1. Two experiments were carried out in an ancient English deciduous woodland to determine the relationship between the occurrence of gaps - unoccupied spaces - in the ground layer and the ability of Narcissus pseudonarcissus to occupy them. In the first experiment, seeds and bulbs of Narcissus were planted into gaps and their fates were followed for 10 years. A second experiment examined the survivorship, mortality, clonal growth and flowering of bulbs and seeds grown in pots at different densities in open and shaded sites. 2. In the first experiment, bulbs planted into gaps in a shaded site performed similarly to controls (naturally established individuals). Plants growing from seeds or seedlings died, as did the controls. It is concluded that the existence and persistence of gaps is a function of the demography of the species rather than of an inherently inimical environment. 3. In the second experiment, survivorship of established bulbs showed a more-or-less constant probability of death, whilst that of seedlings showed declining probability of mortality. Clonal growth rate was density-dependent but establishment of plants from seed was not. Mortality was density-independent, except at high density in open ground. Clonal growth was twice as fast in open ground as in shaded sites. 4. The results of these experiments are brought together in a scheme suggesting that a wide range of densities of N. pseudonarcissus is possible at the site due to the changing balance between reproduction and mortality in a periodically disturbed environment. The implications of this conclusion are explored in relation to ground occupancy and the development of mixtures in the ground flora of ancient British woodlands.
