ARTICLE IN PRESS Perspectives in Plant Ecology, Evolution and Systematics Perspectives in Plant Ecology, Evolution and Systematics 11 (2009) 311–320 www.elsevier.de/ppees Forum Rethinking the common garden in invasion research Kirk A. Moloneya,, Claus Holzapfelb, Katja Tielbörgerc, Florian Jeltschd, Frank M. Schurrd a Department of Ecology, Evolution and Organismal Biology, Iowa State University, 253 Bessey Hall, Ames, IA 50011-1020, USA Department of Biological Sciences, Rutgers University, Newark, NJ 07102, USA c Department of Plant Ecology, University of Tübingen, D-72076 Tübingen, Germany d Plant Ecology and Nature Conservation, Universität Potsdam, D-14469 Potsdam, Germany b Received 18 December 2008; received in revised form 24 April 2009; accepted 31 May 2009 Abstract In common garden experiments, a number of genotypes are raised in a common environment in order to quantify the genetic component of phenotypic variation. Common gardens are thus ideally suited for disentangling how genetic and environmental factors contribute to the success of invasive species in their new non-native range. Although common garden experiments are increasingly employed in the study of invasive species, there has been little discussion about how these experiments should be designed for greatest utility. We argue that this has delayed progress in developing a general theory of invasion biology. We suggest a minimum optimal design (MOD) for common garden studies that target the ecological and evolutionary processes leading to phenotypic differentiation between native and invasive ranges. This involves four elements: (A) multiple, strategically sited garden locations, involving at the very least four gardens (2 in the native range and 2 in the invaded range); (B) careful consideration of the genetic design of the experiment; (C) standardization of experimental protocols across all gardens; and (D) care to ensure the biosafety of the experiment. Our understanding of the evolutionary ecology of biological invasions will be greatly enhanced by common garden studies, if and only if they are designed in a more systematic fashion, incorporating at the very least the MOD suggested here. r 2009 Rübel Foundation, ETH Zürich. Published by Elsevier GmbH. All rights reserved. Keywords: Adaptive evolution; Common garden; Lythrum salicaria; Maron effect; Phenotypic plasticity; Invasive species Introduction Invasive species are those naturalized, exotic species that exhibit runaway population growth once they become established and are generally considered to be of environmental concern due to their impact on native species and ecosystem services (Randall, 1996; Vitousek Corresponding author. E-mail address: kmoloney@iastate.edu (K.A. Moloney). et al., 1996; Kolar and Lodge, 2001; Pimentel et al., 2001; Sakai et al., 2001; Lockwood et al., 2007). Invasion biology aims to explain why certain species that are inconspicuous in their native range become successful invaders after introduction to new non-native ranges. This phenomenon might be explained by purely ecological differences in environmental conditions between the native and the introduced range. However, there is growing interest in quantifying the extent to which differential success of a species in its native and 1433-8319/$ - see front matter r 2009 Rübel Foundation, ETH Zürich. Published by Elsevier GmbH. All rights reserved. doi:10.1016/j.ppees.2009.05.002 ARTICLE IN PRESS 312 Table 1. K.A. Moloney et al. / Perspectives in Plant Ecology, Evolution and Systematics 11 (2009) 311–320 Experimental designs utilized in invasive plant studies incorporating a common garden approach. Garden location Provenance of plants Replication within location Number of Studies Hypotheses Examined (number of studies) N I N&I I No Yes 0 0 NA NA N&I No Yes 5 3 EICA(5)a PPvAE(2)b Misc(1)c I No Yes 7 2 PvAE(2)d Misc.(5)e PPvAE(2)f N&I No Yes 13 1 EICA(8)g NvI(2)h PPvAE(1)i Misc(2)j EICA(1)k I No Yes 0 0 NA NA N&I No Yes 3 2 PPvAE(1)l Misc(2)m EICA(2)n A literature search was conducted on Web of Science to find invasive species studies incorporating common gardens in their approach. We used the search phrase ‘‘(‘‘invasive’’ or ‘‘invasion’’ or ‘‘exotic’’) AND (‘‘Common Garden’’ OR ‘‘Common gardens’’) AND (‘‘plant’’ or ‘‘plants’’)’’, yielding 77 papers spanning the years from 1994 to early 2009 that explicitly incorporated common gardens in the study of invasion biology. We further pared the number of papers down to 33 (for a total of 36 independent experiments) that included only experimental studies explicitly utilizing common gardens to examine causes of phenotypic differentiation between plants from native and invasive ranges (Fig. 1). Garden locations could be in the native provenance (N) and/or in the invaded provenance (I) of the plants used in the experiment. Provenance of the plants used in the experiment was either from only the invaded range (I) or from both native and invaded ranges (N & I). Replication within location was considered to be yes (Y) for studies which included replicate gardens within a specific study location; otherwise it was evaluated as no (N). We also categorize studies with respect to the major hypotheses being tested and break this into 4 major categories: EICA – evolution of increased competitive ability; PPvAe – phenotypic plasticity versus adaptive evolution; NvI – trait comparison between plants of the native versus invasive provenance; and Misc – other assorted hypotheses regarding invasive species. a Wolfe et al. (2004); Stastny et al. (2005); Joshi and Vrieling (2005); Zou et al. (2008a, b). b Leger and Rice (2003); Leger and Rice (2007). c Williams et al. (2008). d Poulin et al. (2007) – 2 experiments e Kollmann and Banuelos (2004); Weber and Schmid (1998;) Weber and D’Antonio (1999); Ebeling et al. (2008); Monty and Mahy (2009). f Byers and Quinn (1998); Geng et al. (2007). g Siemann and Rogers (2001); Siemann and Rogers (2003b); van Kleunen and Schmid (2003); Blair and Wolfe (2004); Buschmann et al. 2005 (2005); Meyer et al. (2005); Rogers and Siemann (2005) – 2 experiments; h Guesewell et al. (2006); Feng et al. (2007). i Caño et al. (2008). j Buschmann et al. (2006); Wolfe et al. (2007). k Siemann and Rogers (2003a). l Maron et al. (2007). m Maron et al. (2004a); Williams et al. (2008). n Maron et al. (2004b); Genton et al. (2005). introduced range results from evolutionary processes causing genetic differentiation (Bossdorf et al., 2005; Wares et al., 2005; Keller and Taylor, 2008). Common garden experiments are ideally suited for disentangling how genetic and environmental factors contribute to the success of invasive species in their introduced range. In common garden experiments, a number of genotypes are raised in a common environment so that the phenotypic variation observed in the field can be attributed to genetic causes or to a non- genetic phenotypic response. The common garden approach has a long tradition in plant evolutionary ecology (Turesson, 1922; Clausen et al., 1940). In recent years, common garden experiments are increasingly being used to address research questions in invasion biology (see Table 1 and Fig. 1). Two of the most commonly addressed questions are (1) tests of the evolution of increased competitive ability hypothesis (EICA, 44% of the 36 experiments listed in Table 1; see Blossey and Nötzold (1995) for a definition of the EICA ARTICLE IN PRESS K.A. Moloney et al. / Perspectives in Plant Ecology, Evolution and Systematics 11 (2009) 311–320 7 Papers Published 6 5 4 3 2 1 0 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 Year Published Fig. 1. Number of papers published that contained a study focusing on invasive, exotic species with a common garden experiment (see Table 1 for details). hypothesis) and (2) an exploration of the relative importance of phenotypic plasticity versus adaptive evolution (PPvAE, 22%) in producing successful invaders. These two areas of research reflect current ideas regarding the most likely processes driving exotic species to become invasive. A long treatise could be written on the relative merits of the various hypotheses and research questions in invasion biology being explored using common gardens, but the intent of this paper is to focus on something that has not yet received serious attention: what, if any, is the ideal design for a common garden study that targets the ecological and evolutionary processes causing phenotypic differentiation between native and invasive ranges, and has this design ever been applied? The answers to these questions will obviously depend to some extent on the exact hypothesis being addressed, but it is clear from reviewing the literature that more critical thought needs to be given in developing appropriate designs (Table 1). Common garden studies with invasive species can be categorized according to their genetic and geographic design, using three basic criteria. The first criterion concerns the genetic composition of the experiment and distinguishes between two basic categories: (1) all plants included in the experiment originate from the invasive provenance (I – 25% of the experiments included in Table 1) or (2) they originate from both the native and invasive provenances (N & I – 75% of the experiments in Table 1). The second criterion refers to the geographic location of the experiment and distinguishes among three categories: (1) gardens located only within the native provenance, i.e., at sites from which the invasive species originates (N – 22% of the studies in Table 1); (2) gardens located solely within the invasive provenance (I – 64%); and (3) parallel studies conducted in both native and invasive provenances (N & I – 14%). The 313 latter category surprisingly comprises only five out of 36 experiments included in a literature search reported here in Table 1. The geographical location of the experiment is important because it is associated with uncontrolled environmental factors (e.g. climate, soil type, nutrient availability, interacting species, etc.) that may influence the outcome of the experiment and the interpretation of its results. The third criterion distinguishes between studies that replicate gardens within provenance (Y) versus those that do not (N). This is primarily a statistical issue that can strongly influence the inference space available to the interpretation of the outcome of common garden experiments. Generally, the experiments in Table 1 that used replicate gardens placed all gardens in only one geographic region. Two exceptions were the studies of Maron et al. (2004b) and Genton et al. (2005), which included replicated gardens within each of the N & I provenances to incorporate a geographic component into the inference space. Maron et al. (2004b) were interested in examining latitudinal trends in the evolution of invasive species between replicates within continent, whereas in Genton et al. (2005) replication was included as a blocking factor in the statistical analysis that compared responses between provenances. The categorization scheme described above results in 12 possible experimental designs. In our literature review (Table 1), we found that 8 of these designs had actually been implemented. Interestingly, the two most common research questions (EICA and PPvAE) were addressed with multiple designs. The EICA hypothesis alone was distributed among four designs, which implies that not enough attention has been paid to the relationship between the research question and the experimental design most appropriate for addressing it. Granted, the individual studies differed subtly in the specific aspects of the EICA hypothesis being tested and in some cases different designs might make sense. However, in general a systematic approach to develop an appropriate experimental design seems to be lacking. We argue that this is one reason for the slow development of a more general theory of invasive species, a situation that has been viewed as problematic by many invasion biologists (e.g., Cadotte et al., 2006). One indication that we need to give critical thought to the geographic design of common garden studies is provided by Maron et al. (2004b). Their study was one of the two encountered in our literature review that conducted parallel common garden experiments in both native and invasive provenances. They constructed four common gardens, two in Europe (N) and two in North America (I). One of the key findings of their study was that the results obtained in the different gardens were inconsistent due to an interaction effect between garden location and genetic provenance of the plants (a form ARTICLE IN PRESS 314 K.A. Moloney et al. / Perspectives in Plant Ecology, Evolution and Systematics 11 (2009) 311–320 of environment genotype, i.e., G E, interaction), a phenomenon that we will refer to in the following as the ‘‘Maron effect’’. The Maron effect can be defined as an interaction between provenance (source of seed) and garden location (a proxy for local environmental factors outside of the control of the experiment). Because of this effect, Maron et al. state that the overall conclusions drawn from their experiment would have been qualitatively different depending upon which subset of the four gardens was considered. Unfortunately, the designs they utilized in the different gardens were not standardized so that it is impossible to explicitly test for a ‘‘Maron effect’’ influencing the outcomes of the experiment. The only other study we found with replicate gardens located in both the native and invasive provenances had a standardized design imposed on all the gardens, but did not test for interactions between garden and the other experimental variables, provenance in particular (Genton et al., 2005). Genton et al. treated gardens within continent as a random effect and explored the effect of ‘‘continent’’ as a fixed effect. The random effect of garden was not significant in their study, but continent exhibited a strongly significant interaction with the seed source (i.e., provenance) treatment (called ‘‘origin’’ in the original paper). The continent by provenance interaction could be due to a number of factors, but Genton et al. (2005) point out that plants from one collection site were the smallest when planted in the invasive continent and the largest when grown in the native continent. They do not discuss other aspects of the study resulting from the highly significant continent provenance interactions. The findings of Genton et al. (2005), along with those presented in Maron et al. (2004b), indicate that we need to consider how important a ‘‘Maron effect’’ might be for common garden experiments with invasive species. If results strongly depend upon the location of common gardens, how do we make sense of experiments that have already been conducted and how should we design future common garden experiments? To address these concerns, we propose an experimental design that explicitly accounts for potential interactions between a ‘‘garden effect’’ and experimentally manipulated variables, thereby allowing an explicit test for the ‘‘Maron effect’’. A minimal design for common garden experiments with invasive species In our view, there are two critical elements in the design of a common garden experiment that examines the genetic and environmental factors potentially causing differences between the performance of a species in its native and introduced range: (1) the location of gardens, which should include both the native and invasive provenance and should be replicated within each, and (2) the choice of genetic source material, which should include appropriate genotypes from both the native and invasive provenance. Many common garden designs take the genetic source material into account, but most often garden location is not appropriately integrated into the experimental design. Fig. 2. Trait response of plants of different provenance planted in different garden locations. (A) No interaction between garden and plant provenance in determining trait expression. (B) ‘‘Maron effect’’, i.e., a significant interaction between garden and plant provenance in determining trait expression. Lower case letters associated with lines indicate provenance of experimental plants that correspond to garden locations indicated by the same upper case letter. The relationships illustrated in panel B represent a situation where adaptive evolution leads to an optimal response by plants to the conditions in their home garden, although they may still be outperformed by a plant originating from a different location. ARTICLE IN PRESS K.A. Moloney et al. / Perspectives in Plant Ecology, Evolution and Systematics 11 (2009) 311–320 Generally, the choice of garden location(s) is based on logistical considerations alone, due to the substantial effort required to successfully conduct common garden experiments. However, serious issues of interpretation may arise if the ‘‘Maron effect’’ is widespread and unappreciated. How important might the ‘‘Maron effect’’ be for common garden experiments exploring the evolution of invasiveness? We believe that the answer is ‘‘Quite important!’’, and explore this in a hypothetical example illustrating the potential problems arising from only including one common garden in a study. Imagine four genotypes sampled from four different provenances (locations). In the absence of a Maron effect, the rank order trait response of all genotypes would remain the same across gardens, independent of where they are sited (Fig. 2A). Interpretation of the results would be independent of garden location. However, if the Maron effect is important, then the trait response would depend upon the location of the garden. For example, if each genotype has its highest fitness in its home site (fulfilling the ‘‘home-away’’ criterion of Kawecki and Ebert, 2004), then we might find a scenario similar to the one depicted in Fig. 2B, which is consistent with the findings of Maron et al. (2004b) and Genton et al. (2005) that the observed response depends critically upon the location of the garden, as described above. The existence of a Maron effect may, in fact, help explain the diverse findings reported in studies of the EICA hypothesis (cf., Colautti et al., 2004; Joshi and Vrieling, 2005; Williams et al., 2008). A primary prediction of the EICA is that plants from the invasive provenance have evolved to become better competitors than plants from the native provenance, due to escape from local herbivores, parasites and disease organisms. However, experimental tests of this hypothesis have shown a broad range of results, ranging from complete agreement (Garden D in Fig. 2B; e.g., Blossey and Nötzold, 1995; Siemann and Rogers, 2001; Wolfe et al., 2004), to equivocal results (Gardens B and C in Fig. 2B; e.g., Willis et al., 2000; van Kleunen and Schmid, 2003) and findings contradicting the hypothesis (Garden A in Fig. 2B; e.g., Bossdorf et al., 2004). The conflicting outcomes could in fact be due to an underlying ‘‘Maron effect’’ with the observed outcome depending upon the site of the experiment (since most experiments are only conducted at one location), as much as upon the underlying relationships between plants of native and invasive provenance. In summary, Maron effects may be important for both fundamental and applied invasion biology: the simple native-versus-invasive hypotheses would fail to explain evolutionary processes, if native genotypes outperform exotic genotypes in parts of the invasive range (as in Garden C in Fig. 2b). This can also be of applied importance: the invasion would be 315 worsened if native genotype b was introduced to site C in the invasive range (Fig. 2b). To address the issues discussed above, we propose a minimum optimal design (MOD) for common garden experiments examining genetic and environmental differences between the native and invasive range (note that similar principles also apply to experiments examining differentiation within the native or invasive range, e.g. Kollmann and Banuelos, 2004). The MOD enables one to test for the ‘‘Maron effect’’ and contains four elements: (1) Garden locations: There should be at least four gardens, two in each provenance (N & I). This provides replication within provenance. In experiments testing adaptive evolutionary hypotheses, available phylogeographic knowledge should be used to identify garden regions that provide proper genetic replication (Keller and Taylor, 2008). Ideally, the gardens should cover a broad spectrum of the environmental variation present in both the native and the invasive range. Other elements of garden location might also be of interest to explore, e.g., in the invasive range sites could be chosen to correspond with different stages of invasion history. The latter could be matched by choosing sites from the native provenance that incorporate other important aspects associated with locations in the invaded range, such as latitude or general climatic conditions. Including more factors associated with garden location will obviously increase the number of gardens needed in the study to allow for replication of the additional design elements. (2) Genetic composition: Genetic material should be obtained using the same, balanced, sampling design in each region. The genetic material should ideally be composed of a representative, random sample of the genotypes within each region. Additionally, genetic composition of plants from each region should be tightly controlled to ensure balanced representation of the genetic makeup from all regions included in the experiment: for example, seeds collected from individual plants representing maternal half-sibs in open pollinated species; or haphazard collection of seeds from a large number of plants that are combined to contain a random sample representative of the region at large. The choice of how to collect and use seed at the level of the region will determine the inference space of the experiment. The genetic material from each region should be equally represented in each garden. To remove maternal environmental effects, material from different origins should ideally be cultivated over multiple generations in a common environment. Less ideally, maternal environmental effects can be accounted for ARTICLE IN PRESS 316 K.A. Moloney et al. / Perspectives in Plant Ecology, Evolution and Systematics 11 (2009) 311–320 by including proxies for the quality of the maternal environment (such as initial seedling size) as covariates in analyses of experimental results. (3) Standardization: The experimental protocols should be standardized across gardens (to ensure, for instance, that response variables are measured in the same way). Standardization is important because differences among gardens may ultimately result from two factors: (1) environmental differences between sites and (2) experimental error due to uncontrolled differences in protocol and design. The more the latter can be eliminated, the more will be learned about environmental variation between gardens. Depending on the exact hypothesis addressed, it may also be important to control certain environmental factors of the experiment (e.g. soils, nutrients, water regimes, interacting species, etc.). Controlling certain aspects of the environment while letting others vary helps to identify those environmental factors that cause between-garden variation (e.g., if soils are kept constant, observed differences should be due to climate and/or biotic interactions). While standardization is perhaps the most difficult requirement to meet, it is indispensable for understanding the causes of a potential ‘‘Maron effect’’. (4) Biosafety: Our proposed design could potentially cause serious biosafety problems because it requires the introduction of foreign germplasm into the garden sites. This may produce even more aggressive plants in the invaded provenance and could lead to back invasion in the native range by aggressive plants that evolved in the invaded range (cf., Genton et al., 2005). These concerns may limit the number of species that can be used in the type of experiment we are proposing. Great care needs to be taken to minimize the potential for gene escape in both the native and invasive provenance. A brief example Based on the principles outlined above, we conducted an experiment with Lythrum salicaria L., an aggressive wetland invader in North America that originated from Eurasia. The experiment was designed to test whether native and invasive genotypes of L. salicaria differ in their response to nutrient availability and water regime (note, however, that the exact hypothesis is not relevant for our argument). It builds upon an experiment that was conducted in a single common garden in Ames, Iowa, USA and is reported in Chun et al. (2007). The experiment involved two environmental treatments (water regime and nutrient level) in a factorial design (see detailed explanation in Chun et al., 2007). We followed the same general protocol as in that experiment, but adapted it to meet the four criteria for the minimal design outlined above: (1) Garden locations: There were four common gardens in our experiment, two in the invaded provenance – Iowa (IA) and New Jersey (NJ), USA – and two in the native provenance – Potsdam (PD) in NE Germany and Tübingen (TÜ) in SW Germany – providing replication within provenance. The gardens within provenance were sited in locations representing substantially different environmental conditions encountered within the distributional range of L. salicaria. (2) Genetic composition: Genetic material was obtained from three populations within the region surrounding each of the four common gardens. Seeds were collected from several maternal parents in each population. From these seed collections, five half-sib families from each population were eventually utilized in the experiment (Fig. 3). Four half-sibs from each parent were planted in each garden, appearing once in each environmental treatment. To Fig. 3. Design of an experiment employing the MOD proposed in this paper for a common garden study examining ecoevolutionary relationships in an invasive species. The common garden sites are indicated by the Region level in the hierarchy depicted in the figure. Genetic sampling was conducted at the level of maternal half-sibs nested within the hierarchy of the experimental design, which includes replicate populations nested within each garden region. See text for further explanation. Regions are Iowa (IA) and New Jersey (NJ) in the United States and Potsdam (PD) and Tübingen (TU) in Germany. Three populations, indicated by two letter acronyms, were sampled within each region. Overview of the impact of the ‘‘garden effect’’ on the outcome of a study utilizing the MOD proposed in this paper. ARTICLE IN PRESS Analyses were conducted using mixed-effects models in R (function lmer, Pinheiro and Bates 2004). After fitting the maximal model, the model was simplified using stepwise backward removal of insignificant terms (P40.05). The table indicates the fixed effect and covariate terms that remained in the resulting minimal adequate model by a solid fill in the corresponding table cell. Significant interactions that involve garden and provenance, thereby indicating a ‘‘Moran effect’’, are highlighted in black. Fixed effects in the model were Water (two levels), Nutrients (two levels), Provenance (Prov in the table; two levels, corresponding to origin of seed with respect to continent associated with N or I provenances) and Garden (four levels associated with the common four gardens in the experiment). Populations nested within garden region, half-sib families nested within populations, blocks and wading pools within blocks were included as crossed random effects (not shown in the table). Further details regarding the design of the experiment and the protocols used can be obtained from the text, Chun et al. (2007) or by contacting the corresponding author. K.A. Moloney et al. / Perspectives in Plant Ecology, Evolution and Systematics 11 (2009) 311–320 Table 2. 317 ARTICLE IN PRESS 318 K.A. Moloney et al. / Perspectives in Plant Ecology, Evolution and Systematics 11 (2009) 311–320 account for maternal environmental effects we included height at transplant as a covariate in statistical analyses. (3) Standardization: The protocols used in all gardens were coordinated to limit variability among gardens. For example, watering was done on exactly the same schedule in all four gardens, after plants were established. We used the same fertilizer formulation from the same manufacturer in all gardens. Soil composition was matched as closely as possible among gardens, using commercially obtained topsoil with similar starting nutrient charges. Measurements of response variables were taken using the same schedule and the same protocols. One team member coordinated all efforts to ensure as much consistency as possible. Still, it must be said that one lesson learned from this experiment is that standardization is the aspect of the MOD that is by far the hardest to achieve. (4) Biosafety: Gardens were located at least 3 km from any known naturally occurring L. salicaria plants to prohibit gene flow out of the experimental garden. Garden sites will be monitored for the foreseeable future to insure that plants do not successfully escape and establish. We have conducted a general analysis of the response variables in our experiment to determine how pervasive the garden effect might be in influencing the outcome of the experiment. Of particular importance are significant interactions between garden and provenance as these indicate the presence of a ‘‘Maron effect’’. Table 2 shows an overview of the analysis and highlights the importance of the ‘‘Maron effect’’ in our experiment: significant interactions involving garden and provenance were found for 8 of 13 response variables analyzed. Moreover, all except one response variable were significantly impacted by at least one interaction between garden and an experimentally manipulated variable. Only the ratio of leaf width to length showed no ‘‘garden effect’’, but also was not affected by any other explanatory variable, presumably because it is a genetically fixed trait. The implications of our finding are that we could grossly misinterpret the differences between plants from different provenances if we only conducted the experiment in one garden. This emphasizes the need for conducting common garden experiments that apply or expand upon the MOD proposed above. the location of common gardens then there is a strong case for applying, at the least, our MOD protocol in common garden experiments with invasive species. In the absence of utilizing multiple gardens situated in the native and invasive provenance, the interpretation of the results of a common garden study should not be extrapolated beyond the context of the individual site at which the garden was situated. This may also explain why it has been very difficult to find universal trends in the processes leading to successful invasion, since most common garden studies do not replicate gardens within or across provenances. The outcome of the experiment may depend on the particular site at which it is conducted, but without gardens sited at multiple locations the implications of the ‘‘Maron effect’’ will go unappreciated and undetected. Adopting the proposed MOD protocol as a minimum in conducting common garden studies is the first step in advancing our understanding of the processes leading to successful invasions. Perhaps the most challenging task is to identify those components of the environment that are involved in genotype-by-environment interactions and thereby cause the Maron effect. To this end, the MOD can be extended by experimentally manipulating the environmental factor hypothesized to be important to the levels observed in each of the provenance regions (see Nuismer and Gandon (2008) for an experimental design suited to do this in the case of tight coevolution with an interacting species). Understanding the environmental factors involved in the ‘‘Maron effect’’ is crucial for extrapolation beyond common garden locations and for developing a more general understanding of the evolutionary ecology of biological invasions. Acknowledgements We would like to thank Jack Chapman, Young Jin Chun, Lina Weiß, and Petra Finkenbein for help in implementing the MOD. KAM gratefully acknowledges support by the Velux Foundation and Iowa State University for his sabbatical at the ETH in Zürich, during which time these ideas came to fruition. No support was provided by the United States National Science Foundation or United States Department of Agriculture. References Conclusion If our experiment and the experiments of Maron et al. (2004b) and Genton et al. (2005) are indicative of a general relationship between experimental outcome and Blair, A.C., Wolfe, L.M., 2004. The evolution of an invasive plant: an experimental study with Silene latifolia. Ecology 85, 3035–3042. Blossey, B., Nötzold, R., 1995. Evolution of increased competitive ability in invasive nonindigenous plants: a hypothesis. Journal of Ecology 83, 887–889. ARTICLE IN PRESS K.A. Moloney et al. / Perspectives in Plant Ecology, Evolution and Systematics 11 (2009) 311–320 Bossdorf, O., Auge, H., Lafuma, L., Rogers, W.E., Siemann, E., Prati, D., 2005. Phenotypic and genetic differentiation between native and introduced plant populations. Oecologia 144, 1–11. Bossdorf, O., Prati, D., Auge, H., Schmid, B., 2004. Reduced competitive ability in an invasive plant. Ecology Letters 7, 346–353. Buschmann, H., Edwards, P.J., Dietz, H., 2005. Variation in growth pattern and response to slug damage among native and invasive provenances of four perennial Brassicaceae species. Journal of Ecology 93, 322–334. Buschmann, H., Edwards, P.J., Dietz, H., 2006. Responses of native and invasive Brassicaceae species to slug herbivory. Acta Oecologica-International Journal of Ecology 30, 126–135. Byers, D.L., Quinn, J.A., 1998. Demographic variation in Alliaria petiolata (Brassicaceae) in four contrasting habitats. Journal of the Torrey Botanical Society 125, 138–149. Cadotte, M.W., McMahon, S.M., Fukami, T. (Eds.), 2006. Conceptual Ecology and Invasion Biology: Reciprocal Approaches to Nature. Springer, Dordrecht, Netherlands. Caño, L., Escarré, J., Fleck, I., Blanco-Moreno, J.M., Sans, F.X., 2008. Increased fitness and plasticity of an invasive species in its introduced range: a study using Senecio pterophorus. Journal of Ecology 96, 468–476. Chun, Y.J., Collyer, M.L., Moloney, K.A., Nason, J.D., 2007. Phenotypic plasticity of native vs. invasive purple loosestrife: a two-state multivariate approach. Ecology 88, 1499–1512. Clausen, J., Keck, D., Heisey, W., 1940. Experimental studies on the nature of species. I. Effect of varied environments on western North American plants. Carnegie Institution of Washington Publications 520, 1–452. Colautti, R.I., Ricciardi, A., Grigorovich, I.A., MacIsaac, H.J., 2004. Is invasion success explained by the enemy release hypothesis? Ecology Letters 7, 721–733. Ebeling, S.K., Welk, E., Auge, H., Bruelheide, H., 2008. Predicting the spread of an invasive plant: combining experiments and ecological niche model. Ecography 31, 709–719. Feng, Y.L., Auge, H., Ebeling, S.K., 2007. Invasive Buddleja davidii allocates more nitrogen to its photosynthetic machinery than five native woody species. Oecologia 153, 501–510. Geng, Y.P., Pan, X.Y., Xu, C.Y., Zhang, W.J., Li, B., Chen, J.K., Lu, B.R., Song, Z.P., 2007. Phenotypic plasticity rather than locally adapted ecotypes allows the invasive alligator weed to colonize a wide range of habitats. Biological Invasions 9, 245–256. Genton, B.J., Kotanen, P.M., Cheptou, P.O., Adolphe, C., Shykoff, J.A., 2005. Enemy release but no evolutionary loss of defence in a plant invasion: an inter-continental reciprocal transplant experiment. Oecologia 146, 404–414. Guesewell, S., Jakobs, G., Weber, E., 2006. Native and introduced populations of Solidago gigantea differ in shoot production but not in leaf traits or litter decomposition. Functional Ecology 20, 575–584. Joshi, J., Vrieling, K., 2005. The enemy release and EICA hypothesis revisited: incorporating the fundamental differ- 319 ence between specialist and generalist herbivores. Ecology Letters 8, 704–714. Kawecki, T.J., Ebert, D., 2004. Conceptual issues in local adaptation. Ecology Letters 7 (12), 1225–1241. Keller, S.R., Taylor, D.R., 2008. History, chance and adaptation during biological invasion: separating stochastic phenotypic evolution from response to selection. Ecology Letters 11, 1–15. Kolar, C.S., Lodge, D.M., 2001. Progress in invasion biology: predicting invaders. Trends in Ecology and Evolution 16, 199–204. Kollmann, J., Banuelos, M.J., 2004. Latitudinal trends in growth and phenology of the invasive alien plant Impatiens glandulifera (Balsaminaceae). Diversity and Distributions 10, 377–385. Leger, E.A., Rice, K.J., 2003. Invasive California poppies (Eschscholzia californica Cham.) grow larger than native individuals under reduced competition. Ecology Letters 6, 257–264. Leger, E.A., Rice, K.J., 2007. Assessing the speed and predictability of local adaptation in invasive California poppies (Eschscholzia californica). Journal of Evolutionary Biology 20, 1090–1103. Lockwood, J.L., Hoopes, M.F., Marchetti, M.P., 2007. Invasion Ecology. Blackwell, Malden, MA. Maron, J.L., Elmendorf, S.C., Vilá, M., 2007. Contrasting plant physiological adaptation to climate in the native and introduced range of Hypericum perforatum. Evolution 61, 1912–1924. Maron, J.L., Vilá, M., Arnason, J., 2004a. Loss of enemy resistance among introduced populations of St. John’s Wort (Hypericum perforatum). Ecology 85, 3243–3253. Maron, J.L., Vilá, M., Bommarco, R., Elmendorf, S., Beardsley, P., 2004b. Rapid evolution of an invasive plant. Ecological Monographs 74, 261–280. Meyer, G., Clare, R., Weber, E., 2005. An experimental test of the evolution of increased competitive ability hypothesis in goldenrod, Solidago gigantea. Oecologia 144, 299–307. Monty, A., Mahy, G., 2009. Clinal differentiation during invasion: Senecio inaequidens (Asteraceae) along altitudinal gradients in Europe. Oecologia 159, 305–315. Nuismer, S.L., Gandon, S., 2008. Moving beyond commongarden and transplant designs: insight into the causes of local adaptation in species interactions. American Naturalist 171, 658–668. Pimentel, D., McNair, S., Janecka, J., Wightman, J., Simmonds, C., O’Connell, C., Wong, E., Russel, L., Zern, J., Aquino, T., Tsomondo, T., 2001. Economic and environmental threats of alien plant, animal, and microbe invasions. Agriculture Ecosystems and Environment 84, 1–20. Pinheiro, J.C., Bates, D.M., 2004. Mixed-effects models in S and S-PLUS. Springer Science+Business Media, LLC, New York. Poulin, J., Sakai, A.K., Weller, S.G., Nguyen, T., 2007. Phenotypic plasticity, precipitation, and invasiveness in the fire-promoting grass Pennisetum setaceum (poaceae). American Journal of Botany 94, 533–541. ARTICLE IN PRESS 320 K.A. Moloney et al. / Perspectives in Plant Ecology, Evolution and Systematics 11 (2009) 311–320 Randall, J., 1996. Weed control for the preservation of biological diversity. Weed Technology 10, 370–383. Rogers, W.E., Siemann, E., 2005. Herbivory tolerance and compensatory differences in native and invasive ecotypes of Chinese tallow tree (Sapium sebiferum). Plant Ecology 181, 57–68. Sakai, A.K., Allendorf, F.W., Holt, J.S., 2001. The population biology of invasive species. Annual Review of Ecology and Systematics 32, 305–332. Siemann, E., Rogers, W.E., 2001. Genetic differences in growth of an invasive tree species. Ecology Letters 4, 514–518. Siemann, E., Rogers, W.E., 2003a. Increased competitive ability of an invasive tree may be limited by an invasive beetle. Ecological Applications 13, 1503–1507. Siemann, E., Rogers, W.E., 2003b. Reduced resistance of invasive varieties of the alien tree Sapium sebiferum to a generalist herbivore. Oecologia 135, 451–457. Stastny, M., Schaffner, U., Elle, E., 2005. Do vigour of introduced populations and escape from specialist herbivores contribute to invasiveness? Journal of Ecology 93, 27–37. Turesson, G., 1922. The species and variety as ecological units. Hereditas 3, 100–113. van Kleunen, M., Schmid, B., 2003. No evidence for an evolutionary increased competitive ability in an invasive plant. Ecology 84, 2816–2823. Vitousek, P.M., Dantonio, C.M., Loope, L.L., Westbrooks, R., 1996. Biological invasions as global environmental change. American Scientist 84, 468–478. Wares, J.P., Hughes, A.R., Grosberg, R.K., 2005. Mechanisms that drive evolutionary change. Insights from species introductions and invasions. In: Sax, D.F, Stachowicz, J.J., Gaines, S.D. (Eds.), Species Invasions. Insights into ecology. evolution and biogeography. Sinauer, Sunderland, MA. Weber, E., D’Antonio, C.M., 1999. Phenotypic plasticity in hybridizing Carpobrotus spp. (Aizoaceae) from coastal California and its role in plant invasion. Canadian journal of botany 77, 1411–1418. Weber, E., Schmid, B., 1998. Latitudinal population differentiation in two species of Solidago (Asteraceae) introduced into Europe. American Journal of Botany 85, 1110–1121. Williams, J., Auge, H., Maron, J., 2008. Different gardens, different results: native and introduced populations exhibit contrasting phenotypes across common gardens. Oecologia 157, 239–248. Willis, A.J., Memmott, J., Forrester, R.I., 2000. Is there evidence for the post-invasion evolution of increased size among invasive plant species? Ecology Letters 3, 275–283. Wolfe, L.M., Blair, A.C., Penna, B.M., 2007. Does intraspecific hybridization contribute to the evolution of invasiveness?: an experimental test. Biological Invasions 9, 515–521. Wolfe, L.M., Elzinga, J.A., Biere, A., 2004. Increased susceptibility to enemies following introduction in the invasive plant Silene latifolia. Ecology Letters 7, 813–820. Zou, J., Rogers, W., Siemann, E., 2008a. Increased competitive ability and herbivory tolerance in the invasive plant Sapium sebiferum. Biological Invasions 10, 291–302. Zou, J., Siemann, E., Rogers, W.E., DeWalt, S.J., 2008b. Decreased resistance and increased tolerance to native herbivores of the invasive plant Sapium sebiferum. Ecography 31, 663–671.