Abstract
This report presents the results of GPS tracking of adult black-legged kittiwakes Atlantic puffins, common guillemots and razorbills breeding on the Isle of May (SE Scotland) in 2018 and assessment of overlap with the consented Neart na Gaoithe offshore wind farm. Locational data were obtained from 16 kittiwakes, 23 puffins, 24 guillemots and 13 razorbills (comprising 71, 175, 207 and 142 trips, respectively) in June and July 2018. The data were partitioned into non-flight behaviours (foraging and resting), relevant to displacement effects, and flight behaviours, relevant to collision risk and barrier effects. A resampling procedure indicated that the sample sizes of tracked birds were adequate to estimate the at-sea area used by the Isle of May populations of all four species during the deployment period.
The at-sea non-flight distributions of the four study species included both inshore and offshore areas, as found in previous GPS tracking studies in 2010-14. Differences among the species were apparent, with guillemots and razorbills using coastal areas more extensively, and puffins and kittiwakes using mainly offshore waters. The core areas used by guillemots were concentrated around the Isle of May and within the Firth of Forth and St Andrews Bay. Razorbills used offshore areas mainly to the east of the colony, and to a lesser extent coastal areas within the Firth of Forth and St Andrews Bay. Puffin distribution was concentrated offshore, from a south-easterly to north-easterly direction from the colony. Kittiwakes had a wider distribution than the three auk species, including mainly offshore areas spanning from a south-easterly to north-easterly direction from the colony. This was reflected in the larger mean maximum range for this species (60.3 ± 7.0 km) compared to guillemot (39.1 ± 2.4 km), razorbill (46.0 ± 2.6 km) and puffin (48.8 ± 2.5 km). The distribution of flight lines matched the distributions of non-flight activities. Guillemots departed from and returned to the colony on bearings ranging from southwest and northwest (for inshore foraging trips) to northeast and east (for offshore trips). A similar pattern was observed in razorbills although flight bearings in an easterly direction were more common. Flight bearings of puffins and kittiwakes spanned from a north-easterly to easterly/south-easterly direction from the colony.
A small proportion (up to 1.5%) of the core areas (50% kernels) used by guillemots, razorbills and puffins for non-flight activities overlapped with the planned Neart na Gaoithe footprint. In contrast, the overlap in kittiwakes was substantially larger (>10%). The proportion of the overall area used at sea (90% kernels) that overlapped with the wind farm footprint was also small (<5% in all species). However, the entire footprint fell within the overall areas used by all four species. The overlap of flight activities with the wind farm footprint was generally higher than the overlap of non-flight activities. At all three levels that we explored (bird, trip and flight), overlap was lowest in guillemots and highest in kittiwakes (and puffins, at the flight level only). The lower overlap observed in the guillemot is likely due to the predominantly inshore distribution of the species during our study.
In the light of past evidence that Isle of May puffins are susceptible to disturbance when captured in burrows, we adopted an alternative deployment method by mist netting breeding birds close to burrows. Despite this, we recorded negative effects of GPS logger deployment on chick provisioning rates and chick survival, in particular in cases where both members of the pair carried loggers. Our results indicate that both handling and device deployment may contribute additively to these effects. We found similar effects in birds fitted with a heavier (8.2g) and lighter (4.1g) logger model, suggesting that the attachment of a device may be a key issue causing disturbance, or there is a threshold mass that puffins will tolerate that is lower than the smaller of the two loggers used. Feeding behaviour was related also to the amount of time that had passed since logger deployment, with feeding rates declining and trip duration and range slightly increasing over time. Such strongly negative impacts of device deployment are rarely observed in seabirds. However, we were able to improve the welfare of chicks through supplementary feeding.
Conclusions: This study demonstrates variation in seabird distributions at sea among species and, when comparing with previous GPS tracking studies undertaken in 2010-2014, variation within species among years. Our study also confirmed overlap between the distributions of guillemot, razorbill, puffin and kittiwake populations breeding on the Isle of May and the planned Neart na Gaoithe offshore wind farm. Our results indicate substantial negative effects of logger deployment on chick feeding rates and chick survival in puffins and demonstrates a particular challenge with using data loggers in this population. We recommend for future GPS tracking studies that puffins are captured at burrow entrances, not in mist nets, to ensure that all chicks of instrumented birds are identified, and that only one member of a pair is deployed with a data logger. The interannual variation in distribution seen in all species indicates that further GPS tracking in future years prior to construction would be very valuable, as well as the development of a structured monitoring plan spanning the periods before, during and after wind farm construction to maximise opportunities for quantifying the impacts of the wind farm on these seabird populations.