Secondary dormancy in Diplotaxis erucoides : a possible adaptative strategy as an annual weed

The germination of stored Diplotaxis erucoides seed was studied under controlled conditions of temperature and light by conducting germination tests over three years after collection. The D. erucoides seed was not dormant at harvest, but secondary dormancy appeared during storage as indicated by reduced germination 12 months after collection (from 92% down to 39%). This could be overcome by prolonged storage or by the use of gibberellic acid. Such shifts between dormancy and non-dormancy can be interpreted in terms of the soil seed bank dynamics in relation to intraspecific competition and/or with its adaptation as an annual weed in periodically disturbed soil in cultivated areas. Additional key words: germination, gibberellic acid, seeds, soil seed bank, storage time.

was to investigate if the germination behaviour of D. erucoides seeds changes over time, by studying the effect of storage duration and environmentally controlled factors of temperature and light on germination.Ripe seed of D. erucoides was collected, at random in October 1996 from a wild growing population, in the Botanical Garden of the Pannonian University of Agricultural Sciences in Mosonmagyaróvár (Hungary).Seed were cleaned and stored at room temperature (about 25°C) for 15 days and then stored in a cold chamber at 5°C in a hermetically sealed container until used for testing.Seed moisture content of seed at room temperature, determined by oven drying two replicates (of 150 seeds each) for 17 h at 103°C before storage at 5°C, was 7.38 ± 0.22% on the fresh weight basis.
Four replicates of 25 seeds each were placed in Petri dishes (7 cm in diameter) on two sheets of filter paper disks moistened with 3.5 ml of distilled water.More water was added, as necessary, during the germination tests.The Petri dishes were incubated at 10, 15, 20, and 25°C constant temperatures and 25/15°C alternating temperatures, under a 16/8 h light/dark regime and an irradiance of 35 µmol m -2 s -1 provided by cool white fluorescent tubes OSRAM L 58W/20.Germination tests were conducted every six months over three years, i.e. at 6, 12, 18, 24, 30 and 36 months after seed collection.
Seed stored for 12 months was soaked in different gibberellic acid (GA 3 ) solutions of increased concentrations (from 0 to 250 mg L -1 ), at room temperature (approx.25°C) for 24 h and then incubated at 25°C with a 16/8 h photoperiod.Seed stored for 12, 24 and 36 months were also soaked in a GA 3 solution at 250 mg L -1 ), or distilled water, under the same conditions and were also incubated at 25°C with 16/8 h light/dark photoperiod.
In all trials, seeds showing radicle emergence were counted every two days and removed from the Petri dishes.The final germination percentage was scored after a 42-day incubation period.For each trial, germination percentages were arcsine transformed and were subjected to analysis of variance (Sokal and Rohlf, 1995).Comparison of means was carried using the least significant difference test (LSD) at the 1% level of probability.
Figure 1 shows the final germination of D. erucoides seed incubated at 25°C measured at six monthly intervals starting immediately after collection over 36 months of storage.The highest germination of 92% was from the first assay with fresh seed.Later assays had lower values between 72% and 85%, except for the assay at 12 months after collection, at which seed germination was significantly lower at only 39% (P < -0.001).Incubation temperatures of 10, 15, 20 and 15/25°C were also tested after storage for 0, 12, 24 and 36 months (Fig. 2).At all assays mean germination was always lower than at 25°C, except for the assay after 12 months storage where germination obtained at 15/25°C was 40%, similar to that obtained at 25°C (39%, Fig. 1).Incubation at 10°C gave extremely low germinations (always lower than 5%).
To try to overcome the fall in germination after 12 months storage, seed was soaked in increasing concentrations of GA 3 (0-250 mg L -1 ) prior to incubation at 25°C (Fig. 3).The germination promoting effect of GA 3 was evident as the germination was always higher than that of the control (41%).Significantly higher  germinations (87-98%) were obtained with GA 3 concentrations ≥ 10 mg L -1 .
Complementary assays were performed on seeds stored for 12, 24 and 36 months, soaking them in water or in a standard GA 3 solution of 250 mg L -1 before incubation at different temperatures (Table 1).As expected from the reduction in germination shown in Figure 1, germination at a temperature of 25°C of seed soaked in water was significantly higher at 24 and 36 months (79% and 65% respectively) than germination after storage for 12 months (41%).Treatment with GA 3 increased germination to 94%.
The fresh seed of D. erucoides was not dormant at harvest as the initial germination at 25°C was 92% (Fig. 1).Germination at lower temperatures tended to be lower (Fig. 2), in accordance with the results of Pérez-García et al. (1995).
The reduction in germination observed 12 months after harvest (39%) suggests that secondary dormancy occurred during storage (Fig. 1).The promoting effect of GA 3 which increased germination from 41% (no GA 3 ) to 87-98% (Fig. 3) conf irmed that the fall in germination was due to seed dormancy.Pérez-García et al. (1995) also found the lowest germination of D. erucoides populations of different geographical origin after 13-16 months storage.
However, this dormancy was also reversed in storage, as indicated by the increased germination after 18 months or longer storage.The recovery of germination after 18 months storage, was not complete, since in no instance did germination reach the values obtained before the onset of dormancy (Fig. 1).This was probably due to a persistent, slight level of dormancy and/or to moderate loss of viability.Pérez-García et al. (1995)   also reported an increase in germination 10 months after an initial decrease.A decrease in seed dormancy level was also detected by Ellis et al. (1993) in other cruciferous seeds after prolonged storage at low temperature.
Laboratory data suggest that, in nature, most D. erucoides seeds might readily germinate shortly after being released.However, those which were not exposed to adequate environmental conditions would temporarily enter the soil seed bank.The build-up of a soil seed bank from which seeds can germinate gradually appears as a proposed adaptive strategy controlling the population demography of D. erucoides and other annual weeds growing in disturbed areas (Sans and Masalles, 1994).
At 12 months after dispersal, however, a fraction of the seed might show dormancy at approximately the same time that a new generation of seed is being released in the population.This strategy may be interpreted as a mechanism that diminishes the number of seeds available for germination and consequently the number of seedlings emerging at that time, which in turn would reduce intrapopulation competition.
In the field, onset of dormancy might be induced by low temperatures during autumn and winter, as suggested by the appearance of experimental dormancy after storage at 5°C in a cold chamber.Burial experiments -not carried out so far-should be conducted to confirm this behaviour.In the laboratory, however, dormancy disappeared during the second year and did not recur during the third year.This suggests that the germination behaviour of D. erucoides seed does not seem to fit the cycling between dormancy and nondormancy model observed in seed of several species in correlation with seasonal temperature cycles (Courtney, 1968;Roberts and Neilson, 1982;Van Hezewijk et al., 1994;Baskin and Baskin, 1998).Thus, some kind of endogenous control could be implied in this mechanism, as proposed by Froud-Williams et al. (1986) in Poa trivialis.According to the model suggested by this data, and the work of Pérez-García et al. (1995), each year recruitment of D. erucoides seedlings would be assured by a stabilized pool of seeds consisting of a fraction of newly dispersed seeds, together with nondormant, one-year-old seed and two-year-old, or older seed, avoiding excessive loss of viability in the soil seed bank.In any case, additional experiments, including burial tests and based on different populations from a wider range of geographical origins, should be conducted to confirm this germination behaviour.

Figure 2 .
Figure2.The germination of Diplotaxis erucoides seed with increased storage duration at different incubation temperatures.Results are mean ± standard deviation after 42 d incubation under a 16/8 h light/dark regime.For each incubation temperature, the germination percentages with the same letter are not significantly different at 1% level using LSD test.

Figure 3 .
Figure 3.The germination of Diplotaxis erucoides seed incubated in GA 3 solutions of increased concentration after 12 months storage.Results are mean ± standard deviation after 42 d incubation at 25°C under a 16/8 h light/dark regime.The germination percentages with the same letter are not significantly different at 1% level using LSD test.
The germination of Diplotaxis erucoides seed with increased storage duration.Results are mean ± standard deviation after 42 d incubation at 25°C under a 16/8 h light/dark regime.The germination percentages with the same letter are not significantly different at 1% level using LSD test.

Table 1 .
The germination of Diplotaxis erucoides seed incubated after 24 h immersion in H 2 O or GA3 solution (250 mg L -1 ) after 12, 24 and 36 months storage.Results are mean ± standard deviation after 42 d incubation under a 16/8 h light/dark regime P: probability.CV: coefficient of variation.