Females of the egg parasitoid Cosmocomoidea annulicornis were able to detect host eggs irrespective of the plant that held them, relying only on common volatile cues. However, they did not discriminate between healthy and parasitized eggs by conspecifics.
The headspace analysis of citrus and Johnsongrass plants with host eggs revealed very few similarities in the blend’s composition. Among the compounds detected, only indole and β-caryophyllene increased in both plant species with oviposition damage. Indole is an aromatic compound that is specifically emitted after herbivore damage in several plants (Erb et al. 2015), although its ecological role as an attractant of natural enemies is uncertain, since some parasitoid species are known to be attracted (James 2005) and others repelled by this volatile (D’Alessandro et al. 2006; Block et al. 2018). Also, it can have a dual effect, as an attractant and a repellent (Ye et al. 2018), which shows that there may not be a straightforward response (Zhuang et al. 2012). β-Caryophyllene is also known to be released during herbivore attack and to act as an attractant for several parasitoids in plants such as maize (Kollner et al. 2008; Tamiru et al. 2017), cotton (Loughrin et al. 1994), beans (Colazza et al. 2004a, b) and elm (Büchel et al. 2011). Some parasitoids find their hosts in distant plant species exploiting similarities in the relative proportions of herbivory induced VOCs in each blend (De Moraes et al. 1998), however, we could not find such a common relation in the ratios between compounds found in both infested plants (indole and caryophyllene).
Based on plant odor alone, there were no signature volatiles or volatile blends that could trustfully indicate the presence of the host herbivore. However, common herbivory induced plant VOCs present in the odor profiles could induce long range foraging decisions, and once on the plant, other sensory inputs could also be very important, particularly visual and physical cues (Aquino et al. 2012; Michereff et al. 2016).
In agroecosystems, parasitoids must interpret very complex chemical environments while searching for their hosts. Herbivory induced VOCs are elicited by conserved plant signaling metabolic pathways (Huffaker et al. 2013). Although terpene synthases are diversified throughout the plant kingdom (Chen et al. 2011), a few induced VOCs are usually released by most plant species, irrespective of their taxonomic affinities (Clavijo McCormick et al. 2012). For example, under herbivore attack many plant species emit the terpenoids (E)-β-ocimene, linalool, α-farnesene, β-caryophyllene and 4,8-dimethyl-1,3,7-nonatriene, although the composition of the blend and relative amounts of compounds may vary among species and even among cultivars or varieties of the same species (Degen et al. 2004; Clavijo McCormick et al. 2012; Coll Aráoz et al. 2020).
Besides indole and β-caryophyllene, citrus plants infested with host eggs displayed differences in the emissions of β-ocimene, undecane, butyl citrate and two unknown compounds. Although D-limonene was the dominant compound detected in citrus plants, there were no variations in the amounts released after herbivore damage. Contrary to this result, Mohammed et al. (2020) found that D-limonene levels sharply increased in several Citrus species after infestation with the pest Aonidiella aurantii (Maskell) (Hemiptera: Diaspididae), and that its parasitoid, Aphytis melinus DeBach (Hymenoptera: Aphelinidae), was attracted to this volatile combined with β-Ocimene. This last compound is released by grapevines plants after feeding and oviposition of the proconiini sharpshooter Homalodisca vitripennis (Germar) and it affects the foraging behavior of the egg parasitoid Cosmocomoidea ashmeadi Girault (formerly Gonatocerus), which positively responded to this volatile when it was combined with α-farnesene (Krugner et al. 2014).
Comparatively, Johnsongrass plants under oviposition damage released lower amounts of α-phellandrene and hemellitol than undamaged plants. Takemoto et al. (2015) reported that broad bean plant emission of α-phellandrene under aphid attack was attractive for the parasitic wasp Aphidius ervi Haliday (Hymenoptera: Braconidae).
The attraction of wasps to volatiles induced by the feeding of adult sharpshooters was subject to the attacked host plant species. Parasitoid females were significantly attracted to citrus plants with feeding damage performed by males, while no effect of this treatment was observed in Johnsongrass plants. The most noticeable difference between the two host plant species under feeding damage was the presence of benzaldehyde in citrus plants. This compound is known to be emitted by other plant species and has proven to be an attractant to the insect pest Sitona humeralis Stephens (Coleoptera: Curculionidae) (Lohonyai et al. 2019), the aphid Rhopalosiphum padi (L.) (Park et al. 2000) and to wild Trichogrammatidae parasitoids in a field study (Simpson et al. 2011). It is possible that other host related chemical cues such as salivary secretions and faeces, would also be important for parasitoids attacking insects with overlapping generations like T. rubromarginata (Freytag and Sharkey 2002), as environments with adults presence are likely to contain eggs (Vet and Dick 1992; Peñaflor et al. 2011).
Headspace samples of plants with parasitized host eggs displayed similar compounds to those detected in plants with non-attacked host eggs, although a number of volatiles present in plants with host eggs were no longer detected in the parasitized samples (citrus: β-Pinene, Undecane and Benzene, 1-ethyl-2-methyl; Johnsongrass: hemellitol, β-caryophyllene, 1,2-benzisothiazole and methyl anthranilate). It is noteworthy that the VOC “Unknown 2” was only detected in citrus samples with eggs parasitized by C. annulicornis.
The fact that C. annulicornis females did not discriminate between healthy and parasitized host eggs during behavioral tests suggests that the presence of a conspecific developing larvae does not influence the attractiveness of the host eggs, because females were not able to detect the differences in the volatile blends. There are no records of superparasitism in this species (Manzano et al. 2021) so it is likely that females rely on probing behavior before ovipositing eggs parasitized by conspecifics. There are several studies demonstrating that for egg parasitoids, probing and subsequent egg marking by females are frequent behaviors that condition host recognition and discrimination of previously attacked eggs (Weber et al. 1995; Keinan et al. 2012).
NMDS analysis highlighted that regardless of the host plant species, plants differentially emitted volatiles after oviposition damage (including healthy and parasitized eggs) compared to untreated and plants with feeding damage alone. Interestingly, despite the qualitative variation in compounds released among different treatments, there were no differences in the total amount of volatiles emitted constitutively and under herbivore damage in the two host plant species analyzed.
It has been generally assumed that natural enemies use herbivory induced plant volatiles to find the plant species where their hosts are likely to be present. For egg parasitoids this seems to be especially true, since insects’ eggs only emit small quantities of volatiles (Fatouros et al. 2008). As demonstrated in the olfactometer assays, C. annulicornis females rely on chemical cues to find their hosts, however, this attraction toward plant volatiles would not necessarily translate into an increase in the parasitoid’s fitness (Clavijo McCormick et al. 2012). Although females might be using volatiles to locate infested plants, there are probably other chemical cues detected at short range that mediate the final decision to oviposit.
Previous studies have shown that plant species influences C. annulicornis oviposition rates. Virla et al. (2019) report significantly higher parasitism rates in the field in host eggs laid on citrus plants, compared to those laid on Johnsongrass. Similarly, laboratory studies support this trend (Manzano et al. 2021), demonstrating that fitness of this parasitoid is highly dependent on plant species. However, in olfactometer assays C. annulicornis females showed no preference between eggs carried by any of the two host plant species. Common volatiles emitted in response to herbivory would be important for orientation toward the host, but oviposition could be mediated by direct contact cues. It is possible that female final acceptance of hosts is based on other traits besides olfactory cues (Colazza et al. 2009), such as leaf morphological structures (Virla et al. 2020), contact cues (Iacovone et al. 2016) and host quality (Hofstetter and Raffa 1998, Zhou et al. 2014).
Our investigation contributes to the understanding on how chemical cues emitted by host plants affect the host searching behavior of an egg parasitoid and constitutes a step forward to finding an attractive blend for the parasitoid of this economically important herbivore species. The manipulation of VOCs released by plants in response to herbivory can be a valuable tool for developing new strategies of integrated pest management (IPM) to increase biological control and its cost efficiency.