Floral nectar composition has been explained as an adaptation to factors that are either directly or indirectly related to pollinator attraction. However, it is often unclear whether the sugar composition is a direct adaptation to pollinator preferences. Firstly, the lower osmolality of sucrose solutions means that they evaporate more rapidly than hexose solutions, which might be one reason why sucrose-rich nectar is typically found in flowers with long tubes (adapted to long-tongued pollinators), where it is better protected from evaporation than in open or short-tubed flowers. Secondly, it can be assumed that temperature-dependent evaporation is generally lower during the night than during the day so that selection pressure to secrete nectar with high osmolality (i.e. hexose-rich solutions) is relaxed for night-active flowers pollinated at night. Thirdly, the breeding system may affect selection pressure on nectar traits; that is, for pollinator-independent, self-pollinated plants, a lower selective pressure on nectar traits can be assumed, leading to a higher variability of nectar sugar composition independent of pollinator preferences, nectar accessibility and nectar protection. To analyse the relations between flower tube length, day vs. night pollination and self-pollination, the nectar sugar composition was investigated in 78 European Caryophylloideae (Caryophyllaceae) with different pollination modes (diurnal, nocturnal, self-pollination) using high-performance liquid chromatography (HPLC). All Caryophylleae species (Dianthus and relatives) were found to have nectar with more than 50% sucrose, whereas the sugar composition of Sileneae species (Silene and relatives) ranged from 0% to 98.2%. In the genus Silene, a clear dichotomous distribution of sucrose- and hexose-dominant nectars is evident. We found a positive correlation between the flower tube length and sucrose content in Caryophylloideae, particularly in day-flowering species, using both conventional analyses and phylogenetically independent contrasts.
Nexus file of Caryophylloideae
Nexus file of the phylogenetic relationships among the investigated Caryophylloideae. Our analyses were based on an unresolved phylogeny constructed using published molecular phylogenies and taxonomic groupings. The file was used in MESQUITE for further analyses.
Figure 1 - model (as MS powerpoint file)
Figure 1. Factors that might affect evaporation of water from nectar. (A) Flower nectar depth (nectar is hidden in long-tubed flowers), (B) hydrolyis of sucrose by invertase into hexoses (increase in osmolality), (C) Relation between activity time of flowers (diurnal vs. nocturnal) and temperature (higher during the day) and humidity (higher during the night), (D) features of the anthophore (an elongated internodium above the calyx found in some Caryophylloideae) like its length and hairs. Note: While for hypotheses (A) and (B) evidence can be found in the literature (see text), (C & D) has been tested in this study.
Figure 1 - model.pptx
Figure 2 - flower morphology (as Ms powerpoint file)
Figure 2. Floral features in Silene: A. Half-drawing of a hypothetical female Silene flower (with sterile anthers and nectar along the anthophore) illustrating where nectar is located and how flower structure (here the protruding corona scales) may affect the functional flower length for flower visiting insects. B. Silene colorata; showing how the petal claws and corona scales may increase the functional flower length. C. Anthophore of Silene dichotoma with nectar.
Figure 2 - flower morphology.pptx
Figure 3 - phylogeny (CorelDRAW file)
Figure 3. Phylogenetic distribution of nectar sugar composition, flower tube length and pollination modes in 78 Caryophylloideae. Nectar sugar composition according to the classification system of Baker & Baker (1983); Flower tube length data (gray box; according to Jürgens (2006) and this study. Pollination modes: blue = day-flowering, black = night-flowering, white = self-pollination.
Figure 3 - phylogeny.cdr
Figure 4 - correlation (MS powerpoint file)
Figure 4. Correlation between functional flower tube length [mm], as the distance from the bottom of the calyx to the end of the calyx tube, or the petal claws, or the corona scales (which ever is extending functional tube length the most), and median amount of sucrose (%) in 78 Caryophylloideae in relation to pollination system (A), and phylogeny (B) (Spearman (all species) rS = 0.542, p < 0.05; Spearman (nocturnal) rS = 0.39, p < 0.05; Spearman (diurnal) rS = 0.84, p < 0.05); Spearman (self-pollinated) rS = 0.31, ns). Dianthus glacialis 1* (mainly selfing), data from Erhardt & Jäggi (1995); Dianthus gratianopolitanus 2* (intermediate between butterfly and moth pollination), data from Erhardt (1990).
Figure 4 - correlation.pptx
Appendix 1: Origin of plants sampled for analysis of nectar sugar composition. Vouchers have been deposited in the herbaria of the Universities of Ulm and/or Bayreuth. Appendix 2 (Data from Jürgens et al., 2002). Pollination mode of Dianthus, Saponaria, Vaccaria, Agrostemma and Silene species based on information from the literature. Appendix 3 (Data from Witt, 2003). Average reproductive success of bagged hermaphrodite flowers of 66 Caryophylloideae species sorted by fruit set, seed set per fruit, and seed set of all bagged flowers. Appendix 4. Flower tube length, anthophore length, hair length and hair density on anthophores of 78 Caryophylloideae. Appendix 5 – SEM of the anthophore region of selected Caryophylloideae. Appendix 6. Species of which a few samples contained traces of possibly other sugars in addition to the main nectar sugars (sucrose, glucose, fructose). Reference list for Appendix.