Data from: Both morph- and species-dependent asymmetries affect reproductive barriers between heterostylous species

The interaction between floral traits and reproductive isolation is crucial to explaining the extraordinary diversity of angiosperms. Heterostyly, a complex floral polymorphism that optimizes outcrossing, evolved repeatedly and has been shown to accelerate diversification in primroses, yet its potential influence on isolating mechanisms remains unexplored. Furthermore, the relative contribution of pre- versus postmating barriers to reproductive isolation is still debated. No experimental study has yet evaluated the possible effects of heterostyly on pre- and postmating reproductive mechanisms. We quantify multiple reproductive barriers between the heterostylous Primula elatior (oxlip) and P. vulgaris (primrose), which readily hybridize when co-occurring, and test whether traits of heterostyly contribute to reproductive barriers in unique ways. We find that premating isolation is key for both species, while postmating isolation is considerable only for P. vulgaris; ecogeographic isolation is crucial for both species, while phenological, seed developmental, and hybrid sterility barriers are also important in P. vulgaris, implicating sympatrically higher gene flow into P. elatior. We document for the first time that, in addition to the aforementioned species-dependent asymmetries, morph-dependent asymmetries affect reproductive barriers between heterostylous species. Indeed, the interspecific decrease of reciprocity between high sexual organs of complementary floral morphs limits interspecific pollen transfer from anthers of short-styled flowers to stigmas of long-styled flowers, while higher reciprocity between low sexual organs favors introgression over isolation from anthers of long-styled flowers to stigmas of short-styled flowers. Finally, intramorph incompatibility persists across species boundaries, but is weakened in long-styled flowers of P. elatior, opening a possible backdoor to gene flow through intramorph pollen transfer between species. Therefore, patterns of gene flow across species boundaries are likely affected by floral morph composition of adjacent populations. To summarize, our study highlights the general importance of premating isolation and newly illustrates that both morph- and species-dependent asymmetries shape boundaries between heterostylous species.Raw data from occurrence records and experiments performed to quantify strength and species- and morph-dependent asymmetries of reproductive isolation between Primula elatior and P. vulgaris.The excel file contains raw data of (i) number of 1x1 km grid cells from occurrence data of Switzerland that was used to estimate ecogeographic isolation (RIecogeo); (ii) number of pollen grains, diameter of pollen grains, anther and stigma positions, and anther-stigma distance from the pollen-transfer experiment that was used to estimate mechanical isolation (RImech); (iii) number of seeds, filled seeds, seedlings, and flowers, and proportion of fruits and survivors from the hand-pollination experiment that was used to estimate F1 seedling formation (RIseedling) and F1 survivorship (RIsurvivorship); (iv) number of plants and open flowers per census day, date of onset, peak, and end of flowering time, number of flowers, number of seeds and filled seeds from the common garden experiment that was used to estimate phenological isolation (RIphenoP), F1 phenology (RIphenoF1), F1 flower production (RIflower), and F1 seed set (RIseed set); and (v) number of seeds, filled seeds, and seedlings from the hand-pollination backcross experiment that was used to estimate F1 male sterility (RImale). Data for occurrence records and each experiment are presented in an individual worksheet. Names of column headings correspond with terms used in the original paper (Table 1).ECE-2016-04-00333.xls