the ornamental ‘Coloratus’ cultivar, but it is still unknown how this species produces its fruit during the fall. We examined the reproductive mode and mating system of wintercreeper by collecting leaves and fruits from 12 wild plants in an urban location of Cincinnati, OH. In this genetic survey, we used microsatellite markers to identify the pollen donor of each embryo within the seeds. Polyembryony was relatively common, with 37.4% of seeds each containing two to four embryos. Many of the 382 embryos extracted were produced asexually through apomixis (50.0%) or were sexual products of outcross fertilization (34.3%) or self-fertilization (15.7%). In seeds with multiple embryos, larger embryos were most likely to be outcrossed, with winged burning bush [Euonymus alatus (Thunb.) Siebold ‘Compactus’] as the most likely pollen donor, and apomixis increasing in successively smaller embryos. Single embryos within seeds were more often outcross fertilized (52%). The fact that all wild adult wintercreeper plants consist of a single genotype is consistent with the production of these apomictic offspring. However, lack of sexually produced wild plants, despite their appearance in the embryonic stage, warrants further study. This is the first report of polygamous apomixis in this species, and research is continuing into how this reproductive strategy may influence invasive spread of the species.

Management Implications

Control and management of invasive populations of Euonymus fortunei (wintercreeper) is notoriously difficult, given the broad sprawling habit of this tough and wiry evergreen vine upon the forest floor. Wild plants consist solely of an escaped ornamental cultivar, the tetraploid ‘Coloratus’. Although the vegetative propagation of wintercreeper vine through stem fragments is unquestioned, fruits and seeds can also be produced on plants that grow upward and flower under high light. We demonstrate here that these typically bird-dispersed seeds often contain multiple embryos, many of which are produced asexually through apomixis and serve to perpetuate the Coloratus genotype in the landscape. In comparison, embryos in monoembryonic seeds are often sexually derived, resulting from outcross fertilization by a common ornamental cultivar of the related Euonymus alatus (winged burning bush, ‘Compactus’). Although we have yet to detect any outcross genotypes in previously sampled wild populations, we recommend that land managers proactively prevent wintercreeper from flowering and producing seeds by preventing plants from climbing upward on trees, poles, or other vertical support and monitoring the presence of E. alatus on their properties. Preventing the production and subsequent bird dispersal of apomictic seeds of the Coloratus genotype is imperative for effectively controlling and managing this species in natural areas.

Introduction

Both the reproductive mode (asexual or sexual) and the mating system (how sexual reproduction takes place: outcrossing or selfing) can impact the spread of an introduced plant species within a new environment. While asexual reproduction through vegetative propagation and/or apomixis typically favors locally adapted genotypes, sexual reproduction can generate novel genotypes through meiosis, genetic recombination, and outcrossing events. These novel genotypes may confer higher fitness with ongoing climate change and during range expansion into new habitats. However, sexual recombination can also disrupt and reduce the frequency of locally adapted genotypes within the home location (Muller 1964; Vrijenhoek 1979). In apomixis, pollen may still be available for occasional outcrossing, although it may have low viability (Hand and Koltunow 2014). Invasive species that can propagate asexually (or clonally) through vegetative propagation or apomixis may also readily spread because they are not constrained by proximity to mates (Baker 1967); the same is true for species that reproduce sexually through self-pollination. In contrast, obligately outcrossing species are relatively less likely to spread into these new habitats and persist without mates, especially if they lack suitable pollinators or optimal environmental conditions (Thurlby et al. 2012). In a best of both worlds situation, an invasive species may be most successful if it can incorporate multiple approaches—for example, local spread via vegetative propagation combined with distant dispersal of fruits with apomictic or sexually derived seeds at the edge of an expansion front where different habitats most likely occur.

Wintercreeper [Euonymus fortunei (Turcz.) Hand.-Maz.; also known as spindle tree, and Fortune’s spindle] is an ornamental perennial groundcover that exhibits asexual and sexual reproduction (Elam 2023) and is increasingly recognized as an invasive species in the United States (Anonymous 2023; Elam and Culley 2023). This popular species is still sold today in plant nurseries and big box stores in the U.S. Midwest as different cultivars, such as ‘White Album’, ‘Gold Splash’, and ‘Moonshadow’. Recently, Elam and Culley (2023) reported that wild populations of E. fortunei across several U.S. states consist of a single escaped cultivar, Coloratus. This cultivar was once popular in the ornamental plant trade, largely desired because of its purple-tinged leaves, which can revert back to a green hue in shaded conditions. Coloratus is also tetraploid (Elam and Culley 2023), the only known polyploid cultivar of E. fortunei mass produced for commercial sale.

While the observation that wild populations of E. fortunei consist of a single genotype is informative, the actual mechanism through which E. fortunei spreads in the natural environment to form satellite wild populations remains unknown. The species is capable of spreading vegetatively, when branches of phanerophytic individuals established in trees are broken off, fall, and are swept away during storm events. If conditions are favorable, these propagules can establish themselves in their new environment through adventitious roots (Weeks and Weeks 2012). In fact, commercial propagators have taken advantage of this trait to increase their stock of individual cultivars through cuttings (Boyer et al. 2008). Older E. fortunei individuals in the wild can also produce hermaphroditic flowers that form seeds after plants climb vertically up trees and poles to reach full sunlight (Conover et al. 2016; Leicht-Young 2014). Fruits mature in late October and early November, releasing their bright orange arils, which encase the seeds and contrast against the winter landscape (Carlo and Morales 2008; Herrera et al. 2011; Noma and Yumoto 1997), attracting birds as seed dispersers (Rounsaville et al. 2018). Seeds may also be dispersed by white-tailed deer (Odocoileus virginianus), which ingest seeds while eating foliage and defecate them later; or seeds may just fall to the ground to germinate (Rounsaville et al. 2018) or be washed away to establish new populations in nearby riparian zones (Weeks and Weeks 2012). Given that new satellite populations of invasive E. fortunei in natural areas most likely to originate from these seeds, it is important to understand the mating system that gave rise to them.

One possibility is that seed production in E. fortunei is sexually facultative, with asexual reproduction through apomixis and sexual fertilization occurring through outcross or self-pollen deposition (Brizicky 1964). This type of mating system is often associated with polyembryony, the production of two or more embryos in a single seed. Polyembryony has already been observed in Celastraceae (Brizicky 1964), including in some Euonymus species such as E. fortunei (Rounsaville et al. 2018). In polyembryony, both sexually and asexually derived embryos can inhabit the same seed (Hand and Koltunow 2014), with the apomictic embryo being genetically identical to the maternal parent (Koltunow and Grossniklaus 2003). In general, apomictic embryos may or may not require a double fertilization event, as they can develop in some species from the nuclear or integumentary cells within the ovule (i.e., sporophytic apomixis or adventitious embryony) or from an unfertilized egg cell in a different embryo sac (gametophytic apomixis; Hand and Koltunow 2014; Spielman et al. 2003; Viana et al. 2021). Within a polyembryonous seed, one or more asexual embryos may be found in different positions alongside the zygotic embryo formed from pollen deposition on the stigma (Brizicky 1964; Lakshmanan and Ambegaokar 1984). All embryos are not guaranteed to germinate, and in species such as Scots pine (Pinus sylvestris L.), programmed cell death eliminates all but one embryo to ensure its survival (Filonova et al. 2002).

In this study, we first determined the frequency of polyembryony in E. fortunei by dissecting embryos from their seeds obtained in an urban area of southwestern Ohio. We then used codominant microsatellite loci to examine the mode and mating system of an invasive population to determine whether embryos produced by wild individuals develop from outcrossing, self-fertilization, and/or apomixis. If outcrossing is present, parentage information could potentially indicate whether intraspecific hybridization among different cultivars of E. fortunei and/or interspecific hybridization with closely related species is occurring within wild populations in Ohio.

Materials and Method

Twelve fruiting E. fortunei plants were sampled throughout the Clifton neighborhood in Cincinnati, OH, during October 2018, Individuals were usually found growing wild on the campus of the Cincinnati State Technical and Community College (39.1506°N, 84.5371°W) and also in and around Burnet Woods (39.1430°N, 84.5234°W), a forested urban city park located directly north of the University of Cincinnati campus. Individual plants of E. fortunei, as well as related winged burning bush [Euonymus alatus (Thunb.) Siebold], were growing near yards, along buildings, or growing wild in parks nearby (Figure 1). For each plant, one to two leaves were sampled to determine the maternal genotype along with 10 to 20 mature fruits to quantify progeny genotypes. After transport to the laboratory, fruit from each maternal parent (referred to as a “maternal family” hereafter) was frozen together in a 15-ml Falcon^® tube (Corning Inc., Corning, NY) at −4 C. In summer 2024, all fruits from 12 maternal plants were opened to extract the enclosed seeds. After the seeds were counted, the aril was removed from each seed, and the seed was carefully dissected, with the embryo(s) removed from the surrounding endosperm and seed coat. Each embryo was then placed in an 1.5-ml Eppendorf tube for DNA extraction; at this point, it was observed that sometimes more than one embryo was found inside a seed. Initially we extracted embryos from all collected seeds from each of six maternal parents. For practicality, we subsequently reduced that number to 15 embryos from each of six additional maternal parents. We separated individual embryos, recording their status as either a single embryo or multiple embryos in one seed (typically two but sometimes up to four). In cases of multiple embryos, there was typically one large embryo, with remaining embryos approximately one-third the size, and these were numbered consecutively by decreasing size as they were removed from the seed. In some cases, two seeds shared the same locule and only had a thin wall of the aril separating them, or there were two or more conjoined embryos in one seed (analyzed separately). Forceps, X-Acto^® blades (ULINE, Pleasant Prairie, WI), and the cutting surface used for seed dissection were regularly rinsed with a 70% EtOH solution to prevent contamination and remove any remaining residue, such as aril tissue or endosperm.

Figure 1. Geographic location of 12 maternal plants of Euonymus fortunei from which fruits were collected within the Clifton neighborhood in Cincinnati. A group of six plants in Rawson Woods are clustered together along McAlpin Avenue. All sampled maternal plants were genotyped as the ‘Coloratus’ cultivar.

To test for potential interspecific pollen donors, we compared the progeny genotypes with those of 14 cultivars of E. fortunei, spreading euonymus [Euonymus kiautschovicus Loes.], and everygreen spindle [Euonymus japonicus L.f.], used in Elam and Culley (2023). We also collected fresh leaves from local wild E. alatus (N = 8) and eastern wahoo [Euonymus atropurpureus Jacq.] (N = 2), and from multiple commercially available cultivars of E. alatus (‘Compactus’, ‘Fireball’, ‘Little Moses’, ‘Rudy Haag’, and ‘Unforgettable Fire’). We also sampled dried leaf tissue from herbarium samples from the Dawes Arboretum for spindle tree [Euonymus phellomanus Loes.] (DAWES Herbarium voucher #1128), American strawberry bush [Euonymus americanus L.] (#1715, #1853), spindletree [Euonymus bungeanus Rupr.] (#4501, #4571), European spindle [Euonymus europaeus L.] (#8356, #292, #288), Hamilton’s spindletree [Euonymus hamiltonianus Wall.] (#307, #308, #1126, #2184, #5776), warted spindle tree [Euonymus verrucosus Scop.] (#4320), and E. atropurpureus (#3484, #4585, #5072, #9143).

The DNA of leaves and embryonic tissue was then extracted using a modified CTAB method (Doyle and Doyle 1987) on either fresh tissue or tissue stored at −4 C overnight (tests revealed no difference with either method). Leaf tissue was ground with ceramic mortars and pestles, but the much smaller embryonic tissue was ground individually in 1.5-ml Eppendorf tubes with plastic mini-pestles (USA Scientific, Ocala, FL). DNA was amplified through PCR for each sample in a 10-μl multiplexed reaction, consisting of 5 μl of GreenTaq Multiplexing Mix (Invitrogen, Waltham, MA), 1 μl of Primer Mix (consisting of 2 μM reverse primer and 2 μM forward-labeled primer for each primer), 3.8 μl of H₂O, and 0.4 μl of DNA. Seven sets of microsatellite primers from Mori et al (2017) were used (ef01, ef03, ef05, ef06, ef07, ef08, and ef09), with samples amplified on an Applied Biosystems SimpliAmp thermal cycler under the following conditions: initial denaturation of 95 C for 15 min; followed by 35 cycles each of 94 C for 30 s, 57 C for 90 s, and 72 C for 60 s; and a final extension of 60 C for 30 min. PCR products were sent to Cornell University’s Life Sciences Core Laboratory Center for fragment analysis on a 3730-xl sequencer (Applied Biosystems, Fortune City, CA) using the LIZ 500 internal size standard.

The resulting fragment data were analyzed using GeneMarker v. 1.85 (SoftGenetics, State College, PA) to identify codominant alleles for each of the seven loci. Maternal parents in this study were confirmed to match the Coloratus genotype (see Results and Discussion) and thus are tetraploid (Elam and Culley 2023), so it was not possible to use standard mating system software programs for diploid individuals. Instead, alleles for each locus were recorded from peaks on the GeneMarker electropherogram and used to construct the multilocus genotype (e.g., ABC) of each embryo. More than one copy of an allele in this tetraploid species (e.g., AABC, ABBC, ABCC) could not be reliably identified, but the same peak pattern was always observed for embryos with the same genotype, indicating that allelic dosage was identical. Allelic presence and absence data were analyzed with GenAlEx v. 6.503 (Peakall and Smouse 2012) to visually compare the match between the multilocus genotype of each maternal parent and its offspring using an Excel spreadsheet (Microsoft, Redmond, WA). We determined whether each embryo was (1) identical to the maternal multilocus genotype (consistent with apomixis), (2) largely matched the maternal genotype but occasionally missing one or more alleles for one or more loci (indicative of selfing of heterozygous loci), or (3) contained one or more novel alleles not found in the maternal parent (indicating an outcrossing event) (see Figure 2). Samples not matching the maternal multilocus genotype typically underwent PCR a second and sometimes a third time to confirm that the loss of an allele was not due to allelic dropout or to confirm the identity of a novel allele. From these data, we calculated the rate of selfing, outcrossing, and apomixis within each maternal family, and then across families. As such, these estimates are only approximations of values within this species, and outcrossing in particular may be underestimated if a separate pollen donor shares the same alleles as the maternal parent (e.g., if both maternal and paternal parents have the Coloratus genotype, which would resemble a self-fertilization event). In addition, selfing could theoretically result in a heterozygous genotype (i.e., identical to the maternal parent) and thus appear as evidence for apomixis for a single locus, but loss of a maternal allele was more likely to be detected across multiple independently assorting loci.

Figure 2. Diagram of how embryos were genotyped for an individual locus (ef05 in this case), compared with the Euonymus fortunei maternal parent, which always exhibited the ‘Coloratus’ genotype. Apomictic embryos were genetically identical to the maternal parent, while selfed embryos exhibited a missing band in any given location (indicated by a circle). Outcrossed embryos always contained a unique allele (168 bp in this case, indicated in black) not present in the maternal parent. The polyploid locus genotypes (in fragments in base pairs) are shown at the bottom, and the size of each band is provided on the left side of the figure.

If embryos were designated as resulting from outcross fertilization, any novel allele not found in the maternal parent was then compared with the genotypes of cultivars of E. fortunei, E. kiautschovicus, and E. japonicus (Elam and Culley 2023), as well as E. alatus and other species of Euonymus (T. Culley unpublished data) to identity the most likely pollen donor. Finally, the genetic composition of embryos was also examined by size to determine whether singleton seeds or the largest seed (in the case of multiple embryos) was more likely to be sexually fertilized, with smaller embryos being produced via apomixis, as suggested in the literature (e.g. Viana et al. 2021).

Results and Discussion

Euonymus fortunei is confirmed as a polyembryonic species, with multiple embryos found in over a third (35.3%) of all individual seeds (Figure 3) sampled from plants growing in southwestern Ohio. Of the 278 seeds from the 12 families (see Supplementary Material), 174 (62.6%) contained only a single embryo, 91 seeds (32.7%) contained two embryos, 12 (4.3%) had three embryos, and only 1 seed contained four embryos. All maternal families contained at least two or more seeds with double embryos. Polyembryony has previously been reported in some fruit and grass species (Xue et al. 1997) as well as E. fortunei (Rounsaville et al. 2018), in which 24.4% of seeds contained multiple embryos, a conservative estimate and lower than the value reported here. Confirmation of polyembryony is consistent with other Euonymus species in which multiple embryos have been detected. For example, Brizicky (1964) reported that polyembryony in E. americanus was frequent, with up to five embryos found in a single seed. Other polyembryonic species include E. alatus, broad-leaved spindle tree [E. latifolius L.(Mill.)], and E. verrucosoides (Brizicky 1964). In contrast, some Euonymus species are monoembryonate (e.g., E. bungeanus, winterberry[E. maackii], and E. hamiltonianus); E. atropurpureus may only produce a single embryo per seed (Brizicky 1964), and further investigation is necessary. There is also some discrepancy as to whether E. europaeus is polyembryonic or not (see Brizicky 1964).

Figure 3. Extracted embryos from a single seed of Euonymus fortunei, sorted by size. The largest embryo is designated as “First,” while the subsequent embryo is “Second.” Bar is shown for scale.

Based on the parentage analysis of maternal sibships using microsatellite data, E. fortunei is sexually facultative, producing multiple embryos asexually through apomixis as well as sexually through outcrossing with occasional self-fertilization, sometimes within the same seed. All wild maternal plants were genetically identical and matched the Coloratus genotype as described by Elam and Culley (2023). In the 12 maternal families in which 382 embryos were extracted from individual seeds, the percentage of apomictic progeny that were an exact match to the maternal genotype was 50.0% (191 embryos; Figure 4A), ranging from 18.8% to 86.7% across families (Figure 4B; Supplementary Material). The remaining embryos resulted either from outcross fertilization (131 embryos, 34.3%) or selfing (60 embryos, 15.7%). Family 10 did not contain any outcrossed offspring, and Family 12 did not produce any self-fertilized offspring (Figure 4B). The type of apomixis in E. fortunei still remains to be determined, but the physical arrangement of multiple embryos within a seed is consistent with sporophytic apomixis (i.e., adventitious embryony, as described by Brizicky [1964]).

Figure 4. Distribution of reproductive origin of embryos sampled from seeds obtained from 12 maternal plants of wild Euonymus fortunei for (A) all embryos grouped together and (B) separated by maternal family. Shown are the percentages of embryos resulting from apomixis (green), outcross fertilization (orange), and self-fertilization (maroon).

The overall high frequency (50%) of apomictic embryos containing the Coloratus genotype in seeds of E. fortunei is consistent with this genotype comprising wild populations (Elam and Culley 2023). Most apomictic embryos were detected in polyembryonic seeds, with the percentage of apomixis increasing with decreasing embryo size (Table 1; Figure 5). For example, the first (and largest) embryos averaged 52% apomixis, increasing to 85% for the smaller, third embryos. In contrast, most embryos within monoembryonate seeds were sexually derived and usually outcrossed, with only 36% of embryos originating from apomixis (Table 1). These results are expected with the model of sporophytic apomixis (Viana et al. 2021) in which the largest embryo in a polyembryonate seed is typically produced through double fertilization with pollen from a different plant, while other smaller embryos in the same seed originate from apomixis. In polyembryonate seeds in E. fortunei, it is still unknown which embryo is most likely to germinate, although it is generally assumed that the largest embryo will be most successful. However, Rounsaville et al. (2018) reported that more than one seedling can germinate from a single seed of E. fortunei and survive. Apomixis is often associated with polyploidy (Hand and Koltunow 2014; Viana et al. 2021; Voigt-Zielinski et al. 2012), which matches with the occurrence of apomictic embryos in seeds of tetraploid Coloratus; it would also be informative to ascertain whether diploid cultivars of E. fortunei can produce seeds through apomixis as well.

Table 1. Estimated frequency of reproductive mode (asexual vs. sexual) and mating system (selfing vs. outcrossing) in seeds of Euonymus fortunei when only a single embryo was present (“Single”); or in the case of multiple embryos per seed, the order of the embryos according to size, with the largest embryo labeled as “First,” the second largest as “Second,” etc. ^a .

^a Embryos extracted from seeds obtained from 12 maternal families were used in this analysis (382 embryos).

Figure 5. Distribution of reproductive origin of embryos, whether as a single embryo in one seed (Single) or in the case of multiple embryos, by their size (largest to smallest, indicated as “First” to “Fourth”). The percentage of embryos resulting from apomixis (green), outcross fertilization (orange), and self-fertilization (maroon) is shown. The number of embryos sampled in each category are indicated above each bar.

What remains perplexing is the fate of sexually derived embryos in seeds of E. fortunei (especially outcrossed embryos), given the absence of these genotypes in previously sampled wild populations (Elam and Culley 2023). In the current study, sexually produced single embryos were identified in 12 maternal families (52.3% outcrossed, 12.1% selfed; Table 1). Embryos defined as outcrossed ubiquitously contained only one unique allele not found in the maternal genotype: allele 168 at locus ef05. This allele is only found in the ornamental cultivar of E. alatus (Compactus, identical to Little Moses and Fireball; T Culley, unpublished data), all tetraploid cultivars. While an allele 169 was detected in E. atropurpureus, it was distinct from allele 168; no other tested Euonymus species shared any similar alleles with the embryos and thus are excluded as potential pollen donors. The presence of allele 168 at locus ef05 in the embryos is consistent with meiotic drive, a condition in some polyploids in which certain alleles are disproportionately transmitted during meiosis (Fishman and McIntosh 2019). Interestingly, there were two other loci (efo7 and ef08) for which the E. alatus Compactus genotype completely differed from that of E. fortunei, yet the presumed hybrid embryos did not contain any other E. alatus alleles. This may not be unusual, given the complex patterns of inheritance within polyploids, especially polysomic inheritance in hybrids between related species (such as E. fortunei and E. alatus). Additional genetic investigation is now needed to confirm such patterns of inheritance.

This study is the first evidence of potential interspecific hybridization between E. fortunei and E. alatus. Although both species are largely geographically separated from one another in their native ranges (POWO 2025), they may have not undergone complete reproductive isolation before being reintroduced into the same areas in North America through the ornamental plant trade. These two species differ in growth habit as a vine and a shrub, but they share very similar flowering times and flower morphology, both are adaptable to a wide variety of environments and often occur in the same area in the introduced environment, and both are tetraploid (Ibáñez et al. 2009; Salihu et al. 1999). There is little information on pollinators of either species, but flies and ants seem to be the main floral visitors (Elam and Culley 2023), attracted to the green-, white-, and cream-colored flowers (Reverté et al. 2016). With prezygotic barriers removed, it is possible that these two Euonymus species may hybridize in their overlapping introduced range, as indicated by the genetic results of this study.

There are several reasons why the outcross genotype observed here in embryos may be absent within wild adult populations of E. fortunei (Elam and Culley 2023). First, it is possible that hybrid adults are just very rare and have yet to be detected in the wild. Another possibility is that postzygotic processes may be occurring, such as hybrid sterility or underdominance (i.e., embryos with allele 168 may be inviable and/or incapable of germinating successfully) and/or linkage drag (Wang et al. 2020). In this case, hybrid seed inviability may be caused by disrupted endosperm development in the seed (Coughlan 2023); this can be tested by growing embryos with allele 168 on a nutritive medium, effectively replacing the endosperm in a process called embryo recovery (Sandstedt and Sweigart 2022). It is unlikely that programmed cell death eliminates all but one embryo in E. fortunei (as in Pinus sylvestris [Filonova et al. 2002]), because visibly healthy and intact embryos can be readily extracted from seed just before germination. In fact, Rousanville et al. (2018) noted that polyembryonic seeds could germinate with two or more radicles and that the resulting seedlings exhibited normal growth and survival; however, the genetic identity of these embryos was unknown (i.e., whether they resulted from apomixis or self- or outcross fertilization). Thus, the most likely explanation for the absence of outcrossed adult individuals in previously sampled wild populations (as reported in Elam and Culley [2023]) is vegetative propagation combined with seeds primarily produced through apomixis following abortion of any recombinant embryos, as first suggested by Rounsaville et al. (2018). This effectively perpetuates the Coloratus clone within existing wild populations and in new satellite populations founded by dispersed seed.

To further refine and confirm the reproductive mode and mating system results of this genetic study, future investigations are being planned to incorporate hand-pollination of bagged and emasculated flowers to confirm apomixis and to observe insect pollinators in locations containing both E. fortunei and E. alatus. Outcrossing could also be tested empirically in E. fortunei using different Euonymus species as pollen donors, with genetic analysis of any seeds that are produced. Furthermore, data are needed on the performance of sexually and asexually produced embryos of E. fortunei (i.e., germination rate and survival). These results should be paired with a genetic analysis of the F₁ generation of the maternal plant to confirm whether the offspring were produced through sexual or asexual reproduction. This type of fine-grained analysis is key to better understanding the spread of E. fortunei in North America.

In summary, many factors likely increase the successful colonization of E. fortunei into new areas, such as polyploidy, bird dispersal of seeds, continual growth as an evergreen vine, an ability to propagate vegetatively, and as reported here in Ohio, a relatively high frequency of apomixis in seeds that favors continued persistence of the Coloratus genotype in the wild. However, the species is still able to retain the possibility of sexual production through outcrossing to potentially generate new genetic variation (including crossing to related E. alatus), particularly represented in the largest embryos in its polyembryonic seeds. Given that cultivars of E. fortunei and related species remain popular in the ornamental trade, this will only contribute further to propagule pressure in the environment. Any plan to slow the spread of this species must be multi-pronged, starting with reduction of new clones introduced to the environment, prevention of flowering in wild individuals, and curbing the commercial sale and distribution of cultivars, especially the polyploid Coloratus cultivar.