Biol Invasions (2011) 13:621–633 DOI 10.1007/s10530-010-9853-1

ORIGINAL PAPER

Relationships among leaf damage, natural enemy release, and abundance in exotic and native prairie plants Eric C. Vasquez • Gretchen A. Meyer

Received: 2 February 2010 / Accepted: 14 August 2010 / Published online: 26 August 2010  Springer Science+Business Media B.V. 2010

Abstract The Enemy Release hypothesis holds that exotic plants may have an advantage over native plants because their specialized natural enemies are absent. We tested this hypothesis by measuring leaf damage and plant abundance for naturally-occurring plants in prairies, and by removing natural enemies in an enemy exclusion experiment. We classified plants as invasive exotic, noninvasive exotic, or native, to determine if their degree of invasiveness influenced their relationships with natural enemies. Our field surveys showed that invasive exotic plants generally had significantly lower levels of foliar damage than native species while there was no consistent pattern for noninvasive exotics compared to natives. The relationship between damage and abundance was different for exotic and native plants: foliar damage decreased with increasing abundance for exotic plants while the trend was positive for native plants. While these results from the field surveys supported the Enemy Release Hypothesis, the enemy exclusion experiment did not. There was no relationship between a species’ status as exotic or native and its degree of release from herbivory.

E. C. Vasquez Department of Biological Sciences, University of Wisconsin–Milwaukee, P.O. Box 413, Milwaukee, WI 53201, USA G. A. Meyer (&) University of Wisconsin–Milwaukee Field Station, 3095 Blue Goose Rd, Saukville, WI 53080, USA e-mail: [email protected]

Pastinaca sativa, the invasive exotic in this experiment, experienced gains in leaf area and vegetative biomass when treated with pesticides, indicating substantial herbivore pressure in the introduced range. These results show that foliar damage may not accurately predict the amount of herbivore pressure that plants actually experience, and that the Enemy Release hypothesis is not sufficient to explain the invasiveness of P. sativa in prairies. Keywords Enemy release hypothesis
Invasive plant
Pastinaca sativa
Enemy exclusion experiment
Noninvasive exotic plant

Introduction Biological invasions have long been of interest to ecologists (e.g. Elton 1958), but in spite of much research effort, the mechanisms that allow some introduced plants to become invasive are still not fully understood (Colautti et al. 2004; Mitchell et al. 2006; Verhoeven et al. 2008). The Enemy Release Hypothesis predicts that exotic plant species achieve increased abundance in the introduced range because the enemies that evolved to recognize them as hosts are absent, whereas the competing native species may be at a disadvantage because of the presence of their evolved enemies. Although the Enemy Release hypothesis is widely accepted and provides the conceptual basis for biological control programs,

123

622

empirical evidence testing the theory is mixed, with a number of studies failing to provide strong support for the hypothesis (reviews by Maron and Vila` 2001; Keane and Crawley 2002; Colautti et al. 2004; Mitchell et al. 2006). However, it has also been recognized that exotic plants will acquire new natural enemies from their introduced range, and that this accumulation of novel enemies has the potential to provide some control over exotic plant populations (Maron and Vila` 2001; Mitchell et al. 2006; Verhoeven et al. 2008). Exotic plant species in natural communities therefore should exhibit a range of damage levels, depending on the net effect of enemies lost versus gained in their introduced range. In addition, exotic plants differ in their ability to spread and invade natural communities, as only a minority of exotics that have naturalized become invasive (Lockwood et al. 2001). Recent studies have shown that highly invasive exotic plants experience less damage by herbivores and pathogens than those that are naturalized but not strongly invasive, suggesting that release from natural enemies may play an important role in promoting invasiveness in some exotic plants (Mitchell and Power 2003; Carpenter and Cappuccino 2005; Cappuccino and Carpenter 2005). The distinction between exotic plants that are highly invasive and those that are noninvasive thus seems to be important to consider for evaluating the Enemy Release hypothesis. Two general approaches have been taken to test the Enemy Release hypothesis: biogeographical studies, which have compared interspecific relationships of exotic plants in their native and introduced ranges, and community-level studies which examined biotic interactions of exotic plants in their introduced ranges relative to co-occurring native species (Colautti et al. 2004). Biogeographic studies have provided stronger support for the hypothesis, with results generally showing less damage and fewer species of natural enemies on plants in the introduced range compared to the native range (Wolfe 2002; Mitchell and Power 2003; Dewalt et al. 2004; Jakobs et al. 2004; Genton et al. 2005). Results of community-level studies are more equivocal, with some studies even showing higher levels of damage on exotic plants than natives (Blaney and Kotanen 2001a, b; Agrawal and Kotanen 2003; Dietz et al. 2004; Agrawal et al. 2005; Parker and Gilbert 2007). It is not currently clear why the general trend for reduced diversity of natural enemies

123

E. C. Vasquez, G. A. Meyer

on exotic plants in their introduced ranges, revealed by biogeographical comparisons, does not translate more consistently into a greater advantage for exotic plants compared to co-occurring native plants in the same communities (Colautti et al. 2004), but study methodologies may play some role. Community-level studies often choose species to compare based on phylogeny, with exotic species paired with native congeners or related native species in the same family (e.g. Blaney and Kotanen 2001a; Agrawal and Kotanen 2003; Agrawal et al. 2005; Parker and Gilbert 2007). Controlling for phylogeny is clearly important, as it allows for the selection of species that are as similar as possible in characteristics other than their status as exotic or native. However, exotic species that lack close relatives in the native community may be more likely to become invasive (Hill and Kotanen 2009). Experimental designs that pair exotics with closelyrelated natives may therefore exclude some highly invasive exotics that might be most likely to demonstrate enemy release. Other studies select a large number of species for comparison, which must be observed or collected from multiple sites because they cannot all be found co-occurring at a single site (e.g. Dietz et al. 2004; Carpenter and Cappuccino 2005; Cappuccino and Carpenter 2005). These designs have the advantage that many species can be included, but the herbivore fauna and associated plant communities are likely to differ across sites, which could affect comparisons between exotics and natives. For example, the plant community context can affect the degree of damage experienced by exotic species (Prieur-Richard et al. 2002). An alternative approach for community-level studies is to select a single site and measure the amount of damage on abundant exotic and native plant species within the site instead of selecting the species in advance. This approach has several advantages. First, the species compared are growing together and are likely to be competing. While there are good experimental reasons to select species based on phylogenetic relationships as noted above, plants in the field compete with both related and unrelated species. Second, all plant species are exposed to the same suite of natural enemies present at the site. Comparisons among exotic and native species are not confounded by possible differences in the number or kinds of natural enemies inhabiting particular sites. Finally, the abundance of each species included can be measured

Relationships among leaf damage, natural enemy release, and abundance

so that the amount of damage can be related to a plant’s success at the site. One drawback of this approach is that fewer species can be included, because each species must be at least moderately abundant at the site chosen to allow sufficient replication for sampling. However, this difficulty can be alleviated by increasing the number of sites examined. This approach complements both phylogeneticallybased experiments and broader-scale studies that examine a large number of species across multiple sites, but is not commonly used. Demonstrating that exotic species experience less damage than natives co-occurring within the same community by itself is not a sufficient test of the Enemy Release hypothesis. It is also necessary to measure some aspect of plant demography, such as plant performance, population density, or population growth rate, to show that plants are actually released from top-down control by natural enemies. Linking the amount of damage to a species’ abundance at a particular site or to its degree of invasiveness is a step in this direction, as it relates herbivore pressure to a species’ success either in a single area or over a broader scale. However, herbivore exclusion experiments provide the most powerful evidence for enemy release. For example, DeWalt et al. (2004) used exclusion experiments to demonstrate that release from natural enemies explained the habitat distribution of an invasive shrub in its introduced range. Exclusion experiments have also been used to investigate the role of soil biota in promoting exotic plant invasions (Beckstead and Parker 2003; Reinhart et al. 2003; Callaway et al. 2004). Very few herbivore exclusion experiments focused on insect herbivores or foliar fungal pathogens have been performed, making it difficult to draw general conclusions (review by Liu and Stiling 2006). Those that have been done generally focus on paired exotic and native species, rather than the larger plant community (e.g. Schierenbeck et al. 1994; Agrawal et al. 2005; Siemann and Rogers 2003, but see Parker and Gilbert 2007). In the study reported here, we measured both leaf damage and release from natural enemies for exotic and native plants residing in the same community. We focused on the herbaceous species growing in prairies and old fields in southeastern Wisconsin. Most of the prairie that originally existed in this region was cleared for agriculture, but a few remnants remain and formerly farmed sites revert to old fields

623

with similar species composition. To determine whether exotic plants experience less herbivory than natives, we estimated percent leaf damage on naturally-growing plants of multiple species at 5 different sites. Based on earlier studies (Carpenter and Cappuccino 2005; Cappuccino and Carpenter 2005), we predicted that invasive exotics would have lower levels of damage than natives, and that the pattern would not be as strong for noninvasive exotics. To measure actual enemy release in terms of plant growth and reproduction, we conducted an enemy exclusion experiment at one of our sites. We applied pesticides to exotic and native plants to release them from naturally-occurring insect herbivores and fungal pathogens, and measured the increase in growth and reproduction resulting from the pesticide treatment. We predicted that the pesticide treatment would cause greater increases in performance for native plants, since under the Enemy Release hypothesis they should be subject to higher levels of herbivore pressure than exotics. Finally, if plants are subject to top-down control by natural enemies, as predicted by the Enemy Release hypothesis, then lower levels of herbivory should lead to greater abundance in the community, regardless of plant origin. We therefore included measures of plant abundance, to determine if there was a relationship between a species’ success at a site and the degree to which that species was enemy-free, and to assess whether the relationship differed for exotic and native species.

Materials and methods Field surveys Field surveys for plant foliar damage and abundance were carried out at 5 prairie and old field sites in southeastern Wisconsin (Table 1). Benedict Prairie is a prairie remnant that was never farmed, while the other sites had all been farmed at some point in the past. Young Prairie includes one of the region’s largest wet-mesic prairie remnants, but the site used for this study was just north of this prairie remnant in an old field that is reverting back to prairie. It has likely been more than 40 years since the sites used at Young Prairie, Bluff Creek, and Wojtkunski Road have been used for agriculture, and the time that Young Road was last farmed is unknown.

123

624

E. C. Vasquez, G. A. Meyer

Table 1 Sites used for field surveys of leaf damage and plant abundance Site Name

County

Description

Invasive exotic

Benedict Prairie

Kenosha

Mesic

Daucus carota

Bluff Creek

Walworth

Wet-mesic

Pastinaca sativa

Wojtkunski Road

Jefferson

Wet-mesic

Pastinaca sativa

Young Prairie

Walworth

Wet-mesic

Pastinaca sativa

Young Road

Walworth

Dry

Centaurea biebersteinii

Young Prairie was also used for the enemy exclusion experiment. All sites are located in Wisconsin. See Vasquez (2008) for detailed locality information and additional description of sites. Species classified as invasive exotics using Wisconsin State Herbarium (2005)

Sites were chosen for this study so that at least one species considered to be an invasive exotic in Wisconsin was present and abundant (Table 1). Pastinaca sativa (Apiaceae) infested three of the five sites. P. sativa is native to Eurasia and was introduced as a cultivated species to North America by settlers in the early 1600’s (Zangerl and Berenbaum 2005). It is considered an herbaceous biennial but may behave more like a monocarpic perennial depending on the end-of-season rosette size (Gross 1981; Baskin and Baskin 1979) and/or the age and disturbance frequency of the location in which it grows (Gross and Werner 1982; Holt 1972). Daucus carota (Apiaceae) and Centaurea biebersteinii (Asteraceae) occurred at the other two sites. Both species are native to Eurasia, with D. carota a biennial and C. biebersteinii a shortlived perennial. Melilotus alba (Fabaceae), another invasive Eurasian biennial, was found at Bluff Creek and Wojtkunski Road in addition to P. sativa, but at much lower abundance. Each field survey included the invasive exotic known to be present and abundant at each site, plus additional exotic and native species. For each site, we compiled a list of relatively abundant species and randomly selected one to three more exotic species and four to seven native species from the most abundant herbaceous plants growing at the site. Exotic species were classified as invasive or noninvasive using the Wisconsin State Herbarium plant list (Wisconsin State Herbarium 2005). All species included in this study are shown in Table 2. Achillea millefolium is treated as native in this study, but both

123

native and exotic subspecies are known to exist in this region and are difficult to separate morphologically (Cochrane and Iltis 2000). Cochrane and Iltis (2000) report that the native subspecies of A. millefolium is ‘‘ubiquitous throughout Wisconsin in a variety of sunny habitats…’’ and that the exotic subspecies is ‘‘rarely escaped’’. Based on this report, the plants in this experiment were most likely native. However, it was not possible to determine the origin of the A. millefolium plants used in this study with complete confidence. Surveys for foliar damage and percent cover of each species were conducted twice at each site over the growing season of 2007, once in early to mid-summer (June or July) and once in late summer (August or September). Plant species composition and percent cover were estimated in 1 m2 quadrats, which were placed randomly along evenly distributed intervals throughout each site. The number of quadrats used per site ranged from 14 to 20, depending on site size. For each quadrat, all herbaceous species were identified and percent cover for each species was estimated. To estimate foliar damage, we haphazardly selected twenty individuals of each species to sample by looking away and pointing a meter stick at evenly distanced intervals while walking along a transect through a population of the target species. In a few cases, 20 individuals of the target species could not be located so fewer plants were sampled. We estimated damage by comparing leaves to species-specific leaf damage templates that represented damage levels equal to 0, 3, 7.5, and 10–100% at 10% intervals. All leaves of an individual plant, up to twenty leaves, were considered in the estimates when possible. When a single plant’s leaves exceeded twenty, we selected a subset of leaves following pre-determined methods (e.g., every 2nd, every 3rd leaf, etc., depending on plant size). For the early season sample at Young Prairie, the control, untreated plants from the enemy exclusion experiment (described below) were used for the estimates of leaf foliar damage. Enemy exclusion experiment This experiment was conducted at Young Prairie (Table 1), and included the invasive exotic P. sativa, two noninvasive exotics (Taraxacum officinale and Vicia cracca), and six native species (Table 3). Species included in the experiment were chosen

Relationships among leaf damage, natural enemy release, and abundance Table 2 Species included in this study

Species

625

Abbreviation

Family

Centaurea biebersteinii DC.

Cb

Asteraceae

Daucus carota L.

Dc

Apiaceae

Melilotus alba Medik.

Ma

Fabaceae

Pastinaca sativa L.

Psa

Apiaceae

Invasive exotics

Noninvasive exotics Plantago major L.

Pm

Plantaginaceae

Trifolium campestre Schreb. Trifolium hybridum L.

Tc Th

Fabaceae Fabaceae

Taraxacum officinale Weber

Tof

Asteraceae

Vicia cracca L.

Vc

Fabaceae

Achillea millefolium L.

Am

Asteraceae

Aster novae-angliae L.

An

Asteraceae

Aster pilosus Willd.

Ap

Asteraceae

Aster sagittifolius Wedem. ex Willd.

Asa

Asteraceae

Asclepias syriaca L.

Asy

Asclepiadaceae

Conyza canadensis (L.) Cronquist

Cc

Asteraceae

Cirsium discolor (Muhl. ex Willd.) Spreng.

Cd

Asteraceae

Cicuta maculata L.

Cm

Apiaceae

Comandra umbellata (L.) Nutt.

Cu

Santalaceae

Erigeron philadelphicus L.

Ep

Asteraceae

Fragaria virginiana Duchesne Geum aleppicum Jacq.

Fv Ga

Rosaceae Rosaceae

Natives

Plant status as invasive exotic, noninvasive exotic or native and nomenclature follow Wisconsin State Herbarium (2005)

Helianthus grosseserratus M.Martens

Hg

Asteraceae

Lycopus americanus Muhl. ex W.P.C.Barton

La

Lamiaceae

Lithospermum canescens (Michx.) Lehm.

Lc

Boraginaceae

Liatris pycnostachya Michx.

Lp

Asteraceae

Monarda fistulosa L.

Mf

Lamiaceae

Maianthemum racemosum (L.) Link

Mr

Liliaceae

Potentilla simplex Michx.

Psi

Rosaceae

Pycnanthemum virginianum (L.) T.Durand & B.D.Jacks. ex B.L.Rob. & Fernald

Pyv

Lamiaceae

Physalis virginiana Mill.

Phv

Solanaceae

Ratibida pinnata (Vent.) Barnhart Silphium laciniatum L.

Rp Sl

Asteraceae Asteraceae

Solidago canadensis L.

Sc

Asteraceae

Solidago missouriensis Nutt.

Sm

Asteraceae

Solidago rigida L.

Sr

Asteraceae

Taenidia integerrima (L.) Drude

Ti

Apiaceae

Tradescantia ohiensis Raf.

Toh

Commelinaceae

Urtica dioica L.

Ud

Urticaceae

Zizia aurea (L.) W.D.J.Koch

Za

Apiaceae

123

626

E. C. Vasquez, G. A. Meyer

Table 3 Final sample size and date of harvest for the enemy exclusion experiment Species

Vegetative biomass

Reproductive biomass

Control

Pesticide

P. sativa

16

16

2

5

T. officinale

20

19

19

20

V. cracca

Control

Date of harvest

Pesticide July 6 June 25

7

8

5

7

July 6

A. millefolium

16

12

0

1

Aug. 13

C. maculata

14

13

6

6

July 24

G. aleppicum

19

18

15

16

July 28

L. pycnostachya

18

20

4

7

Aug. 20

S. rigida

20

19

8

5

Sept. 4

T. ohiensis

16

19

13

13

July 3

Initial sample size was 20 plants/treatment for all species, and all species were harvested in 2007. Sample size for leaf area matches that for vegetative biomass for all species except V. cracca, where N = 4 for both control and pesticide (see text)

using the same methods described for the field surveys. In July 2006, we selected and marked forty individuals of each species at Young Prairie. Twenty plants of each species were randomly assigned to be treated with pesticides, with the other twenty as control plants. The experimental plants were separated by at least 2 m, to prevent accidental exposure of control plants to pesticide treatments. Except for Cicuta maculata, individuals of the same species were never more than 50 m apart to limit any variables caused by the presence of gradients in the environment. We could not locate a group of C. maculata large enough to provide the number of individuals needed so two separate groups, each consisting of twenty individuals, were used. The two groups were approximately 150 m apart but still within the same old field site. Plants treated with pesticides were administered soil drenches of the systemic insecticide imidacloprid (Bayer) and foliar sprays of the fungicide mancozeb (Dow AgroSciences). Imidacloprid is effective against a variety of insect herbivores and has minimal effects on insect pollinators (Diaz and McLeod 2005; Morandin and Winston 2003). We tested for direct effects of imidacloprid on aboveground biomass of all of the species included in this experiment using potted plants, and found no significant differences between control and treated plants (P [ 0.25 for all species, Vasquez 2008). Mancozeb is a non-systemic, broadspectrum fungicide with minimal direct effects on plants or mycorrhizae (review by Mitchell 2003; Trappe et al. 1984).

123

Pesticides were applied according to manufacturer recommendations, beginning on July 5, 2006. Soil drench applications of imidacloprid were adjusted for individual plant size, with at least 120 ml applied to all plants. Mancozeb was applied with a backpack sprayer so that each leaf was sufficiently covered without causing excessive runoff. Control plants were given soil drenches and foliar sprays of water only. Fungicide was applied approximately every 3 weeks and insecticide approximately every 6 weeks, or more frequently if leaves of pesticide-treated plants showed evidence of herbivore or fungal damage. Plants were treated through the 2006 growing season, then marked in the fall with metal tags and located during the early spring of 2007. Treatments resumed upon plant emergence and continued until harvest. Plants were harvested over the growing season of 2007. For each species, individual infructescences were clipped as they set seed. Once seed production had ceased we cut and collected the entire aboveground portion of each plant. Harvest dates differed for each species because they came into bloom at different times (Table 3). At harvest, leaf area of fresh leaves was determined using a flatbed scanner and an image analysis program (ImageJ). All leaves were scanned for each species except T. ohiensis, where a subset of leaves were scanned and total leaf area was estimated from the subset (see Vasquez 2008 for further details). All plant material was dried in a 60C oven for at least 48 h and then weighed. Reproductive and vegetative biomasses were weighed separately. Most of the originally marked plants survived to the conclusion

Relationships among leaf damage, natural enemy release, and abundance

of the study (Table 3). Many of those that did not survive simply did not emerge during the spring of 2007. The largest losses occurred for Vicia cracca. Many of these plants experienced mortality early in the growing season of 2007. Those that did survive began shedding their leaves before the termination of flower production. As a result, these plants were harvested with few or no leaves and were measured for vegetative and reproductive biomass but not leaf area. Final sample sizes are shown in Table 3. Data analysis Field surveys for foliar damage levels and percent cover A one-way ANOVA with species as the independent variable and percent leaf damage as the dependent variable was run for each site and sampling date separately, using natural log transformed data (ln ? 1 was used for datasets that included zeros). A single analysis that combined all of the sites and sampling dates together could not be used because the set of species included for each site and sampling date differed, with some species occurring across sites and others unique to a particular site or sampling date. Contrasts were then used to examine differences in damage among native, noninvasive exotic, and invasive exotic species. There were three a priori contrasts: one tested whether invasive exotics were different from natives, the second tested whether invasive exotics were different from noninvasive exotics, and the third tested whether noninvasive exotics were different from natives. We tested whether the relationship between plant abundance and foliar damage differed for native and exotic species using Analysis of Covariance. We used plant origin (native or exotic), percent cover, and their interaction as predictor variables. The response variable was mean percent foliar damage, calculated for each site and sample date where a species was found. A significant interaction between origin and cover indicates that the relationship between cover and damage differs for native and exotic species. Foliar damage was natural-log transformed for analysis. Achillea millefolium was excluded from these analyses (both the contrast analyses and the ANCOVA), because we did not determine definitely that we were working with the native subspecies.

627

Enemy exclusion experiment We used a two-way ANOVA that included species, treatment (pesticide or control) and their interaction to assess differences in leaf area, vegetative biomass, and reproductive biomass. All plant variables were natural-log transformed prior to analysis. When the ANOVA indicated that species differed in their response to the pesticide treatment (significant interaction term at P \ 0.05), we tested for the effect of plant origin by calculating the mean difference between treated and control plants for each species, and then used a t-test to compare these differences for native versus exotic species. A. millefolium was included in the two-way ANOVA but excluded from the t-test comparing native and exotic species because we did not determine definitely that we were working with the native subspecies. Because emergence and harvest dates varied, the number of pesticide applications that each species received during 2007 varied from species to species. To detect whether the length of time a species was treated with pesticides influenced the degree to which it responded to pesticide treatment we calculated linear regressions between the number of days species were treated in 2007 and the percent increases in leaf area and vegetative biomass resulting from the pesticide treatment. There were no significant effects at P \ 0.05 so these analyses are not shown (see Vasquez 2008 for details). All statistical analyses were performed using SPSS Version 15.0 and 16.0 for Windows (2006).

Results Field surveys for foliar damage and percent cover Foliar damage Invasive exotics generally experienced less foliar damage than natives. Invasive exotics had significantly lower foliar damage than native species for both surveys at three of the five sites (Fig. 1, Benedict Prairie, Bluff Creek and Wojtkunski Road), and for one of the two surveys at the other two sites (Young Prairie—Late and Young Road—early). Invasive exotics also tended to have lower damage levels than noninvasive exotics, particularly in early to

123

628

mid-summer. Early season foliar damage levels were significantly lower on invasive exotics than they were on noninvasive exotics at all four sites where noninvasive exotics were included (Fig. 1), while later in the season invasive exotics experienced significantly less damage at only one site (Young Prairie). There was no consistent pattern for comparisons between noninvasive exotics and natives. At some sites and times, noninvasive exotics had significantly lower damage levels than natives (Fig. 1, Benedict Prairie—Late and Wojtkunski Road—Late), while for other sites and times, damage was significantly higher on noninvasive exotics compared to natives (Wojtkunski Road—early, Young Prairie—early). Relationship between percent cover and foliar damage The relationship between mean leaf damage and percent cover differed for native and exotic species (plant origin*percent cover interaction in ANCOVA: F1,78 = 6.15, P = 0.015). Percent damage increased with increasing cover for native species, while the opposite trend was observed for exotic species (Fig. 2). Enemy exclusion experiment Leaf area The pesticide treatment increased leaf area, and there was also a strong effect of species (Fig. 3a, treatment effect in 2-way ANOVA: F1,265 = 6.7, P = 0.01, species effect: F8,265 = 35.4, P \ 0.001). The effect of the pesticide treatment varied across species (treatment*species interaction in 2-way ANOVA: F8,265 = 2.3, P = 0.02), but was not related to whether a species was native or exotic (t-test, t = -0.87, df = 6, P = 0.42). S. rigida, A. millefolium, P. sativa, G. aleppicum, and L. pycnostachya all showed particularly large increases in leaf area in response to the pesticide treatment (percent increase in leaf area for treated plants, 139, 56, 45, 30, and 28%, respectively, Fig. 3a). Vegetative biomass The pesticide treatment also increased vegetative biomass, and again there was a strong effect of species (Fig. 3b, treatment effect in 2-way ANOVA: F1,272 = 15.3, P \ 0.001, species effect: F8,272 = 40.4,

123

E. C. Vasquez, G. A. Meyer

P \ 0.001). Responses to the pesticide treatment were similar across species (treatment*species interaction in 2-way ANOVA: F8,272 = 1.1, P = 0.37), thus there was no indication that native species benefitted more from pesticide application than exotics. Contrary to our prediction, the invasive exotic P. sativa had the strongest response to the pesticide treatment, with treated plants showing a 113% increase in biomass compared to controls. Reproductive biomass As for leaf area and vegetative biomass, species strongly differed in their reproductive biomass, but in this case there were no significant effects of the pesticide treatment either as a main effect or interaction (Fig. 3c, treatment effect in 2-way ANOVA: F1,135 = 0.47, P = 0.49, species effect: F7,135 = 25.9, P \ 0.001, treatment*species: F7,135 = 0.56, P = 0.79). The lack of an effect of the pesticide treatment on reproductive biomass likely results from the low numbers of individuals that flowered in this study. Only three species, T. officinale, G. aleppicum, and T. ohiensis, had more than ten individuals from each treatment that produced flowers during the summer of 2007 (Table 3). Many of the inflorescences from T. ohiensis plants were lost to deer herbivory. All other species had fewer than ten flowering individuals per treatment.

Discussion Our results add to accumulating evidence that the distinction between invasive exotics and noninvasive exotics is important when testing hypotheses like Enemy Release (e.g. Mitchell and Power 2003; Carpenter and Cappuccino 2005; Cappuccino and Carpenter 2005). Our field surveys of leaf damage show that exotics considered to be invasive consistently experienced significantly less leaf damage than the native species they potentially compete with in a community, while no clear pattern emerged for noninvasive exotic species compared to natives. Invasive exotics also tended to have less foliar damage than noninvasive exotics, particularly early in the season. Escape from herbivory earlier in the season could provide an advantage to the invasive exotic species, as there is evidence demonstrating that herbivore attack

Relationships among leaf damage, natural enemy release, and abundance

629

Fig. 1 Mean percent foliar damage levels for invasive exotic, noninvasive exotic, and native species at 5 prairie and old field sites during early to mid-summer and late summer. Species abbreviations shown in Table 2. Data were ln or ln ? 1 transformed for analysis. IE invasive exotic, NE noninvasive exotic, N native. Mean and SE shown. Significance of contrasts shown as follows: * = P \ 0.05, ** = P \ 0.01, *** = P \ 0.001

123

630

E. C. Vasquez, G. A. Meyer

Fig. 2 Relationship between foliar damage and percent cover at 5 prairie and old field sites for exotic species and native species. Interaction between origin and percent cover significant at P = 0.015, ANCOVA. Foliar damage ln-transformed for analysis

occurring early in a seasonal environment can be more detrimental to population growth than late season attack (e.g. Knight 2007). However, care should be taken when implying any such patterns across multiple species or at the level of community as impacts of early versus late season herbivory can vary considerably even within a single species and such variation may depend on nutrient availability along with other abiotic factors (Maschinski and Whitham 1989). Our study extends previous work by showing that invasive exotic plants experience less leaf damage than native species or noninvasive exotic plants (at least early in the season) for groups of species occurring in the same sites where they compete and are exposed to the same suite of herbivores. Our results suggest a fundamentally different relationship between leaf damage and abundance for exotic and native plants, as foliar damage found on exotic species decreased as the plant’s percent cover at a site increased, while the opposite trend was seen for native species. This result is correlative in nature and should be interpreted cautiously, but is consistent with the Enemy Release hypothesis in that the most abundant exotic plants experienced the least amount of foliar herbivory. The relationship for exotic plants was driven primarily by the responses of two invasive exotics, P. sativa and C. biebersteinii, which both had low levels of foliar damage and attained high cover values. It is not clear why damage should increase with abundance for native plants, but this result suggests that top-down control by natural enemies may not have

123

Fig. 3 Effects of enemy exclusion on a leaf area, b vegetative biomass, and c reproductive biomass for exotic and native species at Young Prairie. Species abbreviations shown in Table 2. All plant variables were ln-transformed for analysis. The pesticide treatment significantly increased both leaf area and vegetative biomass, but pesticide effects did not depend on whether species were native or exotic (see text for statistical analysis). Mean and SE shown

strongly influenced abundance for native plants at our sites. The importance of top-down control for native plant populations has been debated and the limited number of studies in this area has made general conclusions difficult (Maron and Crone 2006). In the tallgrass prairie old field communities of this study, it is

Relationships among leaf damage, natural enemy release, and abundance

possible that herbivores and pathogens responded to the high abundances of certain species by attacking them at greater frequencies, as is predicted by the Resource Concentration Hypothesis (Root 1973). We are unaware of any other studies suggesting that the relationship between plant abundance and damage is different for exotic and native species, and further work will be needed to clarify this result. While the results of our foliar damage surveys in the field generally supported the Enemy Release hypothesis, the enemy exclusion experiment was more contradictory. There was no evidence from this experiment that native species benefitted more from the removal of natural enemies than exotic species. In fact, P. sativa, the invasive exotic species included in the enemy exclusion experiment, showed strong increases in both leaf area and vegetative biomass when treated with pesticides. There was also no evidence from this experiment that lower levels of herbivore pressure were associated with higher abundance at the site. P. sativa had both the highest percent gain in vegetative biomass resulting from the pesticide treatment and was also the most abundant among the species included in the enemy exclusion experiment (Vasquez 2008). These results suggest that the success of P. sativa at Young Prairie may not be explained by the Enemy Release Hypothesis. The interactions between P. sativa and its insect herbivores have been well-characterized in both its native and introduced ranges (Berenbaum 1981; Zangerl and Berenbaum 2005; Berenbaum and Zangerl 2006, and references therein). The aboveground parts of P. sativa produce furanocoumarins, potent secondary chemicals that deter generalist herbivores (Berenbaum 1981). The primary herbivore on P. sativa in North America is the specialist parsnip webworm, Depressaria pastinacella (Lepidoptera: Oecophoridae), which was accidentally introduced from Europe in the mid 19th century (Zangerl and Berenbaum 2005). This insect webs together the reproductive structures of the plant, where it feeds on flowers and developing seeds and greatly reduces the number of viable seeds produced (Lohman et al. 1996). However, we did not observe the parsnip webworm or signs of its damage on any of the plants in our study. In addition, removal of an inflorescence-feeding insect could not account for the increases in leaf area and vegetative biomass that we observed. Therefore the release from herbivory that we observed must have been caused by

631

other insects or fungal pathogens. Based on field surveys in Tompkins Co., New York, Berenbaum (1981) lists 13 species of insect herbivore as feeding on P. sativa. We also observed multiple herbivores feeding on the P. sativa control plants during the course of this study, including aphid colonies and nymphs of what was likely the meadow spittle bug, Philaenus spumarius (Homoptera: Cercopidae). This generalist herbivore is quite common in many parts of the world and can have strong impacts on the growth and fitness of its hosts (Meyer 1993; Meyer and Root 1993). In addition, application of the systemic insecticide imidacloprid as a soil drench could have controlled root-feeding or other below-ground herbivores. While the identities of the specific herbivores responsible for the growth reductions observed for P. sativa in our experiment are not known, it is clear that this species does experience significant pressure from natural enemies in its introduced range. Unfortunately, it was difficult for this study to assess how herbivores and pathogens affected the reproductive outputs of native and exotic species. We did not detect any significant effect of the pesticide treatment on reproductive biomass in this experiment, most likely because many of our experimental plants failed to flower. Although it is often considered a biennial, P. sativa plants are known to spend multiple years in the rosette stage when growing in old fields that have not seen recent disturbance (Gross and Werner 1982; Holt 1972). This appears to be the case at Young Prairie, where only seven plants (5 pesticide and 2 control) produced flowers. Low sample size was also the case for S. rigida and L. pycnostachya, where differences between control and treated plants appeared to be considerable. Without stronger data on the reproductive consequences of enemy exclusion it is difficult to draw conclusions concerning the effects of a reduced enemy load on population size of invasive exotic plants. Taken together, the results of our field surveys and enemy exclusion experiment demonstrate that although foliar damage is a convenient and widelyused measure for herbivory, it may not accurately predict the amount of herbivore pressure that plants actually experience (see Siemann and Rogers 2003 for a similar result). Plants may be attacked by multiple guilds of herbivores, many of which may not cause visible damage to leaves. It is possible that a species which escapes leaf chewers may still experience

123

632

significant reductions in growth and fecundity due to sap suckers, root feeders, stem borers, seed predators, or other types of herbivores. Our observations of aphids and spittlebugs on P. sativa control plants in our enemy exclusion experiment suggest that sap-feeding insects may have had significant impacts on the growth of this exotic in the introduced range, and suppression of root-feeding herbivores may also have positively affected the plants in our experiment. In a common garden experiment using 15 pairs of native and introduced plants, Agrawal et al. (2005) measured impacts caused by natural enemies in a number of different guilds and found that escape from one guild was not necessarily related to escape from other enemies. Plants tended to be released from some guilds but not others. These results and ours indicate that tests of the Enemy Release hypothesis should incorporate multiple guilds of herbivores whenever possible. Our study parallels the contradictory results that have been seen for other community-level tests of the Enemy Release Hypothesis (see Introduction and Colautti et al. 2004), as our field surveys generally supported the hypothesis but our enemy exclusion experiment did not. The reasons underlying this discrepancy are not fully clear, but our results do suggest that while enemy release may play some role in explaining the abundance of invasive exotic plants in prairies, other factors must also be operating for P. sativa at Young Prairie. Our study can offer several insights for future research in this area. First, it is clear that exotic species do accumulate natural enemies in their introduced range, and the degree of foliar damage that they experience in field sites is related to their degree of invasiveness. The question is not whether or not exotic plants have escaped their herbivores in their introduced range, but rather what is the net effect of losses versus gains (Maron and Vila` 2001; Mitchell et al. 2006; Verhoeven et al. 2008). Second, it is important to examine multiple guilds of herbivore, as highly invasive species like P. sativa may show reduced damage from some herbivores such as folivores but still experience significant herbivore pressure from other guilds that may be more difficult to census in the field. Finally, our results show that a species such as P. sativa, which has potent antiherbivore defenses and is known to be attacked primarily by specialists, can experience significant release from herbivory in its introduced range even in a

123

E. C. Vasquez, G. A. Meyer

site where its specialist herbivore is absent. This result illustrates the value of enemy exclusion experiments, yet these experiments have not been widely applied for studies of insects or fungal pathogens in tests of the enemy release hypothesis (Liu and Stiling 2006). A greater emphasis on enemy exclusion experiments in future studies could help to clarify the role of enemy release in promoting plant invasiveness. Acknowledgments We thank Jim Reinartz and Stefan Schnitzer for advice and helpful comments and Kayvon Ali for assistance in the field and laboratory. The Wisconsin Department of Natural Resources granted permission to use many of the sites for this experiment, and in particular we are grateful to Matt Zine for his assistance at Young Prairie. Mike Fort assisted with plant identification. Two anonymous reviewers provided comments that improved the manuscript. This research was supported by a Ruth Walker Grant-in-Aid and an Advanced Opportunity Fellowship, both from the University of Wisconsin-Milwaukee, to ECV.

References Agrawal AA, Kotanen PM (2003) Herbivores and the success of exotic plants: a phylogenetically controlled experiment. Ecol Lett 6:712–715 Agrawal AA, Kotanen PM, Mitchell CE, Power AG, Godsoe W, Klironomos J (2005) Enemy release? An experiment with congeneric plant pairs and diverse above- and belowground enemies. Ecology 86:2979–2989 Baskin JM, Baskin CM (1979) Studies on the autoecology and population biology of the weedy monocarpic perennial, Pastinaca sativa. J Ecol 67:601–610 Beckstead J, Parker IM (2003) Invasiveness of Ammophila arenaria: release from soil-borne pathogens? Ecology 84:824–2831 Berenbaum M (1981) Patterns of furanocoumarin distribution and insect herbivory in the Umbelliferae: plant chemistry and community structure. Ecology 62:1254–1266 Berenbaum MR, Zangerl AR (2006) Parsnip webworms and host plants at home and abroad: trophic complexity in a geographic mosaic. Ecology 87:3070–3081 Blaney CS, Kotanen PM (2001a) Effects of fungal pathogens on seeds of native and exotic plants: a test using congeneric pairs. J Appl Ecol 38:1104–1113 Blaney CS, Kotanen PM (2001b) Post-dispersal losses to seed predators: an experimental comparison of native and exotic old field plants. Can J Bot 79:284–292 Callaway RM, Thelan GC, Rodriguez A, Holben WE (2004) Soil biota and exotic plant invasion. Nature 427:731–733 Cappuccino N, Carpenter D (2005) Invasive exotic plants suffer less herbivory than non-invasive exotic plants. Biol Lett 1:435–438 Carpenter D, Cappuccino N (2005) Herbivory, time since introduction and the invasiveness of exotic plants. J Ecol 93:315–321

Relationships among leaf damage, natural enemy release, and abundance Cochrane TS, Iltis HH (2000) Atlas of the Wisconsin prairie and savanna flora. Technical Bulletin No. 191. University of Wisconsin-Madison, Madison Colautti RI, Ricciardi A, Grigorovich IA, MacIsaac HJ (2004) Is invasion success explained by the enemy release hypothesis? Ecol Lett 7:721–733 DeWalt SJ, Denslow JS, Ickes K (2004) Natural-enemy release facilitates habitat expansion of the invasive tropical shrub Clidemia hirta. Ecology 85:471–483 Diaz FJ, McLeod P (2005) Movement, toxicity, and persistence of imidacloprid in seedling tabasco pepper infested with Myzus persicae (Hemiptera: Aphididae). J Econ Entomol 98:2095–2099 Dietz H, Wirth LR, Buschmann H (2004) Variation in herbivore damage to invasive and native woody plant species in open forest vegetation on Mahe´, Seychelles. Biol Invasions 6:511–521 Elton CS (1958) The ecology of invasions by animals and plants. Methuen, London Genton BJ, Kotanen PM, Cheptou P, Adolphe C, Shykoff JA (2005) Enemy release but no evolutionary loss of defence in a plant invasion: an inter-continental reciprocal transplant experiment. Oecologia 146:404–414 Gross KL (1981) Predictions of fate from rosette size in four ‘‘biennial’’ plant species: Verbascum thapsus, Oenothera biennis, Daucus carota, Tragopogon dubius. Oecologia 48:209–213 Gross KL, Werner PA (1982) Colonizing abilities of ‘‘biennial’’ plant species in relation to ground cover: implications for their distribution in a successional sere. Ecology 63:921–931 Hill SB, Kotanen PM (2009) Evidence that phylogenetically novel non-indigenous plants experience less herbivory. Oecologia 161:581–590 Holt BR (1972) Effect of arrival time on recruitment, mortality, and reproduction in successional plant populations. Ecology 53:668–673 Jakobs GE, Weber E, Edwards PJ (2004) Introduced plants of the invasive Solidago gigantea (Asteraceae) are larger and grow denser than conspecifics in the native range. Divers Distrib 10:11–19 Keane RM, Crawley MJ (2002) Exotic plant invasions and the enemy release hypothesis. Trends Ecol Evol 17:164–170 Knight TM (2007) Population-level consequences of herbivory timing in Trillium grandiflorum. Am Midl Nat 157:27–38 Liu H, Stiling P (2006) Testing the enemy release hypothesis: a review and meta-analysis. Biol Invasions 8:1535–1545 Lockwood JL, Simberloff D, McKinney ML, Von Holle B (2001) How many, and which, plants will invade natural areas? Biol Invasions 3:1–8 Lohman DJ, Zangerl AR, Berenbaum MR (1996) Impact of floral herbivory by parsnip webworm (Oecophoridae: Depressaria pastinacella Duponchel) on pollination and fitness of wild parsnip (Apiaceae: Pastinaca sativa L.). Am Midl Nat 136:407–412 Maron JL, Crone E (2006) Herbivory: effects on plant abundance, distribution and population growth. Proc Royal Soc B 273:2575–2584 Maron JL, Vila` M (2001) When do herbivores affect plant invasion? Evidence for the natural enemies and biotic resistance hypothesis. Oikos 95:361–373

633

Maschinski J, Whitham TG (1989) The continuum of plant responses to herbivory: the influence of plant association, nutrient availability, and timing. Am Nat 134:1–19 Meyer GA (1993) A comparison of the impacts of leaf and sapfeeding insects on growth and allocation of goldenrod. Ecology 74:1101–1116 Meyer GA, Root RB (1993) Effects of herbivorous insects and soil fertility on reproduction of goldenrod. Ecology 74:1117–1128 Mitchell CE (2003) Trophic control of grassland production and biomass by pathogens. Ecol Lett 6:147–155 Mitchell CE, Power AG (2003) Release of invasive plants from fungal and viral pathogens. Nature 421:625–627 Mitchell CE, Agrawal AA, Bever JD, Gilbert GS, Hufbauer RA, Klironomos JN, Maron JL, Morris WF, Parker IM, Power AG, Seabloom EW, Torchin ME, Vazquez DP (2006) Biotic interactions and plant invasions. Ecol Lett 9:726–740 Morandin LA, Winston ML (2003) Effects of novel pesticides on bumble bee (Hymenoptera: Apidae) colony health and foraging ability. Environ Entomol 32:555–563 Parker IM, Gilbert GS (2007) When there is no escape: the effects of natural enemies on native, invasive, and noninvasive plants. Ecology 88:1210–1224 Prieur-Richard AH, Lavorel S, Linhart YB, Dos Santos A (2002) Plant diversity, herbivory and resistance of a plant community to invasion in mediterranean annual communities. Oecologia 130:96–104 Reinhart KO, Packer A, Van der Putten WH, Clay K (2003) Plant–soil biota interactions and spatial distribution of black cherry in its native and invasive ranges. Ecol Lett 6:1046–1050 Root RB (1973) Organization of a plant-arthropod association in simple and diverse habitats: the fauna of collards (Brassica oleracea). Ecol Monogr 43:95–124 Schierenbeck KA, Mack RN, Sharitz RR (1994) Effects of herbivory on growth and biomass allocation in native and introduced species of Lonicera. Ecology 75:1661–1672 Siemann E, Rogers WE (2003) Herbivory, disease, recruitment limitation, and success of alien and native tree species. Ecology 84:1489–1505 Trappe JM, Molina R, Castellano M (1984) Reactions of mycorrhizal fungi and mycorrhiza formation to pesticides. Annu Rev Phytopathol 22:331–359 Vasquez EC (2008) The effects of enemy exclusion and enemy damage on exotic and native plant species: a test of the enemy release hypothesis. MS Thesis, University of Wisconsin-Milwaukee, Milwaukee Verhoeven KJF, Biere A, Harvey JA, van der Putten WH (2008) Plant invaders and their novel natural enemies: who is naive? Ecol Lett 12:107–117 Wisconsin State Herbarium (2005) Wisflora: Checklist of the vascular plants of Wisconsin. University of WisconsinMadison Herbarium web site. http://www.botany.wisc. edu/wisflora/ Wolfe LM (2002) Why alien invaders succeed: support for the escape-from-enemy hypothesis. Am Nat 160:705–711 Zangerl AR, Berenbaum MR (2005) Increase in toxicity of an invasive weed after reassociation with its coevolved herbivore. PNAS 43:15529–15532

123