Impacto dos predadores introduzidos na Ilha do Corvo no sucesso reprodutor das populações de Cagarro (Calonectris diomedea borealis)

ABSTRACT: The Corvo lsland has unique features, in terms of endemic fauna and flora, numbers of human population, introduced predators, size of the island, among others, that gave to this island the status of Biosphere Reserve. Additionally, there are important species of seabirds that only reproduce in restricted areas of Europe, and are classified as prioritary by Annex I of the Birds Directive. Nevertheless, the populations of some species are still to confirm (e.g Pterodroma feae, Oceanodroma castro) and others need more research in terms of numbers and localization of the colonies (e.g Puffinus puffinus, Puffinus assimilis). The more conspicuous of the seabirds here present is Cory's Shearwater, Calonectris diomedea borealis and has the biggest expression in terms of numbers and densities in the Azores archipelago which represents 60-70% of the world population. The Corvo population is one of the major ones but the nidification areas and the actual size of the population is still not well known. One of the major causes of the decline of the seabirds, besides the fisheries, especially those ones who reproduce in cavities on the soil, are the introduced predators in islands, places that originally yield millions of seabirds. The biggest problem is that they evolved in environments that were predator free and lack defense mechanisms to face the predation of eggs and chicks by rats and cats. This causes the reproductive collapse in some species and could force the nesting places to sub-optimal conditions, that can compromise the viability of the populations. Last year, SPEA started a LIFE+ project "Safe islands for seabirds" to assess the impacts of the predators on seabirds and prepare the restoration of habitat, by ways of predator and invasive plants control and reforestation with endemic species. This project has an important ecological relevance as it can represent an important step in the developing of protocols for exotic species eradication in islands with human population, with both cats and rats as the major predators. The objectives of this thesis were to quantify the influence of rodents in the reproductive success of Cory's Shearwater. We analyse the distribution of three species of rodents (Rattus rattus, R. norvegicus and Mus musculus) in different habitats and their dependence with some environmental variables. The method chosen was by means of tracking tunnels in two different seasons, one in April-May and the other in September-October, that wanted to express different food availability and different stages of the reproductive period of Cory's in Corvo. For this analysis, we decide do use an algorithmic model Random Forest for the software R 2.9.2. Our results point out the dependence of the presence of rodents with introduced anthropogenic characteristics as roads, exotic vegetation and distance to village. Also, we noticed some differences between the house mice and the rats in terms of abundance and selection of habitat. We could identify different accessible colonies of Cory's Shearwater in the island, and monitor almost 200 nests, once a week, in eight different places of Corvo. We analyzed and compared times of hatching, fledging, incubation period and breeding success, with other islands (Berlenga, Selvagem Grande and others from Azores, such as Graciosa) where Cory's shearwater reproduces, and finally we tried to understand which are the main causes of failure using the model Nest Survival for the program MARK. This program is an important ecological tool for testing different factors and the correlation between them that could influence the weekly nest survival. Also, give us the weight of the model for explaining the data. Our results point out the synchronization of times of hatching and fledging among the different places in the Atlantic, but there's a major difference in the hatching and fledging successes. Surprisingly the causes of failure don't seem to be related with rat abundance, as most the studies point out, ant the time with less survival probability seems to coincide with the first days after hatching when the chick is more vulnerable. Our assumption is that there is a third variable that wasn't account for, that is influencing the abundance of rats and the reproductive success of Cory's. This variable is the presence of cats. The analise of the Life Table demonstrate the big resilience in fluctuations of the population of Cory's Shearwater in Corvo lsland, a characteristic of long lived species. We recommend that for the planning of the eradication protocol both cats and rats should be target of control at the same time, incorporate prioritary areas of control for rodents and to study more accurately the causes of fail that influences the seabirds success