Damage Levels from Arthropod Herbivores on Lonicera Maackii Suggest Enemy Release in Its Introduced Range

Biol Invasions (2012) 14:863–873 DOI 10.1007/s10530-011-0123-7

ORIGINAL PAPER

Damage levels from arthropod herbivores on Lonicera maackii suggest enemy release in its introduced range

Deah Lieurance • Don Cipollini

Received: 21 March 2011 / Accepted: 17 October 2011 / Published online: 1 November 2011 Ó Springer Science+Business Media B.V. 2011

Abstract The ‘enemy release hypothesis’ argues indicate that levels of herbivory experienced by when a species is introduced to a novel habitat, release L. maackii are relatively consistent across sites, vary from regulation by herbivores results in increased slightly with habitat and branch identity, but are likely vigor, abundance, and distribution. The invasive Asian too low to impact fitness of shrubs. These findings shrub Lonicera maackii appears to benefit from an indicate that low amounts of arthropod herbivory absence of arthropod herbivores in North America. occur for L. maackii across its introduced range, which We assessed the incidence, amount, and type of may contribute to its invasive success. herbivory occurring on L. maackii in forest edge and interior habitats and investigated differences in timing Keywords Herbivory Á Lonicera maackii Á Amur of damage. In October 2008, leaves were sampled honeysuckle Á Enemy release hypothesis from shrubs in forest interior and edge habitat from 8 sites in Ohio. In 2009, sampling was repeated at 3 sites in spring, summer, and fall with a distinction made Introduction between long and short branches. Leaf area removed averaged 1.83% across the 8 populations in 2008 and There are many studies that have tried to determine 3.09% across the 3 populations in 2009, with forest what characteristics (i.e. efficient resource utilization, edge plants receiving slightly more damage than forest high relative growth rates, and phenotypic plasticity) interior plants in 2008. Additionally, long shoots make an invasive species successful (e.g. Bazzaz received more damage than short shoots in 2009. 1986; Feng et al. 2007; Osunkoya et al. 2010a, b). One Damage incidence was also higher in the edge habitat prominent hypothesis explaining the success of inva- and on long shoots compared to short shoots. As sive plants in their new habitats is the ‘enemy release measured in 2009, damage accumulated steadily hypothesis’ (ERH) which suggests that when a plant throughout the season. Chewing was the most invader is introduced to a new habitat it will experi- prevalent type of damage (76. 8%) and low level of ence a reduction in regulation by specialist herbivores pathogen infection was observed (4.81%). Results or pathogens that are present in the native range, which can result in an increase in distribution and abundance (Keane and Crawley 2002). Research has suggested D. Lieurance (&) Á D. Cipollini that enemy release has provided a competitive Department of Biological Sciences and Environmental advantage to such invaders as Norway maple (Acer Sciences PhD Program, Wright State University, 3640 Colonel Glenn Highway, Dayton, OH 4543, USA platanoides) in northeastern United States, butter- e-mail: [email protected] fly bush (Buddleja davidii) in Europe, soapbush 123 864 D. Lieurance, D. Cipollini

(Clidemia hirta) in Hawaii, and whitetop (Lepidium of L. maackii allows the plant to capitalize on such draba) in western United States (Adams et al. 2009; conditions presented in edge habitat including Cincotta et al. 2009; Cripps et al. 2006; DeWalt et al. increased light availability (Hutchinson and Vankat 2004; Ebeling et al. 2008; Morrison and Mauck 2007). 1997; Luken et al. 1995b). However, edge habitats can Results supporting this hypothesis commonly report exhibit increased arthropod abundance and diversity increased growth and reproduction of plants in the relative to understory habitats, and presumably invasive range compared to the native range (e.g. increased herbivore incidence and damage potential Colautti et al. 2004; Ebeling et al. 2008). (Barbosa et al. 2005; Ozanne et al. 2000). The extent to Lonicera maackii (Rupr.) Maxim (Amur honey- which the ERH may explain the success of L. maackii suckle), is one of the most important and prominent in North America has not been determined, and a first invasive plant species in the Midwestern United step requires estimates of herbivory levels on this States. Lonicera maackii, a deciduous woody shrub species in its invasive range. Anecdotal observations native to China, Japan, Korea, and southeastern indicate that L. maackii escapes meaningful levels of Russia, was originally introduced in the late 1880s herbivore and pathogen attack in its invasive range for habitat improvement, erosion control, and horti- (D. Lieurance and D. Cipollini, personal observa- cultural landscaping (Luken and Thieret 1996). Traits tions). Trisel (1997) reported damage amounts from a that L. maackii shares with many other invasive plants combination of frost, deer browse, insect herbivory, include extended leaf phenology, rapid aboveground and drought on L. maackii at Miami University in growth rates, high fecundity, broad phenotypic plas- Oxford, Ohio from 1992 to 1994 that were well below ticity, and tolerance to a variety of habitats (Lieurance various native woody species. Damage levels have 2004; Luken et al. 1995a, b; 1997a, b; Trisel 1997; never been quantified on this species in its native McEwan et al. 2009b). Additionally, L. maackii range. However, observations of L. maackii in its responds to increases in light availability with a native versus introduced range indicate that the shrub plastic response in branch architecture, producing two grows more vigorously and abundantly in its intro- distinct branch types, ‘long’ and ‘short’ (Luken et al. duced range (Luken and Thieret 1996), while being 1995a, b). Long branches extend the crown height in much rarer in its native range in China (J. Ding, response to increased light and are presumably cheap personal communication), and endangered in Japan in construction, poorly defended, and expendable, (http://www.biodic.go.jp/english/rdb/red_plants.csv). while short branches have leaves that are thicker, In an effort to characterize the natural enemies of tougher, and more resistant to herbivory (Ballare Lonicera spp. in their native range and to identify 2009; Guerra et al. 2010; Herms and Mattson 1992; possible biocontrol candidates, 44 insect species and Ishii and Shoko 2010). 21 fungal species were reported on the genus in China Current land use patterns have resulted in disturbed including 4 arthropod species specifically attacking habitats, fragmentation of forests in rural areas, and L. maackii (Zheng et al. 2006). In contrast, observa- creation of suburban woodlots all of which provide tions of L. maackii growing alongside the native appropriate habitat for L. maackii to spread across the Lonicera reticulata in Ohio revealed native vines landscape. In Ohio and Kentucky, L. maackii shrubs suffering substantial early season damage from the comprise up to 50% of the understory species com- specialist honeysuckle sawfly (Zarea inflata)while position in some small woodlots, and near monocul- exotic shrubs remained undamaged by this insect tures along the edges of old fields and roadsides (D. Lieurance, personal observation). (Hartman and McCarthy 2008; Medley 1997; Enemy escape can result from a combination of this Pennington et al. 2010; Watling and Orrock 2010). apparent lack of specialist herbivores and resistance to As anthropogenic habitat fragmentation increases, the generalists that may be present in the invasive range. creation of edge habitat insures the persistence and Resistance strategies employed by plants include the spread of L. maackii in the landscape. The conditions production of secondary metabolites used to defend presented to organisms by edge habitats are often very against herbivores. Previous research has identified different than those of the understory with significant several phenolic metabolites in leaves of L. maackii, microenvironmental differences in temperature, light, including chlorogenic acid, several flavones and their and moisture (Matlack 1993). The adaptive plasticity glycoside derivatives, and iridoids (Cipollini et al. 123 Damage levels from arthropod herbivores 865

2008a, D. Bowers, personal communication). These short branches; and (5) because the early season compounds may act as deterrents to generalist herbiv- phenology of L. maackii provides an abundant food ory and be responsible for some of the allelopathic source for early spring herbivores, the majority of effects attributed to L. maackii (Cipollini et al. 2008a, damage would occur in the spring and then increase b; Dorning and Cipollini 2006; Cipollini and Dorning through time. Results of this study provide informa- 2008; McEwan et al. 2009a; Trisel 1997). It has been tion on the autecology of L. maackii in its introduced postulated that secondary metabolites effective in range and indicate whether enemy release may herbivore defense are a result of coevolution with contribute to the success of this woody invasive. herbivores present in their native habitat (Becerra 2003; Ehrlich and Raven 1964) and for non-native plants the separation from coevolved herbivores is Materials and methods integral to ERH (Keane and Crawley 2002). However these metabolites may be novel and/or effective Sample collection 2008 against generalist herbivores that may be present in the introduced range. Samples were collected in October, 2008 from eight Sometimes a non-native invader can experience sites in Ohio (Table 1). Sites were selected throughout some herbivory from either generalists or from the geographic region to cover 7 counties in south- specialists that may attack closely related native western and central Ohio. The surrounding area plants. However, enemy release can still occur if around the sites were either a rural-suburban interface levels of damage are insufficient to alter the perfor- (Shawnee Prairie Preserve, Germantown Metropark, mance of the plant. Moreover, damage rates experi- Sharon Woods, Taylorsville Metropark), urban green enced by any plant can be influenced by differences in space (Franklin Park Conservatory, Wright State herbivore diversity and exposure between edge habitat Woods), or less disturbed natural areas (Hueston and interior forest (Barbosa et al. 2005; Ozanne et al. Woods State Park, Glen Helen Nature Preserve). 2000), and differences in forage quality between Mature shrubs were sampled in both forest edge and branch type (Cornelissen et al. 1997, 2008; Price interior habitats at each site. Shrubs were considered 1991). Additionally, the timing of herbivory can be to be in the interior if they fell at least 15 meters from influenced by the phenology of the plant. This is true the forest edge and there was no clear evidence of a for L. maackii as it leafs out before majority of native gap in the canopy. Edge shrubs were selected to be on competitors making it an available food source for the physical interface between forest and open field. early season generalist herbivores (McEwan et al. Twenty mature shrubs were selected from each 2009b; Trisel 1997). habitat. Ten branches with several leaves attached We conducted a study to quantify the incidence, were each taken from the exterior (sun leaves) and the amount, and type of herbivory occurring on L. maackii interior (shade leaves) of each plant in the edge across several populations over a two-year period in habitat. Ten total branches were taken from each plant Ohio. We also investigated the role of habitat location, in the interior habitats. Two leaves from each branch timing of removal, and branch preferences by herbi- were randomly selected for assessment for a total of 40 vores. We predicted that (1) overall, damage by leaves from each edge shrub and 20 leaves from each arthropod herbivores on L. maackii would be minimal; interior shrub. Preliminary analysis of herbivore (2) despite the geographic spread of populations damage indicated no significant differences in damage sampled, there would be no difference in the amount levels between sun and shade leaves of edge plants of damage across sites; (3) due to differences in (F1,3163 = 0.59, P = 0.4406) and for this reason all arthropod diversity and exposure between edge and edge leaves were pooled for statistical analysis. interior habitats, edge shrubs would have more leaf A visual assessment of damage was conducted for area removed by arthropod herbivores as well as a all leaves. Damage was scored on a percentage leaf higher incidence of herbivory than interior shrubs; (4) area basis as 1, 2, 5, and then in increments of 5% to a because of possible palatability differences between maximum 100% for leaf area lost or infected. Based the two branch morphologies, leaves on long branches on previous studies of leaf herbivory, categories for would suffer more damage than those growing on leaf damage were established as follows: (1) chewed 123 866 D. Lieurance, D. Cipollini

Table 1 List of sites selected in central and southwestern Ohio for 2008 study Site Symbol County GPS coordinates

Franklin Park Conservatory CB Franklin N 39.96659 W 82.95121 Shawnee Prairie Preserve DK Darke N 40.09863 W 84.64526 Germantown MetroPark GR Montgomery N 39.64136 W 84.40111 Hueston Woods State Park HW Butler N 39.59247 W 84.75365 Sharon Woods SW Hamilton N 39.27597 W 84.40165 Taylorsville MetroPark TV Montgomery N 39.87313 W 84.16598 Wright State University Woods WS Greene N 39.78766 W 84.05665 Glenn Helen Nature Preserve YS Greene N 39.80411 W 83.88058 SW, TV, and WS sampled again in 2009 or skeletonized, (2) scraped, (3) mined, (4) infected was analyzed using nested ANOVA, with site, loca- with pathogen (i.e. fungal infection), and (5) mixed tion nested within site, and branch type nested within (Adams et al. 2009). The incidence, or the percent of location as factors in the model. The effect of site was total leaves experiencing some amount of damage tested over the location nested with site effect, and the (# leaves damaged/# total leaves) was also calculated effect of location nested within site was tested over the for both years. branch nested within location effect. In both models, site was treated as a random effect. Percent herbivore Sample collection 2009 damage data were transformed using inverse square transformation after adding one to all values In 2009, damage levels were sampled in June, August [(x ? 1)-2] and percent incidence data were arcsine and September, at 3 sites studied in 2008 (Wright State transformed to fulﬁll assumptions of normality. To University Woods, Taylorsville MetroPark, Sharon determine differences in the type of damage accumu- Woods) to determine the timing of the accumulation of lated by L. maackii, in 2008, a Chi-squared test of herbivore damage. Because of the difference in independence was performed on data pooled across morphology and potential for differential damage sites and locations. Repeated measures ANOVA was levels, a distinction was also made between long and used to analyze changes in herbivore damage through short shoots. Long shoots had larger, greener leaves time for 2009, with time as the within subjects effect, and elongated branches greater than 15 cm long ([12 and site, location nested within site, and branch nested leaves/branch). Leaves on short shoots were smaller, with location as between subjects effects. Compari- darker in color and the branches ranged between 5 and sons of means were made using Tukey post-hoc tests. 15 cm (B12 leaves/branch). Fifteen mature shrubs All statistical analyses were performed using SAS were selected in both the edge and interior habitats and software (Version 9.2, SAS Institute, Cary, North 5 branches were tagged (3 short and 2 long) on each Carolina). shrub. Herbivory was assessed on a percentage basis using the same rating scale as 2008. Results Data analysis Results of our 2008 survey indicated that the mean Nested analysis of variance (ANOVA) was used to test amount of damage was 1.83% ± 0.09 across sites and for differences in percent leaf area removal (herbivore habitat locations. Nested ANOVA revealed no signif- damage) and incidence of herbivory among sites and icant differences in herbivory among sites, but signif- among habitat location types nested within sites for icant differences between habitat locations (edge vs. 2008. The effect of site was tested over the location interior) within sites (Table 2). Similar results were nested within site effect. End of season data from 2009 found for percent incidence in 2008 (Table 2). With (October only) for herbivore damage and incidence the exception of the Sharon Woods (SW) site, percent 123 Damage levels from arthropod herbivores 867

Table 2 Nested ANOVA analyses evaluating the effect of rolling (0.13% ± 0.09), and mixed herbivory site, habitat location (edge vs. forest interior) in 2008 and site, (1.69% ± 0.42) contributed an insignificant amount habitat location and branch type (long vs. short branches) in of damage (Fig. 2). A low level of pathogen infection 2009 on percent herbivory (% leaf area lost) and incidence (# leaves damaged/total leaves observed) of herbivory on was observed on L. maackii sampled at all sites with a Lonicera maackii in central and southwestern Ohio mean of only 4.81 ± 7.04% of damaged leaves Source DF F value P value positive for the presence of potential pathogen damage. The main symptom of possible infection was leaf Percent damage 2008 necrosis and the majority of these leaves were located Model 15 6.20 \0.0001 in the edge habitat at Hueston Woods State Park Site 7 0.41 0.8715 (HW), Oxford, Ohio and Sharon Woods (SW) in Location (site) 8 8.46 \0.0001 Cincinnati, Ohio. Error 3,149 To address the question of the temporal dynamics Incidence 2008 of herbivory, results of the repeated measures Model 15 5.00 \0.0001 ANOVA for 2009 revealed that herbivore damage Site 7 0.17 0.9836 varied over time and that changes through time varied Location (site) 8 8.13 \0.0001 between habitat locations within sites (Table 3; Error 143 Fig. 3). Herbivore damage occurred early in the Percent damage 2009 season for both locations, increased over time in the Model 11 25.53 \0.0001 edge location, but there was little change through time Site 2 1.65 0.3291 on shrubs located in the interior habitat. Results of the Location (site) 3 1.36 0.3405 between subjects effect showed that site and location Branch [(site) location] 6 15.34 \0.0001 nested within site had no significant effect on herbi- Error 4,022 vore damage while there was a significant effect of Incidence 2009 branch type nested within habitat location and site Model 11 6.60 \0.0001 (Table 3). Herbivore damage was higher for long Site 2 1.34 0.3835 branches for both edge and interior shrubs (Fig. 3). Location (site) 3 1.38 0.3371 Branch [(site) location] 6 5.54 \0.0001 Discussion Error 137

Results of 2 years of observation at multiple sites in damage and incidence were 64 and 63% higher, Ohio indicate the amount of leaf area removed by respectively, in the forest edge habitat than in the arthropod herbivores (and pathogens) from the leaves forest interior habitat (Fig. 1a, b). Analysis of data of L. maackii is less than 3% on average, an amount that collected in 2009 revealed a mean amount of damage is likely too low to impact the fitness of these shrubs. of 3.09% ± 0.16, across all samples but no differ- Because results were largely consistent across multiple ences among sites or habitat locations for damage and sites, and site was considered a random factor, we can incidence, but there were significant differences assume that these results are typical for L. maackii among branch types within habitat locations across the region. Assuming that damage rates are (Table 2). Percent damage and incidence were 66 higher in the native range in the presence of specialist and 76% higher, respectively, on long branches than herbivores and other generalists capable of feeding on on short branches (Fig. 1c, d). L. maackii (J. Ding, personal communication; Zheng Results of the Chi-squared test for independence et al. 2006), these results indicate that L. maackii illustrated significant differences in the type of dam- experiences a release from arthropod herbivory in its age that plants received (P \ 0.0001, v2 = 2808.55). invasive range, a necessary assumption of the ERH. Across the 8 sites in 2008, chewing was the most Additionally, observations of L. maackii co-occurring prevalent form of damage with 76.8% ± 11.7 of with the native L. reticulata at Kiser Lake State Park in damaged leaves receiving this type of damage, while Champagne County, Ohio reveal that the native plants scraping (7.06% ± 3.17), mining (0.13% ± 0.09), receive extensive early season damage from the 123 868 D. Lieurance, D. Cipollini

a b

c d

Fig. 1 Herbivory and incidence of herbivory on Lonicera leaf area removed. Incidence was calculated by the total number maackii growing in edge and interior forest habitats at eight sites of leaves damaged/total number of leaves. Means ± standard in 2008 (a, b) and at three sites in 2009 (c, d) in central and error (SE) are shown. E and I represent edge and interior habitat south-western Ohio. Herbivory was estimated as a percentage of types

from white-tailed deer, a generalist herbivore, but Trisel (1997) reported lower values of deer browse on L. maackii than on native tree species (1.7% vs. 10.6–21.0%), an observation also consistent with the assumptions of ERH. The prevalence of leaf chewing indicates that the small amount of damage that does occur on L. maackii can be attributed to some unidentified generalist herbivores, possibly including Lepidopteran larvae feeding minimally on early season foliage (J. Stireman, personal communication.). The low values of percent leaf area consumed but with a moderate damage incidence could be due to induced changes in palatability of the leaves after small amounts Fig. 2 Percentage of herbivore damage type observed on of tissue removal forcing herbivores to move to other Lonicera maackii in 2008. Percentages were calculated by the leaves or other plants, where they continue to remove number of leaves of a damage type/total number of damaged only small amounts of tissue. leaves. Means ± standard error (SE) are shown and damage Resistance is the ability of a plant to reduce the type was significantly different in v2 test for independence (P \ 0.0001, v2 = 2808.55) preference or health of the herbivore often through mechanical defenses, such leaf toughness, thorns, or specialist Zaraea inflata (Honeysuckle sawfly) while trichomes, or chemical defenses including the pro- L. maackii is left untouched (D. Lieurance, personal duction of secondary defensive compounds (Ashton observation). We did not examine damage resulting and Lerdau 2008; Choong 1996; Strauss and Agrawal 123 Damage levels from arthropod herbivores 869

Table 3 Repeated measures ANOVA evaluating the effect of the question of whether invasive species in this genus site, habitat location (edge vs. forest interior) and branch type possess chemicals novel to the native flora and fauna (long vs. short branches) on percent herbivory (% leaf area (Cipollini et al. 2008b). A number of flavonoids and lost) on Lonicera maackii through the 2009 growing season other phenolics have been identified within the genus Source DF F value P value (Cipollini et al. 2008b; Flamini et al. 1997; Ochmian Between subject et al. 2009; Ren et al. 2008). In the context of Site 2 1.62 0.333 herbivory, iridoid glycosides have been identified in Location (site) 3 0.26 0.851 L. implexa, which deter feeding and decrease growth Branch [(location(site)] 6 8.51 \0.0001 rates of generalist herbivores (Penuelas et al. 2006). Error 128 This same class of compounds has been identified in L. maackii (D. Bowers, personal communication). Within subject Two major flavones, apigenin and luteolin, and their Time 2 10.80 \0.0001 glycoside derivatives, as well as chlorogenic acid, Time 9 site 4 0.50 0.7323 were identified in the leaves of L. maackii, and diet Time 9 location (site) 6 2.45 0.0255 plugs dosed with ecologically relevant concentrations Time 9 branch [location(site)] 12 0.74 0.7069 of crude leaf extracts from L. maackii, or apigenin Error 256 itself deterred feeding of a generalist herbivore Results of within subjects effects presented from Wilks’ (Cipollini et al. 2008b). Additionally, feeding trials Lambda test for multivariate analysis indicated that L. maackii was resistant to the generalist gypsy moth (Lymantria dispar) with caterpillars consuming small amounts of L. maackii foliage in choice tests, and either losing weight or dying on foliage in no-choice tests (McEwan et al. 2009a). Secondary metabolite concentrations vary through time and are often produced in higher concentrations early in the season (Scogings et al. 2004). Bud break for L. maackii typically occurs in March (Trisel 1997) and potentially provides an abundant food source for early season herbivores, but herbivores that do sample this foliage likely suffer deleterious effects on fitness from the higher concentrations of some defensive compounds (Cipollini et al. 2008b; McEwan et al. 2009a). Fig. 3 Herbivory on two distinct morphological branch types Results from 2008 supported the prediction that of Lonicera maackii in both edge and interior forest habitats edge shrubs would have a higher incidence and through time for the 2009 growing season. Herbivory was amount of herbivory than interior shrubs and, although estimated on a percentage of leaf area removed. Means ± stan- comparisons of edge versus interior plants in 2009 dard error (SE) are shown. Mean percent damage presented on data pooled by site were not statistically significant, the same trend is apparent. Typically, edge habitats are a more dynamic system with higher light levels and temperatures, and 1999). Much research has been conducted on second- higher biodiversity where plants experience increased ary metabolites present in plants from the Lonicera exposure to potential herbivores, which may help genus, mostly in the context of their pharmaceutical explain increased herbivory in this habitat location benefits in herbal medicine (Chen et al. 2009; Heinrich (Matlack 1993; Ries et al. 2004). While shrubs et al. 2008; Machida et al. 2002). Many of the experienced higher herbivory in edge habitats, the identified secondary chemicals found in the Lonicera amount was still very low, and the increased amount of genus are known to be present in North America and light available to the plant on the edge might facilitate therefore not novel, but a complete profile of leaf compensatory growth or other tolerance mechanisms, chemistry has not been done for Lonicera spp., leaving which would counteract any effect the herbivores 123 870 D. Lieurance, D. Cipollini might have on the health of this shrub. In a study the amount of foliage consumed was less than 5%, and comparing biomass accumulation and allocation pat- the damage was not regarded by the authors as terns of the invasive congener L. japonica with the biologically significant (Waipara et al. 2007). There is native L. sempervirens, the exotic exhibited greater no consensus for how much damage is required to compensatory growth in response to natural herbivory impact the fitness of a plant as this would vary by (Schierenbeck et al. 1994). These Lonicera species species, but research on various woody trees indicate share similar maximum photosynthesis rates in both between 6 and 12% leaf area loss reduced growth and open and interior habitats and it is likely that L. maackii reproductive output (Crawley 1985; Poorter et al. responds to herbivory in a similar manner (Lieurance 2004; Whittaker and Warrington 1985). While we did 2004; Schierenbeck and Marshall 1993). not measure the cost of herbivory on the fitness of Our results also supported the prediction that leaves L. maackii in the current study, mechanical removal of growing on long branches would experience more 5, 25 or 50% of the leaf area had no significant effect damage than leaves on short branches. The differences on the growth of selected branches of mature shrubs in in damage and incidence between long and short shoots the field. Simulated or arthropod removal of 50% leaf may be attributable to characteristically more tender area was capable of impacting growth of first year and likely more nutritious leaves on long shoots. Leaves plants in the greenhouse (D. Lieurance and on short branches are thicker and tougher indicating D. Cipollini, unpublished data). Our herbivory assess- leaves are more heavily defended through either ment revealed levels of folivory (\3%) that are much mechanical or chemical means (Ballare 2009;Guerra lower than what appears to be needed to the impact et al. 2010;HermsandMattson1992; Ishii and Shoko fitness of L. maackii and would indicate that arthropod 2010). Future research will include a comparative herbivores are currently having no significant impact analysis of leaf nutrition, secondary metabolite pro- on the performance of the plant. In addition, the lack of duction, and leaf toughness between branch types. endophagous herbivore pressure in the form of leaf The 2008 data collection was an end of season mining and various forms of galling also indicates cumulative assessment of leaf herbivory by arthropod enemy release. This type of specialized feeding consumers. This year was a dry year in terms of requires herbivores to adapt specifically to their host summer precipitation and there was also a major plant and it is rare to find generalist endophagous windstorm associated with Hurricane Ike on September herbivores (Frenzel and Brandl 2003). 14 that affected all or part of the sample sites included in this study. Damage from the windstorm was consistent at all 8 field sites and included mechanical Conclusions damage to the leaves, but very little leaf loss. While the end of season assessment did not quantify early The paucity of herbivores, including specialist and leaf abscission due to abiotic or biotic stress, little generalist leaf chewing insects and endophagous abscission was observed and we do not believe that consumers, coupled with evidence of both resistance abiotic conditions dramatically influenced our results. and tolerance strategies within this genus (e.g., Cipollini In addition, herbivory levels in 2009, a season with et al. 2008b; McEwan et al. 2009a; Schierenbeck and more favorable abiotic conditions, were strongly Marshall 1993; Schierenbeck et al. 1994), indicate that comparable to those in 2008, suggesting that results enemy escape is occurring for L. maackii in its in both years are robust. Results of the season long introduced range. ‘‘Invasive’’ attributes including assessment in 2009 were also consistent with the extended leaf phenology, rapid growth rates, high prediction that L. maackii receives the majority of fecundity, broad phenotypic plasticity, and tolerance damage early in the season (June and earlier) with to a variety of habitats (Lieurance 2004; Luken et al. moderate increases thereafter (McEwan et al. 2009b; 1995a, b; Luken et al. 1997a, b; Trisel 1997), in Trisel 1997). combination with the lack of regulation by herbivores, Overall, our results are consistent with those found may contribute to the rapid spread and persistence of in a comparable study of the congener L. japonica in L. maackii throughout its introduced range. Lonicera its introduced range in New Zealand, where approx- maackii will likely not experience significant amounts imately 25% of all plants showed signs of herbivory, of herbivory anytime soon without the introduction of 123 Damage levels from arthropod herbivores 871 a specialist herbivore or pathogen, as is the case with Mooney H, Drake J (eds) Ecology of biological invasions of classical biological control programs. Biological con- North America and Hawaii. Springer, New York, pp 96–110 Becerra JX (2003) Synchronous coadaption in an ancient case of trol has proven to be effective approach with such herbivory. Proc Natl Acad Sci USA 100:12804–12807 invaders as Melaleuca quinquenervia, Eichhornia Center TD, Dray FA (1992) Associations between water-hya- crassipes, and Lythrum salacaria (Center and Dray cinth weevils (Neochetina eichhorniae and Neochetina 1992; Franks et al. 2006; Grevstad 2006). While bruchi) and phenological stages of Eichhornia crassipes in southern Florida. Fla Entomol 75:196–211 having the potential to be an effective and sustainable Chen Y, Feng X, Wang M, Jia X, Zhao Y, Dong Y (2009) control to L. maackii invasion, the presence of several Triterpene glycosides from Lonicera. II. Isolation and Lonicera species in the horticultural trade, and the structural determination of glycosides from flower buds of presence of numerous native Lonicera species will Lonicera macranthoides. Chem Nat Compd 45:514–518 Choong MF (1996) What makes a leaf tough and how this provide obvious challenges to finding an appropriate affects the pattern of Castanopsis fissa leaf consumption by biocontrol agent. The research presented in this paper caterpillars. Funct Ecol 10:668–674 provides the framework for further research we are Cincotta CL, Adams JM, Holzapfel C (2009) Testing the enemy conducting on mechanisms for why herbivore damage release hypothesis: a comparison of foliar insect herbivory of the exotic norway maple (Acer platanoides L.) and the is low (e.g., physical and chemical defenses, ovipo- native sugar maple (A. saccharum L.). Biol Invasions 11: sition signals, etc.), and how much damage is required 379–388 to impact the fitness of L. maackii. We are also Cipollini D, Dorning M (2008) Direct and indirect effects of conducting comparative studies of growth responses, conditioned soils and tissue extracts of the invasive shrub, Lonicera maackii, on target plant performance. Castanea 73: arthropod assemblages, herbivory levels and resis- 166–176 tance mechanisms of L. maackii in relation to a suite of Cipollini D, Stevenson R, Cipollini K (2008a) Contrasting native and exotic Lonicera species. This information effects of allelochemicals from two invasive plants on the in conjunction with results from this study can further performance of a nonmycorrhizal plant. Int J Plant Sci 169:371–375 clarify the role of ERH in the invasion biology of Cipollini D, Stevenson R, Enright S, Eyles A, Bonello P (2008b) L. maackii. Phenolic metabolites in leaves of the invasive shrub, Lo- nicera maackii, and their potential phytotoxic and anti- Acknowledgments The authors thank Wright State herbivore effects. J Chem Ecol 34:144–152 University, the Ohio Board of Regents, and the Ohio Plant Colautti RI, Ricciardi A, Grigorovich IA, MacIsaac HJ (2004) Is Biotechnology Consortium for funding. We thank Tom invasion success explained by the enemy release hypoth- Borgman of Hamilton County Park District, Michael Enright esis? Ecol Lett 7:721–733 of Five Rivers MetroParks, and Kim Landsbergen and the Cornelissen JHC, Werger MJA, Castro-Diez P, Van Rheenen Franklin Park Conservatory for access to field sites and JWA, Rowland AP (1997) Foliar nutrients in relation to facilitation of this research. Comments by two anonymous growth, allocation and leaf traits in seedlings of a wide reviewers substantially improved this manuscript. range of woody plant species and types. Oecologia 111: 460–469 Cornelissen, T, Fernandes GW, Vasconcellos-Neto J (2008) Size does matter: variation in herbivory between and References within plants and the plant vigor hypothesis. Oikos 117: 1121–1130 Adams JM, Fang W, Callaway RM, Cipollini D, Newell E Crawley MJ (1985) Reduction of oak fecundity by low-density (2009) A cross-continental test of enemy release hypoth- herbivore populations. Nature 314:163–164 esis: leaf herbivory on Acer platanoides (L.) is three times Cripps MG, Schwarzlander M, McKenney JL, Hinz HL, Price lower in North America than in its native Europe. Biol WJ (2006) Biogeographical comparison of the arthropod Invasions 11:1005–1016 herbivore communities associated with Lepidium draba in Ashton IW, Lerdau MT (2008) Tolerance to herbivory, and not its native, expanded and introduced ranges. J Biogeogr 33: resistance, may explain differential success of invasive, 2107–2119 naturalized, and native North American temperate vines. DeWalt SJ, Denslow JS, Ickes K (2004) Natural-enemy release Divers Distrib 14:169–178 facilitates habitat expansion of the invasive tropical shrub Ballare CL (2009) Illuminated behaviour: phytochrome as a key Clidemia hirta. Ecology 85:471–483 regulator of light foraging and plant anti-herbivore Dorning M, Cipollini D (2006) Leaf and root extracts of the defence. Plant Cell Environ 32:713–725 invasive shrub, Lonicera maackii, inhibit seed germination Barbosa VS, Leal IR, Iannuzzi L, Almeida-Cortez J (2005) of three herbs with no autotoxic effects. Plant Ecol 184: Distribution pattern of herbivorous insects in a remnant of 287–296 Brazilian Atlantic forest. Neotrop Entomol 34:701–711 Ebeling SK, Hensen I, Auge H (2008) The invasive shrub Bazzaz FA (1986) Life history of colonizing plants: some Buddleja davidii performs better in its introduced range. demographic, genetic, and physiological features. In: Divers Distrib 14:225–233 123 872 D. Lieurance, D. Cipollini

Ehrlich PR, Raven PH (1964) Butterflies and plants: a study in Machida K, Sasaki H, Iijima T, Kikuchi M (2002) Studies on the coevolution. Evolution 18:586–608 constituents of Lonicera species. XVII. New iridoid gly- Feng Y, Auge H, Ebeling SK (2007) Invasive Buddleja davidii cosides of the stems and leaves of Lonicera japonica allocates more nitrogen to its photosynthetic machinery THUNB. Chem Pharm Bull 50:1041–1044 than five native woody species. Oecologia 153:501–510 Matlack GR (1993) Microenvironmental variation within and Flamini G, Braca A, Cioni PL, Morelli I, Tome F (1997) Three among forest edge sites in the eastern United States. Biol new flavonoids and other constituents from Lonicera im- Conserv 66:185–194 plexa. J Nat Products 60:449–452 McEwan RW, Rieske LK, Arthur MA (2009a) Potential inter- Franks SJ, Kral AM, Pratt PD (2006) Herbivory by introduced actions between invasive woody shrubs and the gypsy moth insects reduces growth and survival of Melaleuca quin- (Lymantria dispar), an invasive insect herbivore. Biol quenervia seedlings. Environ Entomol 35:366–372 Invasions 11:1053–1058 Frenzel M, Brandl R (2003) Diversity and abundance patterns of McEwan RW, Birchfield MK, Schoergendorfer A, Arthur MA phytophagous insect communities on alien and native host (2009b) Leaf phenology and freeze tolerance of the inva- plants in the Brassicaceae. Ecography 26:723–730 sive shrub amur honeysuckle and potential native com- Grevstad FS (2006) Ten-year impacts of the biological control petitors. J Torrey Bot Soc 136:212–220 agents Galerucella pusilla and G. calmariensis (coleop- Medley KE (1997) Distribution of the non-native shrub Lonicera tera:Chrysomelidae) on purple loosestrife (Lythrum sali- maackii in Kramer woods, Ohio. Phys Geogr 18:18–36 caria) central New York state. Biol Control 39:1–8 Morrison JA, Mauck K (2007) Experimental field comparison of Guerra PC, Becerra J, Gianoli E (2010) Explaining differential native and non-native maple seedlings: natural enemies, herbivory in sun and shade: the case of Aristotelia chilensis ecophysiology, growth and survival. J Ecol 95:1036–1049 saplings. Arthropod Plant Interact 4:229–235 Ochmian I, Oszmianski J, Skupien K (2009) Chemical com- Hartman KM, McCarthy BC (2008) Changes in forest structure position, phenolics, and firmness of small black fruits. and species composition following invasion by a non- J Appl Bot Food Qual 83:64–69 indigenous shrub, amur honeysuckle (Lonicera maackii). Osunkoya OO, Bayliss D, Panetta FD, Vivian-Smith G (2010a) J Torrey Bot Soc 135:245–259 Variation in ecophysiology and carbon economy of inva- Heinrich J, Svarcova I, Valentova K (2008) Lonicera caerulea: sive and native woody vines of riparian zones in south- a prospective functional food and a source of biologically eastern Queensland. Austral Ecol 35:636–649 active compounds. Chem Listy 102:245–254 Osunkoya OO, Bayliss D, Panetta FD, Vivian-Smith G (2010b) Herms DA, Mattson WJ (1992) The dilemma of plants-to grow Leaf trait co-ordination in relation to construction cost, or defend. Q Rev Biol 67:283–335 carbon gain and resource-use efficiency in exotic invasive Hutchinson TF, Vankat JL (1997) Invasibility and effects of and native woody vine species. Ann Bot 106:371–380 Amur honeysuckle in southwestern Ohio forests. Conserv Ozanne CMP, Speight MR, Hambler C, Evans HF (2000) Iso- Biol 11:1117–1124 lated trees and forest patches: patterns in canopy arthropod Ishii H, Shoko A (2010) The role of crown architecture, leaf abundance and diversity in Pinus sylvestrus (Scots Pine). phenology and photosynthetic activity in promoting com- For Ecol Manage 137:53–63 plementary use of light among coexisting species in tem- Pennington DN, Hansel JR, Gorchov DL (2010) Urbanization perate forests. Ecol Res 25:715–722 and riparian forest woody communities: diversity, com- Keane RM, Crawley MJ (2002) Exotic plant invasions and the position, and structure within a metropolitan landscape. enemy release hypothesis. Trends Ecol Evol 17:164–170 Biol Conserv 143:182–194 Lieurance D (2004) Leaf phenology, fecundity, and biomass Penuelas J, Sardans J, Stefanescu C, Parella T, Filella I (2006) allocation of the invasive shrub Lonicera maackii (rupr.) Lonicera implexa leaves bearing naturally laid eggs of the maxim in contrasting light environments. Masters thesis, specialist herbivore Euphydryas aurinia have dramatically Ohio University greater concentrations of iridoid glycosides than other Luken JO, Thieret JW (1996) Amur honeysuckle, its fall from leaves. J Chem Ecol 32:1925–1933 grace. Bioscience 46:18–24 Poorter L, de Plassche MV, Willems S, Boot RGA (2004) Leaf Luken JO, Tholemeier TC, Kuddes LM, Kunkel BA (1995a) traits and herbivory rates of tropical tree species differing Performance, plasticity, and acclimation of the nonindig- in successional status. Plant Biol 6:746–754 enous shrub Lonicera maackii (Caprifoliaceae) in con- Price PW (1991) Plant vigor hypothesis and herbivore attack. trasting light environments. Can J Bot 73:1953–1961 Oikos 62:244–251 Luken JO, Tholemeier TC, Kunkel BA, Kuddes LM (1995b) Ren M, Chen J, Song Y, Sheng L, Li P, Qi L (2008) Identifi- Branch architecture plasticity of amur honeysuckle (Loni- cation and quantification of 32 bioactive compounds in cera-maackii (rupr) herder)—initial response in extreme Lonicera species by high performance liquid chromatog- light environments. J Torrey Bot Soc 122:190–195 raphy coupled with time-of-flight mass spectrometry. Luken JO, Kuddes LM, Tholemeier TC (1997a) Response of J Pharm Biomed Anal 48:1351–1360 understory species to gap formation and soil disturbance in Ries L, Fletcher RJ, Battin J, Sisk TD (2004) Ecological Lonicera maackii thickets. Restor Ecol 5:229–235 responses to habitat edges: mechanisms, models, and var- Luken JO, Kuddes LM, Tholemeier TC, Haller DM (1997b) iability explained. Annu Rev Ecol Evol Syst 35:491–522 Comparative responses of Lonicera maackii (amur hon- Schierenbeck KA, Marshall JD (1993) Seasonal and diurnal patterns eysuckle) and Lindera benzoin (spicebush) to increased of photosynthetic gas-exchange for Lonicera sempervirens light. Am Midl Nat 138:331–343 and L. japonica (Caprifoliaceae). Am J Bot 80:1292–1299

123 Damage levels from arthropod herbivores 873

Schierenbeck KA, Mack RN, Sharitz RR (1994) Effects of Watling JI, Orrock JL (2010) Measuring edge contrast using herbivory on growth and biomass allocation in native and biotic criteria helps deﬁne edge effects on the density of an introduced species of Lonicera. Ecology 75:1661–1672 invasive plant. Landsc Ecol 25:69–78 Strauss SY, Agrawal AA (1999) The ecology and evolution of Whittaker JB, Warrington S (1985) An experimental ﬁeld study plant tolerance to herbivory. Trends Ecol Evol 14:179–185 of different levels of insect herbivory induced by Formica Trisel DE (1997) The invasive shrub Lonicera maackii (Rupr.) rufa predation on sycamore (Acer pseudoplatanus) III. Herder (Caprifoliaceae); factors contributing to its success Effects on tree growth. J Appl Ecol 22:797–811 and its effect on native species. Dissertation, Miami Zheng H, Wu Y, Ding J, Binion D, Fu W, Reardon R (2006) University Invasive plants established in the United States that are Waipara NW, Winks CJ, Smith LA, Wilke JP (2007) Natural found in Asia and their associated natural enemies vol 1. enemies of Japanese honeysuckle Lonicera japonica,in General Technical Report Number FHTET 2005-15. Chi- New Zealand. NZ Plant Prot 60:158–163 nese Academy of Sciences, Beijing, China

123