Abstract. Population viability analysis (PVA) has become a basic tool of current conservation practice. However, if not accounted for properly, the uncertainties inherent to PVA predictions can decrease the reliability of this type of analysis. In the present study, we performed a PVA of the whole western European population (France, Portugal, and Spain) of the endangered Bonelli's Eagle (Aquila fasciata), in which we thoroughly explored the consequences of uncertainty in population processes and parameters on PVA predictions. First, we estimated key vital rates (survival, fertility, recruitment, and dispersal rates) using monitoring, ringing, and bibliographic data from the period 1990-2009 from 12 populations found throughout the studied geographic range. Second, we evaluated the uncertainty about model structure (i.e., the assumed processes that govern individual fates and population dynamics) by comparing the observed growth rates of the studied populations with model predictions for the same period. Third, using the model structures suggested in the previous step, we assessed the viability of both the local populations and the overall population. Finally, we analyzed the effects of model and parameter uncertainty on PVA predictions. Our results strongly support the idea that all local populations in western Europe belong to a single, spatially structured population operating as a sourcesink system, whereby the populations in the south of the Iberian Peninsula act as sources and, thanks to dispersal, sustain all other local populations, which would otherwise decline. Predictions regarding population dynamics varied considerably, and models assuming more constrained dispersal predicted more pessimistic population trends than models assuming greater dispersal. Model predictions accounting for parameter uncertainty revealed a marked increase in the risk of population declines over the next 50 years. Sensitivity analyses indicated that adult and pre-adult survival are the chief vital rates regulating these populations, and thus, the conservation efforts aimed at improving these survival rates should be strengthened in order to guarantee the long-term viability of the European populations of this endangered species. Overall, the study provides a framework for the implementation of multi-site PVAs and highlights the importance of dispersal processes in shaping the population dynamics of long-lived birds distributed across heterogeneous landscapes.

aves. En el presente estudio se proporciona información detallada sobre los factores que determinan el reclutamiento territorial en una especie amenazada: Aquila fasciata. Se estudiaron dos poblaciones localizadas en el extremo noroccidental de la distribución de la especie, una en Cataluña (NE de España) y la otra en Provence y Languedoc-Roussillon (SE de Francia). La metodología de campo se basó en un estudio a largo plazo que incluyó marcaje de volantones (n = 451 pollos) y seguimiento de territorios en ambas poblaciones durante el período comprendido entre 1980 y 2007. En primer lugar, se estimó una tasa de retorno global del 9.97 % (45 reclutas), que no difirió significativamente entre las poblaciones estudiadas. En segundo lugar, se detectaron efectos significativos del año de nacimiento y del éxito reproductor del territorio de origen sobre la probabilidad de que un volantón fuera posteriormente reclutado: aquellos pollos de territorios con mayor éxito mostraron mayor probabilidad de reclutamiento. En tercer lugar, se utilizaron técnicas de captura-reavistamiento para estimar la probabilidad de que un individuo territorial de una determinada edad fuera un recluta. Este parámetro mostró valores entre 0.985 y 0.999 para individuos de dos años, entre 0.763 y 0.808 para individuos de tres años, y se redujo drásticamente para individuos de cuatro años o mayores, mostrando valores de ente 0.066 y 0.272. En cuarto lugar, las hembras se dispersaron a mayor distancia que los machos. Además, las hembras de Cataluña se dispersaron a mayor distancia que las hembras de Francia. En último lugar, aquellos territorios previamente ocupados, situados en áreas con elevada densidad y situados en áreas de mayor calidad se mostraron más atractivos para individuos sin experiencia territorial previa.

1. Mortality caused by power lines is a conservation problem for many vulnerable bird species. Many large species are especially threatened by electrocution as they frequently perch on pylons leading to electrocution that typically causes death. Electrocution mitigation measures have been implemented to protect several species; however, a resulting decrease in mortality due to these measures has not previously been demonstrated at the population scale. 2. In this study, we used data from a long-term capture-recapture programme (combining resightings of live birds and recovery of dead birds) carried out on the French population of the Bonelli's eagle Aquila fasciata from 1990 to 2009 to estimate the impact of the insulation of power lines on key demographic rates. 3. We found that the survival probability of all age classes increased after the insulation of dangerous power lines, due to a decrease in mortality caused by electrocution. This decrease was partially compensated for by an increase in other causes of death. 4. Our findings show that insulation of power lines has a strong positive impact on juveniles and immature birds and a lesser impact on adults. The overall increase in survival due to power line insulation led to a sharp increase in predicted population growth rates (from 0Á82 to 0Á98), although our findings still suggest that the population is not self-sustaining. Elasticity values indicate that adult survival is the key parameter in the population dynamics of this species, and since adult mortality caused by electrocution seemed close to zero, our ability to act on this parameter is limited. 5. This study demonstrates that insulation of power lines is relevant for the conservation of large bird species at a population scale as it allows the survival rate of all age classes to increase and thus in turn has a strong positive impact on population growth rates. 6. Synthesis and applications. We demonstrated that mortality rates induced by electrocution are considerable and have major consequences for the population viability of birds. We also demonstrated that electrocution mitigation measures can lead to a sharp increase in survival through reducing mortality from electrocution leading to improved population viability. In the light of these results, there is an urgent need that conservationists contact power line stakeholders not only to urge them to generalize retrofitting actions but also, in planning new infrastructure development, to plan for less harmful power lines, since this will be far less costly than developing a posteriori mitigation actions.

Survival typically contributes most to population trends in long‐lived birds and its accurate estimation is therefore vital for population management and conservation. We evaluated the effects of age, territoriality and reproduction on survival in Bonelli’s Eagle Aquila fasciata through multistate capture‐mark‐recapture analyses on a long‐term dataset. Monitoring was carried out in southeast France (1990–2008) and involved the surveying of territorial Eagles, the marking of fledged chicks, and the recording of resightings and recoveries of marked non‐territorial and territorial birds. Survival improved with age, but territoriality was not retained in the best model; yearly survival was estimated at 0.479 for fledglings (to 1 year of age), 0.570 for 1‐ and 2‐year‐olds, and 0.870 for 3‐year‐old and older individuals. The second best model supported a further increase in survival from 3‐year‐olds (0.821) to older individuals (0.880). In the third best supported model, territoriality enhanced survival, but only in 2‐year‐olds (0.632 vs. 0.562 for non‐territorial). We found no correlation between the previous breeding stage and future survival, consistent with the long lifespan of the study species. Nevertheless, 4‐year‐old and older successful breeders were more likely to breed the following year than failed adult breeders (0.869 vs. 0.582), suggesting that the cost of reproduction is small in comparison with the variation in quality among individuals or their territories.

No abstract