Introduction
Hymenoptera (bees, wasps, and ants) are arguably the most species-rich group of animals on earth, in large part due to the sheer diversity of parasitoid wasps within the order as driven by the close associations these wasps have with their host arthropods (Forbes, Bagley, Beer, Hippee, & Widmayer, 2018). Parasitoid wasps lay their eggs eitheron (ectoparasitoids) or in their hosts (endoparasitoids), with some arresting the development of their hosts upon oviposition (idiobionts), whereas others allow their host to continue development during wasp larval growth (koinobionts) (Askew & Shaw, 1986). The level of host specialization varies among parasitoid wasps. Some species are monophagous, specializing on a single arthropod host species (oligophagous), while others utilize a wider, albeit often closely related, range of arthropod hosts, (host-flexible), Koinobiont endoparasitoids are likely to be more specialized because developing wasp larvae must survive through host development and must successfully defend against the host’s immune response to parasitism(Beckage & Gelman, 2004; Vinson & Iwantsch, 1980).
As a host is required for parasitoid survival and reproduction, there are direct fitness consequences to selecting a host (Charnov & Skinner, 1984; Kouamé & Mackauer, 1991; Wang & Messing, 2004). Further, the process of selecting and locating a host differs depending on the level of host specialization of the parasitoid wasp (Bertoldi, Rondoni, Brodeur, & Conti, 2019; Vet & Dicke, 1992; Vinson, 1976). Although there are many factors that may influence host selection in parasitoids (Price, 1971), for host-flexible parasitoids some host species may provide a fitness advantage relative to others. This begs the question: where a host preference exists, how do host-flexible parasitoids locate and select their preferred hosts?
Semiochemical cues (infochemicals) emitted from the host herbivore or plants, also referred to as the plant host complex (PHC), are one way parasitoid wasps may locate a viable host (Tumlinson, Lewis, & Vet, 1993; Vet & Dicke, 1992). Among infochemicals, there are three main classifications: (1) Kairomones that benefit the receiver, (2) allomones which benefit the sender, and (3) synomones that benefit both the sender and receiver (Brown Jr, Eisner, & Whittaker, 1970). For parasitoids, kairomones are often in the form of non-volatile oral secretions or feces from the host (Rutledge, 1996), although volatile pheromones may also be utilized. Synomones are typically plant-emitted volatiles and may be either herbivore-induced or constitutively produced (Hilker & McNeil, 2008; McCormick, Unsicker, & Gershenzon, 2012; Paré & Tumlinson, 1999; Tumlinson, Lewis, et al., 1993). The detection of these infochemicals by parasitoids can be innate or learned, although there are still relatively few studies documenting how host-flexible parasitoids use infochemicals for foraging and host acceptance (Steidle & Van Loon, 2003).
Vet and Dicke (1992) hypothesized that not all infochemicals are equal, and instead suffer from what they term the ‘reliability-detectability problem’. Under this paradigm, host kairomones have high reliability because they are excellent indicators of host presence. However, they also have low detectability because they are typically localized and not detectable at a distance (Vet & Dicke, 1992). Due to their reliability, these infochemicals likely play an important role as a dependable indicator of the presence a specific host for monophagous parasitoids (Vet & Dicke, 1992). Vet and Dicke (1992) also contend that plant-emitted synomones travel much farther than host compounds making them more likely to be detected at a distance by foraging parasitoids, but are likely poor indicators of host presence and therefore unreliable. One hypothesized solution to the reliability-detectability problem is herbivore-induced synomones, infochemicals released from the plant in response to damage from herbivory termed the plant-host complex (PHC). Vet and Dicke (1992) proposed that parasitoids would have strong responses to these synomones if the parasitoid was a specialist on one host species found on a single plant species, or parasitoids that can utilize multiple hosts on a single plant species. In these cases, herbivore-induced synomones of the PHC are not only detectable over a long distance but also reliable indicators of host presence.
The ability for parasitoids wasps to utilize volatile compounds has been well established, but most studies have either focused on specialists (Colazza, McElfresh, & Millar, 2004; de Moraes, Lewis, Pare, Alborn, & Tumlinson, 1998; Du, Poppy, Powell, & Wadhams, 1997; Gols, Bullock, Dicke, Bukovinszky, & Harvey, 2011; McCall, Turlings, Lewis, & Tumlinson, 1993; Röse, Lewis, & Tumlinson, 1998; Tumlinson, Turlings, & Lewis, 1993; Xiu et al., 2019), or on a single host-flexible species (D’Alessandro, Brunner, von Mérey, & Turlings, 2009; Ponzio et al., 2016; Röse et al., 1998; Tumlinson, Lewis, et al., 1993; Turlings, Tumlinson, & Lewis, 1990; Wei et al., 2007) and no other related species. Although Vet and Dicke’s (1992) hypotheses may hold true for parasitoids that specialize on a single host species, it is likely that host-flexible parasitoids utilize these cues differently. An overarching issue with each one of these hypotheses is that they approach the system from a bottom-up perspective, describing infochemical use between the third and second trophic level (parasitoid and host herbivore), and how interactions between the second and first (host herbivore and plant) could impact infochemical use by the third. In doing so, these hypotheses downplay the potential for infochemical cues from the first trophic level, outside of herbivore-induced infochemicals, to function as reliable indicators for the location of a suitable host herbivore. In this study we examine how infochemicals from the plant impact host selection in parasitoids, and whether closely related host-flexible parasitoids exhibit a differential preference for different plant volatile profiles.
An excellent group for testing this question is Eadya Huddleston & Short, 1978 (Hymenoptera: Braconidae: Euphorinae), a small genus of parasitoid wasps endemic to Australia. These wasps attack paropsine leaf-beetles (Coleoptera: Chyrsomelidae: Chrysomelinae) that feed onEucalyptus L’Héritier, 1789 (Myrtales: Myrtaceae). Eadyais currently comprised of 6 described species (Huddleston & Short, 1978; Ridenbaugh, Barbeau, & Sharanowski, 2018), all of which are host flexible to varying degrees. Species such as Eadya daenerysRidenbaugh, 2018 and Eadya spitzer Ridenbaugh, 2018 were documented from four different hosts, while E. paropsidisHuddleston & Short, 1978 has been documented on three hosts between Tasmania and mainland Australia, but only reared from Paropsis tasmanica Baly, 1866 in Tasmania (Peixoto et al., 2018).
Species of Eadya are also of interest as biological control agents targeting invasive paropsine beetles infesting Eucalyptusplantations in New Zealand, with E. daenerys approved to controlP. charybdis Stål, 1860 (T. M. Withers, Allen, Todoroki, Pugh, & Gresham, 2020) and E. annleckieae Ridenbaugh, 2018 as a potential agent to control Paropsisterna ­­ (=Pst. variicollis(Chapuis) in Peixoto et al., 2018) (Nahrung et al., 2020). Leschen et al. (2020) recently synonymized Pst. variicollis under Pst. cloelia (Stål) but referenced no type material.
Given the host flexibility observed within Eadya , plant infochemicals may play a more important role than host cues in host selection. To assess this, we perform cophylogenetic analyses to test for evidence of cospeciation between Eadya and their hosts. If the parasitoids show strong evidence of cospeciation with their hosts, we expect that host infochemicals would have a strong impact on host location and selection. To examine how infochemicals from the plant potentially impact host location and selection in closely related host-flexible parasitoids, we test for distinct chemoprofiles inEucalyptus species and test whether these profiles are different for damaged versus undamaged leaves, suggesting an herbivore-induced response. Then we ask if these plant chemoprofiles are good predictors of parasitoid and/or host species, which would indicate that plants have a strong influence on host location and selection in the parasitoid.