Recent advances toward the sustainable management of invasive Xylosandrus ambrosia beetles

We provide an overview of both traditional and innovative control tools for management of three Xylosandrus ambrosia beetles (Coleoptera: Curculionidae: Scolytinae), invasive species with a history of damage in forests, nurseries, orchards and urban areas. Xylosandrus compactus, X. crassiusculus and X. germanus are native to Asia, and currently established in several countries around the globe. Adult females bore galleries into the plant xylem inoculating mutualistic ambrosia fungi that serve as food source for the developing progeny. Tunneling activity results in chewed wood extrusion from entry holes, sap outflow, foliage wilting followed by canopy dieback, and branch and trunk necrosis. Maintaining plant health by reducing physiological stress is the first recommendation for long-term control. Baited traps, ethanol-treated bolts, trap logs and trap trees of selected species can be used to monitor Xylosandrus species. Conventional pest control methods are mostly ineffective against Xylosandrus beetles because of the pests’ broad host range and rapid spread. Due to challenges with conventional control, more innovative control approaches are being tested, such as the optimization of the push–pull strategy based on specific attractant and repellent combinations, or the use of insecticide-treated netting. Biological control based on the release of entomopathogenic and mycoparasitic fungi, as well as the use of antagonistic bacteria, has yielded promising results. However, these technologies still require validation in real field conditions. Overall, we suggest that management efforts should primarily focus on reducing plant stress and potentially be combined with a multi-faceted approach for controlling Xylosandrus damage.

introductions into areas located outside their native range are increasing (Roques et al. 2009;Haack et al. 2013;Hulcr et al. 2021). The trend toward more introductions is linked to increasing global trade of wood products and climate change (Marini et al. 2011;Meurisse et al. 2019;Biedermann et al. 2019;Grousset et al. 2020;Inward 2020;Lantschner et al. 2020;Zeng et al. 2020), and it is not expected to stop in the near future (Liebhold et al. 2017;Urvois et al. 2021). Some species can also cause severe damage to a broad range of host trees growing in various ecosystems, including plant nurseries, orchards and natural forests (Kirkendall and Faccoli 2010;Sauvard et al. 2010;Hulcr and Dunn 2011;Vannini et al. 2017). Consequently, interest in ambrosia beetles by scientists and plant protection organizations is increasing exponentially.
At present, more than 50 ambrosia beetles are known to be established outside their native range (Lantschner et al. 2020). Most of non-indigenous ambrosia beetles belong to the tribe Xyleborini, including the genus Xylosandrus that contains several important invasive species (Dole and Cognato 2010;Greco and Wright 2015;Ranger et al. 2016a;Rassati et al. 2016). Specifically, Xylosandrus compactus (Eichhoff), Xylosandrus crassiusculus (Motschulsky), and Xylosandrus germanus (Blandford) are three species native to subtropical Asia that have been established in several other countries around the world (Haack et al. 2013;Rassati et al. 2016a;Lantschner et al. 2020;Smith et al. 2020). Attacks of these species can occur in different environments including forests, ornamental plant nurseries, orchards, and urban areas (Ranger et al. 2016a;Agnello et al. 2017;Galko et al. 2019;Gugliuzzo et al. 2019a). Adult females bore galleries into the xylem of twigs, branches, stems, trunks, and exposed roots of a wide variety of woody plants (Greco and Wright 2015;Ranger et al. 2016a). Mutualistic ambrosia fungi are introduced into these galleries by adult females and provide the food source for developing larvae and adult offspring (Kinuura 1995;Hulcr and Stelinski 2017;Biedermann and Vega 2020). The effect of Xylosandrus colonization differs between live and dead trees, as do the symptoms. In both cases, sawdust in the form of frass (for X. compactus) or compacted "noodles" (for X. crassiusculus and X. germanus) is extruded from trees, but attack on living trees also produces sap flow, twig wilting followed by canopy dieback, and branch and trunk necrosis.
The black twig borer, X. compactus, is currently widely distributed in Africa, Asia, Southeastern USA, and South America (Fig. 1). Introductions of X. compactus have occurred in the Pacific Islands, and more recently in Europe (EPPO 2020a). Xylosandrus compactus females are minute and shiny black, approximately 1.4-1.8 mm long and 0.7-0.8 wide (Rabaglia et al. 2006). Adults of this species have been recorded from more than 220 host plants, including agricultural crops, trees, shrubs and nursery plants (EPPO 2020a). In particular, X. compactus has become an emerging pest of several Mediterranean maquis trees and various ornamental shrubs widespread in southern Europe (Pennacchio et al. 2012;Francardi et al. 2017;Vannini et al. 2017;Gugliuzzo et al. 2019b;Contarini et al. 2020;Leza Fig. 1 Current global distribution (as January 11th 2021) of Xylosandrus compactus, X. crassiusculus and X. germanus. Different colors indicate single-or multi-species occurrence of the three ambrosia beetles among countries. The combination of X. compactus + X. ger-manus without X. crassiusculus was not found for any country (EPPO 2020b). Colored areas refer to political country areas, not to biogeographical distribution et al. 2020). Moreover, X. compactus has raised serious concerns as a threat to coffee production in East and Central Africa (FAO 2013;IACO 2013;ICO 2014).
The granulate ambrosia beetle, X. crassiusculus, has been reported in the USA, Central and South America, Africa, Oceania and Europe (EFSA 2020) (Fig. 1). Female X. crassiusculus are on average larger (2.1-2.9 mm in length and about 1.2 mm wide) than the other two Xylosandrus species considered in this review. They are reddish-brown, becoming darker brown from the distal half of the elytra (Pennacchio et al. 2003;Rabaglia et al. 2006;Gallego et al. 2017). Xylosandrus crassiusculus attacks at least 120 woody host species (Schedl 1963). However, thin-barked deciduous species are attacked more frequently, including nursery ornamental (Ranger et al. 2015a) and cultivated fruit trees (Carrillo et al. 2012).
Xylosandrus germanus is now established in 21 European countries (Fig. 1), along with Russia, 35 US states, and six Canadian provinces (Bousquet et al. 2013;Ranger et al. 2016a;Gomez et al. 2018;Galko et al. 2019;Rabaglia et al. 2019;Webster et al. 2020). The beetle has several common names including black stem borer, black timber bark beetle, and tea root borer. Adult females are dark brown to black, shiny, and semi-cylindrical; they are 2.0-2.4 mm long and about 1.0 mm wide (Rabaglia et al. 2006). Over 200 trees and shrubs are among its broad host range (Weber and McPherson 1983a). The most severely attacked hosts are woodland trees, nursery-grown ornamentals, and orchard tree fruit (Cheng and Guo 1994;Yang et al. 2008). Thinbarked deciduous species are preferred (Ranger et al. 2016a). Galko et al. (2019) reported also attacks of X. germanus on fresh lumber.
The term "host range" is not just a list of species. It is important to distinguish host trees that are colonized after their death when the defense mechanisms cease, from tree species that are colonized while still alive. The latter is typically a much more restricted list than the total host range in the literature, but those tree species are more important from the management perspective. Some authors also distinguish reproductive hosts from non-reproductive hosts. This difference has not been explored well in Xylosandrus compared to other ambrosia beetle pests (e.g., Eskalen et al. 2013;Martini et al. 2017;Paap et al. 2020), but the reproductive hosts are likely to be more important for management. Moreover, the geographic distribution of vulnerable host trees and the physio-climatic classification of related native and invaded regions are key factors in the context of the invasion success of these pests.
Males of these three Xylosandrus spp., like other ambrosia beetles belonging to the tribe Xyleborini, are smaller than females, flightless, with a radically reduced thorax (Rabaglia et al. 2006). They are rarely observed outside the brood gallery and sibling mating dominates reproduction (Peer and Taborsky 2005;Keller et al. 2011). Multiple asynchronous emergences of Xylosandrus beetles from the same gallery (Oliver and Mannion 2001) may suggest some internal intersection of galleries within the wood, potentially allowing some non-sibling mating. Larvae develop from eggs laid in the system of galleries excavated by the foundress female. They feed on species-specific fungal mutualists  that are carried by females in a dedicated fungus-spore-carrying structure (mycangium or mycetangium, as suggested by Vega and Biedermann 2020), and inoculated during gallery excavation (Hulcr and Stelinski 2017;Li et al. 2019;Biedermann and Vega 2020). Female oviposition occurs only after feeding on the fungal mutualist that is established within the gallery (Weber and McPherson 1983b;Kinuura 1995).
The key feature often highlighted for the invasion success of Xyleborini is inbreeding-mating among siblings in 1 3 maternal galleries (Jordal et al. 2000;Grousset et al. 2020). Females arriving in a new environment are already fertilized and ready to establish a new population. However, only a fraction of Xyleborini, as well as only a fraction of Xylosandrus, are widespread invasive (Dole and Cognato 2010). There are over 50 species of Xylosandrus, but only these three have colonized most of the world; the remaining ones are rarely collected outside of their native range, and most of them remain rare. There are exceptions, such as, for example, Xylosandrus amputatus and Xylosandrus morigerus whose ranges are also expanding, but not as rapidly as the previous three (Cognato et al. 2011;Rassati et al. 2016).
What allows these three Xylosandrus species to become effective invaders is thus still a matter of speculation. The general fact of feeding on fungal mutualists is an unlikely reason, because most ambrosia beetles are not invasive; on the contrary, the mutualists' microenvironmental demands further restrict the ecological niche breadth. What may provide the key ecological advantage are the main fungal mutualists which these three beetle species carry in their mycetangia. Although there is a different association between specific fungal species and each of these invasive beetles (Skelton et al. 2019), it appears that all three species are abundant in natural areas, which then often spill over into human-made landscapes including nurseries and orchards (Reding et al. 2015).
Another reason for the spread of these three Xylosandrus spp. is climate change which causes unpredictable weather conditions and drastic fluctuations in temperature and precipitation in different parts of the world (Urvois et al. 2021). This creates stress conditions for trees (i.e., flood, drought, etc.) and make them vulnerable for beetle attacks. These invasive Xylosandrus spp., and more specifically X. crassiusculus and X. germanus which are specialized on freshly dead or stressed trees, are therefore able to find suitable tree hosts, and are likely to continue spreading. The majority of the life cycle of Xylosandrus spp., including pupation, occurs within the host plant galleries in the wood. Adults spend very little time outside of trees for dispersal and host selection. However, Xylosandrus beetles readily take flight after emergence and have been shown to be good flyers (Gugliuzzo et al. 2019a). Furthermore, their broad host range, their efficiency at locating and colonizing trees under physiological stress, and the lack of effective natural enemies in newly invaded environments render beetle management challenging and often ineffective. Reviews dedicated to the management of ambrosia beetles have focused on X. compactus with emphasis on coffee in Hawaii (Greco and Wright 2015), as well as X. crassiusculus and X. germanus in ornamental plant nurseries (Ranger et al. 2016a). These have been implemented more recently following the publication of several important advances in understanding of the biology, ecology, and control strategies of these invasive fungus-farming beetles.
However, several aspects concerning the real effectiveness and the possible improvement of some management strategies against Xylosandrus spp. are still under evaluation. The purpose of this review is to summarize, discuss, and explore global advances and current implementations for the sustainable management of emerging ambrosia beetles in the genus Xylosandrus (Fig. 2).

Cultural control
Weakened, dying or physiologically stressed trees are the main hosts of the opportunistic X. crassiusculus and X. germanus (Ranger et al. 2016a). Therefore, maintaining tree health and vigor is a crucial first step for reducing the risk of ambrosia beetle infestations. Maintenance of plant health to prevent attack is particularly true for these two species, which are poor colonizers of vigorous plants, but can locate vulnerable hosts in the early stage of physiological stress (Weber and McPherson 1983b;Ranger et al. 2015a, b). As recommended for other wood borers, reducing the tree density can strengthen the vigor of individual trees (Jactel et al. 2009) and thus their susceptibility to Xylosandrus attacks. The rate of attack of X. germanus and X. crassiusculus on completely healthy trees is statistically negligible . Unlike X. crassiusculus and X. germanus, X. compactus is one of the few exceptions which preferentially colonize healthy or recently stressed trees (Chong et al. 2009;Greco and Wright 2015;Gao et al. 2017;Vannini et al. 2017). Its colonization success, however, has not been examined under experimentally controlled conditions to determine the influence of host quality or host stress ).
Among abiotic factors predisposing trees to attack by Xylosandrus spp., flooding and freezing events are the most representative (La Spina et al. 2013;Ranger et al. 2013aRanger et al. , 2015aRanger et al. , 2019. Trees intolerant of flooding are preferentially attacked over flood tolerant species (Ranger et al. 2015a). Consequently, monitoring soil moisture levels in regions where potential hosts occur could allow risk forecasting for infestations and provide a possible preventative tool. For example, Frank and Ranger (2016) observed fewer ambrosia beetle attacks on flood-intolerant container-grown nursery trees with substrate moisture levels of 30% and 50% compared to those grown in 70% or 90%. These results were confirmed by Frank et al. (2017), whereby infestations of Xylosandrus spp. were reduced on ornamental trees maintained at 50% or lower water holding capacity of the container substrate. Empirical evidence of extreme frost events (i.e., freezing-related necrotic tissues) preceding large-scale ambrosia beetle attacks on European beech (Fagus sylvatica L.) in Western Europe (La Spina et al. 2013), and on thin-barked deciduous trees in the USA, also highlight the possible influence of frost events on tree attractiveness and/ or susceptibility, especially when mild winters were followed by late-spring freezes (Ranger et al. 2015b. La Spina et al. (2013) demonstrated that X. germanus was more attracted to bark tissue of beech trees experimentally injured by freezing compared to undamaged bark. Similarly, in the USA, X. crassiusculus and X. germanus rapidly attacked container-grown trees that were experimentally freezestressed, while no attacks occurred on control trees . Consequently, proper selection and matching of plant species or cultivar with growing region will contribute toward higher plant tolerance to low temperatures and/or to other abiotic stressors and thus prevent secondary beetle injuries.
Along with abiotic factors, biotic factors also can induce the production and emission of ethanol (Kelsey and Joseph 1998;Kelsey 2001;Kelsey et al. 2016), which attracts Xylosandrus spp. . For instance, trees infected by the oomycete Phytophthora ramorum Werres, de Cock and Man in't Veld, the causal agent of sudden oak death disease, were more attractive to certain ambrosia beetles than uninfected trees as a result of a higher ethanol concentration in wood tissues (McPherson et al. 2005(McPherson et al. , 2008Kelsey et al. 2016). Similar observations are recorded for plant fungal pathogens and Xylosandrus spp. . Whether the presence of a fungal pathogen in wood tissue can affect beetle fitness by competing with its main mutualist remains unclear (Fernandez-Conradi et al. 2018). Previous studies showed a mixed response depending on the fungal pathogen and the ambrosia beetle species (Menocal et al. 2017(Menocal et al. , 2018aFraedrich et al. 2018), but no study targeting Xylosandrus spp. has been performed so far. The role of phytopathogens in the associations with Xylosandrus spp.in host selection and colonization success is an important aspect to consider for improving cultural control strategies.
Most initial infestations by X. compactus occur on twigs and small branches rather than larger branches and trunks (Masuya 2007;Greco and Wright 2015;Gugliuzzo et al. 2019c). Therefore, proper tree pruning including removal of tissues exhibiting initial signs of attack (e.g., wilting, necrosis, entry holes, etc.) can be a potential cultural control  Gugliuzzo et al. (2020). Similarly, non-target infested plant material, including trap plants growing in the vicinity of the cultivated crop, or plants becoming infested within ornamental nurseries should be promptly removed and destroyed to prevent adult beetle emergence and re-infestation (Ranger et al. 2016a).

Lures
Ethanol is the most studied semiochemical involved in the attraction of ambrosia beetles to susceptible host plants. Several studies demonstrated that ethanol is attractive to Xylosandrus beetles and serves as a cue for females to locate suitable hosts (Miller and Rabaglia 2009;Ranger et al., 2010Ranger et al., , 2012. Ethanol is released by plants under stress (Kimmerer and Kozlowski 1982;Ranger et al., 2013aRanger et al., , 2019, often as a result of anaerobic respiration, and it is one of the most attractive semiochemicals for many Xyleborini, including some Xylosandrus spp. (Burbano et al. 2012;Reding et al., 2013a;Ranger et al. 2021). Ethanol also is emitted by Fusarium solani (Mart.) Sacc., a fungal symbiont of X. compactus (Egonyu and Torto 2018). Besides its role in attracting ambrosia beetles, ethanol contained in woody tissues influences the colonization success of Xylosandrus spp., notably by promoting the growth of some of their mutualistic fungi, such as A. grosmanniae and A. roeperi associated with X. germanus and X. crassiusculus, respectively . Interestingly, ethanol also has been shown to inhibit the growth of fungal garden competitors Lehenberger et al. 2020). For instance, X. germanus did not produce fungal gardens or broods in trees baited with ethanol, but gardens and brood were established in living stem tissues containing ethanol introduced through irrigation . The concentration of ethanol in wood tissue might also facilitate niche-partitioning. In a field experiment involving ethanol-soaked logs, the number of entry holes decreased with increasing ethanol concentration for X. germanus and increased for Xyleborinus saxesenii (Ratzeburg) ; in addition, the number of galleries with brood was higher for X. germanus than for X. saxesenii, even though the attack rate was significantly higher for the latter species .
Lures for monitoring Xylosandrus species generally consist of sealed polyethylene sachet or pouch-style dispensers, or capped polyethylene vials, filled with 70-95% ethanol. These dispensers are designed to maintain a steady release rate over time, but may possibly be perforated (in the upper part of the pouch, or in the cap of the vials) to facilitate ethanol volatilization and increase release rates. Ethanol release rate is obviously a key variable affecting the efficacy of the lures. Both standard-release (release rate of 65 mg/day at 30 °C) and ultra-high-release (0.6 g/day at 25-28 °C) ethanol lures are commercially available (Miller and Rabaglia 2009;. A positive correlation between ethanol release rate and trap captures or attacks by Xylosandrus spp. was shown in previous studies (Ranger et al. 2012(Ranger et al. , 2015b, although this pattern was not evident for X. germanus when ethanol-soaked logs were used . The optimal release rate of ethanol may vary depending on the beetle species, and it is necessary to conduct further studies to elucidate these mechanisms. Commercially available releasers alleviate the need for lure preparation by the grower and some of the challenges associated with rapid ethanol evaporation (Oliver and Mannion 2001;Reding et al. 2011). Commercial lures of high release ethanol rate have notably been used in Spain, for the monitoring of X. crassiusculus and X. compactus (Gallego et al. 2017;Leza et al. 2020). The addition of conophthorin and α-pinene to ethanol was evaluated for improving the attraction of Xylosandrus spp., but the results were inconsistent and do not justify the increased cost of resultant lures (Miller and Rabaglia 2009;Ranger et al. 2011aRanger et al. , 2014Burbano et al. 2012;Miller et al. 2015aMiller et al. , 2018Miller et al. , 2020. Other experiments comparing single and multi-lure traps showed that X. crassiusculus was trapped exclusively with ethanol-only lures compared to those baited simultaneously with ethanol, frontalin, ipsenol/ipsdienol, and (-)α-pinene (Rassati et al. 2014). More recently, a specific multi-lure trapping strategy for X. crassiusculus and X. compactus is being developed and tested in Europe (Gallego, unpublished).
Traps in combination with lures can be reliably used as monitoring devices for the presence of Xyleborus spp., another genus of important pests belonging to the tribe Xyleborini. Nevertheless, so far they have been ineffective for mass trapping. In fact, for bark and ambrosia beetles in general, mass trapping is only possible with sexual or aggregation pheromones, which are orders-of-magnitude more attractive than tree produced kairomones, including ethanol (Miller and Rabaglia 2009). Furthermore, even in the pheromone-based systems, mass trapping has rarely been shown as cost-effective (Grégoire et al. 2001). In addition, Xylosandrus spp. are not coordinated in their attacks, as the mass-attacking conifer bark beetles are (Blomquist et al. 2010), and instead colonize stressed trees individually.

Traps
Ethanol-baited traps are the predominant method for monitoring ambrosia beetle populations, including Xylosandrus spp. in ornamental nurseries (Reding et al., , 2011(Reding et al., , 2015, forests (Hauptman et al. 2019;Martínez et al. 2019;Marchioro et al. 2020), and urban landscapes and parks (Klingeman et al. 2017;Gugliuzzo et al. 2019a). Several types of traps are commercially available, and homemade inexpensive bottle traps can be easily made and baited with lures (Egonyu and Torto 2018;Owens et al. 2019b). Among most simple models, Baker-type traps can be made by cutting openings (windows) in 2-L soft drink bottles. The ethanol-based lure is usually suspended within the upper part of the bottle and diluted soapy water, or low toxicity antifreeze (propylene glycol), is used as killing and preservative agents in the reservoir created within the bottle.
A single-window bottle trap baited with ethanol-based hand sanitizers effectively attracts and catches X. crassiusculus (Steininger et al. 2015). Moreover, a modified Baker-type trap employing a bottle connector to connect the mouths of two bottles allows easier recovery of trap contents than single bottle designs (Reding et al. 2011). More commonly, Lindgren funnel traps are used for Scolytinae beetles including many invasive Xyleborini (Miller and Duerr 2008). Lindgren traps consist of a variable number of interconnected vertical funnels through which beetles reach a collecting cup attached at the lowest funnel (Lindgren 1983). Lindgren traps traditionally have been black in color since many borers are attracted to dark silhouettes Borden 2006, 2009). However, other Lindgren trap colors (e.g., green, purple) also have been used to trap other woodboring insects like emerald ash borer (Agrilus planipennis Fairmaire), and these alternative colors also will generally collect Xylosandrus spp. as long as the trap is baited with an ethanol lure (Francese et al. 2013;Cavaletto et al. 2020;Marchioro et al. 2020). Indeed, most ethanol-baited traps have been shown suitable for monitoring the presence or absence of ambrosia beetle. However, identifying a most efficient trap type is difficult. As earlier studies have identified forest composition and other factors to affect relative trap efficacy in catching ambrosia beetles (Miller et al. 2018).
Japanese beetle traps were more effective than Lindgren funnel traps, both baited with 95% ethanol, for monitoring X. compactus in Coffea arabica L. plantations and Acacia koa A. Gray nurseries in Hawaii (Burbano et al. 2012). The traditional Japanese beetle trap consists of a yellow-colored funnel with four vertical blades connected to a green plastic collection canister. Miller et al. (2018) caught larger numbers of X. crassiusculus in multiple-funnel traps (baited with ethanol and conophthorin) than in bottle traps in Eastern USA, but the same trials yielded variable results with X. germanus. Also, Gugliuzzo et al. (2019a) showed that red, cross-shaped sticky traps baited with ethanol were effective for monitoring flight activity of X. compactus in infested carob trees (Ceratonia siliqua L.). Werle et al. (2014) found red or opaque corrugated plastic with sticky glue, and ethanol baits were significantly more attractive to multiple ambrosia beetles (including Xylosandrus spp.) than yellow or white, but were not significantly different from other colors like black, blue, brown, clear, gray, green, lavender, orange, or purple. This general preference of Xylosandrus spp. for dark colors over light colors, especially yellow, was confirmed also using colored-panel traps (Cavaletto et al. 2020). Cross-vane sliding and sticky-coated traps (Cross-trap®, Econex, Spain) have been developed by improving the classical traps used for mass trapping of Monochamus galloprovincialis (Olivier), the vector of pine wilt nematode (Bursaphelenchus xylophilus (Steiner & Buhrer) Nickle) (Álvarez et al. 2015). This trap model is currently under evaluation with promising preliminary results for the mass trapping of X. crassiusculus and X. compactus (Gallego, unpublished).
Several other factors can affect catch rates of Xylosandrus spp., including trap placement Klingeman et al. 2017;Marchioro et al. 2020;Miller et al. 2020). Traps placed 0.5 m above the ground capture more X. germanus than those placed at 1.7 or 3.0 m . Similarly, Miller et al. (2020) demonstrated that understory traps capture more X. crassiusculus than canopy traps. Traps hung on trees at a 1.5 m height are effective for X. compactus detection (Burbano et al. 2012). Reding et al. (2015) observed that overwintered X. germanus females occur earlier in proximity to wooded areas adjacent to nurseries and later within nurseries, hypothesizing that forests serve as the reservoir for dispersal of beetles in the spring. Menocal et al. (2018b) showed most X. crassiusculus dispersal in avocado (Persea americana Mill.) orchards occurs near dusk with females initiating flight 30 min before sunset as light intensity and wind speed decrease. Thorough understanding of daily flight periodicity and optimal trap placement should be described for other invasive Xylosandrus spp., to further optimize monitoring and/or mass trapping strategies Ulyshen and Sheehan 2019;Miller et al. 2020).

Trap trees and ethanol-treated tree bolts
Artificial alteration of tree (or selected tissues) attractiveness to modify host location behavior of Xylosandrus spp. is another strategy for beetle monitoring or possible direct management. For example, Ranger et al. (2010Ranger et al. ( , 2012Ranger et al. ( , 2013a enhanced the attractiveness of 4-year-old Magnolia virginiana L. by injecting their stems with a concentrated ethanol solution (e.g., 75 ml of 90% ethanol) or by artificially causing flood stress. Reding et al. (2017) confirmed the effectiveness of such "trap trees" created by ethanol injection (75 ml of 90% ethanol) using a wide range of deciduous trees species. Moreover, Addesso et al. (2019) documented considerably higher attacks in trees injected with high doses of ethanol than in adjacent mildly stressed trees (injected with a lower ethanol dose to simulate common stress occurring in nursery). These studies demonstrate that a wide range of host plant material can potentially be used to attract, monitor and potentially manage exotic ambrosia beetles, including Xylosandrus spp.
Ethanol treatment also increases Xylosandrus beetle attacks on bolts (i.e., tree stem sections). Indeed, Reding and Ranger (2020) observed that tree bolts immersed in 10% ethanol for at least 24 h or filled with 10 ml of 95% ethanol by drilling a chamber in the wood before deployment, both captured more X. germanus than ethanol-baited bottle traps. Similar results were demonstrated for X. crassiusculus; however, drilled bolts filled with ethanol were often less attractive than traps and infused bolts (Reding and Ranger 2020). Furthermore, Rassati et al. (2020) and Cavaletto et al. (2021) demonstrated that the ethanol concentration in tissues of ethanol-soaked bolts affects Xylosandrus spp. host selection and colonization success.
Overall, ethanol-treated bolts, trap logs, and trap trees of selected species can be used to attract and trap Xylosandrus species selectively, i.e., with low non-target catches (Mayfield et al. 2013;Reding et al. 2017;Addesso et al. 2019;Rassati et al. 2020;Reding and Ranger 2020). Manipulation of tree attractiveness to beetles with the aim to develop monitoring strategies or to reduce colonization of the cultivated target crop could emerge as an innovative behavior-based management strategy. For example, selective behavioral modification could be used to reduce beetle infestations on susceptible host plants in nurseries, orchards, and natural tree stands. Another possible use of ethanol-treated bolts could be for insecticide residue studies or evaluation of potential repellents (Brown et al. 2020).

Push-pull strategies
Another semiochemical-based strategy, more recently evaluated for Xylosandrus ambrosia beetles, combines ethanolbaited traps (or trap trees) placed at a distance from the vulnerable host plants and repellents appropriately placed on or in close proximity to the protected crop Addesso et al. 2019;Werle et al. 2019). This management strategy, known as push-pull, manipulates the behavior of insects by combining the two forces: (1) a repellent to 'push' them away from the cultivated host and (2) an attractant to 'pull' them into a trap crop or artificial trap (Cook et al. 2007). However, efforts to use ethanol lures or trap trees (i.e., ethanol-injected or flood-stressed plants) as a 'pull' attractant and synthetic repellents as a 'push' component has had inconsistent results Addesso et al. 2019;Werle et al. 2019;Rivera et al. 2020).
The bark beetle anti-aggregation pheromone, verbenone, has shown promising results as a multi-species repellent, with activity against X. compactus, X. crassiusculus and X. germanus (Burbano et al. 2012;VanDerLaan and Ginzel 2013;Ranger et al. 2014). A 97% reduction in catches was observed for X. germanus when ethanol-baited traps also contained verbenone dispensers releasing 50 mg/day ). Attraction to ethanol-baited traps is also significantly reduced for both X. germanus and X. compactus with the addition of verbenone dispensers releasing 2 mg/day (Burbano et al. 2012;Ranger et al. 2014). Ethanolinjected trap trees had 85% fewer ambrosia beetle attacks when treated with verbenone dispensers compared to trap trees without them, demonstrating the effective repellence of the verbenone . A similar reduction of tree infestation was documented with Xyleborus grabratus Eichhoff after treating host trees or bolts with verbenone dispensers Martini et al. 2019), indicating the broad-spectrum effectiveness of this repellent against ambrosia beetles. Agnello et al. (2017) observed evidence of reduced X. germanus attacks with verbenone in plots that were also treated with permethrin during management trials in apple (Malus spp.) orchards. Moreover, other volatile organic compounds (VOCs), such as limonene and terpinolene, have been shown to reduce attraction of X. compactus and X. germanus to ethanol lures (Burbano et al. 2012;Ranger et al. 2014).
Although both ethanol and verbenone have been widely evaluated as attractants and repellents for Xylosandrus species, respectively, their combined use for push-pull beetle management requires further investigation. Werle et al. Therefore, further optimization of the push-pull strategy for Xylosandrus ambrosia beetles is warranted. The main challenge facing recent attempts to develop push-pull against ambrosia beetle species has been relatively weak activity of the repellents tested to date. In cases where efforts were made to understand mechanisms of efficacy, the effective repellent radius of verbenone-releasing dispensers has been quantified as less than 1 m (Byers et al. 2018;Rivera et al. 2020). Accordingly, it would need a high density of these repellent dispensers per treated area for effective management, which would likely be impractical and not economically sustainable. Identification of more potent repellents and/or development of appropriate formulations that could achieve a high deployment density per area of treated crop should be explored. Another biological challenge is the dispersal capabilities of ambrosia beetles. These minute insects can disperse long distances over brief periods (ca. 20-80 m / 24 h; with maximum distances recorded from 200-400 m) without wind assistance (Seo et al. 2017;Owens et al. 2019a). Therefore, a behavioral manipulation strategy will require potent active compounds to thwart colonizing beetles that could arrive from long distances. While push-pull is unlikely to adequately manage beetle populations as a stand-alone treatment, benefits are possible if it is integrated in association with additional preventive measures such as conservation biological control, removal of infected and stressed trees, or application of protective treatments like insecticides. In this framework, future studies should consider repellent release technology, the use of alternative or synergic repellents, the effect of different ethanol-baited trap designs, and the practical applications within different growing environments, e.g., nurseries, tree fruit, and/or forest stands.

Entomopathogenic fungi
Among promising biocontrol agents of ambrosia beetles, commercial strains of Beauveria bassiana (Balsamo) Vuillemin and Metarhizium brunneum Petch (Ascomycota: Hypocreales) have shown high virulence against X. crassiusculus and X. germanus in different laboratory trials (Castrillo et al. 2011(Castrillo et al. , 2013(Castrillo et al. , 2016. Laboratory studies on comparative virulence using direct spray applications to foundresses showed that strains of B. bassiana Natura-lis® (= ATCC 74040) and M. brunneum F52® (= ARSEF 5198) at a dosage of 600 conidia/mm 2 significantly affected the survival and brood production of X. germanus, resulting in ≥ 60% mortality at 6 days post-exposure compared to controls (Castrillo et al. 2011). Laboratory studies with X. crassiusculus showed that this species is even more susceptible to these pathogens with 50% mortality occurring a day earlier at less than half the dosage required for X. germanus (Castrillo et al. 2011(Castrillo et al. , 2013. Moreover, exposure of X. crassiusculus and X. germanus females to beech (Fagus spp.) stems treated with the same biocontrol agents resulted in lower survival rates, fewer galleries, and reduced brood production (Castrillo et al. 2013(Castrillo et al. , 2016. Some of the progeny produced also became infected by fungal conidia produced on female cadavers, further reducing effective brood size (Castrillo et al. 2013). A natural infection of X. compactus by B. bassiana in the absence of field applied formulations was observed by Balakrishnan et al. (1994) on coffee in India. In the same study, spray applications of B. bassiana spores caused 21% infection of surveyed beetles two weeks after application.
Beauveria bassiana and M. anisopliae (Metch) Sorokin also were found among the fungal flora associated with X. germanus females in hazelnut orchards (Tuncer et al. 2017). Isolates of B. bassiana (TR-217) and M. anisopliae (TR-106) were sampled from X. germanus cadavers and tested against the pest in the laboratory. Eighty and 100% mortality was obtained following B. bassiana and M. anisopliae topical applications, respectively, both at 1 × 10 8 conidia ml −1 dosages (Tuncer et al. 2019). Moreover, another entomopathogenic fungus, Isaria fumosorosea Wize (strain TR-78-3) (Ascomycota: Hypocreales), significantly reduced survival of X. compactus in laboratory trials both after direct application to the beetles and indirect treatment of hazelnut branches . Avery et al. (2018) confirmed the susceptibility of X. crassiusculus to B. bassiana, I. fumosorosea and M. brunneum commercial strains, but spore acquisition by treated beetles was significantly higher after exposure to B. bassiana compared to the other two fungi. Mukasa et al. (2019) observed more than 70% mortality of X. compactus in laboratory and field tests with an atoxigenic L-strain of Aspergillus flavus Link (Eurotiales: Trichocomaceae) isolated from Robusta coffee (Coffea canephora Pierre ex A. Froehner) farms in Uganda.
However, the efficacy of these entomopathogenic fungi against Xylosandrus spp. under field conditions remains to be properly assessed though they produce promising results under the laboratory conditions. Preliminary field studies testing B. bassiana and M. brunneum against X. germanus, X. crassiusculus and Cnestus mutilatus (Blandford) on excised beech bolts suggest that further studies are needed to develop economical strategies and to improve the efficient delivery of these microbials in the field (Castrillo et al., unpublished). Given that insect host-fungal pathogen interactions are determined by variables such as fungal species, strain and dosage, insect life stage, physiology and behavior, and environmental conditions (i.e., relative humidity, UV light, temperature, host plant) (Castrillo 2018; Mann and Davis 2020), field tests are also needed to determine optimal application strategies of these control agents (i.e., formulation type, application mode, rate, and frequency based on beetle phenology). Surveys for naturally occurring pathogens and strains associated with these beetles may offer alternatives that could be more virulent to the beetles or more adapted to their natural habitats rather than commercial ones. However, the cost and length of time required to develop (i.e., mass production and formulation studies) and register these naturally associated strains as new commercial products could be a hindrance to their rapid deployment in the field. In contrast, commercially available strains such as B. bassiana GHA and M. brunneum F52 that have been proven to be pathogenic against these beetles offer readily available products compatible with available application technologies. As with other control options, the use of these fungi needs to be integrated with other control strategies required for optimal beetle control under a given habitat or cropping system.

Mycoparasitic fungi and antagonistic bacteria
Xylosandrus ambrosia beetles perform most of their life cycle within host wood, living in nutritional mutualism with Ambrosiella fungi. Therefore, altering the microbial community associated with beetles, through the release of microbial antagonists, could potentially offer an important tool for developing sustainable management strategies. Targeted growth suppression of beetles' mutualistic fungi has been observed with antagonistic agents like mycoparasitic fungi and antagonistic bacteria, which reduce development of beetle progeny (Castrillo et al. 2016;Eatough Jones et al. 2017;Kushiyev et al. 2020).
Among mycoparasitic fungi, several Trichoderma species and strains (Ascomycota: Hypocreales) that exhibit multiple modes of action (mycoparasitism, spatial and nutrient competition, antibiosis by enzymes and metabolites) are mainly used against plant fungal pathogens (Verma et al. 2007;Vinale et al. 2008). However, little has been documented on the use of Trichoderma spp. against ambrosia beetles. The antagonistic activity of various Trichoderma fungi against Ambrosiella mutualists, specific to each Xylosandrus species, was initially observed in in vitro tests and subsequently evaluated in complementary beetle bioassays (Castrillo et al. 2016). Specifically, Trichoderma harzianum Rifai (commercial strain T-22) significantly reduced brood production in X. crassiusculus and X. germanus after foundresses were exposed to treated beech stems (Castrillo et al. 2016). Beetle galleries in T. harzianum-treated beech stems had sparse symbiont growth and many had no or only a small number of eggs present (Castrillo et al. 2016). Laboratory trials suggest that various Trichoderma-based biofungicides can also affect brood production of X. compactus by outcompeting other fungal mutualists within galleries (Gugliuzzo et al., in prep.). Mycoparasitic Trichoderma spp. obtained from soil samples of the Black Sea Region of Turkey, including isolates of T. harzianum, T. hamatum (Bonord.) Bainier, T. asperellum Samuels, Lieckf. & Nirenberg and T. atroviride P. Karsten, Finl. Mögelsvamp, significantly inhibited growth of A. grosmanniae in both Petri dish assays and within X. germanus galleries (Kushiyev et al. 2020). Moreover, a slight reduction of ambrosia beetle attacks was observed by Brown et al. (2019) under field conditions when evaluating  Gugliuzzo et al. (in prep.) observed significantly higher impact of T. atroviride SC1 on brood production of X. compactus compared to other commercial species/strains tested, irrespective of application rate. The SC1 strain was isolated from decaying hazelnut wood, and it is characterized by rapid growth, germination, and wood colonization (Longa et al. 2008).
Among antagonistic bacteria, Bacillus spp. are known for their ability to dominate within an ecological niche/ substrate via production of antimicrobial compounds, and they can be readily formulated as storage resistant spores (Cawoy et al. 2011;Pérez-García et al. 2011). However, their use against ambrosia beetles has been seldomly investigated. Eatough Jones et al. (2017) (Gugliuzzo et al., in prep.).
Future investigations should identify and enumerate naturally occurring microbial species and strains cooccurring with mutualists that may possess antagonistic properties. Following isolation from naturally infested beetle galleries, evaluations with beetle mutualists can be performed in competitive assays in vitro. Selection of naturally occurring strains that exhibit effectiveness against beetle mutualists could lead to identification of novel biological agents with greater efficacy than current commercial formulations. Moreover, the activity and practical application of newly identified antagonists may yield practical tools for host plant systems where they do not naturally occur. However, as for other biological control agents (e.g., entomopathogenic fungi), practical field applications of mycoparasitic fungi and antagonistic bacteria will have to consider some specific microbial traits (e.g., persistence on tree surface, survival in fluctuating moisture conditions, tolerance to ultraviolet light in sunlight, non-pathogenicity to host plant, and ability to colonize infested galleries), which vary according to the species and strain. Future field applications also need to overcome some of these limitations by means of innovative solutions facilitating the microorganism colonization and persistence (e.g., gel formulations, UV protectants, pressurized sprays, etc.).
Although none of the above-described predators or parasitoids have been reported to effectively suppress populations of Xylosandrus spp., there is a general lack of information about their roles in various ecosystems and within their respective plant-host-predator tritrophic interactions and such investigations are needed. For instance, various clerids are attracted by ethanol-baited traps (Miller et al. 2015b), i.e., the kairomones active for ambrosia beetles, and these predators could have value for managing pest Xylosandrus beetles. Most predator species are difficult to breed in the laboratory and this makes their study challenging. Moreover, integration of ecological services, such as beneficial arthropods for management of Xylosandrus spp., requires thorough investigation of potential new associations that may arise between these beetles and predators or parasitoids following invasion into new environments. It would be of particular interest to test whether the diversity of tree species, which is generally known to confer associational resistance to forest insect pests , could be used to strengthen the control of Xylosandrus spp. by their natural enemies, in the framework of conservative biological control.

Insecticides
Currently, chemical control of Xylosandrus ambrosia beetles consists of preventive insecticide applications to host trees. In addition, direct injection of systemic insecticide into tree trunks has shown efficacy against bark beetle attacks in conifers; however, soil applications of systemic insecticides are generally ineffective (Fettig et al. 2013;Reding et al. 2013a). Chemical control targeting Xylosandrus spp. should be strictly timed with beetle flight activity during spring and performed before observation of the initial attacks on plants by overwintering adults, because these treatments are likely to be more effective if used before beetles have successfully tunneled into the wood (Oliver and Mannion 2001;. Gallery abandonment also has been reported following permethrin applications (Schultz et al. 2002;Reding et al. 2013a;Reding and Ranger 2020), suggesting some potential for its use as a rescue treatment post-attack. Although preventive treatments with conventional insecticides can reduce Xylosandrus spp. attacks, near complete protection should not be expected. All insecticide treatments are likely to be more effective when underlying factors like tree stress and associated ethanol production are minimized (Frank et al. 2017).
The pyrethroids bifenthrin, cypermethrin, and permethrin have been shown as among the most effective insecticides against X. crassiusculus and X. germanus, but results were often inconsistent (Frank and Sadof 2011;Reding et al. 2013a;Ranger et al. 2016b;Reding and Ranger 2018;Brown et al. 2020). Mizell and Riddle (2004) obtained greater reduction of X. crassiusculus attacks on attractive hardwood bolts treated with cypermethrin and bifenthrin than bolts treated with permethrin, chlorpyrifos, and esfenvalerate, while acephate, cyfluthrin, endosulfan, fenpropathrin, imidacloprid, and thiamethoxam were ineffective. Chlorpyrifos also was not effective against X. crassiusculus (Frank and Bambara 2009). Xylosandrus germanus attacks were significantly reduced by chlorpyrifos, lambda-cyhalothrin, and permethrin treatments in a single apple tree trial, but similar treatment efficacy could not be repeated in other locations (Agnello et al. 2017). Moreover, chlorpyrifos caused > 80% mortality of all X. compactus stages infesting twigs of flowering dogwood in Florida (Mangold et al. 1977). Abamectin injection at 1.8% concentration at multiple points in the tree trunk was not completely effective against X. compactus infesting carob trees (Leza et al. 2020). Reding et al. (2013a) observed fewest Xylosandrus spp. attacks on highly attractive ethanol-injected trees treated with permethrin but, conversely, no efficacy was observed following substrate drenches or trunk sprays of certain systemic insecticides (chlorantraniliprole, cyantraniliprole, tolfenpyrad and dinotefuran). Moreover, pre-treatment of trees with permethrin can disrupt X. germanus colonization of flood-stressed trees, but this treatment did not completely prevent beetle attacks (Ranger et al. 2016b). Frank and Sadof (2011) suggested repeated permethrin treatments during peak flight periods of ambrosia beetles for greatest efficacy. However, pyrethroid application timing should also involve consideration of residual activity (Brown et al. 2020). In this regard, Reding and Ranger (2018) observed consistently reduced X. germanus attacks up to 31 days post-application in permethrin residual efficacy trials using ethanol-injected Magnolia trees, and Brown et al. (2020) found permethrin residues ≤ 17 days were optimal for preventing X. crassiusculus attacks using ethanol-filled Liriodendron tulipifera L. bolts. In contrast, bifenthrin produced inconsistent results, and its efficacy lasted only about 10 days; consequently, this compound would require more frequent applications (Reding and Ranger 2018).
Tree coverage is thought to affect efficacy of insecticide treatments and focus has been placed on treatment of tree trunks, which are the main targets attacked by X. crassiusculus and X. germanus Ranger et al. 2013a, b;Gallego et al. 2017). However, similar efficacy was observed when permethrin was applied with a precision sprayer directed to the trunk versus by airblast application (Frank and Sadof 2011). The efficacy and residual activity of precision insecticide applications to the tree trunk against wood-boring beetles vary depending on both active ingredient and formulation of the commercial product. Specifically, emulsifiable concentrates (EC) are generally more effective than wettable powders (WP) (Fettig et al. 2013). Moreover, recurrent rainfall events and water deposition on tree trunks through certain irrigation systems can negatively impact residues of some insecticides depending on formulation, particularly for non-systemic insecticides. Insecticides like permethrin are somewhat resistant to irrigation removal after drying (Brown et al. 2020).
Use of broadcast insecticide applications against ambrosia beetles can cause unintentional negative impacts, including reduction of populations of opportunistic predators and parasitoids (Frank and Sadof 2011). In this regard, more selective compounds should be sought for integration into sustainable programs for Xylosandrus spp. management. Among possible alternatives, some botanical formulations can be effective in reducing ambrosia beetle attacks (Ranger et al. 2011b;Reding et al. 2013a) and their application in organic production would represent one of the few available management options. Specifically, certain commercial plant-derived essential oils have been shown to discourage ambrosia beetles from initiating attacks (Ranger et al. 2011b). Therefore, further evaluation of essential oils in other crops, against other potential target species, and in possible combinations with insecticides could offer interesting advances for developing innovative management strategies. Deterrent particle films, such as kaolin clay, also may represent a biorational alternative to insecticides for ambrosia beetle management. While initial investigations suggest that kaolin clay alone does not reduce beetle galleries, treating trees with the combination of kaolin + bifenthrin has shown improved efficacy (Werle et al. 2017).
The wide variation in efficacy among insecticide products and trials likely indicates some insecticides are not effective at all against these beetles, or that insecticide efficacy may be influenced by a variety of other factors like beetle attraction to the host, pest population at the site, environmental conditions, timing and age of residues, etc. Although insecticides may have some utility when used in an integrated pest management context, especially for high-value tree crops undergoing sporadic periods of environmental stress, they may impact non-target organisms (Desneux et al. 2007;Varikou et al. 2019); thus, the first line of defense against Xylosandrus beetles should remain the maintenance of vigorously growing and stress-free plants.

Long-lasting insecticide netting
An alternative to broadcast insecticide sprays for ambrosia beetle management is insecticide-treated nets. Polyester or polyethylene nets are usually infused or surface coated with insecticides (mainly pyrethroids), and the active ingredient is slowly released over the course of several months (Curtis et al. 2006;Marianelli et al. 2019). Scolytinae crawling on or through these nets dislodge insecticide active ingredient residues that can repel or kill the beetle (Skrzecz et al. 2015). These nets are placed around the main tree trunk or stem to protect the areas where beetles normally form galleries (Franjević et al. 2016;Galko et al. 2019;Ranger et al. 2021). Consistent effectiveness of insecticide-treated nets has been demonstrated for management of a wide variety of agricultural pests (Biondi et al. 2015;Marianelli et al. 2019). The nets may be useful in landscape environments with limited numbers of trees, orchards where long-term tree protection is a goal, or for protection of high-value trees in nurseries, but net and deployment labor costs may be impractical in nursery settings with large numbers of low-value trees.
Cypermethrin-treated nets effectively prevented attacks by bark and ambrosia beetles, including X. germanus, on fresh cut logs evaluated in forests in Croatia, Poland and Slovakia (Skrzecz et al. 2015;Franjević et al. 2016;Galko et al. 2019). More recently, Ranger et al. (2020a, b) obtained variable efficacy by testing two types of deltamethrin-treated nets varying by mesh size (a 'standard mesh' netting of 24 holes/cm 2 and a 'fine mesh' netting of 28 holes/cm 2 ) for protecting stems of flood-stressed eastern redbud (Cercis canadensis L.) from ambrosia beetles. Fewer attacks by X. germanus and other ambrosia beetles were observed across different years and/or locations, but inconsistent results were observed in some cases.
Based on the promising results observed in several locations, insecticide-treated netting appears to be a viable tool for incorporation into integrated management programs for Xylosandrus spp. More research is needed to optimize net attributes including further research on mesh size, as well as net color, and possible alternative active ingredients. Furthermore, such nets require evaluation with a broader range of Xylosandrus spp. in multiple settings.

Fungicides
Fungicides that are used against ambrosia beetle pathosystems target the specific mutualistic relationship between beetles and their mutualistic fungi. In some cases, fungicides also may have repellent properties like insecticides (Brown 2018). Adult females initiate egg laying only after the fungal mutualist is established inside infested galleries and subsequent mycelial growth is necessary for development of the brood (Weber and McPherson 1983b;Kinuura 1995). Fungicides can inhibit fungal growth, which could play a key role in integrated pest management for Xylosandrus spp. (Ranger et al. 2016a, b). Several investigations have demonstrated the potential of fungicide applications as a suitable tool for managing ambrosia beetle infestations (Eatough Jones et al. 2017;Addesso et al. 2018;Mayorquin et al. 2018;Brown 2018;Brown et al. 2019;Grosman et al. 2019).
Several fungicides, including chlorothalonil, dimethomorph + mancozeb, and tebuconazole, were shown to effectively inhibit the growth of the mutualist of X. compactus in vitro (Kagezi et al. 2015). Joseph et al. (2002) demonstrated that the systemic fungicide, propiconazole, significantly inhibits growth of an unidentified ambrosia fungus associated with X. compactus in a laboratory investigation. Significant inhibition of the mycelial growth of the X. germanus fungal mutualist was also obtained by Erper et al. (2018) in laboratory trials with azoxystrobin, captan, iprodione, imazalil, prochloraz, thiram and cyprodinil + fludioxonil. Reddy and Verghese (2006) reported that treating grape vines with the fungicide, carbendazim, reduced infestations of X. crassiusculus and addition of the insecticides, dichlorvos and acephate, further extended treatment efficacy. Ranger et al. (2016a, b) documented that pre-treatment of highly susceptible flooded trees with azoxystrobin and potassium phosphite significantly reduced ambrosia beetle galleries in areas dominated by X. germanus (Ohio and Virginia, USA). Also, Brown et al. (2019) observed a reduction of ambrosia beetle attacks in field trials evaluating the fungicides pyraclostrobin + fluxapyroxad and metalaxyl after flood events. Addesso et al. (2018) demonstrated that a preventative fungicide treatment (pyraclostrobin + boscalid) reduced both Phytophthora root rot disease severity and ambrosia beetle attacks following a simulated flood event, but protection against ambrosia beetles was lower in cases where root rot pathogen was simultaneously infecting plants.
Xylosandrus tree attacks also were reduced by pyraclostrobin + boscalid applied to Magnolia sp. trees 7 days before flood stress or trunk sprays of trifloxystrobin + triadimefon to ethanol-injected trees, but webcam recordings of landing behavior indicated pyraclostrobin + boscalid reduced attacks post-landing (possible short-range toxicity or avoidance) and trunk sprays of trifloxystrobin + triadimefon reduced attacks by pre-landing effects (possible longer-range repellence) (Brown 2018).
More research is needed to identify the most appropriate fungicide classes suitable for the control of fungi specifically associated with Xylosandrus spp. or for preventing adult ambrosia beetle tree attacks by repellence. Key factors that need to be considered for the success of fungicide treatments against these xylomycetophagous pests are: fungal specific mode of action, systemic properties, and bark penetrability of the formulation. Bark-penetrating agents require further investigation and, if optimized, could be a simple attribute of formulations that improves treatment efficacy (Eatough Jones and Paine 2018).

Future research agenda
Despite considerable efforts and investments to discover and develop possible management tools for Xylosandrus spp. and the associated microorganisms (Table 1), significant questions remain on virtually all aspects of the spread, development and management of these beetles, as well as on the role of associated fungal pathogens. In some agricultural contexts, it is still unclear which ambrosia beetles transmit their respective pathogens to specific host crops (see for example Laurel wilt in US avocado, Ploetz et al. 2017). Basic information is required on the identity and biology of these fungal vectors, their attraction to their respective hosts, and ability to obtain and transmit the specific fungi, be it mutualistic or not, which cause disease in living hosts. Improved understanding of the host plant × pathogen interactions, disease host range, and application and efficacy of different chemical measures is essential. Improved control measures will rely on advances in all of these areas. Some of these discoveries may help cure disease in existing trees and allow growers and foresters to maintain their current tree inventory remaining in the ground, and this clearly will require thorough understanding of the microbiomes surrounding fungal gardens. To cite one example, functional inhibition of cellulolytic enzyme secreted by fungal symbionts, such as Fusarium spp. cultured by Euwallacea interjectus (Blandford) attacking fig trees, can be an effective target for host decline (wood degradation) associated with ambrosia beetles (Sakai et al. 2020). Targeted alteration of the beetle's microbiomes with mycoparasitic fungi and antagonistic bacteria could be a critical management tool that circumvent dependence on insecticide treatments as the first line of defense. Alternatively, growers might begin replanting new cultivars or rootstocks, which will require elucidation of factors that may promote host tolerance to beetle infestation and symbiont infection.
The role of environmental biodiversity on Xylosandrus populations and individual fitness should also be carefully evaluated. We postulate that more natural and diverse environments are less suitable for these beetle species due to a higher abundance of adult beetle predators and parasitoids, as well as competitors of the beetle nutritional fungi. For example, Rassati et al. (2019a) found that forest habitat structure strongly influenced the composition of the fungal community associated with X. germanus individuals; in particular, individuals collected in old growth forests were associated with a more diverse mycobiome compared to individuals collected in restored forests, likely reflecting the community of fungi present in these two different habitats. Despite that most of these fungi may represent transient associations, some of them can potentially compete with the primary mutualists of beetles, and this competition can be expected to occur more often in more natural and diverse forests than in managed forests.
Xylosandrus management is greatly hampered given the cryptic nature of beetles that complete most of their lifecycle protected within galleries in wood. The control of these pests is mainly performed with the goal of preventing gallery establishment considering the protection of the beetles after successful entry. The economic damage threshold for a vector of a disease that causes tree death and/or can ruin fruit production is extremely low, if it exists at all. Because a single breakthrough discovery for managing Xylosandrus spp. is unlikely, research should address development of sets of management approaches that can be combined in different ways, optimized and validated for use in different geographic areas, climatic conditions, host availability and type of vegetation, e.g., nurseries, plantations, forests, etc. This approach would allow optimization of management for each situation. The foundation of all management options should be the prevention of host tree stress that initiates the process of Xylosandrus attacks. However, this does not fully apply for X. compactus that can also attack and infest healthy trees.
Nevertheless, there are many aspects to be covered in the future research agenda on alternative control strategies that are being implemented for other beetle pests. Among these, early beetle detection, especially in the entry areas of raw wood products, solid wood packaging material, or live plants, should be the first step for preventing the establishment of non-native species (Rabaglia et al. 2019). Detection programs aiming at finding, containing, and eventually eradicating an invasive species, before it become widely established in a new area, should also carefully consider specific risk and hazard rating systems. To this aim, it could be crucial to estimate the suitable areas worldwide for different invasive species under the current climate by means of appositely developed prediction models, as recently investigated for X. compactus and X. crassiusculus (Urvois et al. 2021). Early detection can also be achieved by employing trained detection dogs able to discriminate between beetle infestation and natural forest odors, as demonstrated by Johansson et al. (2019) for the European spruce bark beetle, Ips typographus (L.) (Coleoptera: Curculionidae). Canines (Canis familiaris (L.)) are an effective proactive management tool because they detect pre-symptomatic avocado trees infested by the invasive redbay ambrosia beetle, Xyleborus glabratus Eichhoff (Coleoptera: Curculionidae)  Balakrishnan et al. (1994); Castrillo et al. (2011Castrillo et al. ( , 2013Castrillo et al. ( , 2016 (Mendel et al. 2018). Based on these promising results, future efforts should consider expanding the utility of canine detection to locate chemical volatiles involved in Xylosandrus ambrosia beetle infestations. Classical biological control is currently considered among the most suitable invasive species control strategies (Avtzis et al. 2019;Giorgini et al. 2019;Salas Gervassio et al. 2019). In this case, it could prove effective if host-specific natural enemies of these pests are identified in the areas of Xylosandrus origin. Indeed, an unexpected degree of target host specificity has been recently observed for parasitoids of other Scolytinae species (Yousuf et al. 2021). Moreover, among other biological control agents, not previously mentioned in this review, entomopathogenic nematodes (EPNs) represent another potential tool deserving further investigation. Indeed, there is already evidence that significant mortality of X. germanus is caused by certain EPN isolates under laboratory conditions (Kepenekci et al. 2018). However, EPNs survival potential, active colonization of beetle infested galleries, and pest control effectiveness under field conditions will have to be carefully evaluated.
Another emerging pest management approach is based on gene silencing through RNA interference (RNAi), which inhibits genes that code for proteins necessary for basic cellular function and stress responses in target insects (Mezzetti et al. 2020). The RNAi approach has proven to be effective in inducing rapid mortality of Scolytinae bark beetles through in vivo laboratory bioassays, including the mountain pine beetle, Dendroctonus ponderosae Hopkins, and the southern pine beetle, D. frontalis Zimmerman (Kyre et al. 2019(Kyre et al. , 2020. Therefore, RNAi could have potential for controlling other problematic Scolytinae, like ambrosia beetles in the genus Xylosandrus. Other genomic manipulations, such as CRISPR/Cas9, have been implemented for mutagenesis studies with beetle pests (i.e., Tribolium castaneum (Herbst) and Leptinotarsa decemlineata (Say)) (Adrianos et al. 2018;Gui et al. 2020) and might represent a possible tool for future investigations focusing on ambrosia beetles. These or other genomic approaches could also be focused on beetle mutualistic fungi or employed to make tree hosts resistant or repellent to beetles (e.g., inducing bark surface to produce Xylosandrus deterrent volatiles).
A holistic management plan is needed that incorporates improved understanding of the potential of biological control, cultural management of pathogen transmission, fungicides and insecticides, all within an overall framework of integrated pest management. Finally, all management interventions are most likely to be successful when integrated with a proper plant maintenance aimed at reducing stressassociated ethanol production.