Time-course observations on 43 flowers of Causonis japonica revealed changes in flower disc colour and sexual expression (Table 1). Temporal changes in floral features showed no difference between diploid (19 flowers) and triploid (24 flowers) individuals. For example, flowering onset times did not differ substantially between ploidy level (diploid: from 07:07 to 13:27, triploid: from 06:58 to 14:49). However, the flowering duration varied significantly from flower to flower, ranging from a minimum of one day to a maximum of six days. Regardless of the ploidy level, all flowers with damaged styles (14 flowers) exhibited brown stigmas after the male phase, then ceased floral development prior to the female phase.
Figure 1 shows the typical time-course changes of C. japonica flower features (flower ID 35 in Table 1) according to the RGB values (representing the activities of nectar secretion: see below). As in the other 42 examined cases shown in Table 1, the initial colour of this flower disc immediately after anthesis (male phase) was orange (Stage 1, Fig. 1a, RGB: 255, 88, 16), as reported earlier7. Immediately after the petals and stamens fell off, the flower disc colour changed to pink (Stage 2, Fig. 1b, RGB: 255, 82, 102). The styles were not yet elongated at this stage, and the flowers were asexual. In 12 cases with damaged styles and brown stigmas, the flower discs remained pink until the flowers fell off (shown as “O–P” in Table 1).
However, in the remaining 31 flowers with normally elongated styles, maturation of the styles (female phase) coincided with the flower discs again exhibiting a distinct orange colour (Stage 3, Fig. 1c, RGB: 255, 88, 16). After the female phase, the flower discs turned pink again (Stage 4, shown as “O–P–O–P” in Table 1), and brownish colouration appeared in the stigmas (Fig. 1d, RGB: 255, 120, 94). Finally, the flower discs turned to a faded pink (Fig. 1e, RGB: 232, 162, 169) just before the flowers fell off. Therefore, the above observations imply that colour-change has a strict correlation with sexual phase.
The timings of the disc colour change to the second orange stage (female phase) varied depending on the onset time of each flower. Most flowers that opened before 10:00 reached the second orange stage (female phase) on the afternoon of the same day (except for two flowers, ID 1 and 2 in Table 1). Conversely, flowers that bloomed after 10:00 reached the second orange stage (female phase) at approximately noon the following day. These flowering processes were not fully synchronised in the same inflorescence; therefore, pink and orange discs often coexisted in the same inflorescence. Indeed, we can collect various stages of flowers at a time point from one population as shown in Fig. 2a.
Figure 1 also shows the typical time-course changes of nectar activities (flower ID 35 in Table 1), which indicates active nectar secretions during both orange colour stages. That is, the flower discs secreted nectar in both male and female phases, with no visible nectar secretion in the pink stages. Moreover, nectar secretion in the female phase of this flower was higher than that in the male phase, a tendency that was also observed in other flowers; however, the total volume of nectar varied among the flowers shown in Table 1. During anthesis, we confirmed that bees, wasps, ants and other insects visited the flowers as previously described7 (Supplementary Fig. 1).
Longitudinal sections of flowers in the pink-coloured and orange-coloured stages (Fig. 2a) revealed that pigmentation occurred only in a subset of parenchymatous cells in both cases (Fig. 2b, c). No structural or cytological changes were observed between the initial orange stage and the pink stages (Fig. 2d, e), suggesting that the observed oscillating colour change depends on the degradation and biosynthesis of orange pigments.
To understand what pigments are involved in the dual colour change of the C. japonica flower disc, we extracted carotenoids and chlorophyll with Acetone. Anthocyanin was also extracted with Methanol-HCl. As a result, while anthocyanin content was not significantly altered throughout the stages examined, we found that carotenoid level strongly correlated with the colour change detected by naked eye. More specifically, in stages 1 and 3 the carotenoid content was high (63.8 and 65.3 µg/g dry weight, respectively), but significantly decreased in stages 2 and 4 (14.3 and 36.5 µg/g dry weight, respectively) (Table 2). Interestingly, an increase in chlorophyll content was confined to stage 4 (Table 2). Together, the observed dual colour change was ascribed to the decrease (stage 2) and the increase (stage 3) of carotenoid contents in the flower discs.
Source: Ecology - nature.com