Heracleum mantegazzianum is one of the most invasive species in the Czech flora. The present study describes its flowering phenology and assess the effectiveness of protandry in preventing selfing in this self-compatible species, describes the timing of flowering in a heavily invaded area of Slavkovský les (Czech Republic) and estimates fruit set in a large sample of plants, which provides reliable data on the often exaggerated fecundity of this species. The study of flowering phenology revealed that protandry is always effective only within individual flowers, where male and female flowering phases are completely separated. In contrast, anther dehiscence in some flowers can occasionally overlap with stigma receptivity in other flowers in the same umbel, providing an opportunity for geitonogamous (i.e. between-flower) selfing. Nevertheless, the potential for selfing in H. mantegazzianum is determined mainly by an overlap in the male and female flowering phases between umbels on the same plant; at least a short overlap between some umbels was observed in 99% of the plants at the Slavkovský les. Although the degree of protandry in H. mantegazzianum favours outcrossing, the opportunity to self may be of crucial importance for an invasive plant, especially if a single plant colonizes a new location. At Slavkovský les, flowering started within one week (from 20 to 27 June 2002) at all 10 sites. The duration of flowering of an individual plantwas on average 36 days,with maximum of 60 days, and increased significantly with the number of umbels on a plant. In the second half of August, the majority of the fruits were ripe and had started to be shed. The beginning of flowering of a plant was significantly negatively correlated with the number of umbels it had – the earlier a plant started to flower the more umbels it had produced. A significant negative relationship was also found between basal diameter and beginning of flowering; plants with large basal diameters started to flower earlier. An average plant at Slavkovský les produced 20,671 fruits. Of these, 44.6% were produced by the terminal umbel, 29.3% by secondary umbels on satellites, 22.6% by secondary umbels on branches and only 3.5% by tertiary umbels. The estimated fruit number of the most fecund plant was 46,470 – compared to an average plant, the proportional contribution of tertiary umbels increased relative to the primary umbel. This study revealed a significant positive relationship between fecundity and plant basal diameter. Although the results of this study indicate that the fecundity of this species is often overestimated in the literature, the number of fruits produced by H. mantegazzianum provides this invasive species with an enormous reproductive capacity.
We investigated the effects of different temperature regimes and dry storage on germination of H. mantegazzianum (Apiaceae, native to Caucasus) seeds in the laboratory and linked the results with studies of seasonal seed bank depletion in a common garden experiment and under field conditions. Seeds were collected at seven sites in the Slavkovský les region, Czech Republic, cold-stratified for 2 months and germinated at seven temperature regimes. Under all temperature regimes, fresh seeds germinated to significantly higher percentages than older (1, 2, 3 years) seeds. For all storage lengths, seeds germinated best at alternating day/night temperatures of 20/5 °C. The length of the germination period had a significant effect only at low constant temperatures of 2 and 6 °C, where germination percentage increased between 2 and 6 months. Seasonal germination exhibited a distinct pattern, with rapid depletion of seed bank by the first spring after seed burial. Non-dormant seeds were present in the soil early in spring and late in autumn. The higher summer temperatures prevented dormancy breaking and another cold period of at least two months below 10 °C was needed to bring non-germinated seeds out of dormancy. The results suggest that (1) seed dormancy of H. mantegazzianum was not completely broken until the first spring, but that some seeds re-enter or retain dormancy during high summer temperatures and that (2) the threshold needed for breaking the dormancy was achieved gradually during the cold autumn and winter months. However, in a small fraction of seeds the dormancy breaking process took several years. Of seeds buried in 10 different regions of the Czech Republic, on average 8.8% survived 1 year, 2.7% 2 years and 1.2% remained viable and dormant after 3 years of burial. The ability of even small fraction of H. mantegazzianum seeds to survive for at least 3 years can result in re-invasion of this species into controlled sites.