Summary 1.Wind power plants represent a risk of bird mortality, but the effects are still poorly quantified. We measured bird mortality, analysed the factors that led birds to fly close to turbines, and proposed mitigation measures at two wind farms installed in the Straits of Gibraltar, one of the most important migration bottlenecks between Europe and Africa. 2. Bird corpses were surveyed along turbine lines and an associated power line to estimate mortality rates. The behaviour of birds observed within 250 m of turbines was also recorded as a putative indicator of risk. The effects of location, weather and flight behaviour on risk situations (passes within 5 m of turbines) were analysed using generalized linear modelling (GLM). 3. Mortality caused by turbines was higher than that caused by the power line. Losses involved mainly resident species, mostly griffon vultures Gyps fulvus (0·15 individuals turbine − 1 year − 1 ) and common kestrels Falco tinnunculus (0·19 individuals turbine. Mortalities were not associated with either structural attributes of wind farms or visibility. 4. Vulture collisions occurred in autumn-winter and were aggregated at two turbine lines where risks of collisions were greatest. The absence of thermals in winter forced vultures to use slopes for lift, the most likely mechanism influencing both their exposure to turbines and mortality. 5. Kestrel deaths occurred during the annual peak of abundance in summer. Carcasses were concentrated in the open habitats around a single wind farm and risk may have resulted from hunting habitat preferences. 6. Synthesis and applications . We conclude that bird vulnerability and mortality at wind power facilities reflect a combination of site-specific (wind-relief interaction), speciesspecific and seasonal factors. Despite the large number of migrating birds in the study area, most follow routes that are displaced from the facilities. Consequently, only a small fraction of birds on migratory flights was actually exposed to turbines. New wind installations must be preceded by detailed behavioural observation of soaring birds as well as careful mapping of migration routes.

The Iberian lynx (Lynxpardinus) has suffered severe population declines in the twentieth century and is now on the brink of extinction 1 . Climate change could further threaten the survival of the species 2 , but its forecast effects are being neglected in recovery plans 3,4 . Quantitative estimates of extinction risk under climate change have so far mostly relied on inferences from correlative projections of species' habitat shifts 5 . Here we use ecological niche models coupled to metapopulation simulations with source-sink dynamics 6,7 to directly investigate the combined effects of climate change, prey availability and management intervention on the persistence of the Iberian lynx. Our approach is unique in that it explicitly models dynamic bi-trophic species interactions in a climate change setting. We show that anticipated climate change will rapidly and severely decrease lynx abundance and probably lead to its extinction in the wild within 50 years, even with strong global efforts to mitigate greenhouse gas emissions. In stark contrast, we also show that a carefully planned reintroduction programme, accounting for the effects of climate change, prey abundance and habitat connectivity, could avert extinction of the lynx this century. Our results demonstrate, for the first time, why considering prey availability, climate change and their interaction in models is important when designing policies to prevent future biodiversity loss.The 9 Rui Nabeiro Biodiversity Chair, CIBIO, University of Évora, Évora, 7000, Portugal, 10 Center for Macroecology, Evolution and Climate, University of Copenhagen, 2100 Copenhagen, Denmark. *e-mail: maraujo@mncn.csic.es towards increasing the carrying capacity of reintroduction sites through habitat restoration, relocating rabbits and limiting direct anthropogenic-related fatalities 4,13 .Although a well-financed effort to avert the extinction of this charismatic species is underway (> e94 million funding since 1994) 4 , non-accounted threats, such asclimatechangeanditsinfluenceonpreyabundance,arenotbeing considered in recovery plans.Here we provide the most comprehensive analysis of the likely effects of climate change yet for a threatened vertebrate. So far, models used to investigate how climate change will

No abstract

No abstract

Aim  To relate the recent Iberian lynx decline to changes in the distribution of the European rabbit after the haemorrhagic disease outbreak of 1989. As Iberian rabbits evolved in two geographically separated lineages, being the recent lynx range practically restricted to the southwestern lineage, we also test if differential range dynamics exists for these lineages, with the consequent implications for lynx conservation and reintroduction planning. Location  The Iberian Peninsula. Methods  We modelled environmental favourability for the lynx based on its distribution before 1989, and for the rabbit using distribution data collected primarily after 1989, and validated them using independent abundance data. We compared both models and combined them in a lynx occurrence forecast. We correlated the prevalence of southwestern rabbit lineage with the environmental favourability for the rabbit. Results  The environmental lynx model correlated with past lynx abundance data, but did not reflect its recent strong range contraction. The rabbit model correlated with recent rabbit abundance, but was negatively correlated with the environmental model for the lynx. The combination of both models forecasted lynx occurrence in a few separated nuclei, which encompass all recent lynx records. The prevalence of rabbit's southwestern lineage correlated negatively with favourability for the rabbit. Main conclusions  The region to which the lynx became confined before 1989 is currently less favourable for rabbits, whereas more favourable areas remain outside lynx reach. This differential favourability correlates with rabbit phylogeographical structure, suggesting that the southwestern lineage is facing more unfavourable conditions or is less resilient to recent diseases. The loss of concordance between lynx distribution and the whole rabbit phylogeographical structure has prevented lynx persistence in northeastern rabbit lineage areas, which should be considered in lynx reintroduction planning. Similar conservation problems could affect other ecologically interacting species whose distributions’ overlapping has sharply diminished.