DescriptionFragmentation and habitat loss are two of the main drivers of biodiversity decline. Therefore, understanding how populations and communities perform in fragmented habitats can help us to gain valuable insights to be applied in conservation, restoration and management of natural ecosystems. Studies have shown that metapopulation and metacommunity dynamics can be modulated by the pattern of interpatch connectivity, yet empirical evidence of this is still scarce. In this dissertation I explore the effects of network topology on metapopulation and metacommunity dynamics using an experimental model system with protists, and computer simulations.In chapter 1, I compared the performance of experimental metapopulations of the aquatic ciliate Paramecium tetraurelia, engineered as random and scale-free networks. I followed occupancy and abundance patterns over 20 protist generations and found that the effect of network topology in metapopulations is modulated by the dispersal/extinction balance. I suggest that the interaction between colonization/extinction rates and network topology alters the likelihood of rescue effects which results in differential patterns of occupancy and abundance in metapopulations.
In chapter 2, I increased the level of biological complexity and evaluated the effects of network topology on three-species metacommunities. After 30 protist generations the effects of network topology on metacommunity dynamics were clear at both regional and local scales. Contrary to findings of chapter 1, in this study scale-free metacommunities performed better than random metacommunities, with higher biomass, occupancy, persistence and diversity. I suggest that the heterogeneity in scale-free networks provides spatial refugia from both competition and predation, resulting in increased metacommunity persistence and function. Altogether these results highlight the importance of biological interactions to predict community dynamics and the need to observe these processes in a spatially explicit context.
Finally, in chapter 3, using simulations derived from findings in chapters 1 and 2, I explore the short-term effects of different scenarios of connectivity loss, in metacommunities arranged as random and scale-free networks. I followed metacommunity occupancy for 30 generations and observed that in most scenarios, patch disconnection had a positive effect on occupancy regardless of network topology. However, the number of disturbance events had opposite effects in scale-free and random networks. Overall, these results suggest that patch isolation is beneficial for metacommunity persistence and highlight the importance of considering spatial network topology in metacommunity studies.