Distyly is a floral polymorphism involving reciprocal herkogamy shaped by selection for pollen transfer efficiency. The variation of the floral organs involved in pollen transfer can be individually affected by environmental and genetic sources of variance, but the organ development will be canalized to minimize reciprocal inaccuracy between anthers and stigmas, as this is the focus of selection. We measured floral organ and cell length of both morphs of distylous Linum tenue (Linaceae) at different developmental stages of field- and glasshouse-grown plants. We analyzed the results to measure reciprocal inaccuracy and identify sources of variance.

To quantify the contribution of different floral traits and other conditions to within and between morph variation, measurements of floral organ length and herkogamy of field and glasshouse open flowers were made from field-collected flowers stored in 70% ethanol or one to three freshly opened flowers per glasshouse grown individual. Flowers were dissected whorl by whorl under a dissecting microscope, and the vertical length of the sepals and petals, petal width, stamens and their component filaments and anthers, pistils and their component styles, stigmas, and ovaries were measured from the base of the ovary, using a combination of vernier callipers and analysis of photos using ImageJ.

The digital photographs of dissected flowers were made against a 1-mm-ruled graph paper background using a Leica M80 light microscope set at 7.5 magnification connected to a computer. Multiple investigators contributed different measures at different times, so to account for observer differences in the glasshouse data, a constant was added to the measures of each morph type for each observer in order to have the same mean as the samples measured using ImageJ, which were considered to be the most accurate measures.

To examine the expression of distyly at a cellular level, whole filaments and styles were separated from freshly harvested newly open flowers from glasshouse-grown individuals. Floral organs were mounted on microscope slides, stained with 0.05 % toluidine blue solution, and viewed at 100x to 400 times magnification using a differential interference contrast Leica DMI2500 microscope with an eyepiece graticule and photographed with a Panasonic GP-US932HAE camera. To account for localised differences in cell length at different positions along the organs, each style and stamen filament was classified into five approximately equal-length regions, R1 to R5, counting from base to tip. Using ImageJ software, up to 20 clearly visible cells were chosen from each image and measured along their longest axis to the nearest 0.1 μm. 

Files are in comma separated values format that can be opened by any text editor.