Arctic Terns are the commonest tern breeding in Britain and
Ireland, but their northerly distribution means they are less
familiar to most observers than other tern species. The population
in Britain and Ireland is concentrated in the Northern Isles, with
73% occurring here. In common with other tern species, Arctic Terns
were probably reduced to low levels by hunting for the millinery
trade and egging, but have probably increased since the 1930s owing
to legal protection. Sandeel stocks in waters around Shetland
increased through the 1970s and early 1980s and this improvement in
food availability may have also contributed to Arctic Tern
population growth. However, a collapse of the sandeel stock around
Shetland between 1984 and 1990 resulted in a reversal in the
fortunes of the Northern Isles Arctic Tern population. This caused
repeated abandonment of eggs and starvation of chicks throughout
the archipelago and the population of Arctic Terns on the Northern
Isles declined between 1980 and 1989 in response to reduced
recruitment and increased mortality or non-breeding. Sandeel stocks
have recovered since 1991 and productivity improved, with evidence
of a population recovery between 1994 and 2000. However, since the
completion of Seabird 2000, a reduction in sandeel availability in
the western North Sea may have contributed to low
productivity of Arctic Terns in the Northern Isles,
particularly in Shetland
.
In western Scotland and the Western Isles, declines and
redistribution of the population have resulted from predation by
introduced American Mink Mustela vison. The population
trends in the future depend on the success of mink eradication and
control projects being implemented in these areas. Many Arctic Tern
colonies at the southern range of the British and Irish population
are increasing, probably in response to site management for
breeding terns. The population along the west coast of Ireland is
increasing at the expense of Common Terns, possibly due to changes
in climatic or oceanographic conditions shifting the balance of
inter-specific competition in favour of Arctic Terns.
Census Methods
Terns show a low degree of site faithfulness from one year to
the next, in response to predation or habitat change and especially
in areas where islands and other suitable habitat are plentiful and
often move en masse between different colonies. This is such a
great problem that in order to gain an accurate national estimate
of tern numbers, a simultaneous census was planned to cover all
colonies in Britain within a single year. Hence in 2000, the
majority of British tern colonies were surveyed including extensive
surveys of Orkney and Shetland, co-ordinated by RSPB and SNH. The
main exception was in the Western Isles where most tern colonies in
Lewis and Harris were surveyed in 1999, while those in the south of
the Sound of Harris to Barra Head were surveyed in 2002. In Ireland
it was decided by the Seabird 2000 Steering Group to make the most
of limited resources and utilise results from the All-Ireland tern
survey conducted in 1995 (Hannon et al. 1997). More recent
counts (with those from 2000 given priority) were included for some
colonies, including all those along the entire east coast.
Movements among these regions are unlikely to have caused severe
bias in trend estimation. During the SCR, counts were made in
different years within regions, and inter-colony movements may have
caused greater inaccuracies.
The proportion of surveys conducted using flush counts of
adult birds was far higher for Arctic Terns than for other tern
species. Arctic Terns breed mainly in the Northern and Western
Isles, where there are a huge number of widely spaced colonies.
These have to be counted within a single breeding season to avoid
biases due to birds moving colonies between years. Counts of Arctic
Tern AONs in the Northern and Western Isles are very time consuming
owing to the low densities at which birds nest, crypsis of their
nests against coastal heath and the lack of vantage points at many
colonies for counts of incubating birds. Therefore, comprehensive
counts within the incubation period of a single year are
impractical, and counts of individuals, that allow rapid coverage
of a large number of colonies within a season, are used instead.
Using counts of adult birds to census Arctic Terns is complicated
by the large variations in attendance among sites and years.
Attendance is strongly related to breeding success, with adults
tending to abandon the colony following breeding failure. Biased
estimates of status, trends and distribution could therefore result
from variations in breeding success among sites and years.
Abandonment of colonies in late incubation or chick rearing was
recorded in all years of census due to poor food availability,
inclement weather or tidal flooding, and this will have caused
underestimates of status. The degree to which breeding failure
reduced counts varied among years. In 1989 complete failure of
colonies occurred throughout Shetland, and consequently the
percentage declines since the 1980 survey were far greater for
those counts conducted late in the breeding season. During the 1994
survey, breeding success was better than in 1989 but was still low
owing to poor food availability, and during the 2000 survey a storm
washed away colonies on the west coast of Shetland in mid June.
This will lead to biases in estimation of trends that examine raw
counts of adults.
The biases associated with variations in attendance can be
overcome by calibration of flush counts against the number of nests
at a sample of colonies. Bullock and Gomersall (1981) estimated
that during 1980, 1.5 adults were present for every nesting pair
during the incubation period, and 1.1 during chick rearing. Counts
of birds were divided by 1.5 to calculate the number of pairs in
the Northern Isles and elsewhere for the SCR Census. During 2000,
calibration studies were conducted on Shetland and Orkney, and
produced season and year-specific correction factor of 1.35 (25 May
– 10 June), 0.88 (11 June – 28 June) and 0.61 (29 June – 15 July).
Adult counts were divided by the appropriate season-specific
correction factor in order to estimate the number of pairs on the
Northern Isles during Seabird 2000, and by 1.5 elsewhere. However,
attendance is still very variable among colonies within years, such
that estimates are subject to a large degree of error. Owing to the
large proportion of Arctic Terns censused as adults and variations
in attendance rates, trends for this species need to be viewed with
greater caution than for other tern species.
Further south in their range, Arctic Terns tend to nest in
discrete colonies on islets or sandy spits that are amenable to
being surveyed in units of AONs by counting nests or incubating
adults. Here, Arctic Terns usually nest sympatrically with Common
Terns and identification can present problems since the two species
look similar and their nests and contents cannot be separated
reliably. In mixed colonies of Arctic and Common Terns, counts of
incubating adults are preferable, but at some colonies vegetation
cover or lack of suitable vantage points make this method
impractical. In such cases, a count of the total number of nests
needs to be made, followed by an estimate of the proportion of
adults of each species circling over the colony. The number of
breeding pairs of Arctic Terns is calculated as the product of the
proportion of adult Arctic Terns and the number of nests of both
species. Species-specific differences in adult colony attendance
rates may lead to biases in such estimates. Alternatively,
observers will sometimes combine counts of the two species and
report them as "Commic" Terns rather than attempting to determine
the numbers of each. This makes estimation of the totals for each
species during Seabird 2000 problematic, but since "Commic" Terns
represent only 3 % and 1 % of the Arctic Tern population the SCR
and Seabird 2000 surveys respectively this is likely produce only
minor bias in the estimation of status and trends.
Survey Coverage
The Seabird 2000 census is likely to have included counts from
the vast majority of Arctic Tern colonies, with coverage being
comprehensive within their British and Irish range, with the
exception of western Ireland. This was rectified by inclusion of
data from the 1995 All Ireland Tern Survey (Hannon et al.
1997). A small number of colonies formed for the first time since
previous surveys may have been missed on Shetland, but these are
likely to represent only a small percentage of the total. During
the SCR, coverage of the Northern Isles and west coast of Ireland
was poor. This was overcome by inclusion of data from the 1980
survey of terns in Orkney and Shetland (Bullock & Gomersall
1981) and the 1983-84 All Ireland Tern Survey (Whilde 1985). The
variation in time-period over which trends are calculated among
regions will produce some bias in the overall national population
trend. There is debate concerning the degree to which coverage,
changes in methods and timing of surveys have contributed to the
changes in status since Operation Seafarer, and so changes over the
last 30 years should be treated with caution. A comprehensive
census of Arctic Terns on Orkney and Shetland was conducted in 1980
(Bullock and Gomersall 1981) that forms the most reliable baseline
for the region prior to the SCR.