In vitro germination of upland cotton pollen : an analysis of parameters affecting germination and tube length

An improved culture medium was developed for the in vitro germination and tube growth of upland cotton (Gossypium hirsutum L. cv. Hancock) pollen. The pollen grains were dusted onto a 3.5% agar medium and then separated with a dissecting needle. The parameters tested to determine their effects on the percent germination and tube length of cotton pollen were pH, carbohydrates, magnesium, manganese, calcium, potassium, boron, gibberellic acid, indole-3-acetic acid, and succinic acid-2,2-dimethylhydrazide.

All of the data were analyzed by a Duncan's Multiple Range test to show significant differences among the treatments. The germination of cotton pollen was more sensitive to the variability of the greenhouse conditions and the in vitro culture conditions than was the pollen tube growth.

A high pH (7.6) significantly increased the percent germination and tube length of cotton pollen in vitro. Magnesium sulfate and gibberellic acid both significantly increased the mean pollen tube length; however, they had no significant effect on the percent germination. Calcium nitrate at all concentrations tested inhibited both germination and tube growth. Manganous sulfate and boric acid were vital for the growth of the cotton pollen tube in I vitro. Potassium nitrate and indole-3-acetic acid did not significantly affect either germination or tube growth, but a positive trend in the percent germination and tube length was associated with their presence in the medium.

The final medium improved by experimentation contained 100 ml distilled water, 25 g sucrose, 5.9 mM manganous sulfate, l.0 mM potassium nitrate, l.0 uM gibberellic acid, 0.l indole- 3-acetic acid, adjusted pH of 7.6, and 3.5 g agar. The mean percent germination and mean tube length of cotton pollen grown on the final medium were 47% and 2.6 mm respectively. Cotton pollen grown on the final culture medium had an increased percent germination and tube length of 74% and 68% respectively over the best previously published medium (Taylor, 1972).