The enemy release hypothesis (ERH) frequently has been invoked to explain the naturalization and spread of introduced species. One ramification of the ERH is that invasive plants sustain less herbivore pressure than do native species. Empirical studies testing the ERH have mostly involved two-way comparisons between invasive introduced plants and their native counterparts in the invaded region. Testing the ERH would be more meaningful if such studies also included introduced non-invasive species because introduced plants, regardless of their abundance or impact, may support a reduced insect herbivore fauna and experience less damage. In this study, we employed a three-way comparison, in which we compared herbivore faunas among native, introduced invasive, and introduced non-invasive plants in the genus Eugenia (Myrtaceae) which all co-occur in South Florida. We observed a total of 25 insect species in 12 families and 6 orders feeding on the six species of Eugenia. Of these insect species, the majority were native (72%), polyphagous (64%), and ectophagous (68%). We found that invasive introduced Eugenia has a similar level of herbivore richness as both the native and the non-invasive introduced Eugenia. However, the numbers and percentages of oligophagous insect species were greatest on the native Eugenia, but they were not different between the invasive and non-invasive introduced Eugenia. One oligophagous endophagous insect has likely shifted from the native to the invasive, but none to the non-invasive Eugenia. In summary, the invasive Eugenia encountered equal, if not greater, herbivore pressure than the non-invasive Eugenia, including from oligophagous and endophagous herbivores. Our data only provided limited support to the ERH. We would not have been able to draw this conclusion without inclusion of the non-invasive Eugenia species in the study.
The enemy release hypothesis (ERH) states that introduced invasive species are successful because they left their co-evolved natural enemies behind. This idea makes intuitive sense and is the theoretical foundation of classical biological control. It is one of the most cited explanations for the undesired success of introduced invasive species worldwide (Williams 1959; Crawley 1997; Maron & Vilà 2001; Keane & Crawley 2002). Although empirical studies testing the ERH on invasive plants are limited in number (Maron & Vilà 2001; Keane & Crawley 2002; Liu & Stiling 2006) and vigor (but see Schierenbeck et al. 1994; Wolfe 2002; Siemann & Rogers 2003; DeWalt et al. 2004), there have been several syntheses to test the predictions stemming from ERH during the last decade (Maron & Vilà 2001; Keane & Crawley 2002; Colautti et al. 2004; Liu & Stiling 2006). One consensus generated from these syntheses and other more recent empirical studies is that the total number of insect herbivores, and the numbers of endophagous and oligophagous herbivores, are all reduced on introduced invasive species compared with conspecific populations in the native range or on co-occurring native congeners (Keane & Crawley 2002; Colautti et al. 2004; Hinz & Schwarzlaender 2004; Torchin & Mitchell 2004; Liu & Stiling 2006). In addition, a modification of the ERH, which states that it is the escape from specialist insects (including endophagous species) that allow the introduced plants to be successful, has received increasing support (Wolfe et al. 2004; Joshi and Vrieling 2005; Stastny et al. 2005; Mitchell et al. 2006).
All the empirical studies reviewed above were performed in one of two ways: first, insect herbivore diversity, load, or insect herbivore impact either on invasive plants in native vs. introduced ranges was examined (e.g., Wolfe 2002; DeWalt et al. 2004), or second, the same comparisons were made between invasive plants and their native counterparts in the new region (Schierenback et al. 1994; Agrawal & Kotanen 2003; Siemann & Rogers 2003). The latter approach is not a direct test of the ERH. Rather, it tests a ramification of the ERH that invasive introduced plants sustain less insect herbivore pressure than their native counterparts. However, all introduced plants, regardless of their abundance or impact, may support a reduced insect herbivore fauna and experience less damage simply because plants tend to lose their associated insect herbivores during the introduction (Colautti et al. 2004) and it takes time, on the ecological and/or evolutionary scale, for a new population to acquire its insect herbivore fauna (Strong et al. 1984). Testing the ERH would be more meaningful if such studies also included introduced plants which do not become invasive, or so-called innocuous species (Colautti et al. 2004; Levine et al. 2004). However, few studies have included introduced non-invasive plants (but see Mitchell & Power 2003; Cappuccino & Carpenter 2005; Carpenter & Cappuccino 2005).
A three-way comparison of insect herbivore faunas in a system in which congeneric native, introduced invasive, and introduced non-invasive (innocuous) plants that co-occur in the same region can provide insightful information on the validity of the ERH. If release from natural enemies is important in determination of the success of an introduced plant species, one would expect that invasive introduced plants escape more from herbivore pressure than do non-invasive introduced plants. One question of particular interest is whether there have been any shifts of oligophagous and/or endophagous herbivores from the native to the introduced plant congeners, and if such shifts occur more onto the non-invasive than to the invasive congeners. Endophagous herbivores are of interest because an internal feeding niche is likely to be correlated with dietary specialization (Frenzel & Brandl 1998). Plants that are closely related phylogenetically (i.e., congeners or confamiliers), as used in many ERH tests, offer a good chance to detect host shifts by herbivores to the introduced plants because herbivore host choice is often determined by plant relatedness.
In this study, we compared insect herbivore faunas among native (two species), invasive (one species), and non-invasive (three species) of Eugenia growing in South Florida. The Eugenia spp. studied here are small-medium sized trees native to Florida and Central-South America (Wunderlin & Hansen 2003; Ruehle et al. 1958). We predict that (1) the total number of herbivore species will be (a) greater on the native Eugenia species than on the introduced invasive and non-invasive congeners; and (b) greater on the introduced non-invasive Eugenia than on the introduced invasive congener; (2) the number and proportion of oligophagous and endophagous herbivores will be (a) greater on the native Eugenia species than on the introduced invasive and non-invasive congeners; and (b) greater on the introduced non-invasive Eugenia than on the introduced invasive congener, and (3) fewer herbivores, particularly oligophagous and endophagous herbivores, will be shared between the native Eugenia and the introduced invasive Eugenia than between the native and introduced non-invasive Eugenia. The first portions of the first two predictions are comparable to predictions made by the usual two-way (native vs. introduced invasive plants) comparisons. For ERH to be supported in the current three-way testing system, the second portion of the prediction should be validated. We believe this study represents the first known comparison of herbivore funna on native, invasive, and innocuous species of the same genus in the same geographic location.
Material and Methods
Study Plants
Eugenia uniflora L. (Surinam cherry), E. aggregata Kiaersk. (cherry of the Rio Grande), E. brasiliensis Lam. (grumichama), and E. luschnathiana Klotzsch (pitomba) are all large shrubs or small trees with potentially animal-dispersed fleshy fruits that were introduced to south Florida from Brazil in the late 1800s or early 1900s for home garden fruit and ornamental purposes (Ruehle et al. 1958; Martin et al. 1987). Eugenia uniflora is a common hedge plant in South Florida, probably due to its robust and rapid growth. Since its introduction, E. uniflora has escaped cultivation and invaded hammocks (evergreen broad-leaved forests) in South Florida, growing side by side in some areas with 2 native congeners, E. axillaris (Sw.) Willd. (white stopper) and E. foetida Pers. (Spanish stopper) (Gann et al. 2001) (Table 1). The other 3 introduced Eugenia spp. still remain in cultivation in many public and private gardens and nurseries.
Study Sites
We carried out most of our sampling at two subtropical hammocks in Broward County where E. axillaris (native), E. foetida (native), and E. uniflora (invasive) co-occur: Hugh Taylor Birch State Park (hereafter referred to as Birch Park), and the Bonnet House Museum and Garden (Hereafter referred to as Bonnet House). Subtropical hammocks in South Florida are evergreen, broad-leaved forests composed predominantly of trees common to the Bahamas and Greater Antilles (Snyder et al. 1990). They occupy limestone outcroppings that are elevated, rarely inundated, and relatively fire-free. In hammocks of both Birch Park and Bonnet House, the canopy trees are primarily composed of Bursera simaruba (L.) Sarg. (gumbo-limbo), Coccoloba unifera L. (sea-grape), Krugiodendron ferreum (Vahl) Urb. (black iron wood), and Ficus aurea Nutt. (strangler fig). The understory is dominated by E. axillaris, E. foetida, and E. uniflora. Sandy soil is characteristic of both sites.
For the introduced non-invasive E. aggregata, E. brasiliensis, and E. luschnathiana, we located up to 14 individuals per species in 4 research, public, and private gardens in Miami Dade and Broward, 2 adjacent counties in South Florida. These gardens include University of Florida, Tropical Research and Education Center, the Fruit and Spice Park, Plantation Heritage Park, and the Fairchild Tropical Garden. These plants are referred to as cultivated aggregata, cultivated brasiliensis, and cultivated lushnathiana (Table 1). In addition, as a control for potential site related differences between these gardens and the natural subtropical hammocks, we also sampled 9, 10, and 28 individuals, respectively, of E. axillaris (native), E. foetida (native), and E. uniflora (invasive) at the above gardens. These individuals were referred to as cultivated axillaris, cultivated foetida, and cultivated uniflora. Sampling frequencies for the cultivated plants were the same as for the wild populations mentioned above.
Determination of Insect Herbivore Faunas
Four and two 5 × 3-m2 plots were established at the Birch Park and the Bonnet House, respectively, for herbivore faunal surveys on wild populations of E. axillaris, E. foetida, and E. uniflora (Table 1). We tagged a total of 182, 202, and 97 wild plants of various sizes of E. axillaris, E. foetida, and E. uniflora, respectively. All these plants were visited every other month during the dry season (Oct to Apr) and monthly during the wet season (May to Sep) from Jan to Dec 2004. Larval and adult insects were hand caught and brought back to the lab for rearing, specimen preparation, and identification. For fruit and seed feeders, we collected random fruit samples from 3-10 trees and 20-100 fruits per tree, depending on availability. Some non-rotten fruits on the ground directly beneath the trees were also included in the samples. Unidentified fruit/seed feeders were reared to maturity for identifications. We sent unknown specimens to specialists in the USA for identification. Information on insect immigration status (i.e., native or exotic) and diet breadth were provided by these insect specialists when possible. Insects were classified as native or exotic, oligophagous or polyphagous, and endophagous or ectophagous feeders. Oligophagous refers to insects which feed only on plants of 1 family while polyphagous indicates herbivores that feed on more than 1 family. Insects were “very important” if they were seen in every census, or were seen to cause 10% or more of leaf or seed damage on average in at least 1 census (Liu, unpublished data). Insects were “important” if they were seen in more than 1 census but caused less than 10% leaf or seed damage. Herbivores were “not important” if they were seen only once during the entire study period or caused very little plant damage. Determination of % damage to plants depended on the nature of the insect. For example, the % damage by a leaf miner was determined by counting the % of leaves with mines, while the % damage by a chewing caterpillar was by counting the % of leaves chewed.
Data Analyses
In addition to the identity of the herbivores, the number of total insect herbivore species on each Eugenia species, the number and percentage of native insect herbivores, the number and percentage of endophagous vs. ectophagous feeders, and the number and percentage of oligophagous vs. polyphagous feeders were determined. The differences in these percentages among the native (average among the 2 species), invasive and non-invasive (average among the 3 species) plants were determined with chi-square tests (Zar 1984) in SPSS 13.0 (SPSS, Chicago, Illinois, USA). Because there may be differences in the herbivore fauna between wild and cultivated populations of the same species as the latter are in artificial settings, 2 sets of the chi-square tests were performed. One was a two-way test that included wild native plants and wild introduced invasive plants. The other was a three-way test that included cultivated native, invasive, and non-invasive plants. We also determined the number of herbivores, particularly oligophagous and/or endophagous, shared between the native, invasive and non-invasive plants. Samples from the two natural area sites were pooled because they had identical herbivore fauna for the three wild Eugenia populations. Samples from the four garden sites were pooled because all gardens did not have adequate sample sizes for among site comparisons.
Results
We observed, collected, and reared a total of 25 insect species in 12 families and 6 orders feeding on the 6 species of Eugenia during the 1-year sampling period (Table 2). Among them, the majority were native (72%), polyphagous (64%), and external feeders (68%). There were 7 additional uncommon species of Lepidoptera reared from bagged branches of various Eugenia spp. that were not included in the results because herbivory by these species was not confirmed. The native wild Eugenia species had higher numbers of herbivore species than the wild introduced E. uniflora and most cultivated Eugenia. The only exception was that the cultivated E. uniflora had more herbivore species than the native Eugenia (Fig. 1A).
The introduced invasive and non-invasive Eugenia recruited fewer oligophagous insect herbivores than the native Eugenia (Fig. 1A). The difference in proportions of herbivore diet breadth (oligophagous vs. polyphagous) among the cultivated native, invasive, and non-invasive Eugenia was marginally insignificant (Pearson χ2 = 5.76, df = 2, P = 0.056). The difference in herbivore diet breadth was not statistically significant between the wild native Eugenia and wild invasive Eugenia (Pearson χ2 = 1.94, df = 1, P = 0.163). In addition, the proportions of herbivore feeding site (endophagous vs. ectophagous) were not different between the wild native Eugenia and wild invasive Eugenia (Pearson χ2 = 0.003, df = 1, P = 0.960), or among the cultivated plants (Pearson χ2 = 1.91, df = 2, P = 0.385) (Fig. 1B). Separate analyses (not reported here) incorporating the excluded uncommon Lepidoptera yielded similar results. Finally, all introduced Eugenia species attracted more exotic insect herbivores than the native Eugenia plants (Fig. 1C). However, the differences in the proportion of native herbivores were not significant between the wild native Eugenia and the wild invasive Eugenia (Pearson χ2 = 1.02, df = 1, P = 0.311), and among the cultivated native, invasive, and non-invasive Eugenia (Pearson χ2 = 0.76, df = 2, P = 0.683) (Fig. 1C).
The native Eugenia shared a total of 6 generalist herbivores, 4 with the invasive Eugenia, 4 with the non-invasive Eugenia, and 2 (the weevil Diaprepes abbreviatus L. and a kerriid scale Paratachardina lobata Chamberlin) with both kinds (Table 2). Among the shared herbivores, only 1 native weevil (Artipus floridanus Dietz) fed on the invasive Eugenia, while two native insects (Ancylis sp. and Elaphrothrips sp.) fed on the non-invasive Eugenia. The insect that caused substantial damage on the invasive Eugenia was an exotic weevil (Myllocerus undatus Marshall), while the insect that caused substantial damage on the non-invasive Eugenia was a native moth (Ancylis sp.). The native Eugenia also likely shared a specialist insect (a leaf blotch mining moth, Chilocampyla dyariella Busck) with the invasive congener (Table 2). However, it was not clear if the leaf miners were able to complete their development in E. uniflora leaves, because these incidents were rare and we were not able to rear any adults.
Discussion
Prediction 1-there will be greater numbers of herbivore species on native Eugenia than on introduced species.
There is limited evidence supporting our first prediction in relation to herbivore species richness on native vs. introduced non-invasive Eugenia because the cultivated native species had more insect herbivore species than 2 of the 3 introduced non-invasive species. This is consistent with the results found in a study comparing insect herbivore fauna between a native Pinus and a co-occurring introduced non-invasive congener (Lindelöw & Björkman 2001). There also was only limited support for the prediction in relation to the native vs. introduced invasive species in this study because the native Eugenia species had more insect herbivore species than the introduced invasive Eugenia in the wild, but not in cultivation. In the only other similar study (Bürki & Nentwig 1997), comparing the herbivore fauna between populations of the native Heracleum sphonylium L. and the co-occurring introduced invasive congeners, H. mantegazzianum Simmier & Levier, there was an equal number of insects associated with both plant species.
Furthermore, contrary to the second part of our first prediction that the invasive Eugenia should have a smaller number of herbivore species than the non-invasive congeners, the invasive Eugenia (E. uniflora), wild or in cultivation, had greater numbers of insect herbivore species than all 3 non-invasive Eugenia. This result is the opposite to that reported in a study on plant pathogens (Mitchell & Power 2003), in which the authors found that more invasive plants tended to have fewer pathogens. Nevertheless, differences in herbivore richness were small among the Eugenia species studied here. In addition, there is always the possibility that high number of herbivore species may not translate into high damage level (Liu, unpublished data).
Prediction 2-There will be greater numbers of oligophagous and endophagous herbivore species on native Eugenia than on introduced species.
The data support the first part of our second prediction that native Eugenia species should have the highest number and percentage of oligophagous insect herbivores. However, the statistical results should be interpreted with caution due to the small number of insect species on each Eugenia species. Our result is consistent with 1 congeneric native vs. introduced species comparison (Bürki & Nentwig 1997), but differs from another (Lindelöw & Björkman 2001). In addition, the native plants had higher number of internal feeders even though the percentage of endophagous herbivore species was not different between the native and introduced Eugenia. However, in contrast to the second part of our second prediction, the invasive Eugenia had as many or more oligophagous and/or endophagous feeders than non-invasive introduced Eugenia. No other studies were found to compare the number of oligophagous and endophagous insects between invasive and non-invasive plants.
Prediction 3-Fewer herbivores will be shared between native Eugenia and invasive Eugenia than between native Eugenia and non-invasive Eugenia.
The third prediction that native Eugenia should share fewer specialist and endophagous herbivores with invasive Eugenia than with non-invasive Eugenia was not supported by the data. While native Eugenia shared no oligophagous or endophagous herbivores with non-invasive Eugenia, they likely shared a leaf miner with E. uniflora (the invasive introduced Eugenia). However, because the blotch mines were only found on the wild individuals, it is possible that the host shift occurred after E. uniflora had invaded the natural areas. In addition, because the mines occurred at such a low rate the biotic resistance from this miner should be small. Host sharing by oligophagous herbivores largely depends on the taxonomic closeness of the host plants (Strong et al. 1984). A phylogeny of the genus Eugenia may help to explain and predict the shifts of specialists from the native to the introduced congeners.
No leaf galls were observed on any of the introduced Eugenia species in this study whereas one specialist galling fly, Eugeniamyia dispar Maia et al. (Diptera, Cecidomyiidae) (Maia et al. 1996) was found on E. uniflora in its native range. All introduced Eugenia studied here have probably escaped specialist insects that may be found in their native ranges. The lack of specialist insect attack may lead to a shift in plant resource allocation to growth (Blossey & Nötzold 1995; Siemann & Rogers 2001, Wolfe et al. 2004) and/or defense to generalist herbivores (Joshi & Vrieling 2005).
Native Eugenia plants in cultivation have a less diverse insect fauna than those in the wild, probably due to the differences in time since population establishment (Strong et al. 1984). Cultivated populations tend to be much younger and have less time to acquire insect fauna. Pesticide treatment in some horticulture or agriculture situations also may cause a decrease in herbivore fauna. However, all cultivated Eugenia individuals sampled in this study were not treated directly with pesticides (Jonathan H. Crane of TREC, Micheal Davenport of FTBG, Chris Rollins of FSP, personal communications). Nevertheless, our analyses and discussions are mostly limited to faunal comparisons among different species of the same source (wild or cultivated).
A result that is not related to the ERH testing but nonetheless interesting is the composition of native vs. exotic herbivores on the 3 categories of Eugenia plants. The native herbivores constituted about half of the insect herbivore fauna acquired by the introduced Eugenia. The numbers of exotic insects attacking the native, invasive, and non-invasive plants are similar (3-5 on each plant species). Since most of these exotic herbivores came from continents other than Central and South America, where the introduced Eugenia are native, it is unlikely that these exotic herbivores were associated with the exotic Eugenia in its native range. We did not observe any native insect herbivores having more importance on the introduced than on the native Eugenia plants. In contrast, it appeared that an exotic weevil (M. undatus), a new comer from Sri Lanka (Schall 2000), fed more heavily on E. uniflora (the invasive Eugenia) than on other congeners (Liu, personal observations). In addition, the only exotic oligophagous weevil (Myctides imberbis, an Australian native) found in this study also was observed on E. uniflora more than on the native or the non-invasive Eugenia. Together, our data suggested that the exotic herbivores provided as much, if not more, herbivore pressure as the native insects to the introduced Eugenia. Our finding was different from that of a recent study which found that the native herbivores, mostly vertebrates, suppressed introduced plants, whereas exotic herbivores, also mostly vertebrates, promoted exotic plants (Parker et al. 2006).
In summary, data on herbivore faunal diversity of Eugenia species provided limited support to the ERH. It is likely that other factors contribute to the success of E. uniflora. If we did not include the non-invasive Eugenia species in the study and only compared the herbivore fauna between the native Eugenia and invasive Eugenia, we would have thought that release from the insect herbivores was an important factor in the success of E. uniflora. We did not include pathogens, also recognized as natural enemies, in this study. Future study should take advantage of this unique three-way system to examine the effects of pathogens and other competing but non-exclusive hypotheses to help explain the success of E. uniflora. For example, competitive interactions of introduced invasive Eugenia vs. native co-occurring plants and non-invasive introduced Eugenia vs. native plants could be examined. The three- way comparison could also be used to examine the importance of relative seed numbers (the propagule pressure hypothesis (Williamson 1996), which states that the species with the greater number of propagules will be the most invasive). Eugenia uniflora, is much more abundant than the non-invasive Eugenia because it has long been used as a hedge plant, and probably produces more potentially invasive seeds.
Acknowledgments
Rita Duncan is acknowledged for helping and teaching Hong Liu insect preparation prior to identifications. Jose Allegria helped with field sampling. The following people helped with specimen identifications: M. C. Thomas (Coleoptera), S. E. Halbert (Homoptera: Flatidae), J. B. Heppner (Lepidoptera), G. S. Hodges (Coccidae), G. B. Edwards (Thysanoptera), and S. E. Halbert (Psyllidae), all at the Division of Plant Industry, Florida Department of Agriculture and Consumer Services; Don Davis (Gracillaridae and Yponomeutidae), Department of Entomology, Smithsonian Institute; David Adamski (Blastobasidae), USNM, C/O Department of Entomology, Smithsonian Institute; Ray Gagne (Cecidomyiidae), Systematic Entomology Laboratory, Agriculture Research Service, US Department of Agriculture. Support for this project came from NSF grant DEB 03-15190 to Peter Stiling.
References Cited
Fig. 1.
Percentage of herbivore species found in different categories on six wild and/or cultivated Eugenia species in South Florida. The vertical dash lines separate wild plants from cultivated ones, with the former on the left. W_axi = wild E. axillaris, w_foe = wild E. foetida, w_uni = wild E. uniflora, c_axi = cultivated E. axillaris, c_foe = cultivated E. foetida, c_uni = cultivated E. uniflora, c_agg = cultivated E. aggregata, c_bra = cultivated E. brasiliensis, c_lus = cultivated E. lushnathiana. “N” indicate native plants, “I” the introduced invasive plant, and “NI” the introduced non-invasive plants. Numbers on top of the bars are the total number of herbivore species found.
Table 1.
Summary of the study system, including the number of plants sampled (n). plants that grow in gardens are cultivated.
Table 2.
Herbivorous insect species found on six wild and/or cultivated Eugenia species in south FLORIDA. Native Eugenia species are in bold and invasive Eugenia are in italics. w_axi = wild E. axillaris, w_foe = wild E. foetida, w_uni = wild E. uniflora, c_axi = cultivated E. axillaris, c_foe = cultivated E. foetida, c_uni = cultivated E. uniflora, c_agg = cultivated E. aggregata, c_bra = cultivated E. brasiliensis, c_lus = cultivated E. lushnathiana, Poly = polyphagous or generalist. Oligo = oligophagous or specialist. Endo = endophagous, Ecto = ectophagous. - does not occur, + not important, ++ important, +++ very important. unid = unidentified. ? indicates unknown or uncertain information.
Table 2.
Continued
