Citation
Thaxter, C.B., Clewley, G., Barber, L., Conway, G.J., Clark, N.A., Scragg, E.S. & Burton, N.H.K. 2017. Assessing habitat use of Herring Gulls in the Morecambe Bay SPA using GPS tracking devices. BTO Research Report 693: British Trust for Ornithlogy, Thetford
Overview
This report examines the movements and habitat use of Herring Gulls Larus argentatus breeding in the Morecombe Bay SPA.
In more detail
- Numbers of coastal Herring Gull populations have reduced markedly in recent years. The breeding gull colony of the South Walney and Piel Channel Flats Site of Special Scientific Interest (SSSI), within the Morecambe Bay and Duddon Estuary potential Special Protection Area (pSPA), is of national and international importance for its numbers of Herring Gull Larus argentatus and Lesser Black-backed Gull Larus fuscus. In recent years, however, the populations of both species at the site have declined significantly, and while this is considered to be largely a consequence of changes in the breeding colony itself, dependencies on habitats away from the colony may also play a role both in the observed decline and in the efficacy of proposed population recovery measures.
- To date, although we know that Herring Gulls are largely resident within the UK after breeding, for instance through ringing records, we have little detailed knowledge of habitat use. A key resource for this species could be mussel bed areas, for which there are several in the local Morecambe Bay area. Data on the movements of Herring Gulls have also been gathered for this site using GPS telemetry, giving a year-round perspective of movements of adult birds from the pSPA.
- We investigated the movements of 24 adult breeding Herring Gulls from the South Walney colony in pre-breeding, breeding, post-breeding and winter periods from the 2014 breeding season through to the 2015 post-breeding season. Delineation of these periods was based on colony-wide estimated mean hatching, fledging and laying dates. After GPS data were downloaded and cleaned, a Hidden Markov Modelling approach was used to identify three ‘states’ of activity: (1) resting stationary, (2) commuting, and (3) other foraging and resting behaviours, based on the turning angle and step length between consecutive GPS points. Given the variation in sampling rates inherent in this dataset, we investigated two sampling rates: (i) 10 minute and (ii) 60 minute, with the former providing a more refined picture of activity but potentially biased to times when GPS devices could sustain such rates. By contrast, the 60-minute dataset gave likely greater error in defining behavioural states but allowed assessment of habitat use through pre-breeding and post-breeding periods. During the winter, GPS rates could often sustain only one or two fixes per day, hence a dataset filtered to 18 hours (1080 minutes) was used to identify resting and commuting location based on step-length alone.
- To assess usage of areas of interest, spatial area utilisation distributions were computed using the activity states of resting and foraging, excluding commuting. Kernel density estimation (KDE) was used to identify core (50% KDE) and total (95% KDE) foraging/roosting areas. The overlaps of these distributions with mussel bed areas and constituent SSSI components of the SPA were then calculated for each bird and the total population. Similarly, we also assessed the time budgets of each bird and the time spent in these areas split by behavioural states through the year.
- A wide variation in habitat use was recorded among birds. Birds remained within the northwest region across the year, travelling no further south than the Mersey Estuary, and with some individuals making substantive use of the colony area even through the non-breeding period. Birds frequented intertidal mudflats, as well as terrestrial habitat such as fields, gravel workings, rubbish dumps and freshwater bodies. Some use of urban areas was seen, although the nearby Barrow-in-Furness was not extensively used. Mussel bed areas were also used, with some clear patterns in the foraging distributions indicating regular movements of the population to certain patches. Areas near to the colony were used most frequently. Birds made most use of the South Walney and Piel Channel Flats SSSI, which encompassed the breeding colony, but some use of other SSSIs within the SPA (with the exception of Roudsea Woods & Marshes SSSI) was also recorded. Substantial individual variation was apparent in the use of mussel beds. Based on mean values across individual birds, up to 11.6% of the 95% KDEs for individual birds overlapped with mussel beds, while up to 4.5% of time budgets were spent in this habitat (varying by period of the year). The largest overlaps with mussel beds were seen during breeding. Herring Gulls spent 50-60% of their time away from the colony during breeding and over 70% at other times of the year. Temporal overlaps with mussel bed areas, based on time spent engaged in foraging/resting behaviours, were up to 9.5% of birds’ time spent away from the colony (Appendix 1).
- Data from GPS telemetry provide a valuable resource to investigate the importance of particular habitats to species. The data collected here for Herring Gulls clearly indicate some use of local mussel beds throughout the year. The analyses carried out here are a useful first step in assessing the relative use of habitat; however, although indications are strong, it is not possible at present to firmly conclude whether mussel bed habitats were positively selected by Herring Gulls above other habitats in the region. Therefore, to further our understanding, more formal statistical assessment of the relative use of these habitats in relation to others should be conducted. For example, the use of resource-selection functions to determine habitat linkages would be a next logical step in the analyses of these data.
