Scent biogenesis and deceptive strategies in Arum maculatum – ARUM

Arum maculatum (Araceae) has long fascinated due to its foetid scent and deceptive pollination via chemical imitation of brood-sites (putatively dung and other degrading organic matter), its thermogenic activity, and its temporary trapping of fly pollinators (mainly midges such as Psychoda phalaenoides). Its pollination cycle takes place over two days: on the first day, pollinators are attracted by VOC (volatile organic compound) emissions of the thermogenic appendix (i.e., exposed and sterile upper part of the spadix), and slip downward into the lower spathe chamber. There, they crawl over the stigma of receptive female florets, depositing the pollen they might carry from another inflorescence. Pollinators are held overnight, and on the following afternoon, mature stamens cover the pollinators with pollen before they escape from the chamber. If such pollen-loaded insects are deceived a second time by a scented female-stage inflorescence, they eventually achieve cross-pollination. This protogynous reproductive strategy ensures that no selfing occurs. Consequently, there is a strong selection pressure to efficiently attract pollinators.

The VOCs frequently identified as putative dung-imitating compounds are indole, p-cresol, and 2-heptanone. Monoterpenes and sesquiterpenes also can represent large proportions of VOC emissions in A. maculatum. Among other compounds, indole, p-cresol, monoterpenes and sesquiterpenes have recently been shown to elicit antennal responses in P. phalaenoides, confirming that pollinators are able to recognize these key compounds. Despite some of these compounds were previously shown to attract psychodid pollinators of A. maculatum (indole, p-cresol, 2-heptanone, a-humulene), recent studies did not confirm their attractiveness as a blend even considering further EAD-active compounds ((E)-ß-caryophyllene, (+)-ß-citronellene, 2,6-dimethylocta-2,6-diene, 3,7-dimethyloct-1-ene, 3,7-dimethyloct-2-ene, 2-heptanol, 6-methyl-5-hepten-2-one, 2-nonanone, 1-octen-3-ol, skatole) when offering them in natural concentrations in field biotests. This suggests that not yet tested scent compounds are needed for pollinator attraction in A. maculatum, among them potentially two unidentified sesquiterpenes representing up to 8% of the floral blend. Male florets within the pollination chamber of A. maculatum release mainly the sesquiterpene bicyclogermacrene, a VOC not emitted by the appendix. Interestingly, the few known psychodid sex pheromones are homoterpenes or sesquiterpenes (9-methyl germacrene B, 3-methyl-a-himachalene, cembrene) produced by males in some biting Lutzomyia species (Psychodidae, Phlebotominae) to attract females. Thus, bicyclogermacrene released from A. maculatum male florets might mimic male sex pheromones of the psychodid pollinators. So far, however, it is not known whether psychodid pollinators of A. maculatum have male sex pheromones. Sexually deceptive plants that exploit chemical mimicry were evidenced so far in orchids only.

This collaborative project aims to study whether Arum maculatum represents in fact a “two deceptive pollination systems in one” with the “outer” scent (i.e. emitted by the appendix) mimicking diverse brood-site scents and the “inner” scent (i.e. released by the male florets) being more pheromone-like. Our three key objectives are to elucidate the physiological and genetic bases of the floral VOC biosynthesis and secretion, identify and test the EAD-active floral VOCs responsible for pollinator attraction, and test whether some of those compounds resemble sex pheromones of psychodid pollinators. These objectives are key to better understand the molecular basis and ecological / evolutionary drivers of the strong and hyperdiverse floral scents in Arum maculatum. We present a multidisciplinary project combining chemistry, chemical ecology, ethology, histology and genetics of the floral scent of A. maculatum.