Evolution and synchronization in ecological networks

Goal To investigate how evolution affects synchronization in ecological networks. Plant and animal populations seldom live in a single homogeneous area, since landscapes are often spatially fragmented into various patches. Migration between different patches may be considerable, especially when patches are not too distant from one another. Such spatially structured populations are described as metapopulations or as ecological networks, where the nodes of the network graph correspond to patches and the graph's arcs to migration flows. Each node may harbor many species and/or various morphs of the same species. If patches have similar geological and climatic characteristics , also the structure of the populations associated with each node are intrinsi-cally similar. In a great number of ecological networks with highly diversified geographical extent and taxa, changes in the abundances of the same population in different patches are highly correlated over time. This property, often called population synchrony, has been documented with reference to many taxa, including aphids and butterflies (Han-Synchrony may arise from two independent mechanisms: (a) migration (dispersal) of populations among patches and (b) dependence of all population dynamics on some common environmental noise (rainfall, water temperature, etc.). The latter mechanism is known as the Moran effect (Moran 1953). (A third mechanism that has been sometimes been mentioned under the notion of entrained synchrony, is based on trophic interactions with populations of other species that are themselves synchronous. This, however, implies the operation of one of the two aforementioned mechanisms and therefore appears to be of lesser interest.) Both mechanisms are almost always present together, and this has generated a large debate about their relative importance. For example , with regard to Canadian lynx, only a few authors believe that the Moran effect is the main one responsible for synchronization (Sinclair et al. 1993, Cazelles and Stone 2003), particularly after the discovery (Shwartz et al. 2002) of high gene flow across distances of 3000 km. By contrast, for Dungeness crabs, even though pelagic dispersal of larvae has been ascertained, Higgins et al. (1997) showed based on a stage-structured model that the Moran effect may lead to large fluctuations with multi-year cycles that seem to be locked. Ecological networks with wild but synchronized local dynamics are so ubiquitous worldwide that one is naturally brought to question whether this might be due to Dar-winian evolution. Since this question, introduced in (Rinaldi 2006), is at the moment 3 addressed only indirectly, and …