P6

Trade-offs between pollinator attraction and florivore repellence in Tanacetum vulgare

Floral volatiles can be highly diverse and serve different functions. While they should attract pollinators, at the same time they may also be attractive for florivores. Similarly, non-volatiles metabolites of flower parts such as pollen contains nutritious but also toxic compounds, serving different functions for insects. Thus, the goal of P6 was to investigate the role of chemodiversity in shaping flower-insect interactions and in particular, to explore potential trade-offs flowers may face in attraction of mutualists and antagonists. A literature review and meta-analysis revealed that in general plants do face trade-offs in attraction of pollinators, because florivores detect and are attracted towards the same chemical signals, and repellent compounds simultaneously negatively affect pollinators. At the same time, chemodiversity of volatile flower metabolites could be connected to floral reward traits or non-volatile nutrients or toxins and should be considered in future studies. For experiments, Tanacetum vulgare was chosen which presents a suitable system to study chemodiversity effects on flower visitors due to its highly diverse terpenoid composition. Chemical analyses revealed a clear distinction among chemotypes based on flower head volatile organic compounds but also pollen micronutrient composition. Laboratory experiments with florivorous Olibrus beetles showed that they are differently attracted to flower heads of distinct chemotypes. In particular, the β-thujone chemotype was more attractive. This chemotype was also more suitable for survival than two other chemotypes. Preliminary trials in a greenhouse suggested that bumblebees might show a similar preference towards the β-thujone chemotype. In a field study using five chemotypes grown in either homogeneous or heterogeneous plots (with P5), we observed several flower visitor species on T. vulgare despite the high terpenoid richness of the flowers. Pollinators distinguished between the plot types and visited heterogeneous plots more frequently, while florivores did not show such a preference. We also found a positive correlation between pollinator visits and germination rate. Thus, in the field chemotypes may minimise trade-offs when growing in heterogeneous plots, leading potentially to higher fitness. Our data are used by P9 to model a virtual Tanacetum.

Taken together, our laboratory and field experiment results suggest that individual flower visitor species can distinguish among chemotypes of T. vulgare and are also affected in their performance, with potential trade-offs for the chemotypes. Overall, the project has answered some key questions about balancing flower-pollinator/florivore interactions, generally across plants using a meta-analysis approach and specifically in the chemodiverse and heavily insect-visited plant system T. vulgare through empirical studies. This project will end but aspects of it will be further investigated in P5.