Knauer, A.C., Kokko H. and Schiestl, F.P.
Why do organisms signal honestly when the interests of signallers and receivers conflict, and a signaller could conceivably benefit from dishonest signals? Plants signal to their animal pollinators to advertise rewards and in return receive directed pollen transfer between individuals. The degree of honesty of floral signals however shows a large diversity between plant species; while some plants honestly indicate reward amounts, others are deceptive in their signalling and do not reward their pollinators at all. Studying the evolution of signal-reward associations in flowers is therefore crucial to understanding how mutualistic relationships between plants and pollinators are maintained.
Mathematical modelling of costs and benefits has brought the general understanding that honest signalling can be an evolutionarily stable strategy when individuals do not profit by deviating from honesty. Thus honesty is expected to evolve when low-quality individuals experience higher costs and/or lower benefits from a ‘high’ signal than do high-quality individuals, which prevents them from cheating. Specific modelling of plant-pollinator interactions has furthermore resulted in the formulation of some conditions that favour the establishment of honest floral signalling in a plant population.
In this manuscript, we study honest floral signals in Brassica rapa (turnip) and address the question of how a signal-reward correlation is maintained. By a series of experiments, we test the predictions made by the existing modelling framework: positive pollinator-mediated selection on honest floral signals and rewards in combination with a trade-off between these floral traits. We subsequently incorporate our findings into an updated model which we finally verify experimentally.
We show that honest signalling is an optimal strategy in B. rapa, and is maintained by a combination of pollinator behaviour and resource limitation. Individuals with low resource availability in flowers optimise their fitness by producing small signals given a trade-off between signals and seed development. Simultaneously, because of resource limitation in seed production, such individuals profit less from large nectar volumes and the resulting long visitation times by bees compared to individuals with high resource availability. Thus, our study provides evidence that differential benefits (low-signalling individuals do not profit from high nectar volumes) maintain signal honesty in B. rapa.