Historical distribution and regional dynamics of two Brassica species


  • Matthias C. Wichmann,

  • Matt J. Alexander,

  • Rosie S. Hails,

  • James M. Bullock

M. C. Wichmann (mwich@ceh.ac.uk), M. J. Alexander and J. M. Bullock, Centre for Ecology and Hydrology, CEH Wallingford, Maclean Building, Benson Lane, Crowmarsh Gifford, Wallingford, Oxfordshire, OX10 8BB, UK. – R. S. Hails, Centre for Ecology and Hydrology, CEH Oxford, Mansfield Road, Oxford, OX1 3SR, UK.


Plants often show a patchy distribution. This can be related to the discontinuous distribution of environmental variables that define suitable habitat. Metapopulation theory suggests that additional patchiness may be caused by the dynamics of local extinctions and recolonisations. However, the contribution of these two mechanisms to explaining the observed patterns, and thus the applicability of metapopulation theory to plants remains controversial partly because population turnover may occur at long time-scales not usually covered by ecological studies. We analyzed the role of environmental variation and population turnover in determining the distribution of two Brassica species, an annual and a perennial, along 44 km of coastline in Dorset, UK. Mapped occurrence and abundance of both species in recent years was compared with distributions from a 70 yr-old dataset and a survey of present-day environmental parameters. Abiotic and vegetation parameters were correlated with the occurrence of both species in binary logistic regression models and explained spatial auto-correlation in the Brassica distributions. These regression models suggest that neither species is occupying all potential habitat in the region studied. The relationship between historical and present distributions differed between the species. While an historical signal was very weak in the annual B. nigra, it had a large influence in predicting the present distribution of the perennial B. oleracea. This suggests local extinction and colonisation events for B. nigra over the 70 yr period, but not for B. oleracea which showed little evidence of population turnover. Our results demonstrate that the consideration of large time-scales can reveal patterns of regional dynamics. We conclude that metapopulation dynamics might be possible for our annual but can be ruled out for our perennial study species over the past 70 yr. We argue that beyond this time-scale possible metapopulation dynamics may be overridden by faster processes of environmental change.