Adaptive diversification of Fly Orchids as a new model to investigate ecological speciation and evolutionary radiations – DiversiFly

The basic idea was to focus on areas where the geographical ranges of species in the Ophrys insectifera group overlap: in the Grands Causses for Ophrys insectifera and Ophrys aymoninii, and in northern Spain (specifically Catalonia) for Ophrys insectifera and Ophrys subinsectifera (10 localities). To these two areas, we added two ‘control’ sites, geographically intermediate: 7-Versols-Et-Lapeyre, which is situated in an area where only O. insectifera is present, and 12-Rodome.

At each site, the individuals were:

- Geolocated and tagged. The sites were then visited twice a year: at flowering time (May) and at fruiting time (July) in order to estimate the probability of survival (or at least of re-flowering) and reproduction (or at least of fruit production).

- Each individual was studied for morphological characteristics relating to the vegetative system and reproductive structures (flowers).

- The measured flower was also photographed in a standardised manner to study other floral traits as well as colouration.

Pour les études génomiques et transcriptomiques, 2 bractées ont été prélevées pour l'extraction de l’ADN et la fleur mesurée a été prélevée et stockée dans une solution de RNAlater à 4°C pour l' extraction de l’ARN.

- Une deuxième fleur, en général celle du dessus, a aussi été prélevée et immergée dans de l’hexane jusqu’à analyse du bouquet d’odeur en chromatographie en phase gazeuse couplée à de la spectrométrie de masse (GC-MS).

Les analyses de génomique des populations ont ensuite été effectuées sur la base d’un génotypage ‘nGBS’ (pour normalised Genotyping By Sequencing). Les transcriptomes floraux ont quant à eux été séquencés en Illumina.

In total, 262 Ophrys individuals from 12 populations were analysed as part of this study, enabling us to generate a novel multi-omic dataset for this system.

The population genomics study conducted where the ranges of species within the O. insectifera group overlap in pairs has thus revealed patterns that are indistinguishable using more traditional sequencing approaches. Thus, it appears that in France, O. aymoninii, endemic to the Grands Causses, is already well differentiated from O. insectifera, with which it is often sympatric and syntopic. The absence of signs of recent hybridisation suggests that reproductive isolation is already well established. Interestingly, however, we have confirmed that O. aymoninii hybridises with another Ophrys species that is phylogenetically more distantly related but with which it shares the same pollinator, Andrena combinata: the late-blooming local form of an Ophrys belonging to the litigiosa/virescens group. The situation is, however, far more complex in Spain, where, although morphologically quite distinct, Ophrys subinsectifera may in fact form a more or less widespread hybrid swarm with the local populations of O. subinsectifera.

Integrative approaches have notably enabled us to identify an island of divergence located on chromosome 10, hitherto unknown in Ophrys. This region includes several outlier SNPs, some of which correspond to genes potentially involved in floral development (LFY) and perhaps scent production (SAT for Stemmadenine O-acetyltransferase and MAS for Malate synthase)? These genes also appear to be differentially expressed between species, which supports a probable functional impact. Other candidates for the floral phenotype have also been identified, including DFR1 (colouration), CCR (for Cinnamoyl-CoA reductase) and paralogues of SAD (odour, well known in Ophrys). Local PCAs subsequently indicated that this region ultimately appeared more similar to the Ophrys of the sphegodes group than to those of the insectifera group in O. aymoninii. We now hypothesise that this genomic region could correspond to an introgressed region, potentially adaptive and perhaps protected from the effects of recombination due to a structural variation (chromosomal inversion?).

Morphometry and colouration – The analysis of calibrated photographs taken in the field (morphology, colouration) has confirmed that any quantitative information derived from them can be highly informative. In this respect, it usefully complements data collected directly in the field, particularly 1D morphometry carried out using a ruler or calliper. However, to carry out more comprehensive phenotyping, to avoid experimental bias and to produce durable data whilst remaining non-destructive, the idea is now to turn to the use of portable 3D scanners to save a complete digital copy of each plant.

Genotyping – By identifying relevant pockets of divergence, the nGBS approach yielded ‘unexpected’ results given the relatively low marker density across such a large genome. The idea now is to take things further with whole-genome resequencing approaches to refine these results. In addition, we are also planning to sequence, assemble and annotate new reference genomes at a truly chromosomal scale.

Transcriptomics – Here too, ‘taking a risk’ paid off. At the start of the project, few contractors were confident that we could obtain mRNA in sufficient quantity and quality from flowers stored in RNAlater under field conditions. Ultimately, the failure rate in library preparation was very low, and for the first time we were able to establish a link between the transcriptomic compartment and other levels of integration within natural Ophrys populations.

Metabolomics – Although, in principle, less expensive and less prone to picking up ‘contaminant’ molecules from the environment, immersion in hexane yielded low concentrations and meant we had to sacrifice one flower per individual. In future, we plan to give preference to SPME fibres for analysing floral odour profiles.

Developing an integrative research allowing to elucidate both the links between genotype, phenotype and fitness and the genomic bases of the variation form a major challenge in our understanding of speciation and the emergence of biodiversity. In DiversiFly, I propose to implement such approach by generating, analyzing and comparing phenotypic (morphometry, coloration, odor), genomic and ecological data on Ophrys orchids. This genus displays one of the highest rates of diversification and hybrid zones formed in the wild by some of its species makes Ophrys a promising model to better understand the causes of adaptive radiations: phenomena of intense and rapid diversification in response to ecological selection pressures. Through its approaches, DiversiFly aims at combining studies led at different evolutionary scales, working on endemic and threatened species.

In a first part, the parallel study of two hybrid zones will allow us to determine what are the phenotypic traits (morphology, coloration, odor) predominantly involved in adaptation and reproductive isolation within the O. insectifera clade. Through cline analyses, outlier research and association studies, we will then confront phenotypic and genomic data (transcriptomes and GBS data) in order to look for the genomic bases of traits of interest. The links between phenotypes and individual fitness will be evaluated based on life history traits related to reproductive success and survival. As all individuals will be marked, phenotyping and fitness will be evaluated each year over the four years of the project to investigate their stability.

In a second part, we will link micro and ‘macro-’ evolutionary scales by working on the whole genus Ophrys. We will use floral transcriptomes on species, representative of the diversity of the genus in order to identify genes involved in shaping floral phenotype, so particular of the Ophrys. Following a comparative approach, we will base our study on both sequence variation and gene levels of expression.

Our results will have a significant impact in the field of ecology and evolutionary biology, contributing to the emergence of a new plant model, with a complex genome, to study speciation and evolutionary diversification. They will also contribute to fields such as genomics, conservation biology, systematics and taxonomy. We also believe that with its flowers mimicking insects, Ophrys forms an excellent model for scientific communication and towards general audience.