Using combined miniature archival light and salt-water immersion loggers, we characterise the year-round individual at-sea movements of Europe's only critically endangered seabird, the Balearic shearwater Puffinus mauretanicus, for the first time. Focusing on the non-breeding period, we show that all of the 26 breeding birds tracked from their breeding site on Mallorca in the Mediterranean Sea successfully made a 2–4 month migration into the Atlantic Ocean, where they utilised well-defined core areas off Portuguese and French coasts. As well as identifying high-risk areas in the Atlantic, our results confirm that breeding birds spend most of the year concentrated around productive waters of the Iberian shelf in the western Mediterranean. Migration phenology appeared largely unrelated to the subsequent (distinctly synchronous) breeding attempt, suggesting that any carry-over effects were compensated for during a long pre-laying period spent over winter in the Mediterranean. Using the light and salt-water immersion data alone we were also able to characterise the pattern of pre-laying visits to the colony in considerable detail, demonstrating that breeding pairs appear to coordinate their over-day visits using a high frequency of night-time visits throughout the winter. Our study shows that geolocation technology is a valuable tool for assessing the spatial distribution of risks to this critically endangered species, and also provides a low-impact method for remotely observing the detailed behaviour of seabird species that may be sensitive to disturbance from traditional study methods.

Summary World‐wide, many seabirds are affected by fisheries in opposing ways: as a source of mortality from bycatch, but also by providing discards as a predictable and abundant food resource. This applies to the Balearic shearwater Puffinus mauretanicus, the most endangered European seabird, whose time to extinction was estimated at only ˜40 years a decade ago. Since the previous assessment, new data and more sophisticated demographic modelling have become available, and new fishing policies from the European Union (Common Fisheries Policy, CFP) will apply, posing different scenarios for the viability of the species. Thus, there is both an urgent need and an opportunity for a more reliable update of the conservation status of the species. Demographic data were collected between 1985 and 2014 at one of the world's largest colonies. Most demographic parameters were estimated using multievent capture–recapture modelling. Some parameters, such as bycatch rate, immature individual survival and recruitment, were estimated for the first time. We incorporated estimates into stochastic population models to forecast time to extinction and assess the viability under different management scenarios, accounting for upcoming fishing policies. Adult survival was much lower than expected (0·809, SE: 0·013) and largely influenced by bycatch, which accounted for a minimum of 0·455 (SE: 0·230) of total mortality. Breeding success was positively correlated with discard availability. Recruitment started at low rates in 3‐year‐old birds (0·030, SE: 0·0455), increasing in following age classes and was almost complete at 6 years. Under the present scenario, we predict a time to extinction of 61 years (95% CI: 55–69). Synthesis and applications. Population projections suggest that the actual impact of fisheries on Balearic shearwaters is unsustainable and the imminent discard ban under the new Common Fisheries Policy may accelerate the declining trend. This study demonstrates that reducing the bycatch rates of fisheries is an unavoidable and urgent conservation measure for avoiding the extinction of the species. We also advise setting up demographic long‐term studies, to allow researchers to diagnose, with reliability, the effectiveness of management actions. These actions will also benefit many other marine top‐predator species affected by this anthropogenic impact.

No abstract

Aim The movement patterns of marine top predators are likely to reflect responses to prey distributions, which themselves can be influenced by factors such as climate and fisheries. The critically endangered Balearic shearwater Puffinus mauretanicus has shown a recent northwards shift in non-breeding distribution, tentatively linked to changing forage fish distribution and/or fisheries activity. Here, we provide the first information on the foraging ecology of this species during the non-breeding period.Location Breeding grounds in Mallorca, Spain, and non-breeding areas in the north-east Atlantic and western Mediterranean.Methods Birdborne geolocation was used to identify non-breeding grounds. Information on feather moult (from digital images) and stable isotopes (of both primary wing feathers and potential prey items) was combined to infer foraging behaviour during the non-breeding season.Results Almost all breeding shearwaters (n = 32) migrated to non-breeding areas in the Atlantic from southern Iberia to the French Atlantic coast, where the majority of primary feather moult took place. Birds foraging off western Iberia yielded feather isotope ratios consistent with a diet composed largely of pelagic fishes, while the isotopic composition of birds foraging in the Bay of Biscay suggested an additional contribution of benthic prey, most likely from demersal fishery discards.Main conclusions Combined application of geolocators and stable isotopes indicates spatial variation in dietary behaviour and interactions with fisheries. Our results imply that both pelagic fish and fisheries discards are important components of diet during the non-breeding period, which may have implications for the at-sea distribution of this migratory species. These findings will contribute to bycatch mitigation in non-breeding areas and provide baseline data that should inform future assessment of seabird responses to changing fishery practices and prey distributions.

Acoustic dataloggers are used for monitoring the occurrence of cetaceans and can aid in fulfilling statutory monitoring requirements of protected species. Although useful for long-term monitoring, their spatial coverage is restricted, and for many devices the effective detection distance is not specified. A generalized additive mixed model (GAMM) was used to investigate the effects of (1) distance from datalogger, (2) animal behavior (feeding and traveling), and (3) group size on the detection probability of bottlenose dolphins (Tursiops truncatus) with autonomous dataloggers (C-PODs) validated with visual observations. The average probability of acoustic detection for minutes with a sighting was 0.59 and the maximum detection distance ranged from 1343-1779 m. Minutes with feeding activity had higher acoustic detection rates and longer average effective detection radius (EDR) than traveling ones. The detection probability for single dolphins was significantly higher than for groups, indicating that their acoustic behavior may differ from those of larger groups in the area, making them more detectable. The C-POD is effective at detecting dolphin presence but the effects of behavior and group size on detectability create challenges for estimating density from detections as higher detection rate of feeding dolphins could yield erroneously high density estimates in feeding areas.