Received: 8 February 2017 | Accepted: 20 July 2017 DOI: 10.1111/1365-2745.12845 RESEARCH ARTICLE Breakdown of a geographic cline explains high performance of introduced populations of a weedy invader Stacy B. Endriss1,2 | Christina Alba3,4 | Andrew P. Norton1,2 | Petr Pyšek3,5 | Ruth A. Hufbauer1,2 1Graduate Degree Program in Ecology, Colorado State University, Fort Collins, CO, Abstract USA 1. What drives the evolution of increased growth and fecundity in plants introduced 2 Department of Bioagricultural Sciences and to a novel range is not well understood. Pest Management, Colorado State University, Fort Collins, CO, USA 2. We investigate between-range differences in performance for Verbascum thapsus, 3Department of Invasion Ecology, Institute a weedy invader known to grow larger in its introduced than native range. of Botany, The Czech Academy of Sciences, Specifically, we question whether adaptation to herbivory or climate best explains Průhonice, Czech Republic 4Denver Botanic Gardens, Denver, CO, USA increased performance of introduced populations. 5Department of Ecology, Faculty of Science, 3. We grew 14 native and 22 introduced populations of V. thapsus in two common Charles University, Prague, Czech Republic garden locations: near Prague, Czech Republic (native range) and in Colorado, USA Correspondence (introduced range). By removing herbivores from half of the plants within each gar- Stacy B. Endriss den we tested the prediction of the evolution of increased competitive ability Email: [email protected] (EICA) hypothesis: increased performance is driven by an evolutionary shift of re- Funding information sources away from defence against herbivory towards growth and reproduction. Akademie Věd České Republiky, Grant/Award Number: RVO 67985939; Grantová Agentura We then investigated whether genetically based clines in performance are ex- České Republiky, Grant/Award Number: 14- pressed along climate gradients within both the native and introduced ranges. 36079G; Colorado State University Graduate Degree Program in Ecology; Colorado State 4. On average, seeds produce larger rosettes when collected from the introduced vs. University Agricultural Experiment Station native range. While this evolution of increased growth in introduced populations in Handling Editor: Mark Rees part matches the prediction of EICA, climate, not herbivory, best explains this be- tween-range difference. Specifically, seeds collected from the native range pro- duce smaller rosettes as the climate of origin becomes cooler and drier, while there is no cline in performance in rosettes grown from seed collected from the intro- duced range, which are large regardless of climate of origin. Thus, a climate-based cline within the native range best explains lower average performance of native compared to introduced populations. 5. Synthesis. The breakdown in a potentially adaptive cline emphasizes the need to more closely investigate the evolutionary processes that shape geographic struc- turing (or its absence) within the introduced range. In addition, EICA is not univer- sally applicable to all invasion scenarios, and our findings underscore the importance of testing the underlying assumptions alongside the predictions of this hypothesis. KEYWORDS abiotic clines, common garden, EICA, invasion ecology, plant–climate interactions, plant–herbivore interactions, Verbascum thapsus Journal of Ecology. 2018;106:699–713. wileyonlinelibrary.com/journal/jec © 2017 The Authors. Journal of Ecology | 699 © 2017 British Ecological Society 700 | Journal of Ecology ENDRISS ET AL. 1 | INTRODUCTION (Strauss & Agrawal, 1999). In addition, evidence for costs of defence is remarkably rare (Neilson, Goodger, Woodrow, & Møller, 2013; Plant populations often evolve increased growth and reproduction Siemens, Keck, & Ziegenbein, 2010), and is by no means ubiquitous after introduction to a novel range (Drenovsky et al., 2012; Whitney across species (Strauss, Rudgers, Lau, & Irwin, 2002). Rigorously test- & Gabler, 2008). However, what drives this evolutionary pattern is not ing both of the assumptions guiding the predictions of the EICA hy- well understood (Burton, Phillips, & Travis, 2010; Dunn et al., 2012; pothesis is therefore critical for determining whether this hypothesis Whitney & Gabler, 2008; Willis, Memmott, & Forrester, 2000). Many is even applicable to the given invasion scenario. hypotheses predict that increased performance of introduced popu- Furthermore, differences in herbivore composition and abundance lations is driven by adaptation to novel abiotic or biotic conditions, are not the only striking disparities between a plant’s native and intro- such as fewer natural enemies or a more benign environment, in a duced ranges. Climate also often varies between ranges (Bocsi et al., plant’s introduced as compared to native range (e.g. Alexander, van 2016; Early & Sax, 2014), and is a known driver of adaptation within Kleunen, Ghezzi, & Edwards, 2012; Blossey & Nötzold, 1995; Bossdorf invasive populations (Alexander et al., 2012; Lee & Kotanen, 2015; et al., 2005; Lee & Klasing, 2004; Mooney & Cleland, 2001; Sakai Maron, Vilà, Bommarco, Elmendorf, & Beardsley, 2004; Montague, et al., 2001). Parsing the relative importance of these different drivers Barrett, & Eckert, 2008; Zenni, Bailey, & Simberloff, 2014). Jointly of adaptation is critical to understanding current patterns of invader investigating EICA and climate- driven variation in performance is evolution as well as predicting which habitats are most susceptible to important for interpreting adaptive evolution for two main reasons. invasion moving forward. First, climate may be the major driver of local adaptation within One important adaptive hypothesis predicts that increased growth introduced populations, not reduced herbivory. For example, many is the evolutionary consequence of plants escaping many of their en- plant invaders rapidly evolve predictable phenotypic variation along emies upon introduction to a new range (the evolution of increased climatic or latitudinal clines, which often mimic pre- existing clines competitive ability [EICA] hypothesis; Blossey & Nötzold, 1995). within the native range (Alexander, 2013; Dlugosch & Parker, 2008b; Specifically, EICA states that reduced herbivory in the introduced Etterson, Delf, Craig, Ando, & Ohgushi, 2008; Hakam & Simon, 2000; range should lead to an evolutionary shift of resources away from Hodgins & Rieseberg, 2011; Keller, Sowell, Neiman, Wolfe, & Taylor, defence against herbivory towards growth and reproduction (Blossey 2009; Konarzewski, Murray, & Godfree, 2012; Leger & Rice, 2007; & Nötzold, 1995). However, despite strong evidence for enemy es- Montague et al., 2008; Novy, Flory, & Hartman, 2013; Roy, Simon, & cape (Colautti, Ricciardi, Grigorovich, & MacIsaac, 2004; Keane & Lapointe, 2000; Weber & Schmid, 1998). Crawley, 2002; Lui & Stiling, 2006), especially from specialists, evi- Second, if between- range differences in adaptation due to differ- dence in support of EICA is mixed (Felker- Quinn, Schweitzer, & Bailey, ences in herbivore pressure do exist, they may be obscured by abiotic 2013; Lowry et al., 2013). While introduced populations do evolve clines that are unaccounted for (Colautti, Maron, & Barrett, 2009). This decreased defence against herbivory in some plant species (Blair & is especially important as many plant populations vary predictably along Wolfe, 2004; Daehler & Strong, 1997), other plant species either climatic gradients in the two traits predicted by EICA to drive invasion invest similarly in defence in both ranges (Cripps, Hinz, McKenney, success: performance and defence against herbivory. Performance Price, & Schwarzländer, 2009; Franks, Pratt, Dray, & Simms, 2008; tends to decrease with increasing latitude in benign environments, as Huberty, Tielbörger, Harvey, Müller, & Macel, 2014), or evolve higher plants evolve decreased biomass, likely driven by strong selection for defence in introduced than native populations (Ridenour, Vivanco, earlier flowering (Abhilasha & Joshi, 2009; Colautti & Barrett, 2013; Feng, Horiuchi, & Callaway, 2008; Stastny, Schaffner, & Elle, 2005). Kooyers, Greenlee, Colicchio, Oh, & Blackman, 2015). However, Furthermore, increased size is often not correlated with a decrease in herbivory too is known to vary predictably along climatic gradients; defence in introduced relative to native populations (Bossdorf et al., herbivory, and thus plant defence, is often hypothesized to diminish 2005; Felker- Quinn et al., 2013). with increasing latitude (Johnson & Rasmann, 2011). Although sup- This inconsistent support for EICA must be explained if it is to re- port for this hypothesis is mixed (Anstett, Nunes, Baskett, & Kotanen, main a useful framework for understanding why many invasive pop- 2016; Moles, Bonser, Poore, Wallis, & Foley, 2011), defences such as ulations evolve greater performance in their introduced habitat. A trichomes (Kooyers et al., 2015) and secondary compounds (Anstett clear way forward is to test the two main assumptions that underlie et al., 2015; Pratt, Keefover- Ring, Liu, & Mooney, 2014; Reudler & the predictions of EICA: (1) that herbivory reduces plant fitness and Elzinga, 2015) do often decrease with increasing latitude. (2) that defences against herbivory are costly (Cipollini & Lieurance, It is also the case that non- adaptive contingencies associated with 2012; Cipollini, Mbagwu, Barto, Hillstrom, & Enright, 2005; Handley, introduction history may best explain between-