Introduction
Chinese privet (Ligustrum sinense Lour.), which is native to southeast Asia, has been introduced to every continent except Antarctica (Figure 1) and is considered invasive in at least 20 U.S. states, 6 Pacific islands, Australia, Italy, Argentina, and Puerto Rico (CABI 2018). Ligustrum sinense is particularly problematic in the southeastern United States, where it is often listed as one of the most threatening invasive species in the region (Miller et al. Reference Miller, Chambliss and Bargeron2004). Introduced for landscaping in 1852, L. sinense (along with the very similar European privet [Ligustrum vulgare L.]) was estimated in 2008 to cover more than 1 million hectares in the southeastern United States alone (Maddox et al. Reference Maddox, Byrd and Serviss2010; Miller and Chambliss Reference Miller and Chambliss2008). In Mississippi and Alabama, L. sinense and L. vulgare were estimated to have occupied 0.12 million hectares (<2%) of forests in 2003, but are projected to reach 2.83 million hectares (31%) by 2023 (Wang et al. Reference Wang, Wonkka, Grant and Rogers2016). Ligustrum sinense control is typically labor-intensive and expensive (US$216 to US$1,820 ha−1; Benez-Secancho et al. Reference Benez-Secanho, Grebner, Ezell and Grala2018; Klepac et al. Reference Klepac, Rummer, Hanula and Horn2007), due to its propensity to form dense stands and resprout following cutting. Understanding the potential ecological impacts of L. sinense invasions can help land managers determine whether control measures are warranted. Accordingly, we reviewed known and potential relationships between L. sinense and native vegetation and wildlife communities. We also provide recommendations for future research priorities based on important research gaps identified in our review.
Figure 1. Global range of Ligustrum sinense. Data for the United States, Australia, and China are displayed at the state/province level. All other data are at the country level. Invaded islands have a circular buffer to aid in visibility. Ligustrum sinense is considered “invasive” in only a portion of its nonnative range. Records retrieved from CABI (2018) and EDDMapS (2019).
Ligustrum sinense is an evergreen to deciduous (depending on local climate) single or multistemmed shrub/small tree that reaches a maximum height of 10 m (Maddox et al. Reference Maddox, Byrd and Serviss2010; Miller and Miller Reference Miller and Miller2005). Ligustrum sinense seeds are encased in ovoid drupes and are spread via endozoochory and hydrochory (Foard Reference Foard2014; Miller and Miller Reference Miller and Miller2005). An Australian study found that L. sinense can produce 1,300 fruits m−2 of canopy, with fruit production positively related to light availability and stem diameter (Westoby et. al. Reference Westoby, Dalby and Adams-Acton1983). Flowering occurs during spring and early summer, and ripe fruit is available during fall and winter (McCall and Walck Reference McCall and Walck2014; Miller and Miller Reference Miller and Miller2005). Local dispersal can also occur via root-sprouting (i.e., new shoots can grow from belowground roots; Miller and Miller Reference Miller and Miller2005). Ligustrum sinense has broad environmental tolerances and can survive in areas with full sun or heavy shade (Brown and Pezeshki Reference Brown and Pezeshki2000; Dirr Reference Dirr1998; Grove and Clarkson Reference Grove and Clarkson2005), although research has shown that seedling survival is higher in areas with high afternoon and low morning light (Kuebbing et al. Reference Kuebbing, Classen, Call, Henning and Simberloff2015). It is found in a variety of sites, including bottomland hardwood forests, upland forests, forest edges (including stream and roadsides; Figure 2), swamps, cedar glades, and disturbed sites such as old fields (Cofer et al. Reference Cofer, Walck and Hidayati2008; Grove and Clarkson Reference Grove and Clarkson2005; Hagan et al. Reference Hagan, Mikhailova, Shearman, Ma, Nankaya, Hart, Valdetero, Bridges and Yun2014; Kuhman et al. Reference Kuhman, Pearson and Turner2010; Merriam Reference Merriam2003; Miller and Miller Reference Miller and Miller2005; Pokswinski Reference Pokswinski2009). Transplant experiments and field surveys have shown that L. sinense can regenerate under a canopy of conspecifics, indicating that invasions can likely sustain themselves over long periods of time (Greene and Blossey Reference Greene and Blossey2012; Grove and Clarkson Reference Grove and Clarkson2005; McAlpine et al. Reference McAlpine, Timmins, Jackman and Lamoureaux2018)—although for how long is unknown.
Figure 2. Example of a Ligustrum sinense invasion along a mixed pine–hardwood forest edge in Alabama. Photo taken in March 2018.
Methods
To evaluate the current understanding of relationships between L. sinense and ecosystems outside its native range, we searched the Web of Science database using the search terms “Chinese privet” and “Ligustrum sinense” (last accessed March 21, 2019). All articles pertaining to relationships among L. sinense and flora and fauna outside its native range were reviewed, and their reference lists were searched for additional relevant articles. This search yielded 48 peer-reviewed papers that contained information on the relationship between L. sinense invasions and native flora and fauna. We also cite additional peer-reviewed sources that contain general background information on L. sinense or invasive species, plus sources such as books, government and environmental organization documents, web pages, and dissertations/theses when important information was unavailable in the peer-reviewed literature. We have focused this review on L. sinense, but occasionally reference research on other Ligustrum species and similar exotic shrubs where relevant L. sinense research did not exist for specific subjects. The vast majority of the studies we found were conducted in the eastern United States, so we only identified the location of the study in our discussion if it occurred outside this area. We generally catalogued papers as pertaining to either vegetation or wildlife, although there were several instances in which papers covered both topics. Further, we reviewed and included papers that may have been less directly focused on vegetation or wildlife (e.g., impact of L. sinense on soil properties) that still contributed to our understanding of the potential impact of L. sinense on native plant and animal communities.
We begin our results and discussion section with a review of relevant invasion frameworks and their implications for interpreting trends and relationships between L. sinense and native species. We then present summaries of documented relationships between L. sinense and native vegetation, split into sections on herbaceous and woody species. Following those summaries we explore the possible mechanisms through which L. sinense could drive these relationships. Next we cover relationships between L. sinense and wildlife, split into discussions on invertebrates, mammals, birds, and herpetofauna. Finally, the potential for native community recovery following L. sinense removal is discussed.
Results and Discussion
Invasion Frameworks
Before we present the results of our literature review, it is important to discuss how observed relationships between L. sinense and native species could be interpreted. In cases where L. sinense is found to be correlated with changes in vegetation communities, there may be uncertainty regarding whether L. sinense is a driver, passenger, or backseat driver of these changes. In other words, did L. sinense cause changes in the plant community (driver), take advantage of changes from other sources (passenger), or take advantage of initial changes and then drive further changes (backseat driver; Bauer Reference Bauer2012; Foard Reference Foard2014; MacDougall and Turkington Reference MacDougall and Turkington2005)? For instance, Wilcox and Beck (Reference Wilcox and Beck2007) observed that native plant species richness was lower in areas with high L. sinense cover, but was this because L. sinense excluded native species or because areas with lower species diversity may be easier to invade (i.e., the “diversity resistance hypothesis”; Kennedy et al. Reference Kennedy, Naeem, Howe, Knops, Tilman and Reich2002)? In another example, Hagan et al. (Reference Hagan, Mikhailova, Shearman, Ma, Nankaya, Hart, Valdetero, Bridges and Yun2014) found that L. sinense was more prevalent in areas with lower overstory basal area, but again is this because L. sinense was more likely to invade sites with those conditions or because L. sinense caused those conditions to develop? Sites with low overstory basal area (possibly due to natural or anthropogenic disturbance) may have more available resources (e.g., sunlight) for L. sinense to take advantage of, increasing the likelihood of it successfully invading (Hagan et al. Reference Hagan, Mikhailova, Shearman, Ma, Nankaya, Hart, Valdetero, Bridges and Yun2014). On the other hand, if L. sinense outcompetes tree species and leads to the reduction of overstory basal area through reduced regeneration or growth, then L. sinense is acting as a driver. There is empirical evidence that, once established, L. sinense can decrease survival of herbaceous and woody plants. Greene and Blossey (Reference Greene and Blossey2012) carried out a transplant experiment that involved planting boxelder (Acer negundo L.), false nettle [Boehmeria cylindrica (L.) Sw.], blunt broom sedge (Carex tribuloides Wahlenb.), and Indian woodoats [Chasmanthium latifolium (Michx.) Yates] seedlings in an area dominated by L. sinense (95% ± 4.2 cover, little herbaceous understory) and an adjacent area with no L. sinense and a herbaceous understory community. At the conclusion of the 63-wk trial, each species, except for B. cylindrica, had significantly lower survival rates in the L. sinense–dominated area. This suggests that L. sinense can drive declines in native plant species (Greene and Blossey Reference Greene and Blossey2012). However, because L. sinense often takes advantage of vegetational or hydrological disturbance during initial invasions, it may be best classified as a backseat driver (Foard Reference Foard2014). It is important to keep the driver–passenger–backseat driver models in mind when interpreting results from observational studies. Although evidence shows that L. sinense can act as a driver/backseat driver, that may not always be the case and the possibility that L. sinense is a passenger in a given example of ecological change should be considered by managers planning restoration actions.
Vegetation
Herbaceous Species
It is common to see very sparse native ground cover under a dense L. sinense midstory (Figure 3). Studies have confirmed that herbaceous species diversity and stem density tend to be lower in areas with high (~100%) monospecific cover of L. sinense compared with those with little or no L. sinense cover (Kittel Reference Kittel2001; Merriam and Feil Reference Merriam and Feil2002). Further, studies that surveyed a range of L. sinense cover classes have documented negative correlations between L. sinense cover and herbaceous species richness, cover, and stem density (Greene and Blossey Reference Greene and Blossey2012; Wilcox and Beck Reference Wilcox and Beck2007). For example, Greene and Blossey (Reference Greene and Blossey2012) found that as L. sinense cover increased along a gradient from 0% to 60%, herbaceous cover decreased from 58% to 25%, herbaceous species richness decreased from 10 to 5 species m−2, and herbaceous species stem density decreased from 225 to 100 stems m−2. Additionally, L. sinense has been noted as a potential contributor to the decline of several plant species of conservation concern, including the green pitcherplant [Sarracenia oreophila (Kearney) Wherry; Schnell et al. Reference Schnell, Catling, Folkerts, Frost and Gardner2000], fringed campion [Silene polypetala (Walter) Fernald & B.G. Schub.; Allison Reference Allison1996], and Schweinitz’s sunflower (Helianthus schweinitzii Torr. & A. Gray; Urbatsch Reference Urbatsch2000; Weakley and Houk Reference Weakley and Houk1994).
Figure 3. Example of the understory of a bottomland hardwood forest in Alabama invaded by Ligustrum sinense. The midstory of this site is dominated by L. sinense, and the understory is comprised primarily of leaf litter and coarse woody debris, with very few native herbaceous or woody plants present. Photo courtesy of Jimmy Stiles (Auburn University), taken June 14, 2019.
Woody Species
Similar to herbaceous species, the density of understory and midstory woody plants, including shrubs, saplings, and seedlings, is often lower in invaded areas compared with uninvaded areas and has been shown to be negatively correlated with measures of L. sinense prevalence (Barksdale and Anderson Reference Barksdale and Anderson2015; Hanula et al. Reference Hanula, Horn and Taylor2009; Hart and Holmes Reference Hart and Holmes2013; Kittel Reference Kittel2001; Loewenstein and Loewenstein Reference Loewenstein and Loewenstein2005; Merriam and Feil Reference Merriam and Feil2002). Specifically, Wilcox and Beck (Reference Wilcox and Beck2007) found that native shrub density in plots with high (90.6 ± 4.6%) L. sinense cover was about one-third that of plots with low (0.4 ± 0.6%) or medium (13.6 ± 1.5%) L. sinense cover. Negative relationships have also been observed between L. sinense and woody species diversity (Burton et al. Reference Burton, Samuelson and Pan2005; Foard et al. Reference Foard, Burnette, Burge and Marsico2016; Hanula et al. Reference Hanula, Horn and Taylor2009; Hart and Holmes Reference Hart and Holmes2013; Kittel Reference Kittel2001; Loewenstein and Loewenstein Reference Loewenstein and Loewenstein2005; Merriam and Feil Reference Merriam and Feil2002; Wilcox and Beck Reference Wilcox and Beck2007).
Given the relationship between L. sinense and relatively low density, diversity, and richness of woody seedlings and saplings, and the potential for L. sinense to be a causative factor in this relationship, multiple authors have expressed concern that invaded forests could convert to L. sinense–dominated shrublands over time due to inadequate woody species regeneration (e.g., Greene and Blossey Reference Greene and Blossey2012; Hart and Holmes Reference Hart and Holmes2013; Loewenstein and Loewenstein Reference Loewenstein and Loewenstein2005; Merriam and Feil Reference Merriam and Feil2002). Several studies have detected a negative relationship between L. sinense cover and tree density, possibly supporting this hypothesis (Barksdale and Anderson Reference Barksdale and Anderson2015; Hagan et al. Reference Hagan, Mikhailova, Shearman, Ma, Nankaya, Hart, Valdetero, Bridges and Yun2014; Hanula et al. Reference Hanula, Horn and Taylor2009; Wilcox and Beck Reference Wilcox and Beck2007). Additionally, Hagan et al. (Reference Hagan, Mikhailova, Shearman, Ma, Nankaya, Hart, Valdetero, Bridges and Yun2014) reported that tree basal area was lower in invaded plots (but recall the earlier discussion on invasion frameworks). However, Hanula et al. (Reference Hanula, Horn and Taylor2009) and Greene and Blossey (Reference Greene and Blossey2014) found no significant relationship between tree basal area and L. sinense cover, and others have found no significant relationship between L. sinense and overstory and subcanopy tree diversity (Kittel Reference Kittel2001; Wilcox and Beck Reference Wilcox and Beck2007). Examples of large-scale forest-to-L. sinense shrubland conversion are lacking in the literature, perhaps because not enough time has elapsed since the introduction of L. sinense or the length of studies has been insufficient to observe such a change. Nonetheless, Hart and Holmes (Reference Hart and Holmes2013) reported that L. sinense can occasionally occupy canopy-dominant positions, which may represent localized areas of tree-to-shrub conversion in their study area.
In addition to potential effects on woody regeneration, L. sinense invasion may also impact forest stands by influencing the growth and survival of mature trees. For example, Foard et al. (Reference Foard, Burnette, Burge and Marsico2016) observed that mature Quercus spp. in an invaded stand had higher rates of self-thinning and slower growth compared with an uninvaded stand (although their sample size was small). In contrast, Brantley (Reference Brantley2008) found no detectible relationship between L. sinense and overstory tree growth, and Hudson et al. (Reference Hudson, Hanula and Horn2014) observed no response in tree growth 5 yr after L. sinense removal. The mixed reports regarding the relationship between tree basal area and L. sinense presence (Greene and Blossey Reference Greene and Blossey2014; Hagan et al. Reference Hagan, Mikhailova, Shearman, Ma, Nankaya, Hart, Valdetero, Bridges and Yun2014; Hanula et al. Reference Hanula, Horn and Taylor2009) further complicate our understanding of how L. sinense may or may not affect mature trees. More research is needed to determine the potential effect of L. sinense on overstory tree health, stand regeneration, and woody community composition. Specifically, replicating Foard et al. Reference Foard, Burnette, Burge and Marsico2016—which looked at tree growth before and after L. sinense invasion based on dendrochronology—with a larger and more spatially diverse data set would improve our understanding of how the presence of L. sinense could affect mature tree species, with implications for both natural area and plantation management. Additional transplant experiments, similar to Greene and Blossey (Reference Greene and Blossey2012), with more tested species and a greater number of replicates would also help clarify the extent to which L. sinense is driving reduced woody regeneration.
Mechanisms of Impact
As previously discussed, it may not always be the case that L. sinense is driving observed relationships between its presence and diminished or altered native plant communities. However, evidence suggests that L. sinense can function as a driver/backseat driver (Foard Reference Foard2014; Greene and Blossey Reference Greene and Blossey2012), and so it is worth considering the mechanisms through which it may do so. Better understanding these mechanisms may help guide efforts to mitigate them.
Competition for light is an important mechanism through which nonnative plants influence native species (Gioria and Osborne Reference Gioria and Osborne2014) and seems the most obvious mechanism by which L. sinense would affect native plants (Greene and Blossey Reference Greene and Blossey2012). Decreased light availability tends to promote shade-tolerant species over shade-intolerant species (e.g., Lin et al. Reference Lin, Harcombe, Fulton and Hall2002), potentially restructuring plant communities. For example, Osland et al. (Reference Osland, Pahl and Richardson2009) documented roughly four times greater light availability in areas where privet had been removed compared with areas where it was still present. However, Brantley (Reference Brantley2008) and Pokswinski (Reference Pokswinski2009) did not detect a significant relationship between L. sinense cover and light intensity. These conflicting results could be due to differences in methodology or the overstory structure of the study sites. An L. sinense midstory may not have a significant effect on understory light availability in closed canopy forests; however, competition for light may play a role when L. sinense is competing against native species in canopy gaps and along forest edges. Further research into the role that light availability plays in L. sinense invasions could help guide management and restoration actions.
In addition to light competition, L. sinense may affect native plant communities by altering soil nutrient dynamics and fungal communities. Ligustrum sinense could affect soil nutrient cycles by altering litter decomposition rates, litter chemical composition, and leaf abscission timing (Mitchell et al. Reference Mitchell, Lockaby and Brantley2011). Mitchell et al. (Reference Mitchell, Lockaby and Brantley2011) found that litter turnover decreased from 7.1 to 2.6 yr as the percent of L. sinense in the leaf litter increased from 0% to 50% (with the remaining litter composed of an equal mix of sweetgum [Liquidambar styraciflua L.], yellow poplar [Liriodendron tulipifera L.], water oak [Quercus nigra L.], and elm [Ulmus spp.] litter). The greater decomposition rate was likely due to the greater nitrogen and lower lignin content, lower C:N ratios, and lower lignin:N ratios of L. sinense litter (Mitchell et al. Reference Mitchell, Lockaby and Brantley2011). Alteration of soil nutrient cycles and nutrient availability can change competitive interactions among plant species (e.g., Aerts and Berendse Reference Aerts and Berendse1988), but this has not been directly studied in relation to L. sinense. Changes in soil microbial and fungal communities have been suggested as another mechanism through which L. sinense influences native and nonnative plant communities (Greipsson and DiTommaso Reference Greipsson and DiTommaso2006; Kuebbing et al. Reference Kuebbing, Classen and Simberloff2014, Reference Kuebbing, Classen, Call, Henning and Simberloff2015, Reference Kuebbing, Patterson, Classen and Simberloff2016), and some of these changes are likely associated and interact with changes in soil nutrient dynamics (Deyn et al. Reference Deyn, Raaijmakers and Van der Putten2004; Ehrenfeld Reference Ehrenfeld2003), which in turn are likely related to changes in litter decomposition. The complexity of the biogeochemical process as it relates to these changes makes it difficult to parse out simple cause–effect relationships. Research by Greipsson and DiTommaso (Reference Greipsson and DiTommaso2006) found that L. sinense can alter arbuscular mycorrhizal fungi (AMF) communities in the soil. Specifically, the mycorrhizal infectivity potential of the soil, as measured by AMF root colonization on a bait plant, was higher in soil from invaded sites compared with noninvaded sites (Greipsson and DiTommaso Reference Greipsson and DiTommaso2006). Changes in AMF communities could affect competitive interactions among native and nonnative species, but was not directly studied (Greipsson and DiTommaso Reference Greipsson and DiTommaso2006). Other research has shown that native species tend to perform better (as measured by shoot and root mass) in soils without a history of invasion by L. sinense, and altered AMF or other soil properties were suggested as a cause but not directly tested (Kuebbing et al. Reference Kuebbing, Patterson, Classen and Simberloff2016). Research by Kuebbing et al. (Reference Kuebbing, Classen and Simberloff2014, Reference Kuebbing, Classen, Call, Henning and Simberloff2015, Reference Kuebbing, Patterson, Classen and Simberloff2016) also points to nonadditive effects on soils when L. sinense and other nonnative plants are present simultaneously, which suggests that studies on L. sinense in isolation may not grasp its full potential impact in mixed native/nonnative plant communities. Additionally, soils conditioned by L. sinense alone or in combination with Amur honeysuckle [Lonicera maackii (Rupr.) Herder] may promote continued or further invasion by those species or other nonnative species (Kuebbing et al. Reference Kuebbing, Classen and Simberloff2014, Reference Kuebbing, Classen, Call, Henning and Simberloff2015, Reference Kuebbing, Patterson, Classen and Simberloff2016), which possibly indicates some form of “invasion meltdown” (Simberloff and Von Holle Reference Simberloff and Von Holle1999). The role played by altered soil nutrients, fungal communities, or other properties in L. sinense invasion deserves further research, particularly as it relates restoration actions. Identifying the specific soil alterations involved, how they may affect native plants and nonnative plants, and how long they last following L. sinense removal may help guide restoration actions or at least guide expectations of the short-term results of restoration actions (Greipsson and DiTommaso Reference Greipsson and DiTommaso2006; Kuebbing et al. Reference Kuebbing, Classen, Call, Henning and Simberloff2015, Reference Kuebbing, Patterson, Classen and Simberloff2016).
A related way that invasive plants may gain a competitive advantage over native plants is through the deployment of “novel weapons,” allelopathic or antimicrobial biochemicals that native species have no evolved resistance to (Callaway and Ridenour Reference Callaway and Ridenour2004). There is limited evidence to suggest L. sinense may affect native plant communities via allelopathy. Several have evaluated the allelopathic potential of L. sinense by watering the seeds of multiple plant species with diluted L. sinense extracts. These extracts had negative effects on the germination and growth of radish (Raphanus sativus L.) and tomato (Solanum lycopersicum L.) seedlings (Grove and Clarkson Reference Grove and Clarkson2005; Pokswinski Reference Pokswinski2009). Barnett et al. (Reference Barnett, Hudack and Dick2016) conducted a similar experiment with species native to the southeastern United States, including common persimmon (Diospyros virginiana L.), red mulberry (Morus rubra L.), soapberry (Sapindus saponaria L.), and American beautyberry (Callicarpa americana L.). They found negative effects of L. sinense extract on S. saponaria and C. americana germination rates and C. americana root growth. These experiments demonstrate the potential for L. sinense allelopathy, but without identification of the specific chemical(s) involved and evidence that such chemical(s) have a significant effect under field conditions, it is not possible to definitively say that L. sinense exhibits allelopathy (Pokswinski Reference Pokswinski2009).
Another mechanism through which L. sinense could impact native plant communities is by altering fuel loads and types in fire-dependent ecosystems. Specifically, L. sinense may reduce fine fuels at ground level by outcompeting herbaceous species and reducing litter accumulation (Faulkner et al. Reference Faulkner, Clebsch and Sanders1989; Mitchell et al. Reference Mitchell, Lockaby and Brantley2011; Stocker and Hupp Reference Stocker, Hupp, Zouhar, Smith, Sutherland and Brooks2008), although Hagan et al. (Reference Hagan, Mikhailova, Shearman, Ma, Nankaya, Hart, Valdetero, Bridges and Yun2014) found increased litter accumulation in invaded sites—which may have been due to the prevalence of L. sinense in flatter microsites in their study area, where litter accumulation was more likely. Dense midstory vegetation also tends to increase fine fuel moisture, further reducing fire frequency and intensity (Nowacki and Abrams Reference Nowacki and Abrams2008). Additionally, L. sinense has low ignitability (Tiller Reference Tiller2015). The potential for reduced understory fuel loads and increased fuel moisture under an L. sinense canopy may suppress fires under some conditions. Fire suppression by L. sinense could alter disturbance regimes and successional processes, but this possibility has not been studied. However, the affinity that L. sinense shows for moist bottomland areas may conflate observations of the ignitability of L. sinense–invaded areas (Batcher Reference Batcher2000; Faulkner et al. Reference Faulkner, Clebsch and Sanders1989). Some land managers have experienced success using prescribed fire to control L. sinense, particularly under dry conditions (Batcher Reference Batcher2000), which indicates that the risk of L. sinense suppressing fires in drier fire-dependent uplands may not be significant. There is even some concern that as the climate becomes warmer and drier, L. sinense could become more ignitable and function as a ladder fuel, resulting in increased incidence of crown fires (Wang et al. Reference Wang, Wonkka, Grant and Rogers2016), although this is a highly speculative prediction.
Finally, invasion by nonnative plants can also affect native plant communities by altering wildlife–plant interactions (Traveset and Richardson Reference Traveset and Richardson2006). For instance, nonnative plants can compete with native plants for pollinators or seed dispersers (Traveset and Richardson Reference Traveset and Richardson2006). An example of such an interaction occurs between the nonnative Himalayan balsam (Impatiens glandulifera Royle) and the native marsh woundwort (Stachys palustris L.), which experiences reduced pollinator visits and seed set when I. glandulifera is present (Chittka and Schürkens Reference Chittka and Schürkens2001). This mechanism has not been directly studied for L. sinense; however, it is a possibility. There is research that suggests L. sinense negatively affects pollinators due, in part, to reductions in the herbaceous understory (Hanula and Horn Reference Hanula and Horn2011a, Reference Hanula and Horn2011b; Hudson et al. Reference Hudson, Hanula and Horn2013), and it is possible that a reinforcing loop exists whereby reductions in pollinator communities further limit pollination potential for remaining flowering species.
Wildlife
Wildlife conservation is a priority for many public and private landowners, so understanding how L. sinense may affect wildlife is important for informing land management decisions. Wildlife diversity is often positively correlated with vegetation structural diversity, likely because structural diversity opens up more ecological niches (Tews et al. Reference Tews, Brose, Grimm, Tielbörger, Wichmann, Schwager and Jeltsch2004). Based on this relationship, it is possible that L. sinense at low to moderate densities could benefit some wildlife communities by providing additional cover and food without effecting native plant communities to a great extent. Some species may also benefit from the conditions created by high-density invasions. Indeed, some landowners may perceive L. sinense as beneficial for wildlife and thus may be reluctant to initiate control measures (Howle et al. Reference Howle, Straka and Nespeca2010). However, our literature review found that L. sinense tends to be correlated with, and is likely a driver of, reduced native plant diversity and density (e.g., Greene and Blossey Reference Greene and Blossey2012). Given the common relationship between wildlife diversity and vegetation structural diversity (Tews et al. Reference Tews, Brose, Grimm, Tielbörger, Wichmann, Schwager and Jeltsch2004), we predict that wildlife communities could be negatively affected by L. sinense invasion (especially high-density invasions), although the effect at the species level would be dependent on the species’ ecological niche. Unfortunately data on relationships between wildlife and L. sinense are sparse, and most of what is available is observational. We present the findings of these studies organized into subsections on invertebrates, mammals, birds, and herpetofauna.
Throughout these sections, we use somewhat arbitrary distinctions between the terms “low density” and “high density” in order to standardize our discussion of potential impacts, even though these terms may not be specifically used by the papers we cite. When we discuss specific studies, we note the L. sinense cover/density of the study plots when applicable. We use the term “low density” to describe situations in which scattered L. sinense individuals are present, with assumed minor appreciable effect on native vegetation. “High density” refers to situations where L. sinense is the dominant understory or midstory species and cover of native species may be limited. Ligustrum sinense individuals in high-density stands tend to have limited foliage at ground level and thus provide limited cover at ground level (Figure 3). It is important to take into consideration that every high-density invasion started off at a lower density, so even if a given invasion does not have obvious impacts now, it may develop into a problem down the road.
Invertebrates
An incredibly diverse group, invertebrates represent the majority of animal life on Earth and play important ecosystem roles as herbivores, predators, prey, decomposers, and so on (New and Yen Reference New and Yen1995). Given this diversity, it is impossible to make broad generalizations regarding how L. sinense invasion may affect these taxa. Even so, there have been a few studies that have directly examined the impact of L. sinense on invertebrate groups. Kuebbing et al. (Reference Kuebbing, Classen and Simberloff2014), for example, did not detect significant differences in ground-dwelling arthropod communities between plots invaded by L. sinense (>75% L. sinense cover) and control plots. However, researchers found that plots with little to no L. sinense present (either naturally or following L. sinense control operations) generally had greater species richness and abundance of native bees and butterflies than invaded plots, and most of these parameters were negatively correlated with L. sinense cover (Hanula and Horn Reference Hanula and Horn2011a, Reference Hanula and Horn2011b; Hudson et al. Reference Hudson, Hanula and Horn2013). In a related study, beetle communities near ground level had greater species richness in plots where L. sinense had been removed compared with invaded plots; however, beetle species richness 5 and 15 m from the ground did not differ significantly between treatments (Ulyshen et al. Reference Ulyshen, Horn and Hanula2010). There was also a high proportion of an exotic beetle (Xylosandrus crassiusculus Motschulsky) at 5 m in plots that still had L. sinense (Ulyshen et al. Reference Ulyshen, Horn and Hanula2010). In the same study area, overall earthworm abundance tended to be lower in plots that were naturally free (or nearly so) of L. sinense, but the proportion of native earthworm species was greater in plots where L. sinense was absent (naturally or following control) than in invaded plots (Lobe et al. Reference Lobe, Callaham, Hendrix and Hanula2014).
There are various mechanisms through which Ligustrum sinense invasion could elicit changes in invertebrate communities. For example, a reduction in native herbaceous species under an L. sinense canopy could limit the availability of native nectar-producing plants and those suitable for hosting pollinator larvae (Hanula and Horn Reference Hanula and Horn2011a). Changes to vegetation structure likely also have an impact. For instance, deer ticks (Ixodes scapularis Say) frequently use young L. sinense stems when questing for hosts (Goddard Reference Goddard1992). However, it seems likely that at high L. sinense densities, where herbaceous ground cover is reduced and the L. sinense canopy may be ≥5 m, platforms for questing may be limited. Either way, these findings could have implications for disease transmission among vertebrate species.
The chemical makeup of L. sinense may also affect invertebrate communities. For example, both the invasive gypsy moth (Lymantria dispar L.) and native lace bug (Leptoypha mutica Say) had relatively low performance when feeding on L. sinense compared with plants native to the southeastern United States (Kalina et al. Reference Kalina, Braman and Hanula2017; McEwan et al. Reference McEwan, Rieske and Arthur2009). A similar species, border privet (Ligustrum obtusifolium Siebold & Zucc.), produces a chemical (oleuropein) in its leaves that inhibits absorption of nutrients by invertebrate herbivores (Konno et al. Reference Konno, Hirayama, Shinbo and Nakamura2009), and it is conceivable that L. sinense has similar traits. This could explain why some have observed decreased invertebrate herbivory on L. sinense compared with native plants and may contribute to its competitive advantage (Greene and Blossey Reference Greene and Blossey2012; Morris et al. Reference Morris, Walck and Hidayati2002). Similarly, L. sinense leaves shed into water sources could negatively impact aquatic invertebrates by changing the water chemistry (Llewellyn Reference Llewellyn2005). Experimental evidence has shown that leaf extracts from L. sinense can reduce survival of at least one aquatic invertebrate (Anisops sp.) in a laboratory setting; however, more research is necessary to determine whether such effects occur in the wild (Llewellyn Reference Llewellyn2005). Ligustrum sinense may also alter the chemical properties of the soil that it grows in, such as increasing pH (Lobe et al. Reference Lobe, Callaham, Hendrix and Hanula2014). An increase in pH was considered a likely reason for the differences in earthworm communities observed by Lobe et al. (Reference Lobe, Callaham, Hendrix and Hanula2014), because one of the nonnative earthworm species in their study prefers the relatively higher pH soils found under L. sinense (Lobe et al. Reference Lobe, Callaham, Hendrix and Hanula2014). Hagan et al. (Reference Hagan, Mikhailova, Shearman, Ma, Nankaya, Hart, Valdetero, Bridges and Yun2014) also found higher soil pH in invaded plots (1.5% to 85% cover, mean 30.4%), but Kuebbing et al. (Reference Kuebbing, Classen and Simberloff2014) did not detect a significant difference. Hagan et al. (Reference Hagan, Mikhailova, Shearman, Ma, Nankaya, Hart, Valdetero, Bridges and Yun2014) acknowledged that their study could not determine whether L. sinense caused the difference in pH, preferred areas with higher pH, or if the difference was due to L. sinense preferring flatter microsites with different soil nutrient dynamics. The finding by Lobe et al. (Reference Lobe, Callaham, Hendrix and Hanula2014) that pH was lower in plots where L. sinense was removed via hand felling suggests that L. sinense was a driver of the pH differences, rather than a passenger (although pre-removal pH measurements were not conducted, and there were some differences in soil type across sites).
Mammals
Low- to moderate-density L. sinense invasions are probably beneficial for some mammalian species. For example, white-tailed deer (Odocoileus virginianus Zimmermann) and American beaver (Castor canadensis Kuhl) show moderate selective preference for L. sinense browse (Rossell et al. Reference Rossell, Arico, Clarke, Horton, Ward and Patch2014; Stromayer et al. Reference Stromayer, Warren, Johnson, Hale, Rogers and Tucker1998), and studies in the United States and New Zealand show that small mammals, such as the white-footed mouse (Peromyscus leucopus Raf ; O’Malley et al. Reference O’Malley, Blesh, Williams and Barrett2003) and common brushtail possum (Trichosurus vulpecula Kerr; Williams et al. Reference Williams, Karl, Bannister and Lee2000), will eat the fruits. Overall, L. sinense is a suitable forage species for O. virginianus, as its browse contains >12% crude protein during winter (Stromayer et al. Reference Stromayer, Warren, Johnson, Hale, Rogers and Tucker1998), which meets the requirements of all age classes during this season (NRC 2007). Browsing on L. sinense is particularly heavy when acorns are limited (Stromayer et al. Reference Stromayer, Warren, Johnson, Hale, Rogers and Tucker1998). However, at high densities, L. sinense may compete with native plants and reduce food availability for O. virginianus during spring and summer (Stromayer et al. Reference Stromayer, Warren, Johnson, Hale, Rogers and Tucker1998).The potential for competition with herbaceous vegetation is particularly concerning, given that forbs are the most nutritious and preferred plants of O. virginianus (Warren and Hurst Reference Warren and Hurst1981). An interesting potential avenue for future research is the possibility that L. sinense provides significant levels of available browse in areas that might not typically have such resources available for O. virginianus, particularly during winter, allowing O. virginianus to maintain high population levels that in turn put too much browsing pressure on more palatable native understory species that may be present (Stromayer et al. Reference Stromayer, Warren, Johnson, Hale, Rogers and Tucker1998). Over the longer term, the potential reduction of oak (Quercus spp.) regeneration by L. sinense is concerning given the dependence of not only O. virginianus but a wide array of wildlife species on acorns, even if L. sinense berries and browse are available as alternatives.
In terms of providing cover, the literature on this is very sparse, but the effect of L. sinense invasion is likely variable depending on the density and spatial extent of the invasion and the habitat requirements of different mammalian species. Christopher and Barrett (Reference Christopher and Barrett2006) found that two similar rodent species (P. leucopus and golden mouse [Ochrotomys nuttalli Harlan]) coexisted in areas dominated by L. sinense in the understory shrub layer, and it is possible that the increased vertical structure provided by L. sinense helped facilitate this coexistence despite the relative lack of cover at ground level. High rodent capture rates in L. sinense patches were also reported by Kittell (Reference Kittel2001). When L. sinense forms hedges along forest edges (Figure 2) or is present in forest interiors at moderate densities, it may provide thermal and escape cover for O. virginianus and other mammalian species, but this has not been studied. Despite the handful of documented or potential ways in which L. sinense may be used as cover by mammals, it also has the potential to reduce habitat structure and cover. As L. sinense forms dense, mature stands, the amount of cover at ground level tends to be lower. Further, others have expressed concern regarding the potential loss of bat roosting habitat from reduced tree regeneration caused by L. sinense (Pallin Reference Pallin2000). The potential for significant vegetation structural changes in areas invaded by L. sinense may thus alter the mammal communities that are able to take cover in those areas; however, further research is needed to clarify what these changes may look like in different regions.
Birds
Invasion by L. sinense has variable effects on bird species, depending on their life history and habitat requirements. For example, the small ovoid drupes of L. sinense are persistent throughout winter and early spring (Greenberg and Walter Reference Greenberg and Walter2010) and may serve as a food source for frugivorous birds during this period of relative scarcity (Lochmiller Reference Lochmiller1978 [L. vulgare]; McCall and Walck Reference McCall and Walck2014; Miller and Miller Reference Miller and Miller2005; Wilcox and Beck Reference Wilcox and Beck2007). Specifically, researchers have documented significant winter use of Ligustrum spp. fruit by northern bobwhite (Colinus virginianus L.; McRae Reference McRae1980), the dusky-legged guan in Argentina (Penelope obscura Temminck; Merler et al. Reference Merler, Diuk-Wasser and Quintana2001), and hermit thrush (Catharus guttatus Pallas; Strong et al. Reference Strong, Brown and Stouffer2005). Preference for L. sinense is species specific: P. obscura appeared to select for L. sinense disproportionate to its availability at some sites (Merler et al. Reference Merler, Diuk-Wasser and Quintana2001), whereas C. guttatus consume L. sinense slightly less often than expected given availability (Strong et al. Reference Strong, Brown and Stouffer2005). Ligustrum sinense fruit was reported to contain 63.66% (±1.10) moisture, 1.91% (±0.06) crude protein, and 7.02% (±3.41) crude fat (McCall and Walck Reference McCall and Walck2014).
Despite the documented use of L. sinense as a winter food source, potential population- and community-level effects on bird species are relatively unknown. It is possible that species reliant on herbaceous groundcover or overstory trees for foraging, nesting, or cover could be negatively impacted by high-density invasions of L. sinense because of its possible effects on those vegetation strata. The potential effects on species that utilize understory and midstory woody plants are likely variable, depending on the needs of the species. For example, differences in the height and branching structure of invasive plants versus the natives they replace can impact bird-nesting success (e.g., Schmidt and Whelan Reference Schmidt and Whelan1999). However, the relative quality of L. sinense as a nesting substrate has not been studied.
We identified only one study that used an observational design with multiple replicates to compare bird use of sites with a range of L. sinense cover. Wilcox and Beck (Reference Wilcox and Beck2007) found significant positive correlations between L. sinense cover and bird species richness and abundance during winter, further supporting the hypothesis that birds will use L. sinense for cover and/or food. The majority of birds observed in high-density L. sinense areas (90.6 ± 4.6% cover) were primarily insectivorous, not frugivorous, possibly indicating that they were using the L. sinense stands for cover or foraging for insects that preferred that habitat or were more easily hunted in that habitat. Wilcox and Beck (Reference Wilcox and Beck2007) also documented differential use patterns of L. sinense–invaded areas that were related to their ecological niches, suggesting that L. sinense invasion could alter local bird community composition. Additionally, Wilcox and Beck (Reference Wilcox and Beck2007) observed that singing (total of all species) during summer tended to be less common in plots with high-density L. sinense, suggesting that mate attraction is more difficult in high-density areas or that high-density areas are not preferred for breeding. McCall and Walck (Reference McCall and Walck2014) reported that birds commonly used L. sinense for cover or perching platforms; however, L. sinense was “nearly ubiquitous” in their study area, so whether individuals select for or against L. sinense requires further study. McCall and Walck (Reference McCall and Walck2014) did not detect nesting or roosting in L. sinense. Woodcock (Scolopax rusticola L.) are known to use areas invaded by L. sinense during fall migration and as winter habitat (Miller and Miller Reference Miller and Miller2005; Myatt and Krementz Reference Myatt and Krementz2007), although whether this is a preferred vegetation type is unclear.
The additional food that L. sinense provides in winter could have complex effects on global ecological patterns by enticing migrating birds to overwinter in areas that they would not otherwise, although such effects are speculative to date. Researchers have suggested that Ligustrum spp. may provide such an incentive for species in southern Argentina that would normally overwinter farther north (Merler et al. Reference Merler, Diuk-Wasser and Quintana2001; Montaldo Reference Montaldo1993). In North America it is possible that supplementary feeding stations (i.e., bird feeders) could affect the timing of migrations (Robb et al. Reference Robb, McDonald, Chamberlain and Bearhop2008), and if this is the case, L. sinense may have complementary or additive effects. However, L. sinense fruits might ripen too late to be much of an enticement for most fall migrants (McCall and Walck Reference McCall and Walck2014). Species such as American robin (Turdus migratorius L.) are known to feed on L. sinense fruits on their northbound trip in spring (Miller and Miller Reference Miller and Miller2005), but it seems unlikely that this would alter spring migration patterns. The effects of additional winter food may also interact with or be swamped out by the effects of climate change on migrations (e.g., Jenni and Kéry Reference Jenni and Kéry2003; Zaifman et al. Reference Zaifman, Shan, Ay and Jimenez2017).
Herpetofauna
The effects of invasive plants on herpetofauna (i.e., reptiles and amphibians) are generally understudied, and there have been no studies, to our knowledge, on the effects of L. sinense specifically. Martin and Murray (Reference Martin and Murray2011) reviewed the limited available literature on invasive plant effects on herpetofauna and developed a general predictive framework of possible changes to herpetofaunal habitat quality, food availability, and reproductive success. They predicted that invasive plants that are structurally different from the native plant assemblage will have the greatest effects and that small-bodied herpetofauna with small home ranges will be most affected (either positively or negatively), because they will have greater difficulty moving to uninvaded areas. The general structure of L. sinense is not necessarily unique to the systems it commonly invades, such as bottomland hardwood forests; however, it seems to reach much greater densities across larger areas than native understory and midstory species.
Many herpetofaunal species are “heliothermic,” meaning they bask in sunlight to aid in thermoregulation (Bogert Reference Bogert1959). Some species, such as freshwater turtles, also choose nest sites with the necessary sun exposure to maintain proper nest temperatures (Bodie et al. Reference Bodie, Smith and Burke1996). The published data are not clear on the degree to which L. sinense invasion reduces light levels at ground level (e.g., Brantley Reference Brantley2008; Osland et al. Reference Osland, Pahl and Richardson2009; Pokswinski Reference Pokswinski2009), but it seems likely that basking and nest-site availability would be affected by L. sinense cover. The first step in exploring how L. sinense could affect herpetofaunal thermoregulation is to better quantify light-level differences among sites with varying levels of L. sinense and native vegetation cover.
In addition to thermoregulation, herpetofauna need cover to aid in moisture retention and predator avoidance. Common cover types include leaf litter, coarse woody debris (CWD), burrows, and herbaceous vegetation. Leaf litter, a particularly important cover type for salamanders, may be reduced by L. sinense invasion due to its ability to intercept leaves from the canopy layer (Faulkner et al. Reference Faulkner, Clebsch and Sanders1989), increase decomposition rates (Mitchell et al. Reference Mitchell, Lockaby and Brantley2011), and possibly suppress overstory regeneration (e.g., Hart and Holmes Reference Hart and Holmes2013). However, Hagan et al. (Reference Hagan, Mikhailova, Shearman, Ma, Nankaya, Hart, Valdetero, Bridges and Yun2014) found that litter depth was positively correlated with L. sinense invasion, possibly because leaf litter tends to accumulate in the flatter microsites that L. sinense occupied in their study area. Limited evidence also shows that CWD cover may be greater in areas with L. sinense, possibly due to increased self-thinning by resource-constrained trees (Foard et al. Reference Foard, Burnette, Burge and Marsico2016). However, this increase in CWD cover may be temporary if overstory tree regeneration is impeded. Potential herbaceous ground cover reductions associated with L. sinense (e.g., Wilcox and Beck Reference Wilcox and Beck2007) could also limit thermal and escape cover for herpetofauna.
Potential for Restoration
Assuming that L. sinense plays a role as a driver/backseat driver of vegetation community change, it is important to determine whether removing L. sinense stands leads to the recovery of plant communities (and thus wildlife habitat). Merriam and Feil (Reference Merriam and Feil2002) took an early pass at this question by removing L. sinense from an invaded area and found that herbaceous species richness, herbaceous stem density, and woody stem density increased the first year following L. sinense control. Hanula et al. (Reference Hanula, Horn and Taylor2009) studied the effects of L. sinense removal more extensively in a study that involved removing L. sinense in 2-ha plots using mulching and hand felling in combination with cut-stump herbicide treatments (2005) and a follow-up foliar herbicide treatment (2006). They then compared the plant communities among control plots (L. sinense present, no treatment), treatment plots (mulching or hand felling), and “desired future condition” (DFC) plots that had little to no L. sinense cover. In 2007, they found that herbaceous cover and diversity in treated plots was higher than in control plots and similar to DFC plots (except for cover in hand-felled plots, which were intermediate between controls and mulched/DFC). The treated, control, and DFC plots developed distinct community compositions by 2007 due to the proliferation of early successional species in the treatment plots. Hudson et al. (Reference Hudson, Hanula and Horn2014) returned to those same plots in 2012 and found that herbaceous cover and diversity were still greater in treatment plots than in control plots and similar to DFC plots. Community composition was still distinct among the control, treatment, and DFC plots, with treatment plots having more early successional species such as pokeweed (Phytolacca americana L.). For woody species, Hudson et al. (Reference Hudson, Hanula and Horn2014) found that L. sinense–removal plots did not have statistically significant differences in shrub/sapling cover or species richness or diversity compared with control plots, although there was a nearly significant difference in cover. Woody seedlings and small saplings appear to have been included in their “herbaceous” plant category, so it is difficult to interpret differences in woody regeneration from their results—although they did state that “woody saplings covered more of the removal plots in 2012 than 2007” (Hudson et al. Reference Hudson, Hanula and Horn2014).
How a site responds to L. sinense control efforts likely depends on site-specific factors such as soil type, climate, overstory canopy closure, disturbance regime, and proximity to native plant propagule sources. More research is needed across a range of site conditions and across longer time periods to better understand how natural systems respond to L. sinense control. Even so, based on the limited research to date and a general understanding of vegetation successional dynamics, it is reasonable to predict that when dense L. sinense midstory thickets are controlled, more resources (e.g., sunlight, water, nutrients) will become available for other species. The species that take advantage of these newly available resources will vary depending on site conditions, particularly in terms of light availability at ground level. In situations where removal of L. sinense substantially increases light availability, it is likely that early successional herbaceous species will initially dominate. In closed canopy forests, light availability may not increase substantially following L. sinense removal, and the species that take its place will likely be relatively shade-tolerant species more characteristic of later successional stages. It is also important to remember that L. sinense may be a backseat driver in many situations, meaning that some underlying disturbance to the natural system allowed L. sinense to invade (Foard Reference Foard2014). Addressing this disturbance may be necessary for long-term control of L. sinense and other invasive species and/or to promote recovery of native species (Bauer Reference Bauer2012; Foard Reference Foard2014; MacDougall and Turkington Reference MacDougall and Turkington2005).
Conclusions and Recommendations
Our literature review revealed many negative correlations between native plant communities and L. sinense invasion, although more research is needed to confirm the degree to which L. sinense is driving declines and the mechanisms involved. Reduced understory cover and diversity are perhaps the most visible today, but long-term reductions in overstory regeneration could dramatically alter future landscapes. Research on the effects on wildlife are limited, but changes to vegetation species composition and structure could affect wildlife in a variety of ways (some positive, but many negative). Low-density L. sinense invasions likely have limited negative consequences for plant species and may have positive effects on some wildlife species (due to increased structural complexity and food resources). However, low-density invasions may develop into high-density invasions, and the negative impacts of L. sinense invasion (i.e., suppression of native plants and the predicted reductions in wildlife habitat quality) tend to be most apparent at higher L. sinense densities.
Low-density L. sinense invasions may be more cost-effective to control than high-density invasions, depending on the method used (Benez-Secancho et al. Reference Benez-Secanho, Grebner, Ezell and Grala2018). For this reason, we recommend taking early action to control L. sinense when possible to prevent further invasion and increasing control costs. Whenever L. sinense control operations are implemented, particularly in high-density invasions, the response of native vegetation should be monitored and reported (when possible) to improve our understanding of the best practices for restoring invaded areas. Additionally, managers whose objectives include increasing coverage of shade-intolerant woody or herbaceous species should be aware that L. sinense control alone may not be sufficient if sunlight is also limited by the overstory canopy or if there are underlying environmental changes that led to the invasion. Maddox et al. (Reference Maddox, Byrd and Serviss2010) and Urbatsch (Reference Urbatsch2000) provide recommendations on control options, and many state Extension agencies also provide guidance (e.g., Enloe and Loewenstein Reference Enloe and Loewenstein2018).
A substantial amount of evidence regarding the ecological relationships between L. sinense and native species has accumulated over the past several decades, but we identified several critical knowledge gaps worthy of future study. There is a particular need for additional research on the effects of L. sinense invasion on wildlife species. With the exception of a few studies on invertebrates, most research has simply documented use patterns of invaded areas. Instead, we recommend future research focus on how L. sinense invasion affects population dynamics and community composition. The effects of L. sinense on wildlife are likely dependent on the scale of the L. sinense invasion, which interacts with various wildlife species in different ways depending on their daily movements and home-range sizes. As such, we recommend future studies consider the appropriate scale for the wildlife taxon in question. There is also a need for additional research on the effects of L. sinense on overstory trees. Although many studies provide evidence that L. sinense may negatively impact the woody regeneration layer, more research is needed to determine the spatial and temporal scales at which conversions from forest to shrubland may occur. Direct negative effects of L. sinense on mature trees are also important to understand, as they could speed forest conversion and have implications for landowners managing for timber production or wildlife habitat.
Future research into the mechanisms by which L. sinense may affect native plants could help managers mitigate those effects, particularly for legacy effects that may persist after L. sinense is removed from a site. More generally, further research into restoration options is needed so that land managers have the best tools at their disposal for managing L. sinense and have realistic expectations for how forest plant communities will respond. Due to the dynamic response of vegetation communities to various management actions, researchers and managers should also report on covariates such as canopy coverage, site type, past management history, and so on to better inform managers concerning the expected response of vegetation communities to L. sinense control across a variety of sites.
Acknowledgments
Research assistantship funding for JSC was provided by the Harry Murphy Dean’s Enhancement Fund for Excellence in the School of Forestry and Wildlife Sciences at Auburn University. No conflicts of interest have been declared.
