RESUMO
During the evolution of angiosperm flowers, some floral traits may undergo certain changes in order to participate in screening. The stamens and pistils of Delphinium caeruleum are covered by two "door-like" staminodes, the evolutionary function of which, however, is quite unknown. In this study, we investigated whether D. caeruleum staminodes acted as visitor filters by assessing the respective strengths of staminodes and visitor insects (six bee species). We measured the operative strength required to open the staminodes and the strength that insects were capable of exerting using a biological tension sensor. Furthermore, we compared the strength required to open staminodes at different phases of the flowering period (male and female phases) and the strength of different visitors (visitors and non-visitors of D. caeruleum). The results showed that the strength needed to open staminodes in the male phase was significantly higher than that in the female phase. There was no significant difference between the strength exerted by visitors and required by staminodes of D. caeruleum in the male phase, but the visitor strength was significantly higher than that required to open staminodes in the female phase flowers. The strength of non-visitors was significantly lower than that required to open staminodes in the male phase. Furthermore, there was a significant positive association between the strength and the body weight of the bees. These results highlighted the observation that only strong visitors could press the two staminodes to access the sex organs and achieve successful pollination. Furthermore, these results revealed the function of pollinator screening by the staminodes of D. caeruleum. The biomechanical approach to the study of flowers allowed us to address relevant ecological and evolutionary questions of the plant-pollinator interaction and explore the functional modules within the flower structure in other plant species.
RESUMO
Plants living at the edge of their range boundary tend to suffer an overall decline in their fitness, including growth and reproduction. However, the reproductive performance of plants in artificially expanded habitats is rarely investigated, although this type of study would provide a better understanding of range limitations and improved conservation of ex situ plants. In the current study, we transplanted a narrowly dispersed species of Gentiana officinalis H. Smith (Gentianaceae) from its natural area of distribution to two different elevations and natural elevation to comprehensively study its pollination biology, including flowering phenology and duration, floral display, reproductive allocation, pollinator activity, and seed production. The findings indicated that the starting point and endpoint of the flowering phenology of G. officinalis were earlier at the low elevation, but the peak flowering periods did not differ significantly between any of the experimental plots. When transplanted, the flowering duration, especially the female phase, was reduced; the floral display, including spray numbers, flower numbers, and flower size (length and width), decreased, especially at high elevations. Moreover, the pollen numbers and pollen-ovule ratio were decreased at both high and low elevations, although the ovule numbers showed no change, and aboveground reproductive allocation was decreased. Furthermore, pollinator richness and activity were significantly decreased, and the seed-set ratio decreased under both natural conditions and with supplemental pollination. Finally, more severe pollen limitation was found in transplanted individuals. These results indicated an overall decrease in reproductive fitness in plants living outside their original area of distribution when the geographical range of G. officinalis was expanded.