Sex Plant Reprod 4: 113-117 (1991)
Influence of season on rose pollen quality
S. Gudin, L. Arene, and C. Bulard


In vitro tests of pollen germination were carried out at different periods during an annual cycle in order to study environmental influence on the quality of Rosa hybrida L. pollen during its maturation process. This quality was evaluated by taking into account the rate of germination as well as the average length of emitted pollen tubes. In addition, during an annual hybridization period, a few pollinations were carried out in vivo with pollen of the same origin in order to study the evolution of the fertilization results, as attested by number of achenes per resulting hips. During the period covered by the experiments, the evolution of the pollen quality detected in vitro can be related to that of seed setting success. Of the two criteria used in vitro to evaluate pollen quality, the factor most liable to influence in vivo fertilization success seems to be the average length of emitted pollen tubes.


The influence of the environmental conditions present at the moment of pollen germination (temperature, relative humidity) on pollen fertilizing ability (evaluated in vitro and/or observed in vivo) has often been investigated. In contrast, only a few studies have been carried out on the influence of environmental conditions during in situ maturation of pollen on its germinability and fertilizing ability. Such studies have been carried out with pollen of Petunia hybrida (Van Herpen and Linskens 1981), Vaccinium spp. (Megalos and Ballington 1987), and Phaseolus vulgaris (Weaver and Timm 1988). Practically speaking, these studies are very important for seed production, as it is a well-accepted fact that pollen quality is a key factor in seed set (Sadras et al. 1985; Struik et al. 1986). With respect to roses, only the observation of Visser et al. (1977) is available, according to which pollen produced in July is less viable (as determined by a tetrazolium test) than pollen produced in May or June.

Seasonal influence on fertilization success has been studied frequently and for a long time in many fruit trees belonging to the Rosaceae family (Dorsey 1919; Marshall 1919; Howlett 1937; Thompson and Liu 1973; Visser et al. 1983). In roses, the seasonal impact of hybridization has been investigated only in relation to its influence on resulting germinability of the achenes (Von Abrams and Hand 1956); its possible relationship with fertilization success was not considered.

As far as in vitro studies are concerned, the pollen-tube length criterion was very seldom used before 1980. Nowadays it and the commonly used criterion of pollen germination percentage are frequently used in studies on the relationship between pollen quality tested in vitro and fertilization results obtained in vivo (Kuo et al. 1981; Asif et al. 1983; Heslop-Harrison et al. 1984; Vasilakakis and Porlingis 1985; Chichiricco and Grilli-Caiola 1986; Knox et al. 1986; Kerhoas and Dumas 1987).

In roses, the few authors who have used in vitro germination assays as a viability criterion have only considered germination rates (Visser et al. 1977; Pearson and Harney 1984). The use of only this criterion for pollen quality evaluation has been criticized by Knox et al. (1986) as methods that enable in situ pollen growth evaluation (see House and Nelson 1958; Kumar and Sarkar 1980) have clearly established that this last factor is better correlated to the achievement of gametic union; Dumas et al. (1988) consider it to be "the only significant measurement of pollen function".

The present work was carried out in order to evaluate the influence of the production season of rose pollen on its germination quality, tested in vitro, and fertilizing ability, observed in vivo.


The evolution of night and day temperature averages during the experimental year is given in Fig. 1.

Figure 2 shows that the fertilization issue (number of achenes per hip) fluctuated considerably during the hybridization period: when results obtained before the end of April are compared with those obtained later in the year, it is on average double after this date. The largest variation occurred between early April and early June; the latter period showing a five times higher number of achenes per hip.

With respect to the quality (tested in vitro) of the pollen produced at different observation periods on the quantitative aspect (number of pollen grains germinated for 50), pollen germination was maximum and statistically similar in early December and between the end of March and of September 1987 (Fig. 3a). The largest variation corresponds only to a two-times multiplication factor (comparison between January and September). Considering the period from the end of February until the end of June (hybridization period), one can see that following the end of April the quantity of germinated pollen grains is on average only 1.2 times that obtained before this date.

On the qualitative aspect, the length of emitted pollen tubes is maximum and statistically similar between the end of March and June and in September (Fig. 3 b). Between these periods, quality decreased in August, which corresponds to the higher temperature conditions of our experiment (see Fig. 1). During autumn and winter the quality was the lowest. The largest variation corresponds to a four-times multiplication factor (comparison between months of February and April). Considering the period from the end of February until the end of June, one can see that following the end of April the quantity of the germinated pollen grains was on average double that obtained before this date.