Estimating pollen fertility in Solanum species and haploids

Summary

Three staining methods (acetocarmine, fuchsin and oxidation of benzidine) and germination in vitro and in vivo were applied to estimate pollen fertility in Solanum species and dihaploids. Pollen was divided into six classes based on shape and contents of the grains. With acetocarmine, fuchsin, peroxidase and germination in vitro 4, 3, 2 and 1 classes, respectively, are supposed to be included in the percentage of ‘good’ pollen as measured by these methods. This percentage therefore, in more than 96% of the cases studied, shows a decrease in the order indicated. Neither aging of pollen at room conditions nor collecting pollen from flowers on 1–9 days after anthesis does influence the percentage of ‘good’ pollen with acetocarmine and fuchsin, whereas this percentage drops sharply to zero with peroxidase and germination in vitro. The latter two methods apparently measure as ‘good’ pollen only the grains with living cytoplasm. When pollen is collected at three successive dates from the same flowers the percentage of ‘good’ pollen drops sharply with all methods used. There is a relation between quantity of pollen per flower and pollen quality (% ‘good’), low-quantity pollen containing significantly lower percentages of ‘good’ pollen than medium- and high-quantity pollen. The latter two are not significantly different in this respect.

From calculations of correlation coefficients it is concluded that only germination of pollen in vitro is significantly correlated with berry and seed set and thus gives a reliable estimate of male fertility. This does not hold true for the two staining methods without due reserve. The peroxidase method is not useful for the Solanum material studied.

After standardized pollination the average number of haploid pollen grains on diploid stigmata was found to be 1625±127, that of diploid pollen on tetraploid stigmata 2863±98.