Peripheral populations have long been predicted to show lower vital rates, higher demographic fluctuations, and lower densities than central populations. However, recent research has questioned the existence of clear patterns across species' ranges. To test these hypotheses, we monitored five central and six northern peripheral populations of the widespread herb Plantago coronopus along the European Atlantic coast during 5 yr. We estimated population density, and calculated mean values and temporal variability of four vital rates (survival, individual growth, fecundity and recruitment) in hundreds of plants in permanent plots. Central populations showed higher fecundity, whereas peripheral populations had higher recruitment per reproductive plant, indicating a higher overall reproductive success in the periphery. Central populations showed a marginally significant tendency for higher growth, and there were no differences between range positions in survival. Fecundity and growth were affected by intraspecific competition, and recruitment was affected by precipitation, highlighting the importance of local environmental conditions for population performance. Central and peripheral populations showed no significant differences in temporal variability of vital rates. Finally, density was significantly higher in peripheral than in central populations, in discrepancy with the abundant-centre model. Density was correlated to seedling recruitment, which would counterbalance in peripheral populations the lower fecundity and the tendency for lower growth of established plants. Such compensations among vital rates might be particularly common in widespread plants, and advise against simplistic assumptions of population performance across ranges. The whole species' life cycle should be considered, since different arrangements of vital rates are expected to maximize fitness in local environments. Our results show also the importance of discerning between geographical periphery and ecological marginality. In a context of climate-induced range shifts, these considerations are crucial for the reliability of niche-models and the management of plant peripheral populations.
2013. Vol. 36, no 2, 136-145 p.