The consequences of partial self-incompatibility on the breeding system of Campanula rapunculoides is considered under conditions of natural pollination. We propagated two partially self-incompatible genotypes of C. rapunculoides and established them in pollinator exclusion cages along with 18 donor plants that differed from the focal plants (recipients) in their genotype at the glucose-1-phosphate isomerase (GPI) locus. Cages of plants were opened for pollinator access for 1 h/d, 24 h every fourth day, or left continuously open (control treatment), and the experiment was carried out in two consecutive years (1997 and 1998). After flowering, recipient plants were scored for percentage fruit set and mean seed number per fruit, and then seeds were screened for paternity (self or outcross). ANOVA analyses showed that the mean number of seeds per flower was significantly greater in 1998 than in 1997 and that it was affected by the amount of pollinator access. ANOVA analyses further show that selfing rates were higher in the treatment in which pollinator access was open 24 h every fourth day, while the open-1-h/d treatment did not differ from control cages in any aspect of the breeding system. Selfing rates were higher overall in 1998 than in 1997. We also find a weak effect of the genotype of the recipient and the position of the flower on the selfing rate, although the first result is not unexpected because the genotypes used in the experiment showed similar levels of self-fertility. Our data show that both between- and within-years selfing rates can respond to environmental variation in pollinator availability in C. rapunculoides.
All Science Journal Classification (ASJC) codes
- Ecology, Evolution, Behavior and Systematics
- Plant Science