Floral scent evolution in the section Pseudophrys: pollinator-mediated selection or phylogenetic constraints?

Sexually deceptive orchid species from the Mediterranean genus Ophrys usually interact with one or a few pollinator species by means of specific floral scents. In this study, we investigated the respective role of pollinator-mediated selection and phylogenetic constraints in the evolution of floral scents in the section Pseudophrys. We built a phylogenetic tree of 19 Pseudophrys species based on three nuclear loci; we gathered a dataset on their pollination interactions from the literature and from our own field data; and we extracted and analysed their floral scents using solid phase microextraction and gas chromatography-mass spectrometry. We then quantified the phylogenetic signal carried by floral scents and investigated the link between plant-pollinator interactions and floral scent composition using phylogenetic comparative methods. We confirmed the monophyly of the section Pseudophrys and demonstrated the existence of three main clades within this section. We found that floral scent composition is affected by both phylogenetic relationships among Ophrys species and pollination interactions, with some compounds (especially fatty acid esters) carrying a significant phylogenetic signal and some (especially alkenes and alkadienes) generating dissimilarities between closely related Pseudophrys pollinated by different insects. Our results show that in the section Pseudophrys, floral scents are shaped both by pollinator-mediated selection and by phylogenetic constraints, but that the relative importance of these two evolutionary forces differ among compound classes, probably reflecting distinct selective pressures imposed upon behaviourally active and non-active compounds.

Olfactory and behavioural responses of tabanid horseflies to octenol, phenols and aged horse urine.

Electrophysiological and behavioural responses of females of two tabanid species, Tabanus bromius L. and Atylotus quadrifarius (Loew) (Diptera: Tabanidae), to ammonia, octenol (1-octen-3-ol), phenols and aged horse urine were compared. Electroantennogram (EAG) responses in both species to octenol, 4-methylphenol (4MP), 3-propylphenol (3PP) and a phenol mixture (4MP and 3PP at a ratio of 16 : 1) increased in a dose-dependent fashion. The most effective stimulus was 4MP and synergism between the two phenols may exist. Aged horse urine also elicited strong EAG responses in both species. Using gas chromatography-mass spectrometry (GC-MS) analysis, we identified 29 compounds in horse urine, which included, in particular, ketones, fatty alcohols and phenols, among which 4MP was the most abundant component (~ 80%). Trapping experiments were carried out using Nzi traps baited with various odours. Octenol and the phenol mixture in combination with ammonia increased catches of tabanids by 1.8-2.8 times relative to ammonia alone. Aged horse urine increased catches of T. bromius and A. quadrifarius by 2.2 and 4.1 times, respectively. The high attractiveness of aged horse urine, especially for A. quadrifarius, is not likely to derive from 4MP alone, but from the mixture of various active compounds used in host location.

Behavioural and electrophysiological responses of females of two species of tabanid to volatiles in urine of different mammals.

Urine volatiles from different ungulates (cows, horses and sheep) were tested as bait for tabanids in southeastern France using Nzi traps during the early summer of 2011. Tabanus bromius Linnaeus, 1758 and Atylotus quadrifarius (Loew, 1874) (both: Diptera: Tabanidae) were the most captured species, respectively representing 57% and 41% of all tabanids collected (all of which were female). Horse urine significantly increased catches of T. bromius (1.6-fold) and A. quadrifarius (3.5-fold), and sheep urine significantly increased catches of A. quadrifarius (2.5-fold). In parallel, an electroantennogram (EAG) study was conducted for the first time on these two species, in which EAGs were recorded using 1-octen-3-ol and extracts of the same urine samples used in the field. For T. bromius, the EAG response to 1-octen-3-ol increased quasi-sigmoidally with dose, with a maximum response at ≥100 µg on filter paper. For both species of tabanid, cow and horse urine elicited larger EAGs than did sheep urine. The behavioural implications in host-seeking and feeding habits are discussed.