Inducible plant defences and the environmental context

Abstract

The term ‘phenotypic plasticity’ describes an organism's ability to express different phenotypes under different environmental conditions. This includes any type of morphological, physiological, metabolic, phenological and behavioural change that can be induced by biotic or abiotic environmental stressors (West-Eberhard 1989). The changes can be reversible or permanent, and have been demonstrated to extend across generations (Rasmann et al. 2012). Initially, ecologists and evolutionary biologists had problems to explain phenotypic plasticity within a Darwinian evolution framework (West-Eberhard 1989; Agrawal 2001), because they were focusing on the genetic basis of phenotypes and thus intentionally on traits that were little affected by environmental variation. With the second half of the 20th century, this view has dramatically changed (Dudley & Schmitt 1996; Agrawal 2001). And it was, to a large extend, the study of rather sessile organisms, such as plants, that characterized phenotypic plasticity as an adaptive trait that allows an organism to cope with changing environmental circumstances while not being able to actively move to another habitat. This is part of the reason why plants became prime choices to study phenotypic plasticity in general and its adaptive value in particular, and why plant phenotypic plasticity is increasingly studied in a behavioural ecology framework (Karban 2008). In the study of plant–organismal interactions, phenotypically plastic responses to attacking herbivores and pathogens are generally found to be major weapons in a plant's defensive arsenal and are thought to allow plants to incorporate the entire interacting community into their defensive strategies (Kessler 2015). For example, herbivory can induce dramatic changes in plant secondary metabolism, which can include an altered production of defensive metabolites and a resulting increased direct resistance that can differentially affect other herbivore species as well as the initial attacker (Viswanathan, Narwani & Thaler 2005; Uesugi, Poelman & Kessler 2013). In consequence, a less damaging herbivore can even become beneficial to the plant by vaccinating against a more damaging herbivore (Kessler & Baldwin 2004; Halitschke, Hamilton & Kessler 2011). Similarly, herbivore-induced production of volatile metabolites can provide information to third- and fourth-trophic-level organisms and so connect the parts and affect the dynamics of complex interaction networks (Poelman & Dicke 2014). In the same way, biotic stressors affect secondary metabolite production, so do abiotic factors, such as light quality (Izaguirre et al. 2006), reactive oxygen species (Blande, Holopainen & Niinemets 2014), metal pollutants (Shahid et al. 2014) and nutrients. Especially nutrient availability has long been intensely studied as a factor that can affect a plant's secondary metabolite production and thus constitutive resistance (Coley, Bryant & Chapin 1985). Because of the generally strong links between resource availability and secondary metabolism and, in extension, plant resistance to herbivores, this correlation has been at the core of many plant defence hypotheses (Stamp 2003). While evidence for multiple of these hypotheses has been accumulated with comparative studies on the species and genotype level for constitutive resistance traits, inducible traits, such as tolerance and inducible resistance, are usually discussed as a function of frequency and probability of herbivory rather than abiotic factors (Karban & Baldwin 1997). However, within the frameworks of adaptive phenotypic plasticity and the optimal defence hypothesis (Stamp 2003), plants can be predicted to express constitutive vs. induced resistance alternatively, depending on environmental factors, such as nutrient availability. In this issue of Functional Ecology, Burghardt (2016) addresses a missing piece in our understanding of the relationship between nutrient availability and resistance and tolerance expression. In a series of experiments with tall goldenrod, Solidago altissima, the author found some remarkable phenotypic shifts in defence strategies as a function of nutrient availability (Burghardt 2016). As in other study systems, increasing nutrient availability for S. altissima reduces herbivore resistance, which has often been associated with reduced secondary metabolite production at high nutrient availability. However, one of the most interesting results of this study is that inducibility of resistance increased with increasing nutrient availability. Moreover, at low nutrient levels most plant genotypes showed induced susceptibility, while high nutrient availability was generally associated with high inducibility of resistance. The higher susceptibility of plants at low nutrient levels was associated with higher tolerance and even overcompensation to herbivory, measured as asexual reproduction, while plants were less able to tolerate damage at high nutrient levels. In other words, plants seem to shift their defence strategy against herbivores from constitutive to inducible and from tolerating herbivory to less tolerance with increasing nutrient availability. The expression of phenotypic plasticity itself is plastic and can vary with environmental conditions. Such dramatic phenotypic plasticity in the expression of alternative defensive modes is likely to contribute to phenotypic defence- and tolerance-level variation in habitats with heterogeneity in nutrient availability and thus will affect population-, community- and ecosystem-level dynamics. Although these data at least partially support general plant defence hypotheses [Coley et al. 1985; Herms & Mattson 1992; Hawkes & Sullivan 2001; Wise & Abrahamson 2007, discussed in Burghardt (2016)], they mostly illustrate the complexity of factors that influence plant responses to environmental stressor and the resulting difficulty to clearly assign any of the available hypothesis frameworks. The study is putting a particular emphasis on how the environment in which species interactions are played out influences plant responses, suggesting that environment-affected phenotypic plasticity can alter relationships between plant traits and so make apparent ecological trade-offs context-dependent. For example, the strong interaction terms between nutrient availability and resistance induction suggest the expression of alternative strategies that allow plants to cope differently (e.g. constitutive vs. inducible resistance, tolerance vs. resistance) with a similar stressor in different environments (nutrient availability). Burghardt (2016) tries to capture this complexity in trait variation by integrating individual traits into multivariate analyses. While this is an increasingly popular approach and while the complexity potentially explains contradicting outcomes in different experiments, the inclusion of multivariate traits into plant defence theory remains an open but very important task. It is the interactions of multiple interacting physiological and ecological trade-offs that make the identification of individual trade-offs challenging, but also has the largest explanatory power when studying species interactions. The same plant genotypes can express different levels of phenotypic plasticity, or differential behaviour, to stressors in different environments whereby, at least in the S. altissima system, the contribution of the environment in affecting phenotype expression can be as large as the contribution of the genotype (Burghardt 2016). These findings have fundamental implications not only for studying ecological interactions but also for studying natural selection on the constitutive expression and inducibility of defence traits. One of the next steps has to be to reveal whether environmental context-dependent expression of different defensive strategies is adaptive and which environmental conditions select for such flexibility in plants. From a more mechanistic point of view, and with an increasing realization of the importance of soil microbial communities affecting nutrient availability and plant phenotypic plasticity (Choudhary et al. 2015), contrasting Burkhardt's manipulations using mineral fertilizers with those that affect magnitude and quality of nutrient availability indirectly seems warranted and timely. Not only would such a comparison reveal the naturally occurring range of phenotypes expressed in response to soil quality variation, but would also link the data from the S. altissima ecological model system to agricultural systems.

Abstract from the paper’s OpenAlex record, shown for indexing and research. Copyright remains with the authors or the publisher. Read at the source ↗

The text the model read, with the source record’s inline markup (italics, sub- and superscripts) rendered and its character entities decoded; quote offsets refer to the raw text.