7 and 5.4%. In addition, there was no significant difference in viabilities of pollen grains collected at 05:00-06:00hrs between ��Jinsenianhua�� and ��Qinhuaihuadeng��. Since pollen grains collected at 5:00-6:00hrs had the highest viability for both cultivars, we therefore carried out artificial fairly pollination experiments using pollen grains collected in this time span.Figure 1Viabilities of pollen grains collected at different times.3.2. Pollen Germination on Stigmas after Artificial PollinationIn both crosses, some pollen grains germinated within 0.5h of artificial pollination. After that, more pollen grains germinated on stigmas and the number reached a peak at 4h after pollination (Table 1; Figures 2(A), 2(B)): there were on average 63.0 and 17.

2 germinated pollen grains per stigma at 4h after pollination in the crosses ��Jinsenianhua�� �� ��Qinhuaihuadeng�� and ��Qinhuaihuadeng�� �� ��Jinsenianhua��, respectively. Then, at 12h, the corresponding numbers of germinated pollen grains on each stigma gradually decreased to 13.8 and 7.4. Although the change pattern of the number of germinated pollen grains on stigmas was very similar for the two crosses, the number in the cross ��Jinsenianhua�� �� ��Qinhuaihuadeng�� was significantly higher than that of ��Qinhuaihuadeng�� �� ��Jinsenianhua�� at each time point after pollination (Table 1). Moreover, there were many pollen tubes with abnormalities such as branching, splitting, coiling, and convolution on stigmas in the ��Jinsenianhua�� �� ��Qinhuaihuadeng�� cross, which resulted in the failure of these pollen tubes penetrating the stigma surface (Figures 2(C)�C2(E)).

In contrast, most pollen tubes grew normally and entered the stigma surface for ��Jinsenianhua�� �� ��Qinhuaihuadeng�� (Figure 2(F)).Figure 2Pollen grains and pollen tubes on stigmas at 4h after pollination and anatomical structure of young embryos. (A): Many pollen grains germinated on stigmas in ��Jinsenianhua�� �� ��Qinhuaihuadeng��. (B): A few …Table 1Number of pollen grains germinating on stigma after pollination.3.3. Percentage of Normal EmbryosZygotes divided very quickly in the two crosses. For example, at 1d after artificial pollination, globular embryos were observed in many ovules, and heart embryos were observed in some ovules at 2d after artificial pollination (Figures 2(G), 2(H)). At 4d after pollination, most embryos had reached the cotyledon embryo stage.

Normal globular embryos were observed in 55.0% of ovaries at 1d after pollination, and normal heart embryos were observed in 37.5% of ovaries at 2d after pollination in the ��Jinsenianhua�� �� ��Qinhuaihuadeng�� cross (Table 2). However, the corresponding values for the ��Qinhuaihuadeng�� �� ��Jinsenianhua�� GSK-3 cross were 21.9 and 14.7%, respectively. As the embryos continued to develop, increasing numbers of embryos degenerated in the two crosses (Table 2; Figures 3(A)�C3(F)). For instance, there were only 20.