Herbicidal interference: glyphosate drives both the ecology and evolution of plant–herbivore interactions

Introduction

Plants have coexisted and interacted with insect herbivores for > 400 million years (Ehrlich & Raven, 1964; Labandeira & Currano, 2013), with such interactions influencing the evolution of plant traits, population dynamics, and even community stability (Myers & Sarfraz, 2017; De-la-Cruz et al., 2020; Agrawal & Maron, 2022). Indeed, interactions between plants and herbivores can reciprocally influence their respective evolutionary trajectories – herbivores exert selective pressure on plants, leading to the evolution and diversification of anti-herbivory defenses; in turn, evolutionary responses to herbivory can alter the ecological conditions in which plants and herbivores interact (Agrawal et al., 2006). Such eco-evolutionary feedback loops can occur in the context of human-mediated forms of stress like those associated with agricultural regimes, urbanization, and climate change (Turcotte et al., 2017; Hamann et al., 2021; Santangelo et al., 2022). However, how these human-mediated stressors alter plant–herbivore interactions and thus the broader eco-evolutionary dynamic between plants and their insect herbivores remains a significant gap in our knowledge.

Chemical herbicides are a relatively novel form of human-mediated selection used in agricultural ecosystems to control and eradicate weedy plants (Shaner, 2014). Unfortunately, plants have quickly adapted to herbicide use; to date there are hundreds of weed species that are considered either resistant or tolerant to some form of herbicide (Baucom & Mauricio, 2004; Vila-Aiub et al., 2009; Délye et al., 2013). Most work addressing natural weeds and herbicide use examines some aspect of the evolution of resistance or tolerance in plant populations – whether focusing on the frequency of resistance or tolerance, the genetic basis of either defense trait, or the potential for fitness costs associated with herbicide defense (Baucom, 2019). Far fewer studies consider the consequences of herbicides on nontarget organisms such as herbivorous insects or pollinators within the weed community (Motta et al., 2018, 2020). While some research has examined how herbicide exposure may directly influence insect development, fitness, and immune responses (Schneider et al., 2009; Baglan et al., 2018; Capinera, 2018; Smith et al., 2021), the potential for indirect effects of herbicide – where herbicide exposure alters or changes some aspect of the plants that subsequently impacts insects or other community members (Fuchs et al., 2021) – is less understood. This is a notable knowledge gap because plant chemistry, physiology, and phenology are directly altered by herbicide exposure (Baucom et al., 2008; Londo et al., 2014; de Freitas-Silva et al., 2022); additionally, aspects of plant size and the mating system have evolved along with herbicide resistance (Van Etten et al., 2016; Kuester et al., 2017). Each of these changes – whether trait changes due to exposure, or trait alterations due to correlated evolution with herbicide resistance – has the potential to interfere with established interactions between plants and insect herbivores, pollinators, and other organisms.

There is evidence that plant herbivory specifically may be impacted by herbicide. For example, in glyphosate-tolerant Beta vulgaris (sugar beet), glyphosate application led to an increase in the herbivore Myzus persicae (green peach aphid) (Dewar et al., 2000). Similarly, Abutilon theophrasti (velvetleaf) exposed to low doses of the herbicide dicamba show an elevated abundance of the phloem-feeding Bemisia tabaci (silverleaf whitefly) (Johnson & Baucom, 2022). Additionally, four herbicides (2,4-D, Command, Newpath, Ricebeaux) applied to Oryza sativa (rice) modestly but differentially affected the abundance levels of the tissue-consuming Lissorhoptrus oryzophilus (rice water weevil) and the herbivory damage inflicted by stem-boring beetles (Kraus & Stout, 2019). The consequences of such altered herbivory experienced by herbicide-exposed plants are generally unknown on both the part of the plant and the insect herbivore. For example, does an increased abundance of an herbivore due to herbicide application lead to greater selective pressure than the plant population would normally experience? Or do different types of feeding result, potentially influencing both the trajectory of herbivory resistance in plants and the herbivore population/community in unexpected ways? It is also unknown if plants exposed to herbivory and herbicide simultaneously may eventually evolve higher levels of resistance to both stressors.

At the same time, there is also the possibility that plants may experience a trade-off between defense to herbicide and herbivory, which could lead to a constraint on the evolution of increased levels of either trait (Simms & Rausher, 1987). Indeed, there have been trade-offs detected between constitutive and induced anti-herbivory defenses (Koricheva et al., 2004), between resistance to different herbivore species (Agrawal et al., 1999), and between the defense strategies of resistance and tolerance to the same threat (van der Meijden et al., 1988; Fineblum & Rausher, 1995; Baucom & Mauricio, 2008). However, the evidence that trade-offs exist between herbivory and herbicide defense is more limited; one example showed that Amaranthus hybridus (smooth pigweed) plants resistant to the herbicide triazine were more susceptible to the specialist herbivore Disonycha glabrata (striped flea beetle) as compared to triazine-susceptible plants (Gassmann & Futuyma, 2005). It is possible that trade-offs between herbicide and herbivory resistance exist but remain undetected only due to the small number of studies on this topic to date.

Here, we consider plant–herbivore interactions in the context of an herbicide-exposed plant population. As human activity has become a primary driver of global change, human-mediated stressors may impact plant–herbivore interactions, thereby potentially altering plant evolutionary responses, which could subsequently feed back to and reshape the ecological context within which plant herbivory is occurring. Our goal is to investigate the effects of a novel but strong selective agent – herbicide – on long-standing plant–herbivore relationships using Ipomoea purpurea (common morning glory). We specifically examine both the ecological effects of herbicide exposure on insect herbivory and the potential for evolutionary consequences of herbicide resistance on the evolution of herbivory resistance. To this end, we ask the following questions regarding ecological impacts: (1) Does glyphosate application affect insect herbivory levels and/or chewing damage patterns? If so, glyphosate has the potential to alter plant–herbivore interactions and thus indirectly moderate the insect herbivore community by favoring the feeding of certain herbivores over others. We further ask: (2) Is there a trade-off between glyphosate and herbivory resistance? The existence of a trade-off would indicate that the two forms of resistance may be mutually exclusive, which may evolutionarily constrain the evolution of either form of resistance. Finally, we examine the potential evolutionary consequences and ask: (3) Is there potential for these resistance traits to further evolve? If so, and a trade-off is present, then not only may plant populations start to diverge based on their resistance strategies, but their associated insect herbivore communities may diverge as well, as herbivores adapt to the evolving plants that form their diet. By answering these questions, our study begins to uncover the eco-evolutionary effects that herbicide may exert on plant–herbivore communities.