Modeling extinction risk for seabirds in Brazil

Abstract

Species extinction risk is associated with intrinsic features such as body mass, geographic range size and diet, and also with external threats, mainly anthropogenic pressure. Current biological knowledge on seascapes still falls behind knowledge from terrestrial ecosystems, and understanding the pathways to extinction for declining groups, such as seabirds, is of paramount importance. Here we provide the first estimates of seabird extinction risk derived from regressive models associating species intrinsic features and external threats, and evaluation of effectiveness of marine protected areas (MPA). We aimed to: (i) understand how intrinsic and external variables interact to set pathways to seabird extinction; (ii) define which variables are most important; (iii) evaluate the geographic structure of extinction risk; and (iv) evaluate the effectiveness of Brazilian MPAs in representing species at high and low risk. We used classification tree to model extinction risk and random forest to evaluate the importance of each variable. Our models were accurate to predict relative extinction risk of threatened species. Breeding range size and threat from fishing vessel were the most important correlates of extinction risk. Brazilian MPAs are not effective to protect seabirds and are short to cover a minimum of 10% of species ranges. We show that correlates of extinction risk act in a synergistic way and suggest that conservation of seabirds in Brazil should be considered high priority given their low current level of protection. © 2013 ABECO.