**1. Introduction**

Pseudocopulation as a means of pollination was first reported over 100 years ago, in two parallel systems [1,2]. Correvon and Pouyanne made observations of European *Ophrys* orchids [1], while in Australia, *Cryptostylis* orchids were reported to use the same sexually deceptive strategy, in which insect pollinators attempt copulatory or courtship behaviour with the flower, thereby transferring pollinia [2,3]. Insect sexual attraction is induced through chemical and physical mimicry of female insects. Pollination by sexual deception is now known to be a phenomenon that has evolved independently multiple times on different continents. There are several hundred confirmed cases in the Orchidaceae, with many more likely to be discovered with future studies [4–6]. There are also single reports of sexual deception in the Asteraceae [7] and Iridaceae [8], indicating that this pollination strategy may be more common than is currently known.

Following the initial observations of pollination via sexual deception in *Ophrys* and *Cryptostylis* orchids, an intensive Swedish research program was launched in 1948 to investigate the chemical cues underlying this bizarre pollination strategy. *Ophrys insectifera* and some southern European *Ophrys* and their solitary bee pollinator species were the main study species [9]. In these early studies, field experiments demonstrated that floral volatiles were the key to pollinator attraction [9,10]. With the use of electroantennography (EAG), it was later shown that two species of male sphecid wasp pollinator, *Argogorytes mystaceus* and *A. fargeii*, unlike their conspecific females, responded to tentatively identified alkanes, alkenes, and terpenes in sorption headspace extracts of *O. insectifera* flowers [11]. A few years later, the first evidence of chemical mimicry of several species of *Andrena* bee pollinators by *O. fusca* and *O. lutea*, was found: aliphatic alcohols, monoterpene- and sesquiterpene alcohols, and aldehydes attracted the patrolling males to varying degrees [12,13].

The first identification of pollinator sexual attractants in the genus *Ophrys* did not occur until the late 1990s, with the successful structural elucidation of attractants from *O. sphegodes* [14,15]. A key to the detection and identification of the semiochemicals from this species was the use of gas chromatography coupled with electroantennogram detection (GC-EAD), which revealed a set of 14 electrophysiologically active compounds to be shared among the orchid and the female of its bee pollinator, *Andrena nigroaenea*. Before being confirmed as attractants in field bioassays, these compounds were identified by GC-MS, including microderivatisation experiments, as a series of long-chained alkanes and alkenes. Furthermore, three (*Z*)-7 alkenes were discovered to be responsible for the attraction of male *Colletes cunicularius* bees to *O. exaltata* [16]. The chemical stimuli for the sexual attraction of various *Ophrys* pollinators also include other types of structures, as shown when a mixture of hydroxy- and keto acids, together with aldehydes and esters, were identified as the attractants in *O. speculum*, which is pollinated by male *Campsoscolia ciliata* scoliid wasps [17].

In Australian sexually deceptive orchids, 1,3-cyclohexanediones (chiloglottones) have been identified as pollinator attractants in *Chiloglottis* [18], as have hydroxymethylpyrazines and a β-hydroxylactone (drakolide) in *Drakaea* [19–22], (methylthio)phenols, acetophenones and monoterpenes in *Caladenia* [23–25], and tetrahydrofuran acid derivatives in *Cryptostylis* [26].

Besides the discovery of a broad range of compounds pivotal for pollination in sexually deceptive orchids, there has also been interest in the biosynthesis of these compounds, with the aim to link biosynthesis to the evolution and speciation of orchids. Schlüter and Schiestl [27] predicted that, in *Ophrys,* the biosynthesis of alkenes would follow the biosynthetic pathway for alkanes [28], but with the addition of an extra desaturation step, potentially achieved by stearoyl-acyl carrier protein desaturases (SAD). Later, three putative SAD genes (SAD1-SAD3) were isolated [29]. Transgenic expression and in vitro enzyme assays revealed SAD2 to be a functional desaturase capable of introducing 18:1 Δ<sup>9</sup> and 16:1 Δ<sup>4</sup> fatty acid intermediates, from which it was hypothesized that (*Z*)-9 alkenes and (*Z*)-12 alkenes are built. Three additional putative SAD genes (SAD4-SAD6) were also identified from an *O. sphegodes* transcriptome [30].

In *O. sphegodes* and *O. exaltata*, SAD1 and SAD2 expression levels were shown to be significantly correlated with (*Z*)-9 and (*Z*)-12-alkene production, while high SAD5 expression was correlated with the (*Z*)-7-alkene production unique to *O. exaltata* [31]. In vitro enzyme activity studies further showed that a putative housekeeping desaturase, SAD3, catalyses the general reactions of stearate to oleate (18:0-ACP to 18:1 Δ9-ACP), and palmitate to palmitoleate (16:0-ACP to 16:1 Δ9-ACP), whereas SAD5 is a specialized 16:0 Δ9-ACP enzyme [32]. Subsequent elongation of a 16:1 Δ9-ACP to a 26:1 Δ19-coenzyme A precursor, followed by decarbonylation, would yield the (*Z*)-7 alkene (25:1 Δ7) that characterizes *O. exaltata*.

In *O. speculum*, the pollinator attractants were also identified as carboxylic acid derivatives [17]. The most attractive compounds from both floral extracts and females of the scoliid wasp pollinator *Campsoscolia ciliata* were (ω-1)-hydroxy- and -oxo acids. However, it is noteworthy that the pseudo-copulation rates in field bioassay experiments more than doubled when aldehydes such

as (*Z*)-9-octadecenal and octadecanal, together with the esters ethyl linoleate and ethyl oleate, were added to the dummy female [17].

The phylogenetic relationships within *Ophrys* are currently under debate [33–37], with some phylogenetic analyses indicating that the *Argogorytes*-pollinated *O. insectifera* group represents a basal taxon, while the latest studies place the *O. fusca* complex, including *O. iricolor*, as ancestral [36,37]. All studies agree that wasp pollination is ancestral to bee pollination in *Ophrys*.

To obtain a broader understanding of the chemical details of semiochemicals in the wasp-pollinated *O. insectifera*, and sex pheromone candidates in its pollinator *A. fargeii*, we used GC-EAD, GC-MS, microderivatisation reactions, and organic synthesis to identify EAD-active compounds. These semiochemicals were compared with previously identified pollinator attractants from the bee-pollinated *O. sphegodes* and wasp-pollinated *O. speculum*, and biosynthetic relationships within *Ophrys* were proposed.
