Positive feedback in species communities

Sometimes the eventual population densities in a species community depend on the initial densities or the arrival times of species. If arrival times determine species composition, a priority effect has occurred. Priority effects may occur if the species community exhibits alternative stable states (or: coexisting attractors), which are in general caused by positive feedback of population density on population growth. This thesis studies four model communities of mainly plankton species that exhibit such positive feedback. The first model is an abstract model in which two species either stimulate or inhibit, depending on population densities, the population growth of the own or the other species. An example is found in communities of plants that compete for scarce resources and at the same time ameliorate harsh conditions. In the model, a priority effect is more likely if the species compete more strongly with the own species than with the other species, or if they h elp the own species more. The reachability of a community state may be limited by the density at which species are introduced and on how often introductions occur. The second model describes the growth of light-limited phytoplankton populations that suffer from photoinhibition, meaning that their photosynthetic rate decreases with increasing light if light is strong. In a single population, photoinhibition may lead to alternative stable states where in one state the population is extinct, and another in which it survives. If several species grow together, there may be several alternative stable states, with at most one species at each state. Species that are less sensitive to photo-inhibition may facilitate the establishment of those that are sensitive. However, sustained coexistence is not possible. Furthermore, if populations start off only at low density, the community species composition does not depend on the arrival times of species. The third model is of a community of two zoo plankton prey species competing for a single resource and sharing a predator. One competitor interferes with the feeding of the other, but is more vulnerable to predation. Given the right trade-offs, a priority effect occurs if food supply is sufficient, and if predation pressure is not too strong. The fourth model describes a planktonic host-parasite system, where nutrient consumed by infected hosts is used for the parasite reproduction. Alternative stable states, where parasite persistence depends on a high enough initial density, arise if nutrient supply is sufficient and host loss rate not too high. Increasing nutrient supply leads to host-parasite cycles and extinction of the parasite through a deterministic Paradox of Enrichment. In all these cases, the occurrence of alternative stable states and priority effects depends on environmental conditions such as resource supply and loss rates. One should realise however, that not all alternative stable states may be reachable if spec ies communities are assembled with species starting off at low density only. If communities are assembled in this way, the number of alternative stable states that are reachable may be limited by the number of niches the environment harbours