Sandwich Terns exhibit the most erratic population trends and
distribution of any seabird breeding in Britain and Ireland. The
population fluctuates dramatically among years owing to large
variations in the proportion of mature birds attempting to breed
and distribution varies owing to mass movements of birds among
colony sites.
The population is distributed widely but patchily around the
coasts of Britain and Ireland and broadly reflects the availability
of Sandwich Tern nesting habitat: low-lying offshore islands,
islets in bays or brackish lagoons, spits or remote mainland dunes.
Despite frequent changes in the colony sites used, the broad
distribution of Sandwich Terns in Britain and Ireland has remained
remarkably similar over the last 30 years. Sandwich Terns are among
the most strongly gregarious of all seabirds, with the population
being distributed in a small number of relatively large colonies in
which birds nest at extremely high densities.
Tern populations in NW Europe were bought to the brink of
extirpation at the end of the nineteenth century by egg collection
for food and hunting of adults for the millinery trade, but
recovered in response to protective legislation in the early
twentieth century. The Sandwich Tern population in Britain and
Ireland increased from the 1920s to the mid 1980s, with protection
from increasing recreational disturbance on beaches as well as from
persecution probably facilitating this recovery. Annual counts of
the main colonies demonstrated that there was a sustained increase
between the two surveys, but that the population fluctuated
erratically around this trend. The population declined between the
SCR census and Seabird 2000, with annual monitoring showing a
stepped decline linked to discrete events at individual colonies
during the late 1980s and early 1990s. The population now appears
to be in a recovery phase, and has returned to levels documented
during the late 1970s and early 1980s.
Survey coverage
Coverage of Sandwich Tern colonies was comprehensive within
their British and Irish range during all three of the national
surveys, and so long-term trends will be real rather than artefacts
of survey coverage.
Sandwich Terns are highly nomadic even compared with other
tern species, and whole colonies may move site within a year or two
in response to changing conditions. Such movements have the
potential to produce severe bias in national population estimates
that rely on summing counts from colonies surveyed in different
years. To minimise such bias, all Sandwich Tern colonies in Britain
and Northern Ireland were surveyed in 2000. In the Republic of
Ireland, the colony in Wexford was surveyed in 1999 and all other
counts were taken from the 1995 All Ireland Tern Survey (Hannon
et al. 1997), and so movements among these regions could
bias estimation of trends and status. Inspection of annual
monitoring data for sites in the Irish Sea suggests that no mass
movemant occurred that would cause severe bias of the population
trends. Furthermore, the proportion of pairs nesting in the
Republic of Ireland in both surveys is relatively small, and so
biases arising from movements between here and the UK will be
slight.
During the SCR census, counts of colonies within regions were
often taken from different years, and counts for the West coast of
Ireland were taken from the 1983 All Ireland Tern Survey (Whilde
1985). As such, some movements among sites probably occurred such
that some pairs may have been double counted and others omitted
altogether.
Census Methods
Sandwich Terns are generally counted in units of Apparently
Occupied Nests (AONs) using transect counts of nests or counts of
incubating adults from a vantage point. The high density of nesting
Sandwich Terns and their irregular distribution within large, dense
colonies can make it very hard to keep track of which birds have
been counted, and so counts of incubating adults tend to be less
accurate than transect counts of nests. During the Seabird 2000
census, equal numbers of pairs were surveyed using each method.
Alternatively, during Seabird 2000 and the SCR Census, 1% and 2% of
pairs respecitively (all of these were in Orkney) were surveyed
using flush counts of adult birds. This method is less accurate
than AON counts, but since flush counts represent such a small
proportion of records this has little effect on the accuracy of
national population estimates and trends.
Counts of Sandwich Tern colonies need to be conducted between the
peak of laying and the start of hatching. Sandwich Tern laying
phenology is highly synchronised within sites and years and,
provided predation or tidal flooding does not cause significant
losses, there is a period of a week or two when the vast majority
of AONs are available for survey. This usually occurs in early to
mid June and this was the recommended count period for sites where
only a single survey could be made. However, since laying phenology
varies among years, surveyors were encouraged to make repeated
counts from mid May to late June and report the highest of these.
During Seabird 2000, 91% of AONs were surveyed within the specified
period. At the remainder of sites, an accurate date was not
reported. During the SCR survey, the census date was reported only
in 47% of colonies, and some small colonies in Orkney were surveyed
in early July. However, it is unlikely that differences in the
timing of counts among surveys will be sufficiently severe to
affect the estimation of trends.
The size of the Sandwich Tern breeding population fluctuates
erratically from year to year. Trends based on comparison of two
widely spaced surveys must therefore be viewed with caution, since
one of them may have coincided with a year of temporarily depressed
population size. Fortunately, all the key colonies in the UK and
Ireland have been surveyed
annually since 1969, allowing long-term
population trends to be discriminated from background levels of
fluctuation with greater confidence.