Nectar production and carbohydrate composition across floral sexual phases: contrasting patterns in two protandrous Aconitum species (Delphinieae, Ranunculaceae)

• Direction of bias in nectar production may be stable or vary between sexual phases.
• Variability in nectar traits may be considered to be evolutionary significant.
• Sexual selection hypothesis is not able to explain gender-biased nectar production.
• Selection against inbreeding appears to explain changes in nectar chemistry.

In dichogamous species, it is necessary that pollinators are attracted both to male- and female-phase flowers in order to achieve effective cross-pollination. We investigated, over a two-year period, how nectar production and composition differs in protandrous flowers of Aconitum lycoctonum L. and Aconitum carmichaelii Debeaux, two species originating from different geographical regions. Flowers of A. carmichaelii secreted approx. 2.5-fold more nectar than flowers of A. lycoctonum. The nectar of A. carmichaelii was sucrose-dominant (87.6:9.5:2.9, sucrose:fructose:glucose), whereas sucrose-rich nectar, lacking glucose, occurred in A. lycoctonum (39.9:60.1, sucrose:fructose). Total sugar concentration was similar (53%, on average) in both species. These species also showed contrasting patterns in nectar production and composition between the floral sexual phases. On average, in A. carmichaelii, male-phase flowers produced 2.4-fold more nectar than female-phase flowers in the two years of study, and nectar was sucrose-dominant, irrespective of sexual phase. By contrast, nectar production in A. lycoctonum was biased towards the male phase in one year of study and towards the female phase in the other, and whereas nectar was sucrose-dominant during the male-phase, it was fructose-rich during the female phase, indicating sucrose re-absorption. Although the characteristics of nectar in these Aconitum species indicate pollination by bumblebees, it is possible that their biogeographical history, and not pollinator selection alone, is important in understanding the lack of glucose in A. lycoctonum nectar. Variability in nectar production and/or carbohydrate composition between sexual phases suggests indirectly that nectar traits complement each of the latter. Nevertheless, further research is required if we are to understand the significance of these disparities in pollination ecology, i.e. for promoting pollinator movements between flowers and plants in order to achieve cross-pollination and avoid inbreeding.