Defining biogeographic provinces to understand the history and evolution of communities associated with a given kind of ecosystem is challenging and usually requires a priori assumptions to be made. We applied network theory, a holistic and exploratory method, to the most complete database of faunal distribution available on oceanic hydrothermal vents, environments which support fragmented and unstable ecosystems, to infer the processes driving their worldwide biogeography. Besides the identification of robust provinces, the network topology allowed us to identify preferential pathways that had hitherto been overlooked. These pathways are consistent with the previously proposed hypothesis of a role of plate tectonics in the biogeographical history of hydrothermal vent communities. A possible ancestral position of the Western Pacific is also suggested for the first time. Finally, this work provides an innovative example of the potential of network tools to unravel the biogeographic history of faunal assemblages and to supply comprehensive information for the conservation and management of biodiversity.

Seagrasses structure some of the world's key coastal ecosystems presently in decline due to human activities and global change. The ability to cope with environmental changes and the possibilities for shifts in distribution range depend largely on their evolvability and dispersal potential. As large-scale data usually show strong genetic structure for seagrasses, finer-grained work is needed to understand the local processes of dispersal, recruitment and colonization that could explain the apparent lack of exchange across large distances. We aimed to assess the fine-grained genetic structure of one of the most important and widely distributed seagrasses, Zostera marina, from seven meadows in Brittany, France. Both classic population genetics and network analysis confirmed a pattern of spatial segregation of polymorphism at both regional and local scales. One location exhibiting exclusively the variety 'angustifolia' did not appear more differentiated than the others, but instead showed a central position in the network analysis, confirming the status of this variety as an ecotype. This phenotypic diversity and the high allelic richness at nine microsatellites (2.33-9.67 alleles/locus) compared to levels previously reported across the distribution range, points to Brittany as a centre of diversity for Z. marina at both genetic and phenotypic levels. Despite dispersal potential of several 100 m, a significant pattern of genetic differentiation, even at fine-grained scale, revealed 'genetic patchiness'. Meadows seem to be composed of a mosaic of clones with distinct origins in space and time, a result that calls into question the accuracy of the concept of populations for such partially clonal species.

No abstract

Theoretically, the dynamics of clonal and genetic diversities of clonal plant populations are strongly influenced by the competition among clones and rate of seedling recruitment, but little empirical assessment has been made of such dynamics through temporal genetic surveys. We aimed to quantify 3 years of evolution in the clonal and genetic composition of Zostera marina meadows, comparing parameters describing clonal architecture and genetic diversity at nine microsatellite markers. Variations in clonal structure revealed a decrease in the evenness of ramet distribution among genets. This illustrates the increasing dominance of some clonal lineages (multilocus lineages, MLLs) in populations. Despite the persistence of these MLLs over time, genetic differentiation was much stronger in time than in space, at the local scale. Contrastingly with the short-term evolution of clonal architecture, the patterns of genetic structure and genetic diversity sensu stricto (that is, heterozygosity and allelic richness) were stable in time. These results suggest the coexistence of (i) a fine grained (at the scale of a 20 Â 30 m quadrat) stable core of persistent genets originating from an initial seedling recruitment and developing spatial dominance through clonal elongation; and (ii) a local (at the scale of the meadow) pool of transient genets subjected to annual turnover. This simultaneous occurrence of initial and repeated recruitment strategies highlights the different spatial scales at which distinct evolutionary drivers and mating systems (clonal competition, clonal growth, propagule dispersal and so on) operate to shape the dynamics of populations and the evolution of polymorphism in space and time. Keywords: clonality; seagrass; spatio-temporal genetic structure; Zostera marina INTRODUCTION Clonality is a life history trait widely distributed among taxa and habitats, particularly in photosynthetic organisms. Partially clonal organisms are characterized by a mixed system allowing the combination of two reproductive strategies: the production of new genetically identical modules through vegetative growth or fragmentation and the production of new genetic individuals through sexual recombination. As a consequence, their population dynamics and evolutionary trajectories are profoundly affected by their rate and mode of clonal reproduction. Populations of clonal plants are composed of genetic individuals, or genets occupying space and dispersing locally through the production of modular shoots, or ramets (Harper, 1977). As genets are able to persist through time and space, the composition and evolution of populations of clonal plants is largely affected by the level of intraspecific competition (Eriksson, 1989(Eriksson, , 1993Pan and Price, 2001;Travis and Hester, 2005).Depending on the turnover of genets and intensity of inter-genet competition for space, two extreme recruitment strategies have been defined (Eriksson, 1993): (i) the 'Initial Seedling Recruitment' (ISR) strategy, characterizing populations originating fro...