Summary1 Plant species invading nutrient-poor ecosystems are likely to have their greatest impact on the native plant community by competing for resources below-ground. We investigated how root competition by an invasive tree, Cinnamomum verum , affects regeneration in nutrient-poor tropical secondary forests, in the Seychelles. 2 We performed three trenching experiments to investigate the effects of severing the root systems of mature trees on the growth of juveniles. These experiments had the following objectives: (i) to compare the responses of native and invasive saplings to release from root competition; (ii) to compare how seedlings (< 50 cm tall) and saplings (> 50 cm tall) of C. verum respond to trenching; and (iii) to compare the response of C. verum seedlings to trenching in forest stands with and without C. verum as the dominant species. 3 The results indicate that the dense topsoil root mat produced by mature C. verum trees suppresses the growth of young trees, mainly by increasing competition for scarce nutrients. Growth responses to trenching were stronger for seedlings than saplings, and stronger for juveniles of invasive than of native species. We conclude that stands of C. verum exert a strong below-ground filtering effect on seedling regeneration. This effect is likely to influence secondary forest succession by selectively reducing the establishment of invasive and small-seeded species. 4 Because of the bias in invasion biology towards relatively nutrient-rich, productive ecosystems, few studies have investigated the role of below-ground resource competition in plant invasions. Our results for an infertile, phosphorus-poor ecosystem show that root competition by an alien species can exert a strong influence on forest regeneration. We suggest that this mechanism may be of general importance in nutrient-poor tropical forests invaded by alien tree species.

Abstract. A method is presented for a quick and easily repeatable evaluation of the state of invasion of alien woody plants and the ecological status (prominence value, diversity, singularity and rejuvenation) of potentially important biodiversity sites in the Seychelles. The usefulness of this method has been tested in a survey on the islands of Mahé and Silhouette. The survey revealed 34 species as invasive alien species, the most prominent being Cinnamomum zeylanicum, Psidium littorale, Adenanthera pavonina and Tabebuia pallida. The greatest intensity of invasion was recorded in Intermediate Forests on moderate or gentle slopes and in stream ravines protected from desiccating winds and insolation. ‘Mountain Mist Forests’, ‘Palm Forests’ and ‘Glacis’ are the least invaded vegetation types. The difference in prominence values of alien and native plant species between leeward and windward sides is significant. Glacis habitats do not show a decline in protection values with decreasing altitude. Prominence values and percentage frequency of invaders are exceeded by the endemic palm Phoenicophorium borsigianum which is able to maintain itself in areas suffering from forest destruction and topsoil erosion.

Lodoicea maldivica, a palm endemic to two small islands in the Seychelles group, has the largest seed in the plant kingdom. We present here an interpretation of the ecological and evolutionary significance of this seed in terms of the island environment where the species grows. We begin by reviewing the available information about the biology and ecology of Lodoicea and present some original data on the growth and development of Lodoicea in its native habitat. A remarkable feature of young plants is the enormous size of their leaves and the great length of their petioles, these being especially elongated when growing beneath the canopy. As a result, juvenile plants can reach a height of 15 m and hold their foliage in the forest canopy. This capacity to produce such an enormous juvenile plant is related in part to the large food reserves in the seed. We suggest that Lodoicea evolved from a more typical borassoid palm (perhaps a plant like Borassus aethiopum which is widespread in the savannas of Africa) and propose two hypotheses to explain why this occurred. According to the ‘shade hypothesis’, increasingly humid conditions on the Seychelles led to strong selection for plants with the tallest seedlings, since these would be the most likely to establish successfully under the low light conditions prevailing on the forest floor of closed forest. The ‘sibling competition hypothesis’ postulates that the island populations of the ancestral palm lost any means for seed dispersal, and their seeds simply fell to the foot of the parent tree. This resulted in a strong selection pressure for reduced fecundity. The two hypotheses are not mutually exclusive, and together they may help us to understand many of the unusual features of this remarkable species.