Introduction

Olfactory perception of food cues and sex signals is intimately interconnected in insects (Reddy and Guerrero 2004; Varela et al.2011; Rouyar et al. 2015; Lebreton et al. 2017; Borrero-Echeverry et al. 2018; Conchou et al. 2019). Deciphering the chemicals encoding food and mates is basic to understanding insect ecology and evolution. Moreover, the knowledge of such behaviour-modifying chemicals can be applied for detection and environmentally safe control of insects (Ridgway et al. 1990; Witzgall et al. 2008, 2010a; Reddy and Guerrero 2010; Suckling et al. 2014; Evenden and Silk 2016; Gregg et al. 2018).
New tools for insect management are needed in the wake of a changing climate that accelerates insect invasions and outbreaks, aggravating food insecurity (Deutsch et al. 2018). Recent efforts to deregulate the most toxic compounds has left growers with few efficient insecticides (Chandler et al. 2011; Jactel et al. 2019). The overwhelming majority of insect species, however, does not feed on human food crops. Including pollination services, insects are integral to all terrestrial food webs. The overuse of synthetic pesticides affects non-target and beneficial insects and other arthropods, and is a contributing cause of the biodiversity apocalypse. This has been a point of debate since DDT (Carson 1962) and nonethless, evidence is accumulating for severe side effects of the currently most widely used family of insecticides, the neonicotinoids (Seibold et al. 2019; Yamamuro et al. 2019; Chmiel et al. 2019; Longing et al. 2020; Wagner 2020).
The establishment of pheromones and other semiochemicals as a species-specific and environmentally safe alternative to conventional insecticides has therefore always been an outstanding rationale for chemical ecology research. Air-permation with synthetic pheromone, for disruption of premating sexual communication, is used against a few key orchard and forest insects (Reddy and Guerrero 2010; Witzgall et al. 2010; Evenden and Silk 2015). Pheromone lures for specific and sensitive detection are available for hundreds of species. Such lures, in combination with traps, insect pathogens or insecticides, may even achieve population control, when the female sex becomes attracted (Ridgway et al. 1990; El-Sayed et al. 2009; Sucklinget al. 2014). In stark contrast to pheromones attracting insects for mating, only few semiochemicals have been identified that attract gravid females for oviposition. Designing female or bisexual lures is therefore a current main challenge towards a more widespread use of behaviour-modifying chemicals for insect control.
Identification of many hundreds of sex pheromones, across all insect orders (El-Sayed et al. 2016), has been facilitated by a mutual coordination of production and response in both sexes. Pheromones are produced in dedicated glands, produce strong antennal responses and immediately trigger a sequence of distinctive behaviours.
Identification of semiochemicals, or kairomones, that mediate oviposition behaviour meets substantial methodological difficulties. Synthetic plant volatile blends that have been found to attract insect herbivores typically build on compounds found across many plant species (Najar-Rodriguez et al. 2010; Tasin et al. 2010; Bruce and Pickett 2011; Lu et al. 2015). The attractant power of such ubiquitous plant volatiles is sometimes faint, compared with sex pheromones.
In comparison, plant compounds that are unique or characteristic for larval food plants have been found to mediate significant attractancy. One such key host plant compound is ethyl (E,Z)-2,4-decadienoate, pear ester, a bisexual attractant for codling moth Cydia pomonella(Lepidoptera, Tortricidae) (Light et al. 2001; Light and Knight 2005). Pear ester is efficient for population monitoring (Knightet al . 2013, 2019) and for behavioural disruption of codling moth larvae and adults, alone or combined with sex pheromone (Light and Knight 2011; Knight et al. 2012; Light and Beck 2012; Knight and Light 2013). The discovery of pear ester demonstrates the potential of kairomones to both improve pheromone-based techniques and to design stand-alone applications. That pear ester is released only in trace amounts from green apples (Gonzalez et al. 2020) underlines that the abundance of volatiles in plant headspace does not correlate with their behavioural saliency. Compounds released in large amounts often stem from main biosynthetic pathways shared by many plants, and cannot encode specific host plant finding.
The most widely employed tool for studying plant compounds mediating host attraction is gas chromatography coupled to electroantennographic detection. GC-EAD measures the response of the entire antenna to odorants (Arn et al. 1975), and biases compounds occurring in large amounts in headspace collections. GC-EAD was designed as an efficient and reliable tool for detection and identification of trace amounts of sex pheromones. GC-EAD suffers, however, from serious bias and produces false positives when screening plant headspace. Ubiquitous compounds present in large amounts, for example short aliphatic acetates or alcohols, farnesenes, linalools and caryophyllenes, invariably elicit an antennal response, generated by the ensemble of olfactory sensory neurons (OSNs) on the antenna, expressing the entire olfactory receptor (OR) repertoire. Typically, ORs respond to large amounts of volatiles that are structurally similar to their cognate ligands. A diffuse signal from many OSNs on the antenna is sufficient to produce electroantennograms, but its behavioural relevance is uncertain. An active compound such as pear ester, on the other hand, has been overlooked in GC-EAD recordings due to its low abundance.
The discovery of the genetic code of insect ORs (Clyne et al. 1999) enables a new approach. The ligand binding specificity of ORs determines the spectrum of volatile chemicals transmitted by OSNs from the antenna to olfactory centers in the brain. Sequencing antennal RNA extracts and annotation provides OR expression data and a first functional differentiation, between pheromone receptors (PRs) and ordinary ORs, responding to environmental odorants. Subsequent phylogenetic analysis groups orthologous ORs from related species and provides leads on putative ligands, through comparison with an accumulating database of deorphaned insect ORs (Fleischer et al. 2018, Robertson 2019). Single ORs are accordingly a tool of choice to interrogate the plant odorscape for bioactive compounds. A powerful experimental approach is to express ORs singly in defined sensilla of the antenna of Drosophila melanogaster (Dobritsa et al. 2003; Hallem et al. 2004), where they can be addressed with single sensillum electrophysiological recordings, coupled to gas chromatography (GC-SSR).
In codling moth Cydia pomonella (Lepidoptera, Tortricidae), Cpom OR3 has been deorphaned, following transcriptome analysis (Bengtsson et al. 2012, Walker et al. 2016) and heterologous expression (Bengtsson et al. 2014, Cattaneo et al. 2017, Wan et al. 2019). The main ligand of CpomOR3, which belongs to the PR clade, is the plant volatile pear ester (Light et al. 2001; Light and Knight 2005, Bengtsson et al. 2014). A recent assembly of the codling moth genome reveals presence of two copies of CpomOR3, which, according to functional characterization in Xenopus oocytes, respond to a lesser extent also to codling moth sex pheromone, codlemone (Wan et al. 2019). A seemingly conserved response in a closely related species underscores this deeply rooted interconnection of pheromone and plant volatiles. Green budworm moth Hedya nubiferana (Lepidoptera, Tortricidae) is attracted to codlemone (Arn et al. 1974, El-Sayed, 2019) and to pear ester (Schmidt et al. 2007, Jósvai et al. 2016).
We have investigated the response of green budworm moth H .nubiferana to codling moth sex pheromone and to pear ester, in laboratory and field bioassays. Comparative phylogenetic analysis of ORs in the antennal transcriptome of green budworm and codling moth confirm the behavioural evidence and suggest the presence of a conserved olfactory channel dedicated to these compounds, in both species. This demonstrates how functional characterization of ORs in model species such as codling moth (Bengtsson et al. 2014; Gonzalez et al. 2016), followed by in silico studies of antennal transcriptomes in the taxonomically related species will advance the identification of insect kairomones, and the development of insect management.