Effect of hydration and dehydration cycles on seed germination of Aster kantoensis (Compositae)

Seeds of Aster kantoensis Kitamura (Compositae) were experimentally exposed to different cycles of hydration and dehydration: 3H1D (cycles of 3-d hydration and 1-d dehydration periods), 2H3D, 2H1D, 1H3D, 1H2D, and 1H1D. Under continuous hydration (control), all viable seeds germinated within 9 d. However, all viable seeds exposed to the 3H1D, 2H3D, and 2H1D cycles germinated within 36, 50, and 36 d of the start of the experiment, respectively. Not all viable seeds exposed to the 1H3D, 1H2D, and 1H1D cycles germinated during the experimental period. Compared with the control, the number of days required for 50% germination increased in seeds exposed to the hydration–dehydration treatments except for those seeds exposed to the 3H1D and 2H1D cycles. In addition, seeds treated with a 1-d hydration period required a larger number of cumulative hydration days for 50% germination than those seeds exposed to one of the other three treatments or the control. These results indicate that sporadic germination of A. kantoensis seeds during an extended period of several months in the gravelly flood plain where they are found results because the dispersed seeds are exposed to various cycles of hydration and dehydration depending upon the rainfall pattern and spatial heterogeneity of evaporation rates at the sand surface resulting from shading by gravel.