Assessing movements of Lesser Black-backed Gulls using GPS tracking devices in relation to the Galloper Wind Farm

Abstract

  1. A programme of Lesser Black‐backed Gull Larus fuscus tagging and tracking work was initiated within the Alde‐Ore Estuary Special Protection Area (SPA) during the 2019 breeding season, and continued throughout the 2020 breeding season, in order to fulfil requirements of the Galloper Wind Farm (GWF) Ornithological Monitoring Programme (OMP) and test key predictions of the Environmental Statement (ES).

  2. The results from both the 2019 and 2020 breeding seasons of tracking are summarised within this report. Specifically, in order for GWF Ltd (GWFL) to be able to achieve the objectives of the OMP, the project aimed to produce data and analyses to provide an assessment of:

    i. The area use of Lesser Black‐backed Gulls breeding at the Alde‐Ore Estuary SPA and thus the relative time spent within the GWF;

    ii. Their relative behaviours within and outside the wind farm.

    iii. Their relative flight heights within and outside the wind farm; and

    iv. Their movements within and outside the wind farm.

  3. Methods (Chapters 2 and 3). In total, 30 breeding adult Lesser Black‐backed Gulls were caught at the nest at Havergate Island within the Alde‐Ore Estuary SPA in May 2019 and fitted with GPS tracking devices – 15 with University of Amsterdam (UvA) GPS devices and 15 with Movetech GPS‐GSM devices. Data collection covered both the 2019 and 2020 breeding seasons and the 2019/20 non‐breeding season (and therefore the species’ full annual cycle). Comparison is provided with data from pre‐construction studies undertaken between 2010 and 2015 (Thaxter et al., 2014b; RSPB, unpublished) at Havergate and Orford Ness, also within the Alde‐Ore Estuary SPA.
  4. All tracked birds were monitored throughout the 2019 and 2020 breeding seasons (although there was limited access to the colony in 2020, due to restrictions associated with the Covid‐19 pandemic). Comparison of the breeding success and return rates of tagged and control birds suggested no detrimental effects of the tagging procedure and devices.
  5. Data summary (Chapter 4). From the 30 birds tracked during the 2019 breeding season, 2187.1 days of data were collected that were of sufficiently high quality for analysis. On average 78.8 ± 27.6 days of data were collected from each individual. The mean date of departure from the colony was 9th August, with the final tracked bird departing on 20th October. In the 2020 breeding season, 19 birds were tracked, providing 2085.1 days of analysable data, with an average of 128.0 ± 41.5 days per individual. The mean departure date was 6th August, with the final bird departing on 28th September.
  6. Trip statistics. In 2019, data were collected on 4,340 complete trips away from the colony. Birds had an average offshore foraging range of 31.5 ± 27.0 km, and an overall average foraging range (including onshore trips) of 12.4 ± 14.5 km, with trips covering an average total distance of 31.1 ± 47.6 km and lasting 5.2 ± 16.2 hours. In 2020, data were collected on 4,266 complete trips, with birds having an average offshore foraging range of 21.3 ± 19.1 km, and an overall average foraging range of 8.3 ± 9.8 km. Trips covered an average total distance of 19.5 ± 26.8 km and lasted on average 3.6 ± 5.4 hours.
  7. Connectivity. In 2019, 19 (63%) of the tracked individuals displayed some connectivity with operational offshore wind farms (OWFs), with 17 (57%) individuals showing connectivity with the GWF. The majority of interactions occurred with Galloper and Greater Gabbard Wind Farms, with only single individuals interacting with five other OWFs. In 2020, 11 (59%) individuals displayed connectivity with both the Galloper and Greater Gabbard Wind Farms, with only one individual displaying connectivity with another OWF.
  8. Area use. Area use was assessed through a Time‐In‐Area approach (TIA), which defines areas representing the birds’ 50% (core), 75%, 95% and 100% utilisation distributions (UDs). In 2019, 4.30% of the 95% UD calculated for all birds overlapped with OWFs, and 1.87% with the GWF. In 2020, the percentage overlap was much lower, with 0.98% of the 95% UD for all birds overlapping with OWFs, and just 0.33% with the GWF. Differences in trip statistics and the relative use of offshore areas and OWFs, including the GWF, may relate to the poorer breeding success seen at the Havergate colony in 2020, as Lesser Black‐backed Gulls tend to forage offshore more when provisioning chicks. In both 2019 and 2020, offshore area use and thus overlap with the GWF and other OWFs was greater during trips that began in the daytime than those that began at night.
  9. Behaviours within the GWF and other OWFs. Three methods (Hidden Markov Models (HMMs), Expectation Maximisation Binary Clustering (EMbC) and a Random Forest (RF) model) were used to understand behaviour based on data collected at a standard five minute sampling rate and at a fast‐sampling rate of 5‐10 seconds. Initial assessment showed a commuting corridor between the colony and the Galloper and Greater Gabbard Wind Farms, with offshore substations/service platforms within these sites being used regularly as resting/perching locations. Offshore foraging locations were identified between the colony and the GWF, within the GWF and beyond the GWF to the south‐east. A ‘foraging/searching’ and ‘floating’ area was also identified within the northern section of the GWF, next to the GWF offshore substation/service platform, where two turbine rows were more separated. Generally, most time spent within OWFs was spent ‘commuting’ or foraging/searching and less time was spent ‘stopped’ or floating, which was in contrast to other areas offshore outside OWFs, where generally birds spent most time floating and commuting. This pattern, however, varied between method of classification used, and year and also between the Galloper and Greater Gabbard Wind Farms.
  10. Altitudes within the GWF and other OWFs. Analysis of offshore altitudes was carried out using data acquired through both GPS and barometric pressure sensors and based on data collected at both five minute and 5‐10 second rates. Analyses considered (i) all fixes, whether birds were in flight or on the sea, and (ii) only those fixes when birds were in flight (foraging, commuting), as based on behavioural classifications produced by the EMbC models. Estimates of the altitudes and the proportions of fixes within the RSZs of the Galloper and Greater Gabbard Wind Farms, and altitudes in areas outside of OWFs, are presented for all periods, the day‐time and nighttime. Mean altitudes estimated from GPS and barometric pressure sensor data were similar, when based on data collected at a five minute resolution. Mean overall altitudes were 8.91 m and 10.06 m respectively within the GWF, with an estimated 17% of all fixes within the GWF being within the RSZ. In contrast, at the 5‐10 second resolution, mean altitudes estimated from GPS were consistently higher than those from pressure sensor data. Mean overall altitudes were 13.41 m and 8.54 m respectively within the GWF, with an estimated 23% of all fixes within the GWF being within the RSZ according to data from GPS, compared to 14% according to pressure sensor data. Altitudes were higher during the day than at night across both methods.
  11. Movements within the GWF and other OWFs. While Lesser Black‐backed Gulls showed little macro‐avoidance of the Galloper and Greater Gabbard Wind Farms, and there was evidence of attraction to structures within these sites, analyses of movements within the Galloper and Greater Gabbard Wind Farms suggested significant meso‐avoidance of the turbine rows. No GPS fixes were recorded in the turbine rotor swept zones and the distribution of observed fixes within the three‐dimensional space was significantly different from a random distributions of points.
  12. 2019/20 Non‐breeding season movements. Additional data were collected on the wintering destinations and migration strategies of Lesser Black‐backed Gulls over the 2019/20 nonbreeding season. Birds reached destinations up to 3800 km from the Havergate colony in Western Sahara and Mauritania, with other wintering destinations including Morocco, Spain, Portugal; only one bird spent the non‐breeding season in the UK.
  13. Comparison with historical datasets (Chapter 5). Comparative analyses of the foraging trips, wind farm connectivity, area use and altitudes of Lesser Black‐backed Gulls tracked during preconstruction studies undertaken between 2010 and 2015 (Thaxter et al., 2014b; RSPB, unpublished) at Havergate and Orford Ness are also presented. These results indicated that trip metrics, such as foraging distance offshore and trip duration, were similar in these earlier studies, albeit with variation between years and according to the duration of tag deployments (reflecting the different devices used). Further, while data collected in 2019 and 2020 provided evidence of attraction to structures within the Galloper and Greater Gabbard Wind Farms, the overall use of the GWF was also similar in these earlier studies.
  14. Conclusions (Chapter 6). The tracking study undertaken in 2019 and 2020 revealed that Lesser Black‐backed Gulls from the Alde‐Ore Estuary SPA showed significant use of both the Galloper and Greater Gabbard Wind Farms, although overall use of the GWF was similar to preconstruction studies. While Lesser Black‐backed Gulls showed little macro‐avoidance of these sites, and there was evidence of attraction to substations/service platforms, analyses of movements within the Galloper and Greater Gabbard Wind Farms suggested significant mesoavoidance of the turbine rows.