Seasonal Dominance of Exotic Ambrosia Beetles Compared to Native Species within Deciduous and Coniferous Woodlots

Ambrosia beetles (Coleoptera: Curculionidae: Scolytinae) are among the most successful invaders of trees on a global scale. Exotic species can establish large populations within forested habitats and disperse into tree nurseries and orchards with the potential for substantial economic losses. Our objective was to assess the seasonal dominance of exotic ambrosia beetles compared to native species by characterizing their �ight phenology, abundance and species diversity. Weekly sampling using ethanol-baited traps was conducted within deciduous and coniferous woodlots in Ohio, USA from March/April to September/October in 2014, 2015, 2016, and 2019. Over the course of the study, 16 native and 11 exotic species of Scolytinae were identi�ed. No difference was detected in the number of exotic Scolytinae species or their abundance, Shannon’s index (H), and evenness (Eh) between the coniferous vs. deciduous woodlots. On average, initial �ight occurred at 188 degree days (DD) for exotic species compared to 273 DD for native species. Seasonal �ight duration of exotic species averaged 49 days compared to 10 days for native species. Of the 145,882 total Scolytinae captured over the four years, only 622 were native beetles. Captures of exotic Scolytinae were 341-times greater than native species across the four trapping seasons, including captures of the exotic ambrosia beetle Xylosandrus germanus being 450-times greater than the most common native species Xyloborinus politus. These results provide insight into the invasion success of ambrosia beetles and will aid in predicting and monitoring key species.


Introduction
Bark and ambrosia beetles (Coleoptera: Curculionidae: Scolytinae) are wood-boring beetles that excavate brood galleries within woody plant tissue.Although ambrosia beetles are distinguished by xylomycetophagy, or fungus-farming, within these galleries and bark beetles often ingest wood as well as fungus (Kirkendall et al. 2015), we are focusing on the broader Scolytinae to encompass both bark and ambrosia beetles within this paper.Scolytinae in the tribe Xyleborini are among the most successful invaders of novel habitats due their ability to infest a large variety of host tree species, elusive nature, haplodiploidy, and mutualism with nutritional fungal symbionts (Weber and McPherson 1983a;Atkinson et al. 1990; Kirkendall et al. 1993 Hulcr and Stelinski 2017).Many ambrosia beetle introductions are through packing materials, ports of entry, or imported lumber (Haack 2001(Haack , 2006;;Rassati et al. 2016a; Olenici et al. 2022).In addition to being a forestry pest (Gossner et al. 2019;Økland et al. 2011), exotic Xyleborini beetles disperse into nurseries and orchards from adjacent woodlots and infest horticultural trees (Werle et al. 2015;Ranger et al. 2016;Agnello et al. 2017).Adult females bore tunnels and brood chambers in trees and cultivate their nutritional fungal symbiont that serves as a food source for the brood (Weber andMcPherson 1983b, 1984;Werle et al. 2015;Hulcr and Stelinski 2017).
A select few exotic species of ambrosia beetles are problematic as tree pests because attacks can result in high economic loss in ornamental nurseries and orchards due to tree death, branch dieback, and negative impacts on growth and aesthetics (Ranger et al. 2016;Brockerhoff and Liebhold 2017).Xylosandrus germanus (Blandford) has been established in the U.S. for decades after it was detected in the 1930's in New York state (Felt 1932).Xylosandrus germanus has shown rapid population increases in the eastern United States (Rassati et al. 2016b) and Europe over just a few years (Henin and Versteirt 2004;Galko et al. 2018;Gossner et al. 2019;Olenici et al. 2022).Xylosandrus crassiusculus (Motchulsky), which is also well-established in the eastern U.S., was found on peach trees in South Carolina in the 1970's (Anderson 1974) and is also currently established in Europe and other parts of the world (CABI 2021).Anisandrus maiche (Stark) is becoming an increasing concern in the U.S. since its detection in Ohio, Pennsylvania, and West Virginia as early as 2005 (Rabaglia et al. 2009).These three species exhibit a broad range of host trees with additional host species being updated as these pests expand their distribution and host range (Weber and McPherson 1983ab The ability of exotic ambrosia beetles and their fungal mutualists to colonize a wide range of genera in the absence of coevolutionary history likely plays an important role in their invasion success.Different ambrosia beetles may be attracted to or prefer speci c tree species (Gossner et al. 2019;Rassati et al. 2016a), as well as experience limitations with how well their symbiotic fungi grow within different host species (Castrillo et al. 2012).Furthermore, ight patterns and establishment of prominent Scolytinae pests may be in uenced by beetles' attraction to ethanol emitted by stressed trees, particularly trees that are ood stressed (Ranger et al. 2015) or freeze stressed (La Spina et al. 2013); attacks are more prevalent on stressed trees emitting ethanol due to higher rates of ethanol accumulation in tree tissues (Ranger et al. 2015;Ruzzier et al. 2021).A monoterpene, α-pinene, common in pine trees, has been shown to attract bark beetle species, and to attract some species of ambrosia beetle species and repel others when combined with ethanol (Miller and Rabaglia 2009).The chemical pro le of trees can be complex and may vary between different stands of trees (Olenici et al. 2022), ultimately altering which Scolytinae species are better able to thrive in coniferous and deciduous woodlots.Although native species of ambrosia beetles are found throughout North America, they are more geographically isolated by trends in forest vegetation compared to exotic species.Ambrosia beetle species are less selective in host tree preference and climate limitations than their bark beetle counterparts (Rassati et al. 2016b).Although still in greater abundance in broadleaf forests, X. germanus thrived in pine forests in association with high management intensity (Gossner et al. 2019).Trap captures of X. germanus also occurred in conifer stands in Romania but were higher in broadleaf and mixed stands with a majority (> 50%) of beech trees (Olenici et al. 2022).Thus, the ability of exotic ambrosia beetles and their fungal symbionts to utilize a wide range of host tree species could contribute to a few exotic species increasing to much greater populations than native species (Rassati et al. 2016b).
The seasonal ight phenology of exotic ambrosia beetles compared to native species may also contribute to their invasion success.The in uence of temperature on initiating earlier spring ight after overwintering and longer duration of ight activity of exotic ambrosia beetles compared to native species could allow exotic species to be prime invaders, either by earlier access to more vulnerable stressed trees, increasing range and rates of infestation, or potential for additional generations.Exotic species such as Xylosandrus saxesenii (Ratzeburg) and X. germanus have been shown to have ight in early April and as late as October in Missouri (Reed and Muzika 2010) and other exotic species, such as Xylosandrus compactus (Eichhoff), have been shown to have multiple peaks of ight activity occurring as early as April and as late as September in Italy (Gugliuzzo et  To better understand why exotic ambrosia beetle species are prime invaders and able to obtain higher populations than native species, we monitored weekly captures of ambrosia beetles in coniferous and deciduous stands to evaluate seasonal ight patterns.We hypothesized exotic ambrosia beetles would dominate in both coniferous and deciduous habitats; that exotic ambrosia beetles would require fewer degree days (DD) until rst trap detection and exhibit a longer duration of ight activity compared to native species.Ultimately, by better understanding when pest species rst begin ying and are at their peak ight, management of these species will become more attainable.

Area of study and experimental design
Experiments were set up over 4 years (2014,2015,2016 and 2019) at six locations in Wayne County, Ohio (Table 1).Three locations were coniferous woodlots comprised of Pinus spp., three were deciduous woodlots comprised of Acer spp., Quercus spp.and other hardwoods (Table 1).All woodlots were well-established with trees planted 30-50 years ago (Fig. 1A-C).Woodlots were not maintained with re or pesticides, but had disturbance from collection of rewood and selective logging every 5-10 years (Stuart Courtney, personal communication).Within each woodlot, four traps were deployed (n = 24 each year) in March/April and checked weekly through September/November (Table 1).Traps were placed 10 m apart in a diamond con guration with each trap placed at a cardinal direction (Fig. 1D).The collected bottles with beetles were brought back to the laboratory where beetles were sorted and identi ed to species.We focused on damaging pests and therefore excluded Hypothenemus spp.from our analyses (Monterrosa et al. 2022).

Data Analyses
Weather data was downloaded from "Daymet: Daily Surface Weather Data on a 1-km Grid for North America, Version 4 R1" for each year using the longitude and latitude for each location (Thornton et al. 2022).We calculated cumulative degree days (DD) starting on January 1 by adding the maximum and minimum temperature each day, dividing the sum by two and subtracting the base temperature of 10°C.Longitude and latitude coordinates were determined using Google Earth Pro (Version 7.3.6.934;2022 Google LLC.).
Mean beetle capture was the average of the four traps per site.Total trap captures were averaged for each site within each year.Number of captured beetles for each species was used to calculate a diversity index (H, Shannon's Index) for each site within each year (n = 24).Shannon's Index (H) for diversity is as follows: where p i is the proportion of individuals for i-th species within a community (Shannon 1948; Shannon and Weaver 1949).
Evenness (E h ) was calculated from Shannon's Index and total number of species captured.Evenness (E h ) was calculated as: where H is Shannon's Index calculated in (Eq. 1) and k is the number of total species within a community (Shannon 1948; Shannon and Weaver 1949).
To compare ight patterns among species, we used a negative binomial generalized linear model (package 'MASS') to account for non-normal count data for DD for initial capture, duration of capture and number of captured beetles in R Statistical Software (R Core Team 2021).Our model included year, site, stand type, and species with stand type being excluded when considering beetles captured in either coniferous or deciduous woodlots.We tested Shannon's Index (H) and evenness (E h ) using a generalized linear model.The model was initially tested with year, site, stand type and all their interactions, but because no interactions were signi cant, these were pooled.The package 'agricolae' was used for Tukey's HSD mean comparison test with = 0.05.

Seasonal Flight Phenology of Native and Exotic Scolytinae
Initial ight, measured by cumulative degree days (DD), was not affected by year (χ 2 = 0.29; df = 1; P = 0.5891) or stand type (χ 2 = 3.77; df = 1; P = 0.0522), but was affected by site (χ 2 = 16.46;df = 2; P = 0.0003) and species (χ 2 = 485.72;df = 26; P < 0.0001; Table 2).Native species had a later average initial ight at 273 DD compared to only 188 DD for exotic species (P < 0.05).One of the earliest beetles to begin ying in spring was X. germanus at 41.1 DD, but with slightly earlier initial ight each year; mid-May in 2014 and 2015 and the last week of April in 2016 and 2019, respectively (Fig. 2), corresponding to a slight increase in degree days each year.Interestingly, A. maiche and X. crassiusculus were later yers.X. crassiusculus typically began ying mid-to late-May (Fig. 3), around 423.5 DD (Table 2), but the variability was relatively high due to some location-years having low overall captures.A. maiche initial ight occurred consistently within the rst week of June (Fig. 4) or 309.9 DD (Table 2).
628.31; df = 26; P < 0.0001; Table 2), but not by stand type (χ 2 = 0.59; df = 1; P = 0.4418).Exotic Scolytinae species had an average ght duration of 49 d compared to only 10 d ight duration for native beetles (P < 0.05).Anisandrus sayi and X. politus had the longest ight duration for native species at 55.4 and 33.3 d respectively.In comparison, exotic species X. germanus had the longest ight duration of 167.8 d, followed by X. saxesenii at 86.2 d and A. maiche at 81.5 d (P < 0.05; Table 2).
Generally, two ight peaks were observed for X. germanus-late April to late May or 50-360 DD and then a smaller peak in mid-July or 680-970 DD (Fig. 5B).X. crassiusculus peak ight is di cult to pin-point because overall numbers were quite low in 2014 and 2015, but the largest ight peak was 550-710 DD and smaller peaks occurred as early as 120 DD and as late as 1250 DD (Fig. 5C; Fig. 3).Captures of X. crassiusculus in 2016 were much higher and a peak can be clearly observed at the end of June.In comparison, peak ight occurs in mid-August in 2019, but this could be explained by a warmer spring in 2016 compared to 2019.Peak ight duration of A. maiche was much longer -from approximately 250-1100 DD (Fig. 5D), occurring mid-July in 2014, but with two peak ights observed in 2015 and 2016, from mid-or late-June and again in the rst week of August.An earlier peak ight was observed in 2019 during the last week in June, with no large second peak (Fig. 4).
Overall ight patterns of all native ambrosia beetle species combined showed low numbers of captured beetles, typically under 1 beetle per day with one peak of almost 4 beetles per day (Fig. 5A).The two most abundant native species in both coniferous and deciduous woodlots (P < 0.05), Xyleborinus politus and Anisandrus sayi, represented 3.0 and 1.9 cumulative individuals captured per trap per year, respectively; whereas, the two most abundant exotic species in the coniferous and deciduous habitats, X. germanus and A. maiche, represented 1347.8 and 132.2 cumulative individuals per trap per year, respectively.X. germanus had signi cantly higher cumulative captures per site per year than any other Scolytinae species in either deciduous or coniferous woodlots (P < 0.0001); with up to 238 beetles per day (Fig. 5B).Across the four trapping years, X. germanus represented a mean of 82.8% and 82.4% of the cumulative beetle captures in the coniferous and deciduous woodlots, respectively (Fig. 6) but were inconsistently higher in deciduous and coniferous woodlots across years (Fig. 2).Number of captured X. crassiusculus were up to 59 beetles per day with similar captures in coniferous and deciduous woodlots, albeit, with very low overall numbers, in 2014 and 2015 and slightly more captures in deciduous woodlots in 2016 and in coniferous woodlots in 2019.It is notable that the Metz location captured zero X. crassiusculus in 2014, 0.25 beetles per trap in 2015 (and much later in the season) and up to just 1 beetle per trap in 2016, then 20 beetles per trap in 2019.However, these numbers were similar in both the coniferous and deciduous Metz woodlots indicating X. crassiusculus likely was introduced into this area sometime between around 2015 and was initially slow to increase in population.Anisandrus maiche represented a mean of 14.6% and 14.3% of the cumulative beetle captures in the coniferous and deciduous woodlots, respectively (Fig. 6).Deciduous woodlots yielded slightly higher captures of A. maiche, with a few exceptions in 2015 and 2019 (Table 2; Fig. 4).

Scolytinae Community Diversity
Over the course of the study, 16 native species and 11 exotic species of Scolytinae were captured (Table 2).Eight of the 27 species were bark beetles of which seven species were native and one species (Hylastes opacus E.) was exotic (Table 2).Of the 19 species of ambrosia beetles, nine were native and 10 were exotic (Table 2).

Discussion
Our aim was to analyze ight patterns of native and established exotic ambrosia beetles in coniferous and deciduous woodlots to better understand ight phenology.Identifying initial and peak ight patterns and woodlot preferences allows for better monitoring and trapping to mitigate attacks.As we hypothesized, exotic Scolytinae exhibited earlier ight, longer ight duration and much higher total beetle captures in both coniferous and deciduous woodlots.

Seasonal Flight Phenology of Native and Exotic Scolytinae
The early initial ight and longer ight duration exhibited by exotic beetles in our study is likely another contributing characteristic to the invasion success of exotic Scolytinae.Initial ight for exotic species occurred around 188 DD compared to 273 DD for native species, which may allow for exotic species to better target stressed trees with early spring ood events as well as select for damaged trees if a spring frost occurs (La Spina et al. 2013;Ranger et al. 2015).Flight duration of native species was much shorter, about 10 days, compared to 49 days of ight for exotic species.The longer ight duration may allow exotic species to thrive because they have greater opportunity to infest further distances, attack at greater rates, have better selection of vulnerable trees and more adaptability timing their ight to changes in climate patterns.
Native ambrosia beetles had low peak captures, up to 4 beetles per day, around 40-170 DD with a second, smaller peak around 690-790 DD.In comparison, X. germanus ight had two distinct peaks, a larger initial peak from 50-350 DD with up to 235 beetles captured per day and a second smaller peak with up to 89 beetles captured per day from 680-970 DD, indicating potential for a second generation as has been observed as early as March/April in warmer climates within the US ( ).Fight patterns of A. maiche sustained high numbers throughout a longer period -from 250-1100 DD.This sustained duration of comparatively high ight activity also makes it di cult to determine if later ight could be the result of a second generation of A. maiche or a long period of emergence from overwintering.Furthermore, because A. maiche is a more recent introduction, fewer studies have shown ight patterns of this species, so a second generation could also be possible in warmer climates.Compared to native species, X. germanus had an earlier and longer seasonal ight pattern and A. maiche had a later and longer seasonal ight pattern, which may help explain why competition is minimal and why these two species are so successful in Ohio.Although there is overlap in ight activity with A. maiche and X. crassiusculus, the windows of ight do seem to allow for some strati cation between these species, allowing them to coexist rather than compete.
Abundance Native vs. Exotic number of native species captured (16) was higher than that of exotic species (11), the difference of actual beetle captures over the four years of this study was alarmingly much higher for exotic (145,260) than native beetles (622).
The proportion of exotic beetles captured in our study was higher (99.6%) compared to other studies which have shown as few as 60% adventive beetles captured (Gandhi et al. 2010), 86.9% exotic beetles captured (Reed and Muzika 2010) and up to 89.3-97.2%adventive beetles captured (Miller et al. 2015).The differences could be due to different combinations of trapping lures in other studies (compared to just ethanol in our study) or due to the longer trapping duration in our study.
Furthermore, climate may affect the range of Scolytinae; temperature has been shown to limit the geographic range of nonnative ambrosia beetle species within the US and rainfall is more limiting to ambrosia beetles than bark beetles when colonizing new areas (Rassati et al. 2016b).It is likely exotic beetles captured in our study are already established and have a lower species replacement component, meaning they are likely to be found across a larger geographic range, have lower host tree preference, and may be able to overcome changes in temperature or precipitation (Rassati et al. 2016a(Rassati et al. , 2016b)).
This, coupled with their high numbers can cause severe damage and economic loss to trees and exhibit their increasing potential as a severe tree pest.Conversely, native bark beetle species have been shown to be more harmful than neonative or alien species, and are more limited in establishing in a non-native range by both biotic and abiotic conditions their environment (Forgione et al. 2022).However, if the environment is adequate, ambrosia beetles are less selective of their host compared to bark beetles (Rassati et al. 2016b).Indeed, studies have shown regional differences in ambrosia beetle species composition (Reding et

Ambrosia beetle community
There was no difference in community diversity (Shannon's Index) nor was there a difference in evenness between exotic and native species.However, both Shannon's Index and evenness decreased each year and Shannon's Index varied by site.
This indicates that, although the community diversity and evenness is decreasing each year-likely due to increasing populations of a few species, there is no indication of competition or displacement of native species from the highly successful exotic species.Indeed, although infrequently (i.e., 3

Conclusion
Our study documented initial ight, duration of ight, number of captured beetles, community diversity and evenness to be similar in both coniferous and deciduous woodlots.By comparing exotic species to native species, we were able to show minimal competition, and strati cation of ight phenology that allow these species to coexist.We've shown that initial ight begins mid-to late-April and peaks in May and July for X. germanus and in June for A. maiche in Ohio, which allows us to better predict, detect and manage for these pest to mitigate tree damage in nurseries and orchards.We've also shown that exotic Scolytinae y for a longer duration, which in conjunction with earlier ight phenology is at least partially responsible for their thriving populations compared to native species.

Declarations Figures
windspeed, and therefore, greater ight ability(Pasek 1988;Mahroof et al. 2010).Werle et al. (2015) attributed differences in ight between adjacent woodlots and nurseries to woodlots having less wind and allowing for easier ight.Alternatively, availability of breeding substrate(Gossner et al. 2019) could have been similar within sites in our study and produced similar progeny and therefore no statistical differences in trap captures.Strati cation of ight patterns over a season may alter the Scolytinae community by allowing species to coexist, but at different times rather than compete for the same resources, especially if stressed trees are of a lower abundance in any given year.Another hypothesis may be that host trees are not as e cient at defending against exotic Scolytinae -that there is simply a lack of or delay of host tree defenses against these beetles(Forgione et al. 2022).Additionally, the resource allocation hypothesis poses that exotic beetles may be better able to shift their energy into reproduction rather than enemy defenses(Blossey and Nötzold 1995; Doorduin and Vrieling 2011; Forgione et al. 2022) and Darwin's naturalization hypothesis suggests less competition for exotic species in the U.S. compared to their native range (Darwin 1895, Forgione et al. 2022).

Figure
Figure 3

Figure 5 First
Figure 5

Table 1
Summary of trapping sites and their tree species composition in order of decreasing predominance (Ranger et al. 2010;Reding et al. 2010nt to last collection date for all sites within a given year were as follows: 21 Apr.2014-22Sept.2014;26Apr.2015-21Sept.2015(exceptMetz,whichbegan 26 Apr.2015); 8 Mar.2016-27 Sept. 2016; 21 Mar.2019-4Nov.2019.Traps and MonitoringTo trap ambrosia beetles, we used bottle traps baited with ethanol lures as described byRanger et al. (2010).Traps were designed from a 1 L plastic bottle hung upside down with two windows cut out (11 cm x 7 cm) and the ethanol lure suspended within the open bottle with a twist tie(Ranger et al. 2010;Reding et al. 2010).A 0.5 L bottle with ~ 28 mL killing solution (1:1 propylene glycol: tap water; Sierra Antifreeze/Coolant; Old World Industries, Inc., Northbrook, Illinois) was connected to the 1 L bottle via a Tornado Tube® (Steve Spangler Science, Englewood, Colorado).Traps were suspended vertically 0.6 m above ground level by securing the inverted 1 L bottle to a metal rod.We used low-release lures in 2014 and 2019 (10 ml of 95% ethanol released at 65 mg/d at 30°C; AgBio, Westminster, Colorado, USA) and made low release lures in other years by adding 8.5mm of 95% ethanol and heat sealing.Lures were replaced if damaged, leaking or the ethanol was depleted.

Table 3
Species diversity of Scolytinae captured within coniferous and deciduous woodlots in Ohio, USA a Different lowercase letters within years and rows indicate signi cant differences in total number of native vs. exotic Scolytinae species using a general linear model and lsmeans ( = 0.05; df = 1).
Hudson and Mizell 1999; Oliver and Mannion 2001; Reding et al. 2010; Werle et al. 2012; Werle et al. 2015; Viloria et al 2021).Peak ight of X. crassiusculus was observed much later, at 550-710 DD, with up to 59 beetles captured per day.Similar to our study, only one generation was observed for X. crassiusculus in Ohio and Virginia (Reding et al. 2010), but a second peak, and potentially a second generation occurred in Tennessee (Reding et al. 2010; Viloria et al. 2021 (Colombari et 2022)022.2018)-likelyindicative of their preferred climates rather than host type or host range.Our study quanti ed much higher numbers of exotic Scolytinae -341-times more total captured beetles per species than native beetles.Furthermore, we were able to illustrate how prominent X. germanus has become, with 1348 beetles captured years, A. maiche has increased to alarming numbers; our study showed an average capture of 132 beetles each year with a minimum capture of 37 beetles at one location in 2014 and a maximum of 1642 at one location in 2016.Although it is not currently found in large numbers in Europe, A. maiche has been reported in Ukraine in 2007 and Russia in 2009, and more recently in the Veneto Region of Italy in 2021(Colombari et al. 2022).Based on DNA sequencing, the specimen collected is a closer match to the population in the Eastern U.S. and Canada than the specimen found in Ukraine, indicating a potential introduction from shipments from the US rather than expansion from Eastern Europe(Colombari et 2022).However, as show in this current study, its population growth has occurred much more rapidly in North America than it has in Europe and should be continued to be monitored to prevent further population growth.
per trap and show that A. maiche populations have increased in recent years, becoming a growing concern as a tree pest.As of 2013 and 2014, 19 and 9 specimens of A. maiche were collected in three counties within Ohio (Reding et al. 2015) but within a few (Zach et al. 2001;Økland et al. 2011;cies have been shown to emerge from the same galleries (Oliver and Mannion 2001).There are numerous studies on alien species colonization in new habitats(Zach et al. 2001;Økland et al. 2011; Rassati et al. (Elton 1958al.2022.2022) and evidence of increasing rates of alien introductions (Kirkendall and Faccoli 2010).Elton's diversity-invasion hypothesis predicts species to be less successful colonizers when introduced to more diverse communities (ie, deciduous woodlots) because there are fewer unoccupied niches available(Elton 1958).Although initial ight, duration of ight, total captured beetles and ambrosia community diversity and evenness were similar in both coniferous and deciduous woodlots in our study, other studies have shown a more diverse woodlot to have greater ambrosia beetles because of increased availability of niches, more shrubbery and canopy strati cation and potential for reduced