Seabird 2000
Atlantic Puffin Fratercula
arctica
Introduction
The Atlantic Puffin is the most instantly
recognizable and popular of all North Atlantic seabirds. It breeds
in the North Atlantic and the adjacent Arctic Ocean, with
strongholds in Iceland and Norway, with around 10% in Britain and
Ireland, where it is the second most abundant breeding seabird.
Puffins are pelagic and we are still largely ignorant of how they
spend their time away from the colony. Atlantic Puffins from north
western Britain and Ireland disperse widely outside the breeding
season, as far as Newfoundland in the west and the Canary Islands
to the south and even into the Mediterranean as far east as Italy.
In contrast, most birds from eastern Britain remain within the
North Sea though in recent decades some have dispersed as far as
the Bay of Biscay.
Atlantic Puffins typically nest underground in
burrows dug in the soil of offshore islands, where such habitat is
in short supply, they nest among boulder screes or at low densities
in cracks in sheer cliffs rather like Black Guillemots or
Razorbills. Puffins are highly colonial and most colonies occur
where the nesting birds are safe from mammalian predators. However,
even in mid-summer, a colony can appear deserted during the middle
of the day since most birds are either down burrows or out at sea
feeding. At other times awe-inspiring numbers can be seen standing
on the slopes, bobbing around on the sea or flying in vast wheels
over the colony. Chicks are fed on small fish that the adult
carries cross-wise in its beak. In Britain and Ireland the
commonest prey is the lesser sandeel, followed by sprat, herring
and a wide range of small juvenile gadoid fish. Fish are caught by
underwater pursuit and a bird can catch several during a
dive.
Census Methods
Puffins are censused by counting apparently
occupied burrows or individual birds attending the colony area. The
former is the preferred method and during both Seabird 2000 and the
SCR Census (1985-88) most of the larger colonies were estimated in
this way. An apparently occupied burrow (AOB) is one with signs of
current use such as fresh digging, squirts of droppings radiating
from the entrance, hatched eggshells, broken vegetation or dropped
fish in the entrance. Puffin burrows can generally be distinguished
from rabbit burrows since the latter are usually larger and have
piles of soil and pellet-like droppings at the entrance. However,
there is no easy way to separate a burrow of a Manx Shearwater from
that of an Atlantic Puffin. Mixed colonies of these species are
relatively uncommon, but where the species coexist, for example in
southwest Britain and Ireland, tape playback of Manx Shearwater
calls has recently been used to distinguish burrows of the two
species. Counts of AOBs are best made before or during the laying
period, when birds are digging or cleaning out their burrows and
when the vegetation is short. However, acceptable counts can be
made at any time from late April to early August as long as the
vegetation does not obscure burrow entrances.
At some colonies such as those on the Farne
Islands and Coquet (both in Northumberland), the Isle of May (E.
Fife), North Rona (Western Isles) and Ynys Gwylan Fawr (Gwynedd)
all, or most, of the AOBs were counted and here the main source of
error would be the misclassification or overlooking of burrows (On
the Isle of May, this was found to result in an underestimate of
2%). For logistical reasons such complete coverage is impossible to
achieve at many colonies. In such situations the density of burrows
must be determined in sample plots, the area of the colony
estimated and the measures combined to get an estimated population
size. This approach is associated with some major sources of error.
The first is a statistical problem resulting from scaling-up to
estimate population size since this assumes that the sampled areas
were representative of the whole colony. An attempt was made to
assess the precision of the estimates for St Kilda and Sule Skerry
(Orkney), in the former there was a 95% chance that the true
population was within 12% of the estimate, in the latter the figure
was 20%. The second error is associated with determining the area
of the colony, either by direct measurement or by the use of aerial
photographs. To date no rigorous check of this error has been
attempted at any colony in Britain or Ireland.
Where birds nest under boulders (e.g The
Shiants, Western Isles), in mixed colonies with Manx Shearwaters
(Skomer and Skokholm, Dyfed), in completely inaccessible places
(Foula, Shetland) or at low densities along stretches of cliffs
(mainland colonies), counts of burrows are impractical. In these
cases there is no alternative but to count birds attending the
colony. Several different approaches were used during Seabird 2000.
On the Shiant Islands and at Hermaness (Shetland) counts of birds
were converted to AOBs by making simultaneous calibration counts of
birds ashore in areas where the number of AOBs was known. On Fair
Isle (Shetland) counts were converted to pairs (taken as equal to
AOBs) by concurrent observations of known numbers of colour-ringed
adults seen ashore. The ratio of birds to AOBs obtained on Fair
Isle was also used to convert counts of birds to AOBs on Foula
(Shetland). On Skomer and Skokholm, all adults present on land and
on the sea below the colony were counted on several evenings of
peak attendance early in the season, when only breeding individuals
would have been present, and the maximum count was taken. Where
time constraints prevented such a detailed approach, observers were
asked to make counts of birds on land, close inshore or in the air
during April and May, before substantial numbers of immatures begin
to attend the colony, preferably in the evenings or during foggy
conditions when maximum numbers of breeding adults are usually
present. The high variability of such counts both within and
between days, and the lack of any obvious factor influencing
attendance, means that such counts are of rather limited value in
assessing breeding numbers, but they do at least give some idea of
colony size.
In Seabird 2000 the majority of colonies were
surveyed during 1999-2002. However, the large colonies on Sule
Skerry (Orkney), Mingulay (Western Isles) and Isle of May (Harris
& Wanless 1998) were surveyed in 1998 and that on the Farne
Islands was surveyed in 2003. The Seabird 2000 population estimate
for Atlantic Puffins on Eilean Mor, the largest of the Flannan
Isles (Western Isles) consisted of a sample count of AOBs in the
densest part of the colony conducted in 1998, added to a complete
count of AOBs on the rest of the island conducted in 2001.
For the calculation of total populations, some
arbitrary decisions had to be made to allow the combination of
counts of individuals and AOBs. The previous practice of assuming
that one individual corresponded to one AOB (Cramp et al.
1974, Lloyd et al. 1991) was continued and applied to
counts from Seabird 2000, Operation Seafarer (1969-70) and the SCR
Census (1985-88). However, this approximation may well result in a
serious underestimate of the number of AOBs. On Skomer and
Skokholm, counts were made during known peaks of attendance prior
to egg laying, during both Seabird 2000 and the SCR Census. Counts
made at both colonies during both censuses were divided by 1.5
assuming that 75% of all breeding adults were present.
In both the SCR Census and Seabird 2000, 83% of
the total population estimates came from counts or estimates of
AOBs. In the SCR Census 65% of the counts of individual birds came
from the preferred counting months and in Seabird 2000 the figure
was 73%. The overall estimates of the two surveys should be broadly
comparable, however, detailed comparisons of numbers in the largest
colonies (Table 2) recorded in the SCR Census and Seabird 2000 were
restricted to colonies where counting methodologies was similar in
both censuses.