Main content area

Effects of genotype and stigma development stage on seed set following intra- and inter-specific hybridization of Dianthus spp.

Fu, XiaoPeng, Zhang, JingJing, Li, Feng, Zhan, PinTing, Bao, ManZhu
Scientia horticulturae 2011 v.128 no.4 pp. 490-498
Dianthus barbatus, Dianthus caryophyllus, Dianthus chinensis, aniline, anthers, cross pollination, cultivars, dehiscence, developmental stages, field experimentation, flowering, genotype, hybridization, ornamental plants, pollen, pollen tubes, seed set, seeds, selfing, stigma, wilting
The caryophyllaceae is a family of flowering plants commonly known as the pink or carnation family. Members of this family are widely grown as ornamental plants. The species Dianthus chinensis, Dianthus barbatus, and Dianthus superbus include notable examples of such ornamental varieties. Intra- and inter-specific hybridization is important for the breeding of improved varieties and, therefore, it is desirable to optimize the efficiency of cross-pollination procedures within this family. Here, we conducted intra-specific pollination using four lines of D. chinensis, and inter-specific pollination between D. chinensis, D. barbatus and D. superbus genotypes. For each of these crosses we recorded information of pollen viability, stigma development stage at time of pollination, pollen germination on the stigma surface and pollen tube elongation within the stigma, and final seed set. Pollen viability levels were determined at the start of anther dehiscence, which occurred on day 1 of flower opening. The genotypes D. chinensis var. ‘Bz2’, D. barbatus var. ‘Xb’, and D. superbus wild species ‘Qm’ all displayed good levels of pollen viability (i.e. between 68% and 75%); D. chinensis vars. ‘H58III’, ‘H68I’ and ‘H6II’ showed lower levels of pollen viability (i.e. between 32% and 51%). In each Dianthus line, the stigma organs were observed to undergo morphological changes during the 1–7d following flower opening. We have characterized these changes as five distinct development stages, namely, ‘column’ (day 1), ‘Y-like’ (day 2), ‘Y-shaped with slight curve’ (days 3–4), ‘Y-shaped horn-like’ (days 5–6), and ‘wilting’ (day 7). Pollen was artificially applied to stigma organs at each of these development stages and aniline blue staining was used to follow pollen tube germination on the stigma surface and pollen tube elongation down the length of the stigma. This showed that the ‘Y-shaped slightly curved’ stigma stage (3–4d) was optimal for pollination. The ‘Y-shaped horn-like’ stage (5–6d) was also receptive to pollination. By contrast, the earlier ‘Y-shaped’ (2d) and ‘column’ (1d) stages, and also the later ‘wilting’ (7d) stage, were far less receptive. Parallel field experiments were conducted in which seed set was measured following the artificial pollination of flowers at different development stages. These data also indicated that the optimal time for pollination was during 3–4d of flower opening. This conclusion held true for both intra- and inter-specific crosses of the various Dianthus lines. Thus, stigma receptivity in Dianthus does not become optimal until 3–4d after flower opening. Since anther dehiscence occurs at 1d (i.e. when stigma are at the unreceptive ‘column’ stage), this suggests that Dianthus employs a delayed self-fertilization strategy. We also found a significant effect of genotype on pollen tube growth and seed-set values. The highest numbers of seed per flower were achieved when the D. superbus wild species ‘Qm’ was the female parent in inter-specific crosses. Thus, genotype and stigma development are identified as key factors that determine the success of cross-pollination in Dianthus lines.