Phylogeny does not predict the outcome of heterospecific pollen-pistil interactions in a species-rich alpine plant community

Premise: Co-occurring plant species that share generalist pollinators often exchange pollen. This heterospecific pollen transfer (HPT) impacts male and female reproductive success through pollen loss and reductions in seed set, respectively. The resulting fitness cost of HPT imposes selection on reproductive traits (e.g. floral color and shape), yet we currently lack strong predictors for the post-pollination fate of heterospecific pollen, especially within community and phylogenetic contexts. 

Methods:  We investigated the fate of heterospecific pollen at three distinct stages of plant reproduction: 1) pollen germination on the stigma, 2) pollen tube growth in the style, and 3) fertilization of ovules. We experimentally crossed eleven naturally co-flowering species in the sub-alpine meadows of the Colorado Rocky Mountains, across a spectrum of phylogenetic relatedness. Using generalized linear mixed models and generalized linear models, we evaluated the effect of parental species identity and phylogenetic relatedness on pollen tube growth at each reproductive stage. 

Key Results: We found that heterospecific pollen tubes can germinate and grow within pistils at each reproductive stage, even when parental species are >100My divergent. There was no significant effect of phylogenetic distance on heterospecific pollen success and no evidence for a mechanism that suspends heterospecific pollen germination or pollen tube growth within heterospecific stigmas or styles.

Conclusions: Our results show that even in communities where HPT is common, pre-zygotic post-pollination mechanisms do not provide strong barriers to interspecific fertilization. HPT can result in the loss of ovules even between highly diverged plant species.

Methods
This is the raw data from the experimental crossing study associated with Cohen et al. 2024 published in the American Journal of Botany. One of the experiments was a semi-in vivo crossing experiment, where we observed pollen tubes growing through sections of styles in petri dishes. We did an 11x11 cross, using 5 North American Pedicularis species and close- and distant-relatives that grow in the alpine plant communities they are a part of. We additionally did hand-pollinations in the field, then observed pollen tubes growing in styles by staining the styles with aniline blue. Finally, we measured the change in size of ovaries following the field-based hand pollinations.

In our study, we used generalized linear models and generalized linear mixed models to assess whether or not phylogenetic distance was a significant predictor of the outcome of heterospecific pollination at each of the three stages described above. For the semi-in vivo study we used the raw data to calculate a proportion of trials in which pollen tubes were observed growing through styles. In the field-based study, we used an ordinal regression to assess the impact of phylogenetic distance on how far down the length of the style pollen tubes were observed. Finally, we normalized ovary area using the average ovary area from our control treatment (no pollen), and used a generalized linear model (gaussian distributed) to investigate the relationship between change in ovary size and phylogentic distance between crossing parents.