My PhD is about...

Dissertation Chapter 1- Introduction

Plant biochemistry fundamentally influences the function of all terrestrial ecosystems on our planet.

Photosynthesis splits water in the presence of light in order to form energy-rich organic bonds from atmospheric carbon dioxide. Carbon assimilation plays a foundational role in all food webs, making plants the target for a wide diversity of mammals, insects, fungi, and other primary consumers. Despite being rooted in place, plants are hardly defenseless as they respond to a vast diversity of herbivore attackers with an array of biochemical processes, resulting in compounds that are either directly toxic to herbivores (Zenk, 1967; Seigler & Price, 1976) or recruit predators for protection (Turlings et al., 1990). Both direct and indirect defenses influence herbivore behavior (Hermann & Thaler, 2014), nutrition (Ballhorn et al., 2007; Veyrat et al., 2016), and survival (Tigreros et al., 2017). Therefore, the food web, from producer to consumer depends on plant chemical biology.

Over time, herbivores attack, plants defend, herbivores develop resistance, and plants defend with greater intensity or novel strategies in a dynamic coevolutionary arms race between plants and consumers (Ehrlich & Raven, 1964). Increasing resource allocation to defense may lead to greater reproductive success in the face of strong herbivore pressure, but such investment can also reduce plant fitness due to the resource and energetic costs of defense to the plant (Ballhorn et al., 2014a). Different defense traits require varying amounts of resources such as precursor molecules, complex enzymatic pathways, ATP and reducing power—all of which may conflict with plant growth and primary metabolism (Nunes-Nesi et al., 2010). High levels of defense are less beneficial in the absence of herbivore attack, but invaluable when effective against the correct enemy.

Herbivore-specific defense strategies have led to a staggering diversity of chemicals. Plants produce chemical defenses that are either toxic in small concentrations (qualitative defenses), such as cyanogenic glucosides (Poulton, 1990), glucosinolates (Brown et al., 1991), and alkaloids (Nuringtyas et al., 2014), or cumulatively increase in toxicity or inhibit digestion when ingested in greater quantities (quantitative defenses), such as phenolic compounds (Faeth, 1985), or various peptides including protease inhibitors (Broadway & Duffey, 1986). Plants reinforce chemical defenses by producing varying degrees of mechanical defenses such as trichomes, spines, prickles, and thickened leaf cuticles. Some plants tolerate high degrees of defoliation with only minor fitness consequences, or invest in rapid growth patterns to spatially “escape” attack, minimizing the need to invest in synthesizing costly defense compounds (McNaughton, 1983). Many defenses are not constitutively produced, but are inducible in response to herbivore-associated cues, which are often transmitted via volatile organic compounds (VOCs).

In addition to direct chemical and physical defenses, plants can deploy volatile chemicals advertise localized herbivore attack to other organisms that benefit from finding the same herbivores. Natural enemy recruitment, or indirect defenses, can effectively protect plant tissues from damage by attracting predators that kill or aggressively evict herbivores from the plant surface. Plants draw the attention of the third trophic level through specific VOCs (Turlings et al., 1990) and rewards such as sugar-rich extrafloral nectar (EFN) (McLain, 1983).

Herbivore feeding behavior is simultaneously shaped by chemical defenses, mechanical defenses, and host plant quality (i.e. bottom up effects) as well as top-down pressure from predators (Read et al., 2009; Moles et al., 2013). Combinations of interacting plant traits emerge that most effectively reduce fitness costs from herbivore attack, resulting in plant defense syndromes, or suites of traits (Agrawal & Fishbein, 2006). Direct interference between compounds or hormonal regulation can drive trait tradeoffs in addition to resource allocation shifts (Ballhorn et al, 2010b; Wei et al, 2014a). Observing which traits appear in the same plants at the same time can reveal information about each trait’s role in the greater ecological backdrop shaping plant defense expression. Describing patterns and underlying processes that explain the context-dependency of plant defense interactions has been an interesting challenge in chemical ecology. Lima bean (Phaseolus lunatus, Fabaceae) is a model system in chemical ecology with well-characterized suites of defenses (Frehner & Conn, 1987; Kost & Heil, 2005; Ballhorn et al., 2005, 2008a; Ballhorn, 2011b)

Lima bean cultivars and wild-type genotypes express a gradient of cyanogenic potential, the capacity to release toxic hydrogen cyanide from pre-formed precursors (Nienhuis et al., 1995). Using these consistent and well-documented cyanogenic-specific genotypes (cyanotypes) with either high cyanogenesis, HC, or low cyanogenesis, LC, the relationship can be examined between cyanogenesis as a direct chemical defense and the expression of other defense traits in lima bean. High cyanogenic potential interferes with anti-pathogenic polyphenol oxidases (Ballhorn et al., 2010b), and HC plants produce relatively lower amounts of indirect defense traits (Ballhorn et al., 2008b). Low cyanotypes show the opposite pattern, being more effectively protected against fungal pathogens, and more proficient at recruiting predators than their highly cyanogenic relatives. Tradeoffs between cyanogenesis and various other defenses compromise one mode of defense in favor of another. All plants employ a combination of several protection mechanisms, and examining the intersectionality of such defense syndromes reveals covarying strategies that may have a synergistic influence on fitness (Agrawal & Fishbein, 2006). To test the role of mechanical defense in the lima bean defense syndromes with either high or low cyanogenic potential, I quantify hook-shaped trichomes in various HC and LC genotypes in Chapter 2. Building a more comprehensive picture of plant defenses that includes mechanical defenses contributes to understanding how both cyanotypes functionally defend against multiple feeding guilds.

Whether a defensive trait increases or decreases plant fitness depends entirely on the plants’ ecological context. Plant competition could pose a challenge to plants investing heavily in defensive enzymes and precursors, as is the case with cyanogenesis (Frehner & Conn, 1987). In Chapter 3, I test the fitness costs of producing high or low levels of cyanogenic potential when plants are faced with competition across a gradient of intensity, with and without herbivore pressure. Leading hypotheses aim to describe expression patterns based on each defensive trait’s and trait combination’s fitness costs and advantages (Stamp, 2003).

Fitness costs and benefits associated with any particular defensive trait or combinations of traits can be challenging to quantify, yet are essential in developing plant defense theory. In Chapter 4, I test one component of the Optimal Defense Hypothesis (ODH) (McKey, 1974). Within-plant allocation, where within the plant defenses are allocated, assumes that plant organs most critical for fitness will be most highly defended, leading us to test the cyanogenic potential of various lima bean organs, including leaves, flowers, and pods. By creating a spatial map of within-plant allocation, and subsequently testing the fitness effects of removing varying percentages of those organs, a central tenet of this hypothesis can be tested to gain insight into how herbivore damage in nature impacts on plant survival.

Identifying the optimal timing and allocation for growth and defense in response to a diversity of interspecific interactions is a dynamic process for plants, both individually and over evolutionary time. Soil microbial symbionts add to the incredible complexity of interspecific interactions that influence plant resource allocation by priming induced defenses, (Van der Ent et al., 2009; Pineda et al., 2012), acting as pathogens, or provisioning raw materials to the plant. Nitrogen-fixing bacteria, such as rhizobia that associate with legume roots in the family Fabaceae, provide access to plant-available nitrogen assimilated from atmospheric dinitrogen. Direct access to this otherwise limited nutrient implies a reduction in the costs of protein-rich biological processes and nitrogen-demanding defenses. In lima bean, rhizobia positively increase both growth and cyanogenic potential (Thamer et al., 2011). In Chapter 5, I tested the relationship between quantitative defense level and plant symbiotic association by counting nodules and measuring the corresponding chemical phenotypes. I wanted to understand whether nitrogen obtained from greater numbers of root nodules would directly influence quantitative defense expression, or if genotypically-determined defense levels could force plants to maintain greater colonization to meet cyanotype-specific nitrogen demands. Costs of a larger carbon sink resulting from maintaining symbionts for more nitrogen-fixation in root nodules may have previously unmeasured impacts on the food web (Pringle, 2015a).

Predators and parasitoids respond to specific plant chemistry, and are sensitive to changes in plant volatile signals and extrafloral nectar (EFN): sugar-rich nectar reward not associated with pollination. Both plant traits require significant unrecoverable quantities of assimilated carbon. Rhizobia also have heavy carbon requirements; ca. 16-30% is directed to root nodules (Peoples et al., 1986; Kaschuk et al., 2009). The strong carbon sink associated with root nodules may explain why lima bean VOCs decreased overall and shifted in composition in plants with rhizobia (Ballhorn et al., 2013c). The influence of rhizobia on indirect defense traits and the predator/parasitoids’ response to rhizobia-mediated plant chemistry from the bottom up is not well-characterized (Rasmann et al., 2017). To understand whether rhizobia alter investments in VOCs and natural enemy attraction, I quantify recruitment of parasitoid wasps in natural communities in Costa Rica in Chapter 6, ultimately aiming to understand whether rhizobia have food web impacts that extend to the third trophic level, in this case, to parasitoid wasps.

Ants are another predator with a widespread and intimate evolutionary history with plants. Harnessing the aggressive power of ant colonies can be an effective shield against herbivores. Ant-plants range from obligatory myrmecophytes, complete with a unique set of symbiont-specific traits, to facultative myrmecophiles, most of which rely on secreting EFN. Most plants that secrete EFN do so via the facultative model, with EFN production being induced in response to signals of herbivore pressure in order to conserve carbohydrates until needed. Given the carbon-rich nature of this trait, I tested whether rhizobia, which act as a carbon sink, impose a constraint on EFN secretion in Chapter 7. Using a structural equation model, I tested the influence of rhizobia on several plant traits, including soluble protein, cyanogenic potential, below- and aboveground biomass, and EFN to assess which traits mediated by rhizobia could affect the recruitment of ant patrollers.

Facultative EFN as a reward for predators only functions as a plant defensive trait if the predators can find this resource and respond with aggressive behavior towards plant antagonists at the appropriate times. Despite being an extensively-studied interaction, ant responses to EFN volatile components, or scents, have not been examined. In Chapter 8, I sampled the volatile profiles of several distantly-related species’ EFN, and compare ant attraction. Volatile-based communication fills a gap in our understanding of how EFN rewards mediate ant-plant interactions in response to induced signals, and on a broader time scale in maintaining non-obligate mutualistic interactions.

Plants evolved their diverse suite of defensive traits in the context of interacting enemies and mutualists, above and belowground. By employing natural enemies, and maintaining nitrogen-fixing root nodules, plants factor cooperation into growth and defense allocation dynamics. Cooperative relationships and resource costs benefitting another species present an interesting challenge to evolutionary biology. To explain how mutualistic relationships persist, hypotheses based on host sanctioning and partner choice have been extensively tested in both ant-plant and legume-rhizobia relationships. Examining how forces that facilitate persistence of cooperation through evolutionary time will contribute to our understanding of plant-herbivore interactions and food web energy flow.