X-Rays vision for uncovering reproductive investment in plants as a function of breeding system

Thomas Kreisberger, Susanne Pamperl, Marion Chartier, Sara Manafzadeh, Stephan Handschuh, Susanne Sontag, Hannes Paulus, Jürg Schönenberger, Yannick Städler

Flowering plants are stationary organisms that must rely on external agents for mating. Therefore, the ratio of male to female reproductive investment, the pollen/ovule ratio (P/O) depends on the efficiency and behaviour of external agents in transferring pollen. Most flowers pollinated by biotic vectors offer a reward (usually pollen and/or nectar) to their floral visitors. However, in various flowering plant lineages deceiving, i.e., non-rewarding flowers have evolved. Experience of deceit often causes pollinators to avoid deceptive flowers, thus leading to unequal visitation rates between the first flowers to open in a inflorescence and the subsequent ones. Does such a behavioural difference lead to adaptation in intra-individual reproductive investment?
In order to answer this question we focussed on the Orchidinae (orchids), which exhibit variation between rewarding/deceptive reproductive strategies at a low taxonomic level. We developed a new set of methods to precisely quantify the notably hard to acquire pollen and ovule counts in orchids. Flowers from different positions within the inflorescences were collected, contrasted and scanned via X-Ray Computed Tomography (CT). Since contrasting agents for CT tend to selectively enrich in ovules and pollen, these structures could be isolated from their surrounding tissues by image processing; pollen and ovules were then counted via an efficient, semi-automatic procedure using the ImageJ and AMIRA software.
In all taxa studied, the number of ovules decreased from the bottom to the top of the inflorescence, whereas the number of pollen grains fluctuated around its mean. The P/O was thus highest in the upper regions of the inflorescences whereas the lowest values could usually be found in the lower regions. Rewarding species displayed stronger variance in pollen grain numbers than deceptive species, possibly reflecting the higher chances of visitation of rewarding flowers across the whole inflorescence. Finally, we compiled P/O ratios in Orchidinae from the literature and from unpublished data and reconstructed the evolution of this character in association with the presence or absence of a pollinator reward. Our new CT-based methods provide a robust way of P/O measurements even for the numerous and minute pollen grains and ovules of orchids. The possibilities of the methods we developed include the measurement of phenotypical traits and reproductive investment on the same material, thereby opening new alleys in the study of reproductive investment in plants.

Department of Botany and Biodiversity Research, Department of Evolutionary Biology
External organisation(s)
Universität Zürich (UZH), Veterinärmedizinische Universität Wien
Publication date
Austrian Fields of Science 2012
106042 Systematic botany, 106008 Botany, 106012 Evolutionary research
Portal url