Corre. Trophic roles of black rats and seabird impacts on tropical islands: Mesopredator release or hyperpredation?. Biological Conservation, Elsevier, 2015Elsevier, , 185, pp.75-84. 10.1016Elsevier, /j.biocon.2014 a b s t r a c tRats contribute to the decline of tropical seabird populations by affecting their breeding success through direct predation of eggs and chicks. When they coexist with other predators, invasive rats may also generate indirect interactions via the changes they impose on the structure of communities and trophic interactions following invasion ('hyperpredation process'), or when apex predators are eradicated from the ecosystem ('mesopredator release effect'). Understanding these effects is necessary to implement restoration operations that actually benefit threatened seabird populations. We investigated these processes on two French tropical seabird islands of the western Indian Ocean, Europa and Juan de Nova, where black rats coexist with two different apex predator species (introduced cats and potentially native barn owls). The parallel use of several methods (diet analysis, stable isotopes, seabird monitoring) to identify trophic roles of rats revealed that the direct impact of rats on seabirds was particularly high on Europa where only rats and owls occur, with high consumption of chicks resulting in low breeding success for several seabird species. We also suggested that hyperpredation associated with top-down regulation of cats is occurring on Juan de Nova, although territoriality of cats may buffer this process. Conversely we found evidence that mesopredator release effect is unlikely, irrespective of the apex predator identity. Considering the most likely effects on both islands we provided recommendations on eradication priorities to mitigate the risk of local extinction that seabirds are currently facing.

The island syndrome predicts directional changes in the morphology and demography of insular vertebrates, due to changes in trophic complexity and migration rates caused by island size and isolation. However, the high rate of human-mediated species introductions to some islands also increases trophic complexity, and this will reduce the perceived insularity on any such island. We test four hypotheses on the role of increased trophic complexity on the island syndrome, using introduced black rats (Rattus rattus) on two isolated coral atolls in the Mozambique Channel. Europa Island has remained relatively pristine and insular, with few species introductions, whereas Juan de Nova Island has had many species introductions, including predators and competitors of rats, anthropogenically increasing its trophic complexity. In the most insular environments, the island syndrome is expected to generate increases in body size and densities of rodents but decreases in the rates of reproduction and population cycling. Morphology and reproduction were compared using linear regression and canonical discriminant analysis, while density and population cycling were compared using spatially explicit capture-recapture analysis. Results were compared to other insular black rat populations in the Mozambique Channel and were consistent with predictions from the island syndrome. The manifestation of an island syndrome in rodents depends upon the trophic composition of a community, and may not relate to island size alone when many species additions, such as invasions, have occurred. The differing patterns of rodent population dynamics on each island provide information for future rodent eradication operations.

Seabirds are notoriously sensitive to introduced mammalian predators and eradication programs have benefitted seabird populations and their habitats on numerous islands throughout the world. However, less evidence is available from the tropics as to the benefits of rat eradication. Here, we report the seabird recovery and vegetation dynamics on a small coralline island of the tropical western Indian Ocean, eight years after Norway rat (Rattus norvegicus) eradication. Two species of seabirds were breeding before rat eradication (red-footed and masked boobies, Sula sula and Sula, dactylatra) and, in both species, the number of breeding pairs had an apparent increase of 22-23% per year after rat eradication. Such a high annual growth rate cannot be achieved by auto-recruitment only and our data suggest that immigration from other source populations never occurred in at least one of these species. We suggest that it is rather due to a rapid increase in breeding success, which rapidly increased the observed number of breeders since birds remained in the available-for-counting-as-breeders group for much longer. Two other species, the white tern (Gygis alba) and the brown booby (Sula leucogaster) were recorded breeding in 2014. The former species has not bred on the island since 1856 and the latter has never bred on the island. Plant cover (monospecific formation of the ruderal herb Boerhavia diffusa) dramatically increased from less than 30% of surface coverage to more than 70%. Although the initial restoration project was to eradicate all introduced mammals of the island simultaneously, house mouse (Mus musculus) eradication failed. Mouse density was high 8 years after rat eradication (32 mice/ha in dry season and 52 mice/ha in rainy season) but not higher than at a comparable tropical island of the region (Juan de Nova) where mice coexist with introduced black rats (Rattus rattus) and feral cats (Felis catus). These results are discussed in terms of the direct positive effects of rat eradication on seabirds and plants and the indirect positive effects of post-eradication seabird increase on soil manuring and vegetation recovery. Overall, our results show that on tropical islands, seabird and habitat recovery can be very rapid after rat eradication and should be implemented as a restoration tool wherever possible.