| ||||||||||||||||||||
Falling flowers foil self-fertilization
As in the previous commentary, we focus here on late-acting self-incompatibility (SI) mechanisms. The species investigated by Bittencourt et al. (Sao Paolo, Brazil and St Andrews, Scotland, pp. 827-834) is Spathodea campanulata, a tree originating in tropical west Africa. The effectiveness of its SI mechanism is very clear. Self-pollinated flowers abscind within 4 d and no fruit is set. So, where does SI operate? There is little or no difference in pollen tube growth rates between self and non-self pollen but there is a clear difference in the speed of ovule penetration, which takes up to 24 h longer with self than with non-self pollen. Nevertheless, once ovule penetration has occurred, self pollen achieves a normal double fertilization giving rise to an endosperm cell and a zygote. At this stage further differences become apparent. Although an endosperm is effectively initiated in a self-fertilization, subsequent cell divisions occur only rarely, so that by 72 h the number of two-cell endosperms in selfed ovules is little more than half that in out-crossed ovules, and there is no evidence at all of subsequent cell divisions in the selfed ovules. This difference arises because of the lateness of ovule penetration by the self pollen; there are no indications of endosperm malformation. A similar situation exists with the embryo. Again, because of slow penetration of the ovule, formation of a zygote with self pollen occurs around 24 h later than with non-self pollen and there is no post-zygotic embryonic development following self-fertilization. There is no obvious embryo malfunction; embryogenesis is simply terminated by floral abscission. The latter is thus a major controlling factor in SI in S. campanulata and the authors make the interesting suggestion that it is the slow penetration of the ovule by self pollen (presumably dependent on a recognition event) that initiates the floral abscission pathway.
j.a.bryant{at}exeter.ac.uk
| ||||||||||||||||||||