Discussion
Our results make two key conceptual contributions to understanding allocation of plant defenses. First, we showed that non-native species, which presumably have escaped from specialist natural enemies, had lower constitutive defenses and higher induced defenses, which correlated with higher growth rates. Second, non-native plant species, ostensibly exposed primarily to generalists, showed a strong positive relationship between herbivore pressure and constitutive defense, and a strong negative relationship between herbivore pressure experienced by a species and the intensity of induced defenses exhibited by that species. These findings expand our understanding of constitutive vs. induced defense allocation associated with herbivore pressure on non-native plants.
Integrating these results suggests that reduced herbivore attack (Figure 1C and 1F; i.e., escape from enemies) favored allocation to induced defenses (Figure 2B) instead of constitutive defenses (Figure 2A). Our results also suggest that induced defenses are a “cheaper” overall strategy which favors higher growth rates (Figure 5). It appears that the generalist-dominated herbivore community that attacked non-native plant species drove a strong positive relationship between the intensity of herbivore pressure and allocation to constitutive defense among our target species (Figure 2A and 2C). This supports long-held theory that has had minimal empirical support - intense or consistent herbivory should select for more expensive constitutive defenses over cheaper induced defenses (Ito & Sakai, 2009).
Generalist herbivore pressure
Native generalist herbivores often avoid non-native plant species or do not suppress them (Schaffner et al., 2011; Sedio et al., 2020). For example, Cappuccino & Carpenter found that non-native invasive plants experienced 96% less leaf damage than non-native species that were not classified as invasive (Cappuccino & Carpenter, 2005). One hypothesis for this pattern is the novelty of biochemicals, in their new ranges, produced by non-native plants (Schaffner et al., 2011; Callaway & Aschehoug, 2000). Cappuccino & Arnason (2006) found that a large suite of invasive species had more unique secondary chemicals than exotic plants, relative to native species. These, and other studies (see review by Inderjit et al. 2021), indicate that unusual or novel biochemicals of some non-native species may be why native herbivores avoid them (Inderjit et al., 2021).
Defense strategies
Intense and predictable herbivory is thought to select for constitutive plant defenses (Kalske & Kessler, 2020). Induced defenses appear to be adaptive to less intense and infrequent herbivory (Agrawal & Karban, 1999; Ito & Sakai, 2009) as a cost-saving alternative to constitutive defense (Detto & Xu, 2020). Relevant literature on invasives is minimal, but the invasive genotypeLespedeza cuneata exhibited greater induced defense but lower constitutive defense than ancestral native genotypes (Beaton et al., 2011) as might be predicted by reduced or more sporadic herbivory in the non-native range. In contrast, non-native genotypes ofAlternanthera philoxeroides have higher constitutive defenses and lower induced defenses (Liu et al., 2020). Clearly, the literature on evolutionary changes in constitutive or induced defenses by exotic plants is limited.
Other recent results indicate that secondary metabolism plays key roles in the interactions between non-native plants and herbivores (Tian et al., 2021; Yu et al., 2022). Phenolics inhibit the digestion of protein (Rehman et al., 2012) and have been central in many studies of plant-defense mechanisms (e.g., Kumar et al., 2020). We showed that weight gain of a generalist herbivore was negatively correlated with phenolic concentration across all species and herbivory treatments, suggesting that defense might derive from variation in constitutive and induced phenolic concentrations. Triterpenes are terpenoids, also abundant in plants (Kumar et al., 2020), and larval herbivore weight gain was negatively correlated with triterpenoid content across our native species. However, triterpenoid content did not correspond with herbivore pressure. Triterpenes have many other functions (González-Coloma et al., 2011), and the functional diversification of triterpenes may have masked simple responses to herbivores.