Systemic insecticides

The ability to protect ash trees from EAB with systemic insecticides has advanced considerably in recent years. In the early years following identification of EAB, some landscapers and arborists applied cover sprays of broad-spectrum insecticides to protect landscape ash trees on golf courses and private property. Field studies showed relatively persistent pyrethroid or carbamate insecticide products, which targeted leaf-feeding adult EAB, could be effective if sprays were timed correctly (McCullough et al. Reference McCullough, Cappaert and Poland2005; Herms et al. Reference Herms, McCullough, Smitley, Sadof and Cranshaw2014). Sprays were unpopular, particularly in residential and urban areas, however, because of concerns about drift and possible environmental contamination, potential effects on pollinators, beneficial predatory insects and other non-target organisms, and possible applicator exposure. Other issues included the difficulty of accessing and adequately covering the upper canopy of large trees.

Only a few systemic insecticides, primarily imidacloprid products applied either to the soil or as trunk injections, were available when EAB was first identified. While many problems associated with cover sprays were avoided, treatment with these products yielded inconsistent results (McCullough et al. Reference McCullough, Cappaert, Poland, Lewis and Molongoski2006; Herms et al. Reference Herms, McCullough, Smitley, Sadof and Cranshaw2014). Annual applications sufficiently protected some trees at some sites from EAB, but the same treatments failed at other sites. Treated trees sometimes lived a few years longer than untreated trees, but still eventually succumbed to EAB infestation despite multiple years of investment in treatments (McCullough et al. Reference McCullough, Cappaert, Poland, Lewis and Molongoski2006).

Fortunately, new systemic insecticides have been developed, application technology has improved, and our understanding of how to optimise these treatments has advanced considerably since the mid-2000s. Numerous field trials have shown a product sold as TREE-äge™ (ArborJet Inc., Woburn, Massachusetts, United States of America) with the active ingredient emamectin benzoate applied via trunk injection, consistently provides two to three years of nearly 100% EAB control, even where local EAB density is high (Smitley et al. Reference Smitley, Doccola and Cox2010; McCullough et al. Reference McCullough, Poland, Anulewicz, Lewis and Cappaert2011a; Herms et al. Reference Herms, McCullough, Smitley, Sadof and Cranshaw2014). In a recent two-year study, an azidirachtin product sold as TreeAzin® (BioForest Technogies Inc., Sault Ste. Marie, Ontario, Canada) was highly effective if applied annually and at least moderately effective for two years following application (McKenzie et al. Reference McKenzie, Helson, Thompson, Otis, McFarlane and Buscarini2010; Kreutzweiser et al. Reference Kreutzweiser, Thompson, Grimalt, Chartrand, Good and Scarr2011; D.G.M., personal observation). Trials have also shown Safari™ (Valent U.S.A. Corp., Walnut Creek, California, United States of America) (dinotefuran), a highly soluble neonicotinoid often applied as a basal trunk spray to the lower 1.5–2.0 m of the tree and improved imidacloprid products can provide effective control if applied annually (McCullough et al. Reference McCullough, Poland, Anulewicz, Lewis and Cappaert2011a; Herms et al. Reference Herms, McCullough, Smitley, Sadof and Cranshaw2014).

Girdled ash trap trees

Ash trees that are stressed by removal of a 15–20 cm wide band of bark and phloem around the circumference of the trunk can play multiple roles in efforts to manage local EAB populations. In North America, EAB will colonise and develop on healthy trees of most native ash species, but beetles preferentially attack stressed trees over healthy trees (Cappaert et al. Reference Cappaert, McCullough, Poland and Siegert2005; Siegert et al. Reference Siegert, McCullough, Williams, Fraser and Poland2010b; Herms and McCullough Reference Herms and McCullough2014), primarily in response to changes in volatile profiles and perhaps hyperspectral reflectance elicited by the stress (Rodriguez-Sanoa et al. Reference Rodriguez-Saona, Poland, Miller, Stelinski, Grant and de Groot2006; Crook et al. Reference Crook, Khrimian, Francese, Fraser, Poland and Sawyer2008; de Groot et al. Reference De Groot, Grant, Poland, Scharbach, Buchan and Nott2008). In multiple sites in Michigan, United States of America, larval densities were at least three to four times higher on girdled ash trees than on adjacent or nearby non-girdled ash trees (McCullough et al. Reference McCullough, Poland, Anulewicz and Cappaert2009a, Reference McCullough, Poland and Cappaert2009b, Reference McCullough, Siegert, Poland, Pierce and Ahn2011b; Tluczek et al. Reference Tluczek, McCullough and Poland2011). Differential attraction to girdled trees is most pronounced at low EAB densities and declines as increasing larval densities stress more trees (McCullough et al. Reference McCullough, Poland and Cappaert2009b; Mercader et al. Reference Mercader, Siegert, Liebhold and McCullough2011b).

Girdled ash trees remain the most effective tool for detection and delineation of low-density EAB populations (McCullough et al. Reference McCullough, Siegert, Poland, Pierce and Ahn2011b; Mercader et al. Reference Mercader, McCullough and Bedford2013). From 2004 to 2006, state regulatory officials established grids of small ash trees, typically along roads and highways that were girdled in spring (Rauscher Reference Rauscher2006; Hunt Reference Hunt2007). Numerous previously unknown EAB infestations were identified in autumn when larval galleries were found as survey crews debarked the trees. An intensive study to evaluate girdled trees and three artificial trap designs, all baited with host volatiles, was established across a 16 ha forested site with a very low-density EAB infestation. Results showed 100% of the girdled trees were infested by EAB, while 81% of purple double-decker traps, 56% of green double-decker traps, and 25% of purple canopy traps captured at least one EAB adult (McCullough et al. Reference McCullough, Siegert, Poland, Pierce and Ahn2011b). Mercader et al. (Reference Mercader, McCullough and Bedford2013) compared the effectiveness of 1710 girdled ash trees and 1035 baited artificial traps suspended in the canopy of ash trees by analysing three years of data collected from a >390 km² area with varying EAB densities. Results showed girdled trees were more likely to be positive (i.e., detect EAB) than baited artificial traps, even when traps were placed on girdled trees. Differences were particularly evident when EAB densities were low. For example, at densities of 3.5 larvae per m² of phloem, the probability of detection was 80% for girdled trees but only 15% for artificial traps (Mercader et al. Reference Mercader, McCullough and Bedford2013).

Along with providing information on EAB distribution and density, girdled ash trees can be used as trap trees or population “sinks” (McCullough et al. Reference McCullough, Siegert, Cappaert, Poland and McDonald2007; Siegert et al. Reference Siegert, McCullough, Poland, Heyd, Pierce and Xue2010a). This tactic involves girdling trees to attract adult EAB, including ovipositing females, then destroying the girdled trees by debarking, chipping, or burning before the new cohort of larvae can complete development. Using girdled ash trees to reduce EAB population growth could be an effective option in forested, rural, or even urban areas. Virtually any tree that is not suitable for treatment with systemic insecticides, for example, can be girdled in spring, then destroyed in fall or winter, eliminating the developing larvae as well as their potential offspring.

Moreover, intriguing evidence from forested sites with low or very low-density EAB infestations indicates girdled trees may also influence EAB dispersal and subsequent distribution. Ovipositing beetles were directed towards and laid more eggs on ash trees in 4 ha blocks where four clusters consisting of three girdled trees were established, than in similar blocks without girdled trees (Siegert et al. Reference Siegert, McCullough, Poland, Heyd, Pierce and Xue2010a). In the second year of the study, the EAB population remained concentrated in the blocks with clusters of girdled trees. Densities of EAB were again higher on girdled than on non-girdled although differences were not as great as in the first year. In an extensive, multi-year evaluation, non-girdled trees near girdled trees had larval densities higher than otherwise expected (R.J.M. and D.G.M., personal observation), indicative of a “spillover” effect. This pattern could potentially be applied operationally in an integrated management programme by using girdled trees to direct beetles to an area where other trees have been treated with a systemic insecticide or, alternatively, to an area scheduled for harvest the following season.

Ash removal and phloem reduction

Reducing the potential production of EAB by harvesting ash trees for timber or firewood, or removing low vigour or unattractive landscape trees, represent additional options for some sites. Harvesting merchantable ash trees can provide value for forest landowners, and depending on site and stand characteristics, may facilitate long term silvicultural objectives. White ash (Fraxinus americana Linnaeus) is a desirable species for timber or veneer production (e.g., Wright Reference Wright1959; Northeast Timber Exchange 2014) and virtually all ash species are valued for firewood because of their straight grain and relatively high BTU production (e.g., Chimney Sweep 2014). Municipal foresters and arborists may elect to replace problematic, diseased, or unwanted ash trees, including those that have outgrown their sites or those interfering with power lines.

Potential effects of ash removal on EAB population growth depend on the abundance and size class distribution of the harvested ash trees, which in turn, determine the area of ash phloem available for EAB larval development. McCullough and Siegert (Reference McCullough and Siegert2007) developed equations to estimate ash phloem area from tree diameter at breast height (dbh) and reported that, on average, 89 EAB could potentially develop and emerge as adult beetles per m² of green ash (Fraxinus pennsylvanica Marshall) or white ash phloem. For example, a tree that is 15 cm in dbh can potentially produce at least 305 adult beetles while a tree that is 45 cm in dbh can produce more than 3325 beetles. In second growth forests, which characterise much of the area invaded by EAB, at least 75–80% of the ash stems are typically < 15 cm in dbh but collectively account for roughly 15% of the potential EAB production. In contrast, merchantable ash trees (e.g., dbh>25 cm) comprise only 5–6% of the ash stems but would collectively produce 55–65% of the EAB adults (McCullough and Siegert Reference McCullough and Siegert2007).

Wholescale removal of ash trees to simply reduce ash phloem, however, is not an optimal strategy for EAB management. Mercader et al. (Reference Mercader, Siegert, Liebhold and McCullough2011a) compared effects of treating merchantable ash trees with TREE-äge™, using those same trees as girdled trap trees, or simply destroying those trees to reduce ash phloem. Their simulations showed ash removal was less effective at slowing EAB population growth than either of the other options and would ultimately accelerate EAB spread. Harvesting ash for timber or for locally used firewood, however, can be an appropriate component of an integrated management approach in rural or forested areas. In urban areas, unwanted landscape ash designated for removal or replacement can be girdled in the spring, allowed to function as an EAB trap tree to attract ovipositing EAB females during the summer, then removed in the fall or winter. Although ecological services provided by these trees are eventually lost, at least some value is derived by using the trees as an EAB population sink before the trees are removed.

SLow Ash Mortality (SLAM) pilot project

A large scale effort to develop, implement and assess effects of an integrated management strategy for EAB began in 2008 in the eastern Upper Peninsula of Michigan (McCullough et al. Reference McCullough, Siegert and Bedford2009c; Poland and McCullough Reference Poland and McCullough2010). The SLAM pilot project involved numerous cooperators from two universities and two federal and two state agencies. Because the difficulty of accurately detecting and delineating low densities of EAB was recognised, collaborators made a conscious decision at the onset of the project to emphasise protection of the ash resource, rather than attempting to slow the spread of the EAB population. The primary goal of the SLAM pilot project was to evaluate whether local EAB population growth could be reduced to delay the onset and progression of ash decline and mortality.

Much of the focus of the SLAM project revolved around how best to use two management tactics; girdled ash trees and treatment with TREE-äge™, the highly effective systemic insecticide. Application of both tactics was supported by results from simulation models, developed as part of the SLAM effort (Mercader et al. Reference Mercader, Siegert, Liebhold and McCullough2011a, Reference Mercader, Siegert, Liebhold and McCullough2011b; McCullough and Mercader Reference McCullough and Mercader2012; Kovacs et al. Reference Kovacs, Haight, Mercader and McCullough2014). Potential effects of varying ash abundance, distribution, and condition on adult EAB dispersal and EAB population growth were evaluated by Mercader et al. (Reference Mercader, Siegert, Liebhold and McCullough2011b) using a coupled map lattice model. This model, which links spatially explicit ash and EAB density, was parameterised with empirical data from several field studies. Components of the model include parameters for adult EAB dispersal, larval phloem consumption, adult dispersal, and population growth. As larval density builds and trees become stressed, larval development rates shift from predominantly two-year to one-year life cycles (Siegert et al. Reference Siegert, McCullough, Williams, Fraser and Poland2010b; Tluczek et al. Reference Tluczek, McCullough and Poland2011). Simulations demonstrated the influence of dispersal and host selection by adult females, including their response to host stress elicited by girdling, and the important effect of EAB population growth on ash mortality rates.

Subsequently, when Mercader et al. (Reference Mercader, Siegert, Liebhold and McCullough2011a) applied the coupled map lattice model to simulate potential effects of alternative management tactics, results showed treating trees with TREE-äge™ had the greatest effect on reducing EAB population growth and expansion over time. Girdling ash trees also slowed population size and radial spread rate, but this tactic was roughly half as effective as TREE-äge™ injections. Effects diminished as EAB density increased and trees available for girdling were exhausted. Harvesting ash to reduce phloem affected EAB density locally, but did little to slow overall population growth or radial spread. Treating trees with TREE-äge™ was also the only tactic that protected ash trees, and the ecological services provided by those trees, over time (Mercader et al. Reference Mercader, Siegert, Liebhold and McCullough2011a; McCullough and Mercader Reference McCullough and Mercader2012). Furthermore, when the economic costs of managing EAB in urban environments were simulated, results consistently showed treating ash trees with TREE-äge™ provided considerable economic benefits compared with removing trees as they succumbed (McCullough and Mercader Reference McCullough and Mercader2012; Kovacs et al. Reference Kovacs, Haight, Mercader and McCullough2014).

However, in contrast to simulation models which require assumptions and simplified environments, the extensive spatial scale and heterogeneous nature of the SLAM project area provided an opportunity to deploy and evaluate tactics under real-life, operational conditions. The SLAM project was centred on a low-density EAB infestation, identified in fall 2007 when a regulatory survey crew debarked a girdled ash detection tree growing along a road (McCullough et al. Reference McCullough, Siegert and Bedford2009c). The area surrounding the infested tree included a mix of national forest land and private forests, rural and vacation residences, and small municipalities.

To facilitate surveys of EAB and ash, systematic grids were overlaid on the project area, which encompassed >390 km². Grid densities ranged from one grid cell per 2.6 km² in the periphery of the project area to as many as 16 grid cells per 2.6 km² in the core of the infestation. At least one small (10–20 cm dbh) girdled ash detection tree was established in each grid cell and supplemented with baited artificial traps to monitor EAB distribution and density annually, from 2008 to 2012 (Mercader et al. Reference Mercader, McCullough and Bedford2013). A systematic inventory to characterise ash distribution and density in urban and forested areas progressed each summer. Ash trees were tallied by diameter class and ash phloem area was estimated for each grid cell following McCullough and Siegert (Reference McCullough and Siegert2007) as described in Mercader et al. (Reference Mercader, McCullough and Bedford2013). Surveyors recorded species for 1236 ash trees, representing ~18% of the ash inventory. Black ash (Fraxinus nigra Marshall), green ash (F. pennsylvanica), and white ash (F. americana) comprised 32%, 50%, and 18%, respectively, of the trees that were identified to species in the inventory records.

Outreach activities were undertaken to raise awareness of EAB and build local support for the SLAM activities. Outreach and regulatory efforts were also geared at reducing risks of people transporting potentially infested ash firewood, logs, or trees into or out of the project area. A network of forest health plots was established to enable resource managers to monitor condition of ash and other overstorey species over time.

Carefully developed maps of the data collected from the 2008 girdled detection trees, artificial traps, ash inventory, and land-use classes were evaluated to develop a strategy for the placement of treatments in 2009. Cooperators determined the ideal treatment configuration involved the strategic establishment of girdled trees in the core of the infestation surrounded by insecticide treated trees. The rationale underlying the strategy assumed that clusters of two to three girdled ash trees would be established in the central core of the infestation, where EAB density was highest, to attract and retain adult EAB emerging from infested trees. Additional ash trees within 400–800 m around the perimeter of the girdled tree clusters would be treated with TREE-äge™ to control leaf-feeding and potentially dispersing adult beetles. Similarly, in a few areas outside the core, where isolated detection trees were found to be infested, two to four ash trees would be girdled and at least a few additional trees within 400 m would be treated with TREE-äge™. The combination of girdled trees and treated trees was designed to both slow EAB population growth and to reduce the number of adult EAB dispersing beyond the core.

Unfortunately, operational implementation of the plan, particularly on the national forest, was limited by restrictions and logistical challenges. Several areas where the insecticide was to be applied had to be dropped from consideration when national forest personnel unexpectedly withdrew approval, thus excluding a considerable portion of the project area. At the same time, the area designated as critical habitat for Hine’s emerald dragonfly (Somatochlora hineana Williamson; Odonata: Corduliidae), a federally listed endangered species (Federal Register 2010) was substantially expanded, excluding another portion of the project area from treatment. Additional challenges included lengthy contract approval processes and difficulties of identifying and contacting absentee landowners in winter and early spring, when many vacation properties were vacant. Despite these problems, 444, 603, 748, and 855 girdled “sink” trees were established in 2008, 2009, 2010, and 2011 and 229 and 358 ash trees, primarily in right-of-ways along a few roads, were treated with TREE-äge™ in 2009 and 2010, respectively. In contrast to the original plan, however, insecticide treatments and girdled ash trees were not coupled and the insecticide applications were concentrated in only few areas, mostly on private property. These conditions provided a markedly different, but equally valuable and unique opportunity to assess effects of these tactics on the local EAB population and to evaluate whether the realised treatments affected EAB spread.

Evaluation of the SLAM Pilot Project is still in progress but a few key points are evident from analyses to date. While only a tiny fraction of the ash trees in the project area were treated with either the systemic insecticide or girdled, both treatments exerted detectable effects on EAB population growth and reduced the rate of ash decline that would be expected in the absence of any treatment (Mercader et al. Reference Mercader, Siegert, Liebhold and McCullough2011a, Reference Mercader, Siegert, Liebhold and McCullough2011b). Despite the less than optimal distribution of trees treated with the TREE-äge™ insecticide, the treatment significantly reduced overall EAB population growth for at least two years post-injection, a result consistent with previous field studies (Smitley et al. Reference Smitley, Doccola and Cox2010; McCullough et al. Reference McCullough, Poland, Anulewicz, Lewis and Cappaert2011a). SLAM results to date suggest the number of trees injected with TREE-äge™ may have been more important in slowing EAB population growth than the total area of ash phloem within treated trees. In practical terms, this indicates injecting as many trees as possible and distributing treatment across a given area would likely be a more effective use of resources than focussing insecticide treatment on a few large trees. Effects of the girdled ash trees on the EAB infestation were more complex. Preferential oviposition on girdled trees was associated with an overall reduction in the EAB population the next year, coupled with higher than expected EAB density on trees in the immediate area. In effect, girdled trees acted as a sink for EAB progeny of adult beetles, but also generated a “spillover” onto nearby trees. Collectively, results suggest girdled trees exerted a small, but noticeable, effect on EAB spread, effectively concentrating or retaining beetles in areas with girdled ash trees.

Integrating tactics

Developing an integrated management strategy for any pest, including EAB, requires consideration of the potential interactions among pest control tactics. Obviously, it is important to avoid combining tactics that are antagonistic (i.e., one tactic interferes with or reduces the efficacy of the other), or redundant, which wastes limited resources (Suckling et al. Reference Suckling, McCullough, Herms and Tobin2012). Any combination of tactics should minimally yield additive results and tactics that yield synergistic or super-additive effects (i.e., one tactic enhances effects of another) are even more desirable (Berec et al. Reference Berec, Angulo and Courchamp2007; Suckling et al. Reference Suckling, McCullough, Herms and Tobin2012).

Highly effective systemic insecticides now available for EAB control can be combined with other tactics to produce additive or possibly even synergistic effects. For example, results from the SLAM pilot project suggest that coupling girdled trees and trees treated with a highly effective systemic insecticide, like TREE-äge™, will likely produce synergistic effects for a given area. Strategies that optimise distribution of those treatments across a given area can potentially increase benefits, perhaps substantially. Recently, the potential to combine these strategies within a single tree by injecting ash trees with TREE-äge™, then subsequently girdling the trees to create “lethal trap trees” has been demonstrated (D.G.M., personal observation). Girdling serves to enhance EAB attraction to treated trees, but the insecticide will control leaf-feeding adult beetles as well as young larvae. Although girdling will eventually kill the tree, this combination of tactics does not require that a girdled tree be felled and destroyed the year it is girdled, extending the longevity of the girdled tree. Moreover, combining “lethal trap trees” with non-girdled trees that are treated with insecticide could perhaps provide an effective “attract and kill” tactic for reducing EAB population growth and delaying the progression of ash decline and mortality.

Coupling an effective systemic insecticide and biological control should also be compatible and minimally yield an additive effect. Trees treated with TREE-äge™ will harbour few, if any, late instar EAB larvae and thus will not attract woodpeckers, who prey on fourth instars and prepupae (Lindell et al. Reference Lindell, McCullough, Cappaert, Apostolou and Roth2008; Flower et al. Reference Flower, Long, Knight, Rebbeck, Brown and Gonzalez-Meler2014) or native or introduced larval parasitoids that attack only live larvae (Cappaert and McCullough Reference Cappaert and McCullough2009; Duan et al. Reference Duan, Bauer, Abell and Van Driesche2012, Reference Duan, Bauer, Abell, Lelito and Van Driesche2013). Egg parasitoids of EAB, such as the introduced Oobius agrili Zang and Huang (Hymenoptera: Encyrtidae) (United States Department of Agriculture 2013), will not encounter the systemic insecticide and should be able to complete development in EAB eggs on treated or untreated trees. On trees treated with TREE-äge™, EAB mortality caused by egg parasitism may be redundant; larvae from those eggs would presumably not survive. Those eggs, however, will produce adult egg parasitoids, who can subsequently attack more EAB eggs on treated or untreated trees, increasing overall EAB mortality rates. Moreover, simulations have suggested that using insecticides to reduce pest density in combination with specialised natural enemies that engage in non-random searching for their prey or hosts may even yield synergistic effects (Barclay and Li Reference Barclay and Li1991; Suckling et al. Reference Suckling, McCullough, Herms and Tobin2012).

Parting thoughts on the integration of existing tactics

Research conducted since the discovery of EAB in North America has generated a considerable amount of information about this invasive pest. Using this knowledge to develop practical, integrated management strategies for specific areas is clearly possible, although many challenges remain. In urban and residential areas, there is an increasing emphasis on the value of protecting desirable landscape ash trees with systemic insecticides, rather than merely attempting to remove trees as they succumb to EAB infestation. Such a strategy is strongly supported by economic analyses of costs and benefits (McCullough and Mercader Reference McCullough and Mercader2012; Van Atta et al. Reference Van Atta, Hauer and Schuettpelz2012), and given the SLAM pilot project results, these efforts may additionally slow overall EAB population growth in a given area. Managing EAB and maintaining ash as a significant overstorey component in forested areas will likely remain a difficult proposition. Strategic use of girdled trees, ash harvests, and systemic insecticides, where practical, can help to manage EAB populations and mitigate damage and economic losses. Continued research is needed, however, to identify factors affecting ash resistance among and within native ash species, improve methods to detect, delineate and monitor EAB, and develop practical options to enhance biological control of EAB by native and introduced natural enemies. Whether such technical advances will be achieved in time to preserve the functional role of ash in North American forests remains to be seen.