Management Implications
Elaeagnus pungens (thorny olive) is an aggressive invasive shrub in North America that can take over natural areas and impact the ecology of the local environment. Material and labor costs associated with controlling this shrub can be high, and land managers need effective methods to reduce or eliminate E. pungens growth. Management of E. pungens can be uniquely challenging for several reasons. The thorns of the felled plant material or sprawling branches can puncture tires of equipment used in management activities, and cause injury to people working in these areas. The thick growth of particularly aggressive stands of E. pungens can result in a damp, shady forest floor and a lack of adequate fuel, precluding prescribed burning treatments. We evaluated three herbicide application methods using triclopyr as the free acid formulation (Trycera®): foliar, basal bark, and cut stump, in central South Carolina, USA, to determine the fastest, easiest management approach with maximum control efficacy. We prioritized treatment methods that could be applied with limited personnel (two to three applicators) and without motorized vehicles. We evaluated these factors to help land managers more effectively use their resources, limit the introduction of herbicides into the environment, and determine a reliably effective method to kill E. pungens plants. Basal bark and cut stump treatments were much more effective in reducing E. pungens biomass in our plots than the foliar treatment. Between basal bark and cut stump treatment methods, we recommend basal bark due to its relative ease of application compared with the cut stump application method and equal efficacy in terms of controlling E. pungens. This approach should be used as part of an integrated management strategy, and further studies should evaluate other management tactics to reduce seedling germination success and control the vigorous resprouting associated with E. pungens growth.
Introduction
An expansion in global trade and travel has increased the number of invasive species impacting natural and managed systems worldwide (Essl et al. Reference Essl, Latombe, Lenzner, Pagad, Seebens, Smith, Wilson and Genovesi2020; Liebhold Reference Liebhold2012; Seebens et al. Reference Seebens, Essl and Blasius2017) and led to a drastic rise in global costs associated with invasive species management (Crystal-Ornelas et al. Reference Crystal-Ornelas, Hudgins, Cuthbert, Haubrock, Fantle-Lepczyk, Angulo, Kramer, Ballesteros-Mejia, Leroy, Leung, López-López, Diagne and Courchamp2021). In the United States alone, invasive species costs ranged between US$18.2 billion and US$78.9 billion between 1970 and 2020 (Diagne et al. Reference Diagne, Turbelin, Moodley, Novoa, Leroy, Angulo, Adamjy, Dia, Taheri, Tambo, Dobigny and Courchamp2021; Zenni et al. Reference Zenni, Essl, García-Berthou and McDermott2021). Additionally, natural and managed systems can be negatively impacted, as native species are often outcompeted by unchecked invasive species populations, which can lead to a reduction in biodiversity and the alteration of entire trophic cascades (Beschta and Ripple Reference Beschta and Ripple2009; Kimbro et al. Reference Kimbro, Grosholz, Baukus, Nesbitt, Travis, Attoe and Coleman-Hulbert2009; Schmitz et al. Reference Schmitz, Hamback and Beckerman2000). Economic and ecological impacts associated with introduction of invasive species will only continue to escalate over time without significant regulatory intervention.
The southeastern United States, with its long growing season and warm, humid climate, is an ecologically diverse region with a rich forestry history, economy, and production potential (Carter et al. Reference Carter, Allen, Fox, Albaugh, Rubilar, Campoe and Cook2021; Napton et al. Reference Napton, Auch, Headley and Taylor2010). However, forests in this region are also impacted by many invasive plant species (Oswalt et al. Reference Oswalt, Fei, Guo, Iannone, Oswalt, Pijanowski and Potter2015). Invasive plants are adept at establishing and flourishing in areas where management activities (e.g., logging, clearing, burning, fire suppression, and reforestation) and the indirect effects of climate change disrupt forest ecosystems (Holmes et al. Reference Holmes, Aukema, Von Holle, Liebhold and Sills2009). Several invasive shrub species in the region are known to outcompete native flora and dominate the forest understory (Maynard-Bean et al. Reference Maynard-Bean, Kaye, Wagner and Burkhart2020); these species include Chinese privet (Ligustrum sinense Lour.), bush honeysuckles (Lonicera spp.); and silverberry or olives (Elaeagnus spp.).
Elaeagnus pungens (Thunb.), colloquially known as thorny olive, silverthorn, thorny elaeagnus, spiny oleaster, or silverberry, is a broadleaved evergreen shrub native to Japan and China (Figure 1). It was introduced as an ornamental species in 1830 and later promoted for wildlife (Davison Reference Davison1942) and now occurs throughout the southeastern (and in parts of the northeastern) United States (Miller Reference Miller2006). Elaeagnus pungens is a multistemmed, freely branched, dense shrub that can reach 7.5-m tall and 4.5-m wide; it can use other species to support branch growth and climb opportunistically (Serviss and Tumlison Reference Serviss and Tumlison2021; Figure 1A and 1F). Once established, this shrub produces copious, fast-growing branch “whips” that allow the plant to quickly increase in size and overtake neighboring vegetation (Figure 1A). Additionally, root suckering and prolific stem sprouts can lead to dense understory growth in a forested setting (Miller Reference Miller2006). The stems produce thorns, and the leaves are tough and waxy with a dark green surface and an underside that is silvery with brown scales (Figure 1B, C, and E). In the southeastern United States, E. pungens benefits from an especially lengthy growing season that allows it to outcompete native plants (Riffe Reference Riffe2018). Specifically, E. pungens flowers and fruits for approximately 10 mo out of the year (Dirr Reference Dirr1990; Miller Reference Miller2006). The fruit—fleshy red drupes—is readily consumed by birds and other animals that disperse it across large areas (Davison Reference Davison1942), allowing E. pungens to quickly spread across the landscape (Figure 1C and 1D). Elaeagnus pungens readily grows in a variety of environmental conditions (open sun, forested settings, frequently flooded areas, disturbed sites, etc.) and is often a problem in fencerows, roadside margins, waste areas, and open woodlands.
Figure 1. Characteristics of Elaeagnus pungens in Calhoun County, South Carolina, USA: (A) dense, sprawling growth; (B) leaf surface is dark green and waxy; (C) leaf undersides are silver and reflective; (D) fruit is red drupes; (E) thorns 2.5–5 cm in length grow on branches; (F) growth is multistemmed and freely branched. Photos A, D, E, and F by MND; photos B and C by DRC.
Despite demonstrated deleterious ecological impacts, Elaeagnus spp. are still commonly sold in nurseries and online and are cultivated for hedges, screening, natural barriers, bank stabilization along highways, and landscape uses (Beaury et al. Reference Beaury, Patrick and Bradley2021; Fertakos et al. Reference Fertakos, Beaury, Ford, Kinlock, Adams and Bradley2023). Elaeagnus spp. have been shown to reduce the abundance of native species, facilitate the establishment of other non-native plants, and cause long-lasting impacts on local soil characteristics and flora (Collette and Pither Reference Collette and Pither2015; Katz et al. Reference Katz, Tuttle, Denslow and Norton2020). Their presence impacts soil microbial communities (Malinich et al. Reference Malinich, Lynn-Bell and Kourtev2017) and alters stream biogeochemical cycles (Mineau et al. Reference Mineau, Baxter and Marcarelli2011). Elaeagnus spp. can be aggressive invaders in forests (Yates et al. Reference Yates, Levia and Williams2004) and negatively impact native tree seedling and sapling abundance (Lázaro-Lobo et al. Reference Lázaro-Lobo, Lucardi, Ramirez-Reyes and Ervin2021). Even though birds feed on Elaeagnus fruit (Davison Reference Davison1942), one study showed a decrease in cavity-nesting birds in areas invaded by Elaeagnus, presumably due to a lack of overstory trees (Fischer et al. Reference Fischer, Valente, Guilfoyle, Kaller, Jackson and Ratti2012). Further, planting Elaeagnus in highway medians led to an increase in bird mortality by automobile strikes (Watts and Paxton Reference Watts and Paxton2000).
In his book Manual of Woody Landscape Plants, Dirr (Reference Dirr1990) described E. pungens in its natural form as “a genuine horror” and went on to say, “Fast does not adequately describe the speed at which it grows.” Because of E. pungens’ prolific and hardy nature, employing a long-term, ecosystem-wide strategy prioritizing prevention, active monitoring, and prompt eradication would likely be more successful than site-specific, local control measures. However, ecosystem-wide strategies are not always feasible or possible. Several local control methods have been evaluated with varying efficacy. Prescribed fire can be an effective management tool to reduce seed viability and kill young Elaeagnus plants (Muscha et al. Reference Muscha, Vermeire and Angerer2023), but established plants will typically resprout following a fire (Michielsen et al. Reference Michielsen, Szemák, Fenesi, Nijs and Ruprecht2017) or cutting (Corns and Schraa Reference Corns and Schraa1965).
Herbicides are a common, established, and often effective management tactic for invasive plants (Pile Knapp et al. Reference Pile Knapp, Coyle, Dey, Fraser, Hutchinson, Jenkins, Kern, Knapp, Maddox, Pinchot and Wang2023), and formulations with the active ingredient triclopyr are known for their efficacy on woody plant species (e.g., Bovey Reference Bovey1965; Bovey and Whisenant Reference Bovey and Whisenant1991; DiAllesandro Reference DiAllesandro2012; Enloe et al. Reference Enloe, O’Sullivan, Loewenstein, Brantley and Lauer2016, Reference Enloe, Leary, Lastinger and Lauer2023), including Elaeagnus (Edgin and Ebinger Reference Edgin and Ebinger2001). Although most research on Elaeagnus management has occurred in the western United States and focused on Elaeagnus control in the context of rangeland management, a combination of cutting and stump treatment with triclopyr was shown to be effective for Elaeagnus management in former coal mines in the Appalachian region of the United States (Franke et al. Reference Franke, Zipper and Barney2018). However, relatively few studies have examined best management practices for forested lands in the southeastern U.S. We conducted informal interviews with land managers in the southeastern U.S. to determine the most common practices used to control woody understory growth in this region, whether motorized vehicles were used, and the number of personnel most often available, along with the operators’ assessment of the tools at their disposal. We determined that many land managers lacked information that incorporated time, effort, and cost of application with treatment efficacy when controlling woody understory growth. Our objective was to determine an effective chemical management method for a dense E. pungens infestation in a southern hardwood-dominated forest. We prioritized methods that would be easily accepted and utilized by land managers. Using the active ingredient triclopyr in the free acid formulation, we tested foliar spray, basal bark, and cut stump application methods with the goal of quantifying local Elaeagnus control and ease of application.
Materials and Methods
Site Description
Our study was conducted on a 214-ha (529-acre) forested tract in Calhoun County, South Carolina, USA (33.63684, -80.70592). More than 26 ha (65 acres) of the understory in this area is dominated by E. pungens (voucher specimens are deposited in the Clemson University Herbarium, accessible as Molly Darr #1 [CLEMS0083037, CLEMS0083038, CLEMS0083039], Molly Darr #2 [CLEMS0083040, CLEMS0083041, CLEMS0083042, CLEMS0083043], and Molly Darr #3 [CLEMS0083044, CLEMS0083045]). The region’s humid subcontinental climate has long, warm summers and mild winters with a mean maximum temperature from 2018 to 2023 of 37.3 C (99.2 F), a mean minimum temperature of 7.8 C (46.0 F), and an average annual precipitation of 110 cm (43 in.) (National Oceanic and Atmospheric Administration: =https://www.weather.gov/wrh/Climate?wfo=cae). Soils in these areas are classified as “southern Coastal Plain” and consist of Faceville fine sandy loam and Ailey-Vaucluse complex soil series (USDA-NRCS 2019; Table 1). The study site consisted of a hardwood-dominated riparian bottomland stand and an adjacent upland old-field sweetgum (Liquidambar styraciflua L.) stand. The overstory of the bottomland area was composed primarily of sweetgum, with winged elm (Ulmus alata Michx.), loblolly pine (Pinus taeda L.), white oak (Quercus alba L.), and ash (Fraxinus spp.) interspersed. The overstory of the upland area was a mixture of white oak, southern red oak (Quercus falcata Michx.), loblolly pine, American beech (Fagus grandifolia Ehrh.), hickory (Carya spp.), and sweetgum (Table 1). Treatment plots were assigned in a randomized complete block design and grouped into three 0.4-ha (1-acre) blocks, each at a separate geographic location on the study site (Zar Reference Zar2010; Table 1). Each block contained sixteen 1-m2 quadrats, and each quadrat within a block was randomly assigned one of three treatments (foliar spray, basal bark, and cut stump) or an untreated control (UTC). All 4 treatments had 4 replicates per block; therefore, each treatment had 12 replicates across the entire experiment.
Table 1. Site characteristics for each experimental block in the Elaeagnus pungens management study conducted in Calhoun County, South Carolina, USA.
Pretreatment Measurements
Pretreatment measurements were conducted to obtain a baseline measurement of Elaeagnus abundance and biomass to confirm that no preexisting differences occurred among treatment assignments. On March 11, 2020, we measured the total number of E. pungens plants, along with the basal circumference (cm) of every E. pungens plant within each 1-m2 quadrat (16 quadrats per block, 48 total quadrats across 3 blocks). Basal stem circumference was measured at 15 cm (6 in.) aboveground level. Ideally, pretreatment measurements would have used direct measurements of biomass, as was performed for posttreatment measurements (see “Posttreatment measurements” section below). However, a true measurement of biomass requires destructive sampling, which could not be performed before treatment applications for this experiment. Instead, because basal stem diameter has a strong relationship with total aboveground biomass for many woody shrubs and trees (Reeves and Lenhart Reference Reeves and Lenhart1988; Telfer Reference Telfer1969) we employed a nondestructive sampling method to estimate pretreatment quadrat biomass by using basal stem circumference as an indicator of biomass. To verify that basal stem circumference was an accurate indicator of biomass for a species with such variable branching, we destructively sampled E. pungens plants outside our study locations to test this relationship. We cut down representative samples of all the circumference size classes we had collected in the baseline measurements (3 to 50 cm). Basal stem circumference of 45 plants was measured at 15 cm (6 in.) aboveground level, with 15 plants collected outside each study location (block). After the basal stem circumference of each individual plant was recorded, plants were cut, placed in paper bags, and transported to the laboratory, where they were dried at 65 C to a constant weight. We calculated the relationship between stem circumference and biomass using a power function regression analysis and observed a strong relationship between basal stem circumference and biomass (y = 1.173x 3.1079; R2 = 0.813), suggesting that basal stem circumference is an adequate indicator of biomass for pretreatment quadrat comparisons.
Herbicide Treatments
We applied triclopyr herbicide using the free acid product formulation (Trycera®, 343.92 g ai L−1, Helena Agri-Enterprises, Collierville, TN, USA) as foliar spray, basal bark, and cut stump application methods between January and September 2021 (Table 2). We used E. pungens’ annual cycle of growth initiation, timing of flowering (October to December), and fruiting (March to July) to determine the time of year that is most effective for each treatment in this environment (Dirr Reference Dirr1990; Ferrell et al. Reference Ferrell, Langeland and Sellers2019; Miller Reference Miller2006; Miller et al. Reference Miller, Manning and Enloe2013).
Table 2. Field application details for experimental treatments: foliar spray, cut stump, and basal bark.
a Formulated product: Trycera®, 344 g ai triclopyr L−1.
The foliar spray treatment was applied when E. pungens was actively growing new leaves, but the plant was not yet flowering. We used a 15-L backpack pump sprayer to apply the solution (Table 2) directly to the leaves and stems and attempted to spray to runoff. The basal bark application was made during fall, and the herbicide solution (25% Trycera®; Table 2) was applied to the lower 30 to 40 cm (12 to 16 in.) of every E. pungens stem using a 15-L backpack pump sprayer. For cut stump applications, stems were cut with hand clippers or a chain saw about 15 cm (6 in.) above ground level and treated directly per label directions. We used a 15-L backpack pump sprayer to apply the solution (100% Trycera®; Table 2) directly to the cut stump, covering the entire wood surface. No action was taken in the untreated control plots.
Posttreatment measurements
On May 16, 2022, the total biomass of surviving E. pungens plants was destructively sampled in each treatment plot. These posttreatment biomass measurements were taken the season following treatment application, 12, 8, and 16 mo after the foliar spray, basal bark, and cut stump treatments, respectively. Because treatments were administered at different times of year (as is necessary for each treatment to be effective), the time window between application and sampling was not equal among treatments (Ferrell et al. Reference Ferrell, Langeland and Sellers2019; Miller et al. Reference Miller, Manning and Enloe2013). We ensured that plants in each treatment had a full overwintering cycle to respond to herbicide application and that the time allotted was adequate to show the desired effect (Shaner Reference Shaner2014). We prioritized sampling all treatments on the same day to produce measurements that are not biased by seasonality.
On the day of sampling, all living E. pungens plants within each quadrat were clipped 15 cm (6 in.) above ground level, including live foliage and woody material. Dead E. pungens material was left in the field. Living plant material was identified by retention of leaves and woody tissue with living cambium. Dead material was colorless, brittle, and leafless. The freshly clipped E. pungens stems were bagged and transported to the Clemson University Forestry and Environmental Conservation shop room (Clemson, SC, USA) the following day, where the drying process began immediately. Once all bags were dried at 65 C to constant weight, the plant material was weighed to determine posttreatment biomass for each quadrat using the same process described earlier. All measurements for all treatments occurred on the same day.
Data Analysis
Statistical analyses were performed in SAS v. 9.4 using the PROC GLIMMIX procedure (SAS Institute Inc. 2023). All tests were performed using generalized linear models with treatment (UTC, foliar spray, basal bark, and cut stump) as the fixed-effect independent variables. For the pretreatment assessment, two response variables were tested, mean stem circumference per 1-m2 quadrat and mean number of stems per 1-m2 quadrat. For the posttreatment assessment, the response variable tested was mean biomass per 1-m2 quadrat. The model included block (i.e., location) as a random effect and treatment as the fixed effect. Various distributions (Gaussian, Poisson, negative binomial, or lognormal) were examined for each response variable and selected based on optimal qualities: random spread in residual/ predicted plots, linear pattern in residual/ quantile plots, and low corrected Akaike information criterion values. A lognormal distribution was used for both pretreatment tests and the posttreatment test. Treatment effects within each model were considered significant at P < 0.05. Significant models were then analyzed by Tukey’s honest significant difference to determine whether differences occurred among individual treatments, and significance was accepted at P < 0.05.
Results and Discussion
Before application of treatments, neither the mean basal circumference (cm) nor the mean number of E. pungens stems per 1 m2 (F(3, 42) = 1.29; P = 0.292) was different among treatment plots (F(3, 42) = 0.42; P = 0.738) (Figure 2A and 2B), demonstrating that no preexisting bias existed among plots in terms of plant size or abundance before treatment applications. After treatment applications, biomass (kg m−2) of E. pungens was significantly different among treatments (F(3, 42) = 26.63; P < 0.001). Both cut stump and basal bark treatments resulted in significantly lower E. pungens biomass (0.012 ± 0.004 kg m−2 and 0.006 ± 0.005 kg m−2, respectively) compared with the foliar spray and untreated control (2.27 ± 1.39 kg m−2 and 2.97 ± 1.41 kg m−2, respectively) (Figure 2C). The posttreatment biomass in the cut stump and basal bark treatments did not differ from each other, nor did those in the foliar spray and UTC treatments (Figure 2C).
Figure 2. (A) Stem size pretreatment comparison from an Elaeagnus pungens management study in Calhoun County, South Carolina, USA: mean (±SE) stem circumference of E. pungens per 1-m2 plot for each experimental treatment. (B) Stand density pretreatment comparison: mean (±SE) number of E. pungens stems per 1-m2 plot for each experimental treatment. (C) Aboveground plant biomass posttreatment comparison: mean (±SE) E. pungens dry weight per 1-m2 plot for each experimental treatment. Elaeagnus pungens biomass was significantly greater in the untreated control (UTC) and foliar spray treatment compared with the basal bark and cut stump application methods. Means sharing the same letter are not significantly different from each other.
Herbicides are used globally for vegetation management in forest ecosystems, and responsible use requires a constant refinement of application techniques to ensure the most efficient and effective management methods are being used (Little et al. Reference Little, Willoughby, Wagner, Adams, Frochot, Gava, Gous, Lautenschlager, Örlander, Sankaran and Wei2006; Pile Knapp et al. Reference Pile Knapp, Coyle, Dey, Fraser, Hutchinson, Jenkins, Kern, Knapp, Maddox, Pinchot and Wang2023). One of our goals was to evaluate the usability of these application techniques from a land manager’s perspective (Kettenring and Adams Reference Kettenring and Adams2011). Natural resource land managers typically consider several different facets of invasive plant management techniques when determining which is most appropriate or useful for their specific situations (Kerr et al. Reference Kerr, Baxter, Salguero-Gómez, Wardle and Buckley2016; Lindenmayer et al. Reference Lindenmayer, Wood, MacGregor, Buckley, Dexter, Fortescue, Hobbs and Catford2015). In our study, we considered treatment effort, duration, and associated cost in addition to E. pungens mortality and related reduction in biomass to identify the optimal treatment method, and to that end, our study provided immediate and useful results for managers. These data could be included in a decision tree to help guide management activities in similar areas (e.g., Lindenmayer et al. Reference Lindenmayer, Wood, MacGregor, Buckley, Dexter, Fortescue, Hobbs and Catford2015).
Each treatment method we evaluated had pros and cons. While foliar herbicide application is often one of the fastest and least physically demanding application methods, it was the least effective treatment method in our study (Table 2; Figure 2C). Foliar applications of triclopyr as the free acid formulation had the most immediate and dramatic visual effect, but the resulting superficial visual crown reduction was misleading. Upper foliage was killed within 5 mo of treatment (TLE, personal observation), and resprouting was evident around the plant base within 8 mo of treatment (TLE, personal observation). In many cases, the bottom half of the plant remained healthy, and growth continued normally. Both cut stump and basal bark application methods significantly impacted E. pungens mortality (Figure 2C). Cut stump application took less time but was more physically demanding than the basal bark application method and required three people to be on-site while both foliar spray and basal bark applications were completed with two people (TLE, personal observation; Table 2). The cut stump method required the applicators to cut through the base of the plant and drag the plant material out of the way to reach additional plants in other quadrats and move through the treatment area. This method left a great number of large, sprawling, dead E. pungens branches on the forest floor with intertwined sprouts, making physical navigation difficult. Further, the untreated tops of recently felled E. pungens may still hold viable seeds for a period of time posttreatment. Some studies have shown the use of triclopyr through cut stump applications yielded high percentages of resprouts in plants prone to root suckering (DiTomaso and Kyser Reference DiTomaso and Kyser2007; Fogliatto et al. Reference Fogliatto, Milan and Vidotto2020). Anecdotally, more resprouting appeared to be present in quadrats treated with the cut stump method than those treated with basal bark applications, but more research is needed to determine sprouting potential for E. pungens while using this method. While the basal bark application method required a slightly longer application time than cut stump or foliar spray, basal spray application was equally effective and less physically demanding on the applicator than cut stump applications and could easily be performed by one person if necessary (TLE, personal observation; Table 2).
Herbicide application should be conducted in a manner that minimizes negative impacts to non-target flora and fauna. To do this requires a combination of empirical data (e.g., Gibson et al. Reference Gibson, Shupert and Liu2019) and knowledge of how active ingredients work. Selective treatments with triclopyr instead of a broad-spectrum herbicide (e.g., glyphosate) were used to avoid non-target impacts to native flora. Other commonly used herbicides in forested settings (e.g., picloram) are soil active, and other non-target species may be injured through root absorption. Additionally, Trycera® carries an aquatic label allowing basal bark and cut stump applications in aquatic sites, making triclopyr a logical choice for this invasive species removal effort.
Recently, Yannelli et al. (Reference Yannelli, Bazzichetto, Conradi, Pattison, Andrade, Anibaba, Bonari, Chelli, Cuk, Damasceno, Fantinato, Geange, Guuroh, Musa Holle and Küzmič2022) listed 15 emerging challenges and opportunities for vegetation science, one of which was halting forest degradation by targeted restoration in prioritized ecosystems. Accomplishing such a goal requires a thorough knowledge of methods to reduce or eliminate unwanted vegetation to facilitate the restoration of desired species. The details surrounding management costs (including finances, time, and labor) all factor into a land manager’s decision-making process when determining when or whether to engage in management activities for invasive plants (Zhang et al. Reference Zhang, Zhai, Ervin and Coyle2023). Although some may posit that straightforward studies like these which evaluate a single active ingredient on a single target species are too basic to have broader global applicability, we argue that these studies are crucial in the restoration of degraded forest ecosystems worldwide. Our study provides several essential details for land managers dealing with Elaeagnus spp. in temperate systems.
The use of E. pungens in ornamental, hedgerow, wildlife, and roadside plantings are the primary causes for its current widespread distribution in North America. Restricting its sale and use for landscape and roadside plantings would contribute positively to reducing its spread. As with any invasive flora, maintaining healthy natural landscapes through the cultivation of native plant communities and weed prevention and control is more effective than strictly attempting to control invasion. However, cost-effective management options are necessary for land reclamation and ecosystem restoration in invaded areas (Kimball et al. Reference Kimball, Lulow, Sorenson, Balazs, Fang, Davis, O’Connell and Huxman2015; Meli et al. Reference Meli, Herrera, Melo, Pinto, Aguirre, Musálem, Minaverry, Ramírez and Brancalion2017). This study addressed a common management question and provided strong evidence for two effective application methods for Elaeagnus management in a forested setting. As many invasive shrubs in this region typically fill the same ecological niche and can be managed with similar techniques, our findings likely apply to other woody invasive species in North American natural areas (e.g., Lonicera, Ligustrum, Rhamnus spp.) for which triclopyr is already known to be an effective herbicide (Bisikwa et al. Reference Bisikwa, Natukunda and Becker2020; Delanoy and Archibold Reference Delanoy and Archibold2007; DiAllesandro Reference DiAllesandro2012; Enloe et al. Reference Enloe, O’Sullivan, Loewenstein, Brantley and Lauer2016, Reference Enloe, O’Sullivan, Loewenstein, Brantley and Lauer2018; Harrington and Miller Reference Harrington and Miller2005; Hogan et al. Reference Hogan, Baker, Back and Barber2024; Mervosh and Gumbart Reference Mervosh and Gumbart2015). We advocate for additional research to further develop management techniques for Elaeagnus and other invasive woody flora, as this knowledge will directly and positively impact natural area restoration efforts worldwide.
Acknowledgments
We thank Audubon South Carolina and the Columbia Chapter of the Audubon Society for project support. This work could not have been completed without the help of an amazing volunteer crew (Dave Schuetrum, Scott Weitecha, and Hank Stallworth, Audubon South Carolina Staff, Mark Musselman, Rick Armstrong, and Jessica Sharp-Miner).
Funding statement
Helena Agri-Enterprises LLC graciously provided the product used in this study. This is Technical Contribution No. 7345 of the Clemson University Experiment Station. This material is based upon work supported by the NIFA/USDA, under project number SC-1050622,1700622. Any opinions, findings, conclusions, or recommendations expressed in this publication are those of the author(s) and do not necessarily reflect the view of the USDA.
Competing interests
The authors declare no conflicts of interest.
Open access
