Long-term assessment of seed limitation in plants:: results from an 11-year experiment

1 Seed limitation of plant species distribution has received much attention in recent years. The empirical evidence of seed limitation is, however, still weak, and mostly correlative. Unambiguous evidence can only be provided by long-term transplantation experiments. 2 We used an 11-year experiment, in which seeds of six perennial forest herbs were sown into 43 patches that were either occupied or unoccupied by the target species, to investigate the importance of seed limitation. In addition, we determined whether short-term (1-4 years) and long-term (11 years) experimental assessments provided similar results. Given that long-term experiments are generally not possible, we examined whether environmental factors and vegetation composition can correctly identify suitable habitats. 3 Our results show that the distribution of the forest herbs investigated were limited by the availability of seeds, and that this pattern was qualitatively consistent, over time. However, the number of initially unoccupied patches with successful recruitment decreased with time. 4 The fact that a similar trend in recruitment success was found in sowing experiments in initially occupied patches suggests that the decrease in recruitment success over time cannot solely be attributed to habitat suitability. Environmental factors, such as soil nutrients and vegetation composition, were in some cases correlated with recruitment success, but the strength of these relationships was often weak. 5 Our results support the notion that seed availability is an important limitation of plant species abundance and distribution. The results also point to several difficulties related to the empirical assessment of seed limitation. We therefore suggest a protocol that should enhance the quality of future experiments.