Aims: Zoochory is a prominent dispersal process in a wide range of plant species.However, the extent to which the trait composition of dispersed seed assemblages differs from the regional assemblage remains untested, as does the extent to which sympatric native herbivores disperse traits non-randomly.Location: Lorris and Montargis forests, central France. Methods:We compared the taxonomic and functional composition of seed assemblages dispersed by three wild ungulates to the characteristics of the regional flora in an agro-forested landscape. We collected roe deer (Capreolus capreolus), red deer (Cervus elaphus) and wild boar (Sus scrofa) faeces in two forests, and subjected the samples to germination under controlled conditions. We first examined how different vectors influence the composition of the dispersed plant assemblages by comparing abundance and species richness of the seedlings emerging from the faeces of the three ungulates. We then compared the functional characteristics of the dispersed pool with those of the regional flora in a multivariate functional space built from 20 relevant plant traits.Results: A total of 754 seedlings and 46 plant species germinated from 300 faeces samples, with higher plant species richness and abundance for red deer faeces. All three ungulates widely use forest habitat, but the proportion of non-forest plants was higher in red deer and wild boar faeces than in the regional species pool. Traits such as seed shape, seed size or seed bank longevity affected dispersal probability, but their effects were overshadowed by habitat effects.Conclusions: Endozoochory acts as an indirect functional filter, filtering species according to the vectors' feeding habitat. This could affect the composition of forest plant communities by allowing plants from open habitats to colonize forested areas.

ABSTRACT1. Diving surveys were undertaken to investigate the effects of marine reserve protection on spiny lobster (Jasus edwardsii) populations at Tonga Island Marine Reserve, northern South Island, New Zealand over a 2 year period from December 1998 to December 2000.2. Spiny lobsters were 2.8 times more abundant overall, and mean size was 19 mm carapace length larger in shallow transects and 28 mm carapace length larger in deep transects, in the marine reserve than at adjacent fished sites. That pattern was evident despite very high variability within sites, and among sites within reserve and fished areas.3. Large reproductive males were 10 times more abundant within the reserve compared to adjacent fished areas, suggesting that more eggs would be fertilized in the reserve than on the adjacent fished coast.4. Estimates of size-specific fecundity, combined with abundances of females, suggested that almost nine times more eggs would be produced in the reserve than in fished areas.5. Diving and commercial trapping investigations found weak trends in spiny lobster abundance consistent with gradients at reserve borders. Tagging investigations near those borders led to only one resighting of a tagged spiny lobster.6. We estimate that the mean abundance of spiny lobster in the reserve has increased by 22%, 5 years after its establishment, indicating an annual population increase of 4.4%. Over the same period, abundance of spiny lobster outside the reserve has declined by 2.9% per annum.7. Based on known spiny lobster movements, we suggest that marine reserves of more than 10 km length should be given priority over smaller reserves. Smaller reserves will, however, protect part of the population for at least a portion of their lifespan.8. Previous studies of movement of J. edwardsii suggest that spillover from the reserve should occur, and as population density increases we predict that more spiny lobsters will move out from the reserve.

Big game populations have been rapidly increasing in temperate and boreal ecosystems worldwide, which has prompted the need to study the socio-ecological issues related to their management. In addition, the use of game-fencing has been spreading in certain rural areas, especially in French Sologne. These two parallel trends raise the issue of the new socio-spatial arrangements developing in rural areas. We set up an in-depth case-study combining stakeholder-interviews and ecological surveys (red deer trails, pellets and bed sites) to understand how game fences may affect both animal and human mobility, and why this may raise human conflicts. The study site was a 17.52 km² fenced landscape comprising a public forest bordered on the north and south by two privately owned fenced lots. Both the interviews and the ecological data indicated that the fences were crossable for red deer. We did not find any fencing effect on the location of red deer trails, though red deer pellet groups were significantly more frequent outside fenced lots. On the other hand, we recorded more bed sites within fenced lots, in accordance with the stakeholders’ statements that fenced lots had a high sheltering value for red deer. This implies that stakeholders have contrasting opportunities to encounter red deer because the fences deter non-owners from trespassing. Although any private owner has the legal right to fence his or her land under French law, our finding calls for more attention to how the fencing is designed, since its effects may impact the surrounding landscape. For instance, fences can allow a single owner to impose game management policies in the areas surrounding his or her estate.