The role of preadaptations or evolutionary novelties for the evolution of sexually deceptive orchids.

Abstract

Most flowering plants use visual and olfactory cues to attract their pollinators (Raguso, 2004), floral odors are, however, often of key importance in specialized pollination systems (Schiestl, 2005; Brodmann et al., 2008; Dötterl & Vereecken, 2010; Ayasse et al., 2011; Jürgens et al., 2013). They are produced in large numbers by various groups of plants, and so far c. 2000 components have been identified in the headspace of flowers (Knudsen et al., 2006). Despite this knowledge on the diversity of floral scents, less information exists about the factors that affect the evolution of scent patterns and the associated plant–pollinator relationships. One way to approach this knowledge gap is to study specialized pollination systems, such as those of deceptive plants and their betrayed pollinators. For example, recent studies focused on mimicry of insect oviposition sites (Jürgens et al., 2013), prey (Brodmann et al., 2008) or mating partners (Schiestl, 2005; Ayasse et al., 2011). The latter, sexual deception, is a highly specialized pollination mechanism that mainly occurs in orchids. In these systems floral volatiles have been shown to play a key role in pollinator attraction, species isolation and speciation (Schiestl, 2005; Ayasse et al., 2011). Sexually deceptive orchids mimic the sex pheromone of females of their pollinator species to attract male insects that then pollinate the flowers. Chemical analyses have demonstrated that flowers of sexually deceptive orchids produce complex and species-specific mixtures of mostly widespread compounds or just a single or a few unusual chemical compounds (reviewed in Ayasse et al., 2011). This pollination strategy has independently evolved in European, Australian, South African, and South American taxa (Gaskett, 2011). In this issue of New Phytologist, Bohman et al. (pp. 939–952) introduced a new chemical class of pollinator attractants in sexually deceptive orchids and studied whether preadaptations have been important in the evolution of sexual deception in Australian orchids. Preadaptation was found to be important in the orchid genus Ophrys, which is mainly distributed in the Mediterranean, and where many species lure their male pollinators by long-chain saturated and unsaturated hydrocarbons (Schiestl, 2005; Ayasse et al., 2011). A phylogenetically controlled study by Schiestl & Cozzolino (2008) indicated that production of alkenes that do have a key function in pollinator attraction evolved earlier than sexual deception and thus seems to be a primitive character state in Ophrys. ‘The new data by Bohman et al. suggest that sexually deceptive orchids adapt to locally abundant insect species with an effective sex pheromone communication system.’ Overall, the three sexually deceptive orchid genera, from which behavior mediating compounds were identified, use quite different compound classes for pollinator attraction (Fig. 1). Ophrys uses ubiquitous components of fatty acid metabolism (alkanes and alkenes), Chiloglottis uses uncommon compounds (chiloglottones), likely derived from ubiquitous intermediates in fatty acid formation (3-ketoacids and 2,3-unsaturated acids; Franke et al., 2009), and Drakaea uses uncommon nitrogen-containing compounds (pyrazines) derived from amino acid metabolism (Dickschat et al., 2010). The differences in signaling between Drakaea and Chiloglottis are especially unexpected as they show that two closely related Australian plant genera evolved different pollinator attraction strategies, although both use thynnine wasps as pollinators. A second aspect of the paper by Bohman et al. was to study the occurrence of pyrazines and chiloglottones in both closely related and more distantly related plant species with the aim to assess whether, comparable with Ophrys, preadaptation can explain the evolution of sexual deception in Australian orchids. Both compound classes were only found in a few species outside Drakaea and Chiloglottis, and the authors did not find evidence for ‘either chiloglottones or pyrazines as preadaptations with other functions that were subsequently coopted for pollinator attraction during the evolution of sexual deception’ (Bohman et al.). Another interesting issue brought up by this chemical survey was that plants pollinated by thynnine wasps produce either pyrazines or chiloglottones but no taxa produced both classes of compounds. The production of either pyrazines or chiloglottones can only partly be explained by phylogenetic constraints (e.g. closely related Paracaleana species emit different classes of compound), which allows us to hypothesize that thynnine wasp pollinators exert strong selective pressures to not emitting both classes of compounds. Behavioral assays with natural or synthetic scent samples will allow this assumption to be tested. Although it is clear that both strategies of insect sex pheromone mimicry, using blends of ubiquitous compounds (hydrocarbons in Ophrys) or unusual semiochemicals (pyrazines, chiloglottones in Australian orchids), can achieve species-specific attraction of male pollinators, the questions remain as to why sexually deceptive orchids evolved such diverse chemical strategies to deceive their pollinators, and why Australian orchids use compounds of much lower molecular weight and higher volatility than Ophrys species. A possible explanation is that typical pollinators of Ophrys orchids in Europe do not occur in Australia and vice versa. Many of the bee genera involved in Ophrys pollination such as Andrena, Eucera and Osmia do not occur in Australia (Michener, 2007), and thynnine wasps do not occur in Europe. In general, the bee fauna is particularly rich in the Mediterranean basin (Michener, 2007) where there is the highest diversity of Ophrys. The diversity is lower in Australia, although certain species of native bees are known to be important pollinators in Australia (including food deceptive and food rewarding orchids). Thynnines, however, are a diverse group of predominantly Australian wasps that are very abundant in many areas (Peakall, 1990). The new data by Bohman et al. suggest that sexually deceptive orchids adapt to locally abundant insect species with an effective sex pheromone communication system. Differences in the chemical nature of the pheromones in deceived insects explain the different chemical strategies exploited by the different orchids. A reason for differences in molecular weight and volatility of compounds between European Ophrys and Australian orchids may be that these compounds likely have different functions in the pheromone and mating system of the pollinators. While the chiloglottones and pyrazines in the Australian systems attract the male wasps from both long and short distances, the long-chain hydrocarbons in Europe only work at close range. Indeed, some of the European pollinators seem to be attracted from distance to their mates by compounds other than long-chain alkenes (e.g. short-chain aliphatic compounds, terpenoids), which, interestingly, also occur in Ophrys (Borg-Karlson, 1990). These short-chain compounds, however, do not elicit the copulatory behavior in males needed for pollination to occur in Ophrys. The Australian orchids, therefore, can use the same compounds for attracting the pollinators from distance and close range, whereas Ophrys orchids need to produce a different set of compounds for distance and close-range attraction. The mating systems of the pollinators possibly influence the chemistry of the sex pheromones and consequently the semiochemicals released by the orchids as well. For example, thynnine wasps fly long distances to find their females by volatile sex pheromones (Peakall, 1990). In many bees, however, males search for females in the close surroundings of nests (Michener, 2007), and in these cases very volatile compounds, which might even attract parasites and predators, may be of minor importance. In conclusion, chemical mimicry as a mechanism of sexual deception was only found in a small number of orchid taxa from a few regions of the world. Moreover, the chemistry of this special kind of mimicry is not well investigated in most of the species, in particular in species which are pollinated by insects other than those so far studied (e.g. ants, saw flies; Gaskett, 2011). The chemistry and biochemical/genetic pathways involved in sexual deception systems seem overall to be highly complex. Indeed, even in a few Ophrys species, pollinated by Bombus or a scoliid wasp, do chemicals other than hydrocarbons play a role in pollinator attraction (reviewed in Ayasse et al., 2011), and preliminary analyses suggest that this also is the case in Ophrys species pollinated by long-horned Eucera bees (M. Ayasse et al., unpublished). Identification of the key signals in many more sexually deceptive orchids together with a well-resolved phylogeny will be needed to learn more about the evolution of sexual deception and whether preadaptation or evolutionary novelty was more important for the radiation of this species-rich group of plants. Therefore, as in the study of Bohman et al., a combination of chemical and electrophysiological methods, as well as bioassays, should be used to identify pollinator-attracting compounds. As shown in O. speculum, where flowers produce large amounts of saturated and unsaturated hydrocarbons that do not have a function in pollinator attraction, a pure description of the compounds produced without confirming their importance by means of electrophysiological and behavioral assays may not be meaningful (Ayasse et al., 2003). The authors thank Andre Kessler for helpful comments on a previous version of the manuscript and Rod Peakall for providing photographs of orchids and pollinators.