Seed Bank Type its Size and Seasonal Dynamics

What is clear from the above is that no-one has previously conducted a serious study of seed bank dynamics or controlled burial experiments with H. mantegazzianum seeds. To clarify the behaviour of this species in terms of seed bank type and its dynamics, several studies were carried out within the framework of the GIANT ALIEN project. The amount of seeds in the seed bank, its vertical distribution and seasonal changes were studied by Krinke et al. (2005) at seven sites in the Slavkovsky les region, Czech Republic. The study spanned two growing periods with censuses made in autumn after most seeds were released, in spring before seed germination, and in summer after spring germination and before seed release in the following autumn. The seeds were classified into three categories (dormant, non-dormant and dead) after collection (for explanation see Fig. 5.4).

The total number of seeds significantly increased with mean density of flowering plants at a site. The numbers of living, dead and total seeds were high in autumn, remained at the same level until spring, but from spring to summer, they all decreased. The number and proportion of dormant seeds was significantly higher in autumn than in spring or summer, and the number of non-dormant seeds was highest in spring (Fig. 5.4A). Proportions of dormant, non-dormant and dead seeds exhibited considerable seasonal dynamics. The percentage of dead seeds consistently increased from autumn to the following summer (Fig. 5.4A). The percentage of living seeds in the total seed bank decreased during winter from 55.9% in the autumn sample to 41.7% in spring to 14.8% in summer (Fig. 5.4B). The percentage of non-dormant seeds among living seeds was 0.3% in autumn, over winter it increased to 87.5% in the spring sample, and decreased to 3.0% in summer (Fig. 5.4B). After massive fruit release in autumn, nearly all living seeds (99.7%) were dormant. As almost no non-dormant seeds were found in autumn, this supports the observations that germination and population recruitment from seedlings in this species occur exclusively in spring (Krinke et al., 2005; see Pergl et al., Chapter 6, this volume).

Of the total variation in seed bank size, about four-fifths was attributed to variation among sites, and one-fifth to that within sites. Expressed per m2, the average value pooled across localities was 6719 ± 4119 (mean ± sd) in autumn, 4907 ± 2278 in spring and 1301 ± 1036 in summer for the total number of seeds, and 3759 ± 2906, 2044 ± 1198 and 192 ± 165, respectively, for living seeds (Krinke et al., 2005 and their Table 5). These data (Krinke et al., 2005) represent the first quantitative estimate of a seed bank in H. mantegazzianum, because numbers reported previously were based on

Fig. 5.4. (A) Changes in representation of dormant, non-dormant and dead seeds in the seed bank of H. mantegazzianum from autumn (after seed release) through spring (before germination) to the following summer (before new seeds are shed). Mean values shown are pooled across nine localities in the Slavkovsky les region, Czech Republic. (B) The proportion of non-dormant among living seed is close to zero in autumn, reaches a peak in spring after dormancy has been broken by cold and wet stratification over winter, and decreases to very low values in summer, after the vast majority of non-dormant seed germinated in spring. The proportion of living seed among the total number of seed steadily decreases in the course of the 'seed-cycle year', as part of the seed population gradually decays. Based on data from Krinke et al. (2005). Seeds germinated up to 1 month were considered as non-dormant, non-germinated living seeds were considered as dormant and decayed seeds found in the soil sample or seeds found dead after germination were considered as dead.

Fig. 5.4. (A) Changes in representation of dormant, non-dormant and dead seeds in the seed bank of H. mantegazzianum from autumn (after seed release) through spring (before germination) to the following summer (before new seeds are shed). Mean values shown are pooled across nine localities in the Slavkovsky les region, Czech Republic. (B) The proportion of non-dormant among living seed is close to zero in autumn, reaches a peak in spring after dormancy has been broken by cold and wet stratification over winter, and decreases to very low values in summer, after the vast majority of non-dormant seed germinated in spring. The proportion of living seed among the total number of seed steadily decreases in the course of the 'seed-cycle year', as part of the seed population gradually decays. Based on data from Krinke et al. (2005). Seeds germinated up to 1 month were considered as non-dormant, non-germinated living seeds were considered as dormant and decayed seeds found in the soil sample or seeds found dead after germination were considered as dead.

estimates from seedlings germinating in the field (Andersen and Calov, 1996) or on multi-species seed bank studies (Thompson et al., 1997). Such numbers of seeds found per m2 of the soil exceed the average value in the family Apiaceae by an order. Only two species of Apiaceae (Ammi majus L. and Torilis japónica (Houtt.) DC.) exhibit seed density values comparable with

H. mantegazzianum (Thompson et al., 1997). The reproductive potential of H. mantegazzianum is enormous and seems to be a crucial feature making invasion possible to the extent observed in the region (Krinke et al., 2005) and elsewhere in Europe (Ochsmann, 1996; Tiley et al., 1996).

Quantitative data of Krinke et al. (2005) allow some extrapolations to the landscape level. From knowledge of H. mantegazzianum population size in the largest study site (99,000 m2) and the average number of non-dormant seeds present in the spring, it can be calculated that each year in spring there are 386 million seeds, ready to germinate, in a single site.

So which seed bank type best fits what we now know about H. mantegazzianum? The species was considered to have a transient soil seed bank, i.e. missing from the seed bank or present only in the surface layer (Thompson et al., 1997). However, Krinke et al. (2005) classified the soil seed bank of H. mantegazzianum as a short-term persistent soil seed bank sensu Thompson et al. (1997) as in their samples 95% of seeds were concentrated in the upper soil layer and some living seeds were also present in lower soil layers. Moreover, the data reported in the following sections clearly indicate that seeds of H. mantegazzianum do persist in the soil for some years and that a short-term persistent seed bank is the case here.

Was this article helpful?

0 0

Post a comment