Beware: alien invasion. Where to next for an understanding of weed ecology?


In this issue of New Phytologist, Dawson et al. (pp. 859–867) take an important step in assessing the role of plant traits, and plasticity therein, in determining invasiveness. They ask whether alien plant species that have larger invasion ranges also have greater plasticity in three functionally important traits: biomass, root/shoot ratio (R:S ) and specific leaf area (SLA). This approach is novel in that it effectively treats invasiveness as a continuum, rather than a categorical (invasive/noninvasive) state. In addition it examines plasticity in ecologically significant plant functional traits, which is of interest given the association of these traits with global species distributions (see Wright et al., 2004; Nicotra et al., 2010).

This approach is novel in that it effectively treats invasiveness as a continuum, rather than a categorical (invasive/noninvasive) state.

The ability of a species to respond to changes to environmental conditions, particularly increased resource availability, is often proposed to facilitate invasions (Baker, 1965; Davis et al., 2000). ‘Phenotypic plasticity’ describes an organism’s morphological, anatomical and developmental response to the environment (Schlitching, 1986). Analyses of the broad literature tend to support the concept that invasive plants display higher plasticity (Daehler, 2003; Davidson et al., 2011; but see Palacio-López & Gianoli, 2011).

When asking a question about movement and potential diversification of species, the characteristics of species’ lineages, and the shared evolutionary history of the species in these lineages, is important. Dawson et al. explicitly incorporate phylogeny into their meta-analysis and find differing degrees of phylogenetic structure in the traits they consider (little in biomass and SLA but quite a lot in R:S ratio). Overall their analyses shows that widespread invasive species have greater plasticity in biomass, but plasticity in both R:S and SLA is not correlated with number of regions invaded.

Dawson et al. conclude that focusing on the ‘endpoint of trait plasticity’– the performance or fitness of the plant under different environmental conditions – may be a more fruitful approach than continued studies of plasticity in functional traits themselves. They recommend detailed field experiments on multiple co-occurring native and alien species. We agree that the endpoint of plasticity is critical, but as we discuss, we are reticent to discourage further consideration of plasticity in underlying functional traits.

When is plasticity important to the invasion process?

An invasion can be divided into two stages; (1) introduction and naturalization and (2) expansion of species (Williamson, 1996). Phenotypic plasticity may be involved directly in either or both stages. Dawson et al. analysed whether plasticity in functional traits is associated with stage two. It is possible that plasticity is more important for stage one, establishment in novel environments, than for spread or competiveness (Palacio-López & Gianoli, 2011). Such a role for plasticity is consistent with suggestions that pioneer species may be more plastic than nonpioneer species (Bazzaz, 1979). Although invasive plants are generally pioneering species, in that they often establish in novel environments with very different climates from their home range (Gallagher et al., 2010), not all pioneer plants are invasive. As such, there is a need for further investigation to separate whether plasticity is associated with pioneering characteristics and/or plays a more direct role in providing a competitive advantage of invasive species over the native flora.

With regard to stage two, it is likely that plasticity in functional traits would be related not simply to expansion of geographic range, but to increasing environmental range. This is a subtle but important difference. Dawson et al. assessed whether the number of regions invaded was correlated with levels of plasticity. This approach could be extended to address the more complicated issue of whether plasticity in functional traits is associated with the number of biomes/habitat types an invasive covers. The question then becomes one of how much larger an environmental range an invasive species can occupy and whether plasticity in functional traits facilitates this expansion.

It has also been proposed that instead of being a characteristic that increases the likelihood that a species will become invasive, plasticity may be a by-product of selection on changes to mean values of traits during invasion (rapid evolution of plasticity post-colonization; Agrawal, 2001). However, there is only limited support for the latter hypothesis: Colautti et al. (2009) found no consistent evidence of evolution for changes in mean trait values between native and invasive populations in a meta-analysis of 28 species. Likewise, glasshouse experiments paired with genetic analyses of invasive Senecio inaequidens plants in Europe revealed the native population that was most closely related to the invasive populations also had the greatest phenotypic plasticity (Bossdorf et al., 2008). Thus, it seems that high trait plasticity is a pre-existing characteristic of invasive species.

Patterns of plasticity in functional traits may differ under stressful vs favourable conditions

In evaluating the relative response of invasive and noninvasive species it is important to consider the environmental context. As described in Richards et al. (2006) and further discussed in Davidson et al. (2011) plasticity in underlying traits would be adaptive in an invasive species if it enabled a genotype to maximize fitness under optimal conditions (‘master-of-some’ response to increased resources) or maintain homeostatic fitness under poor conditions (‘jack-of-all-trades’ response to decreased resources) or both (jack-and-master). The master-of-some response provides a mechanism by which higher plasticity of invasive species could enable invasive species to out-compete native species and thus facilitate the invasion process.

Dawson et al. examined plasticity in response to an increase in resources; however, the range of conditions examined inevitably varied across studies within the meta-analyses. Especially in the context of climate change (e.g. increased frequency of drought events in many areas) it remains topical to differentiate between responses to a reduction in a resource below average conditions, and responses to an increase in that resource. For example, Davidson et al. (2011) found that, in response to a decrease in resources from average to deficient, native species were better able to maintain fitness homeostasis than co-occurring invasive species.

To put this in a slightly different context, Poorter et al. (2012) advocate the use of dose–response curves because measures of plasticity depend strongly on the conditions under which the plastic response (or reaction norm) is assessed. Rather than quantifying plasticity at a discrete interval (noting that Dawson et al. did include magnitude of resource level increase as a covariate for the chosen interval), a dose–response curve integrates over a wide range of conditions and investigates changes in the response across the conditions (Fig. 1; Poorter et al., 2012). Thus, the question of whether invasive or noninvasive species differ in their ability to respond adaptively to decreased resource availability remains somewhat open.

Figure 1.

Dose–response curves of the absolute response of the fraction of 10 whole plant mass represented by roots (RMF, analogous to the scaled root/shoot ratio (R:S) in Dawson et al., this issue pp. 859–867) to (a) nutrient 8 availability, (b) water availability and (c) of the response of specific leaf area (SLA) to light availability. Data are a compilation based on many species. For each environmental factor, a reference condition was chosen (indicated by a vertical line), and data for each species in each experiment were subtracted from the allocation values observed or interpolated for that reference level. The shaded area indicates the interquartile range (between 25th and 75th percentile) of the observed ratios in that part of the response curve. The dotted lines indicate the 10th and 90th percentiles. The bold continuous line within the shaded area indicates the median value. Both traits show plasticity, but in the case of the response of RMF to water availability in particular, the position of sampling points along the response curve will dramatically affect the estimate of plasticity. Figures modified with permission from; see Poorter et al. (2009, 2012) for further information.

The importance, and the challenge, of assessing relative performance

The adaptive value of a plastic response depends on whether it increases average net fitness (ideally taken as multigenerational fitness measures). Measuring fitness, however, is not a simple matter and relies on proxies, which must be selected with care (see Box 1 in Davidson et al., 2011). Measurements of adaptive plasticity should also consider the representation of different conditions in the environment and assess the impacts of altering the frequencies/likelihoods of encountering these different resource conditions. Furthermore, one can consider adaptive plasticity at different time scales: plasticity which was adaptive under past conditions and may be of neutral or maladaptive importance now; plasticity which is currently adaptive; and plasticity which may now be neutral or maladaptive but that could represent hidden adaptive potential under novel environments. Each of these is potentially important: the first for understanding the history of diversification of lineages, the second for understanding current selective pressures and the last for determining responses to future environments. The latter two are therefore relevant for understanding invasion biology and predicting responses to climate change.

In the context of invasions, however, the important question is not simply whether plasticity in a given trait is adaptive, but how the performance of invasive compared to native plants differs under the same conditions. Superior performance may be underpinned by greater plasticity, higher average mean trait values, steeper trait–fitness relationships or a combination of these (e.g. van Kleunen et al., 2010; Godoy et al., 2011; Fig. 2). For example, van Kleunen et al. (2010) found greater mean trait values for several functional traits in invasive compared to noninvasive species.

Figure 2.

Theoretical trait–fitness relationships across two environments (A and B) for a native/non-invasive and invasive species demonstrating three mechanisms (a–c) by which the invasive species may achieve greater average fitness than a native species. Closed symbols, mean trait values in each environment; open symbols, mean fitness for each species, across environments. (a) The invasive species displays greater plasticity than the native species (seen as a greater distance between the closed points) and higher average fitness. (b) The invasive displays higher average mean trait values than the native species which translates to higher average fitness despite identical levels of plasticity in both species. (c) The invasive displays a steeper relationship between the focal trait and fitness in environment B and therefore exhibits higher average fitness despite having the same mean trait values and plasticity as the native species.

Conclusions and future directions

We agree with Dawson et al.’s recommendations: to understand the role of plasticity in the spread of species (be it now, historically or in the future) will depend on detailed field experiments on multiple co-occurring native and alien species that include direct fitness measurements (including mutigenerational fitness where possible). Such multi-species studies, when conducted across a broad environmental range, also provide an ideal opportunity to examine under what circumstances plasticity in functional traits is important to fitness. So doing will enable us to examine the relationship between traits and their plasticity, and to identify when plasticity in one trait provides for homeostasis in another.

Understanding the role of plasticity and detecting patterns in adaptive plasticity of key functional traits and species types is important not only for managing invasions but also for managing populations under climate change (Sax et al., 2007) and for improving modelling of species/community responses to climate change (Ghalambor et al., 2007). Dawson et al. have shown that plasticity in biomass in response to increases in resources may be important for spread of invasive species, however many more questions remain. For example, what effect does position on the dosage response curve have? Does the response vary dramatically among species of different ecological, as well as evolutionary history? What effect does varying the likelihood of encountering different conditions have on the adaptive value of plasticity? Hypothesis driven meta-analyses, such as conducted by Dawson et al., are useful to establish broad patterns regarding the likely role of plasticity in the invasion process or in responding to novel environments. These analyses provide a more informed starting point for essential empirical enquiries.


The authors extend a big thanks to Hendrik Poorter for permission to use the figures from the website in this commentary.