603

The flowers of the earliest-diverging clades of angiosperms have a large and variable number of tepals (undifferentiated sepals and petals), carpels, and stamens (Figure 28.11A). Evolutionary change within the angiosperms has included some striking modifications of this early condition: reductions in the number of each type of floral organ to a fixed number, differentiation of petals from sepals, and changes in symmetry from radial (as in a lily or magnolia) to bilateral (as in a sweet pea or orchid), often accompanied by an extensive fusion of parts (Figure 28.11B).

According to one hypothesis, the first carpels to evolve were leaves with marginal sporangia, folded but incompletely closed. Early in angiosperm evolution, carpels fused with one another, forming a single, multichambered ovary (Figure 28.12A). In some flowers, the other floral organs are attached at the top of the ovary rather than at the bottom as in Figure 28.4B. The stamens of the most ancient flowers may have been leaflike (Figure 28.12B), with little resemblance to the stamens of the generalized flower seen in Figure 28.4B.

Why do so many flowers have pistils with long styles and anthers with long filaments? Natural selection has favored length in both of these floral organs, probably because length increases the likelihood of successful pollination. Long filaments may bring the anthers into contact with insect bodies, or they may place the anthers in a better position to catch the wind. Similar arguments apply to long styles.

A perfect flower represents a compromise of sorts. On the one hand, by attracting a pollinating bird or insect, the plant is attending to both its female and male functions with a single flower type, whereas plants with imperfect flowers must create that attraction twice—once for each type of flower. On the other hand, the perfect flower can result in self-pollination, which is usually disadvantageous. Another potential problem is that the female and male functions might interfere with each other—for example, the stigma might be placed so as to make it difficult for pollinators to reach the anthers, thus reducing the export of pollen to other flowers.

Might there be a way around these problems? One solution is seen in the bush monkeyflower (Mimulus aurantiacus), which is pollinated by hummingbirds. Its flower has a stigma that initially serves as a screen, hiding the anthers (Figure 28.13). Once a hummingbird touches the stigma, however, one of the stigma’s two lobes is retracted, so that subsequent hummingbird visitors pick up pollen from the previously screened anthers. Thus the first bird to visit the flower transfers pollen from another plant to the stigma. Later visitors pick up pollen from the now-accessible anthers, fulfilling the flower’s male function. Figure 28.14 describes the experiment that revealed the function of this mechanism.