Premise of research. Self-compatibility is a requirement for reproductive assurance via selfing and may therefore be beneficial in environments with infrequent or unpredictable pollinator service. However, self-compatible plants may spatially separate anthers from stigmatic surface within flowers (herkogamy), potentially preventing autonomous self-pollination while promoting outcrossing. We investigated the effects of herkogamy on patterns of mating in a self-compatible, homostylous alpine herb, Primula halleri, which evolved from obligately outcrossing, heterostylous ancestors. Primula halleri displays diminishing herkogamy during anthesis, affecting its capacity for self-pollination. Specifically, we tested whether higher herkogamy at the end of anthesis (terminal herkogamy) promotes higher outcrossing rates at the cost of lower seed set, while lower terminal herkogamy ensures seed set via self-pollination at the cost of inbreeding.
Methodology. We estimated family- and population-level genetic outcrossing rates in four populations using microsatellites and derived an estimate for inbreeding depression. We tested effects of different levels of terminal herkogamy on means and variances of outcrossing rates. We also assessed whether seed set was resource or pollen limited. We interpreted results in light of previously published data on seed set with and without pollinator-exclusion treatment.
Pivotal results. Population-level outcrossing rates ranged between 0.5 and 0.8. Outcrossing rates of plants with higher terminal herkogamy had similar means but lower variances than those of plants with lower terminal herkogamy. This result demonstrates that separation between sexual organs larger than 1 mm in mature flowers favors outbreeding, while below this threshold, delayed selfing ensures reproduction.
Conclusions. Contrary to long-held notions about the association of homostyly and selfing, we provide new genetic evidence that the homostylous P. halleri adopts a mixed mating system due to variable herkogamy. Moreover, our results support a trade-off between seed number and seed quality effected through variation in herkogamy, illustrating that herkogamy affects multiple aspects of reproductive fitness simultaneously.