Náttúrufræðingurinn - 1979, Síða 96

scanty), and there are old breeding
records from the Sog and from Lón.
Outside tliese areas there are records
of passage migrants and occasional
wintering birds (Eyjafjördur, n. Ice-
land; Straumur and the Reykjavík
area, sw. fceland), and vagrants in a
few other localities.
Census results are summarized in
Table 3. Gensuses in May 1976 (total
1796) and Jan-Feb 1977 (2023) pre-
sumably accounted for the entire po-
pulation. The males moulted on the
Mývatn-Laxá where they were censu-
sed in 1975 (1022), 1976 (1026) and
1978 (1273). Census figures and ob-
served age ratios in spring indicated
a mean apparent annual survival rate
of 0.94 for full-grown males. The
real survival rate may well be lower,
if some juveniles migrate out o[ Ice-
land and return as adults. It is more
diflicult to estimate fernale survival
as only two complete censuses were
made (May 1976, Jan-Feb 1977) and
censuses restricted to the Mývatn —
upper Laxá may be biased towards
females, making observed sex ratios
doubtfully applicable. A niean appa-
rent annual survival rate ol' 0.88 is
suggested for the females, again this
may be liigher than real survival if
immigration/emmigration occurs. A
model of the Icelandic population in
April 1975—78 (Table 4) suggests a
total population of about 1800 in
spring 1975 and 1976, about 2000 in
1977 and about 2100 in 1978. Ghanges
in the adult population were closely
correlated with the nuntbers of ju-
veniles in the preceding spring in
males, but not in females, suggesting
that migration or dispersion was im-
portant in regulating female numbers
on the breeding grounds.
The agc of first breeding in the wild
is not known. l’robably no males breed
in their 2nd yr and certainly at least
some 2nd yr females do not lay. The
number of females older than 2 yr
(i. e. the potential breeding popula-
tion) was estimated at about 550—570
each spring 1975—77 and perhaps up
to 640 in 1978.
6. l’rocluction of young to the fledging
stage (Table 5) was estimated from
young/fentale ratios in August 1976—
78. The young occurred mainly on
the upper Laxá, where tliey could
all be censused, and on Lake Mývatn
where census efficiency was presuma-
bly much lower. The proportion of
young on the lake and river varied be-
iween years and was associated witlt
changes in the food supply in these
two habitats. Age ratios of Barrow’s
goldeneye were about 1.9 young/9 in
1976 and 1977 and about 1.0 in 1978.
Age ratios of liarlequin duck (Hislri-
onicus histrionicus) varied similarly:
1.7 (1975), 2.4. (1976), 1.3 (1977) and
0.6 (1978). Low production in both
species in 1978 was associated wilh
a reduction in the blackfly (Simulium
vittatum) populalion of the river
(Table 2). Food conditions in Mý-
vatn were unusually poor in 1975
and 1976 and then improved greatly.
This was reflected in production of
diving ducks on the lake; the age
ratios of scaup (AytHya marila) and
tufted duck (A. fuligula) combined
were: 0.5 (1975), 2.0 (1976), 3.6 (1977)
and about 3 (1978). 'l'hus food and
not weather probably determined the
production of young of these íour
species. Production in the Barrow’s
goldeneye may be more even between
years than in the other species because
of its ability to use the resources of
Lake Mývatn when those of the Laxá
fail and vice versa.
7. Prior estimates of the Barrow’s golden-
eye population of the Mývatn region
are briefly reviewed, these are more
fully dealt with elsewhere (Gardars-
son 1979, Gudmundsson 1979) This
population has remained approxi-
mately stable since 1960 and probably
since abput 1920. During 1900—15 it
may have been larger, at least at the
lake itself. Mass mortality of duck-
190