Seeds of oilseed rape (Brassica napus L.) can persist in the soil over several years by becoming secondarily dormant and can then germinate to create volunteer plants in following crops. As well as agricultural impacts caused by volunteers, gene dispersal in time – particularly from genetically modified cultivars – can be another undesirable consequence. Conventionally bred and transgenic seeds were tested in 2001 and 2002 in laboratory experiments, and in a field experiment, by burying seeds in the soil to determine the variation in dormancy and persistence capacity.

In the conventional group of cultivars tested in the laboratory, the level of dormancy was 13–76% in 2001, and 3–76% with an extended group in 2002. The transgenic group of cultivars was 1–31% dormant. In the burial experiments the number of viable seeds recovered in the conventionally bred cultivars ranged from 34–90% in 2001, and 7–68% in 2002. In the same studies the transgenic cultivars developed persistence levels from 12–79% in 2001, and 46–67% in 2002. Since dormancy levels of conventionally bred cultivars from 2 harvest years in the laboratory tests correlated significantly (r=0·71), it appears that there is a genetic background to secondary dormancy. There was also a significant correlation (r=0·61 in 2001 and 0·80 in 2002) between the results from laboratory and burial experiments. This indicates that the laboratory approach can simulate the situation in the field. Ageing over 6 months decreased the capacity for seed persistence to about a fifth of the level shown when freshly harvested. As a consequence of ageing and environmental impacts on persistence, only seeds from the same location and harvest year should be used for testing genetic variability. The high genetic variability among currently available rape seed cultivars gives breeding strategies a good chance of ideotyping low persistence genotypes and minimizing the risk of gene dispersal in time.