When differentiated lineages come into contact, their fates depend on demographic and reproductive factors. These factors have been well-studied in taxa of the same ploidy, but less is known about sympatric lineages that differ in ploidy, particularly with respect to demographic factors. We assessed prezygotic, postzygotic, and total reproductive isolation in naturally-pollinated arrays of diploid-tetraploid and tetraploid-hexaploid population mixes of Campanula rotundifolia by measuring pollinator transitions, seed yield, germination rate, and proportion of hybrid offspring. Four frequencies of each cytotype were tested, and pollinators consistently overvisited rare cytotypes. Seed yield and F1 hybrid production were greater in 4X-6X arrays than 2X-4X arrays, while germination rates were similar, creating two distinct patterns of reproductive isolation. In 2X-4X arrays, postzygotic isolation was near-complete (3% hybrid offspring), and prezygotic isolation associated with pollinator preference is expected to facilitate the persistence of minority cytotypes. However, in 4X-6X arrays where postzygotic isolation permitted hybrid formation (44% hybrids), pollinator behavior drove patterns of reproductive isolation, with rare cytotypes being more isolated and greater gene flow expected from rare into common cytotypes. In polyploid complexes, both the specific cytotypes in contact and local cytotype frequency, likely reflecting spatial demography, will influence likelihood of gene exchange.

Pollinator visitation observations were made visually and were collected by hand on paper by T. Miranda-Katz and B. Sutherland. Data were then transcribed into a spreadsheet by B. Sutherland. Visitations in which only one plant were visited were then removed from the dataset to be analyzed. Both raw bout information and processed data for each bout (ratios of intra- and intercytotype visitations) are reported.

Fruit set, Seed set, and Germination data were assessed visually and transcribed into a spreadsheet.

F1 cytometric data was collected using a BD FACSCalibur flow cytometer. Modal value for fluorescence for each sample was recorded and compared to either an internal or external standard. Modal value/standard ratios were then compared to the same ratio for known diploid plants to determine ploidy. F1 ploidy, maternal ploidy, ploidy pair, and maternal frequency are recorded.

Missing values are recorded as "."

Tabs containing encoded information (i.e. weather type, insect type, population, etc) is included in each Excel file as relevant.