Theory predicts that life history traits that reduce the impact of environmental variation will exhibit a pattern of negative covariation (trade-offs). In plants, such traits are seed size, spatial seed dispersal and seed dormancy which interact to reduce risk in a variable environment. These life history traits also affect the colonisation ability of species, and, hence, the distribution of local populations over a larger spatial scale. The main objective of this thesis concerns the process of dispersal of plants living in fragmented landscape. Capacity for adhesive seed dispersal, dormancy and seedling establishment were investigated for several grassland species under natural conditions. Population dynamics of two common perennial grassland species, Agrimonia eupatoria and Geum rivale, both possessing adhesive dispersal attributes were studied. Relationships between seed-related traits affecting seedling establishment, adhesive seed dispersal and seed dormancy were examined by combining experimental and comparative approaches, with and without phylogenetic information. The studies on adhesive seed dispersal comprise 21 species and offer experimental data on seed retention times and potential dispersal distances in fur of three animal species. The results suggest that adhesive dispersal is an effective mode of dispersal, even for species that lack adhesive structures on their seeds. Several species, out of the 11 investigated within Rosaceae, possessed seeds that were able to germinate after being buried in the soil column for three years. Overall, small-scale disturbances were beneficial for seedling emergence and recruitment. The results confirmed an expected positive effect of seed size on establishment ability. The results also suggested a negative relationship between seedling establishment and the inherent capacity of adhesive dispersal and seed dormancy (the latter with use of a phylogenetic approach). The hypothesised negative relation between seed size and spatial dispersal, or seed dormancy, was not detected. However, correlations without phylogeny revealed a significant negative relationship between seed size and dormancy. The demographic study showed that the populations of the two perennial species exhibited very stable and slow dynamics, even though the demographic parameters varied considerably between populations and among years. Stasis and growth in the adult stage classes contributed most to population growth. This thesis gives support to the suggestion of road verges as potential dispersal corridors for grassland plants in a fragmented landscape. However, the occurrence (occupancy) of the species in road verges was not related to their recruitment features or dispersal capacity
Stockholm: Stockholm University , 1999. , 40 p.