An Experimental Study of the Costs of Reproduction in the Kittiwake Rissa Tridactyla

Author(s): Karl-Otto Jacobsen, Kjell Einar Erikstad and Bernt-Erik Sæther

Source: Ecology, Vol. 76, No. 5 (Jul., 1995), pp. 1636-1642
Published by: Wiley on behalf of the Ecological Society of America
Stable URL: http://www.jstor.org/stable/1938164
Accessed: 14-11-2017 18:51 UTC

REFERENCES
Linked references are available on JSTOR for this article:
http://www.jstor.org/stable/1938164?seq=1&cid=pdf-reference#references_tab_contents
You may need to log in to JSTOR to access the linked references.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
http://about.jstor.org/terms

Wiley, Ecological Society of America are collaborating with JSTOR to digitize, preserve and
extend access to Ecology

This content downloaded from 134.117.10.200 on Tue, 14 Nov 2017 18:51:16 UTC
All use subject to http://about.jstor.org/terms
Ecology, 76(5), 1995, pp. 1636-1642
© 1995 by the Ecological Society of America

AN EXPERIMENTAL STUDY OF THE COSTS OF

REPRODUCTION IN THE KITTIWAKE
RISSA TRIDACTYLA1

KARL-OTTO JACOBSEN AND KJELL EINAR ERIKSTAD2

Norwegian Institute for Nature Research, Troms0 Museum, University of Troms0,
N-9005 Troms0, Norway, and
Department of Ecology/Zoology, Institute of Biology and Geology,
University of Tromso, N-9037 Tromso, Norway

BERNT-ERIK SAETHER
Norwegian Institute for Nature Research, Tungasletta 2, N-7005 Trondheim, Norway

Abstract. The costs of reproduction in Kittiwakes Rissa tridactyla were studied from
1990 to 1993 on Horn0y in northern Norway by enlarging and decreasing brood sizes of
two chicks to three and one during the first week after hatching. No parents were able to
raise the enlarged broods of three to fledging. Most chick mortality occurred in the 19-22
d age group. Chick body mass and fledging success were lower among enlarged broods
than among reduced and control broods. Body mass of females, but not males, was lower
at the end of the chick-rearing period in enlarged broods. Only those females rearing
enlarged broods experienced a higher mortality during the next breeding season. Our results
support Lack's hypothesis of clutch size regulation: Kittiwakes laid the maximal number
of eggs they were able to raise successfully to fledging. Furthermore, as female Kittiwake
mortality increased when clutches were enlarged, females have a trade-off between in-
vestment in current reproduction and investment in future reproduction. Because males and
females responded differently to brood enlargement, fitness estimates showed that Kitti-
wakes may have a sex-specific optimal clutch size.
Key words: brood size manipulation; clutch size; cost of reproduction; Kittiwake; life history;
Rissa tridactyla.

INTRODUCTION Dijkstra et al. (1990), the parents were able to raise

more nestlings than normal through the addition of eggs
Optimal life history theory is based on the assump-
or chicks. Thus, the most common clutch size was
tion of trade-offs between different life history traits
smaller than the most productive, so the experiments
(Stearns 1992). Lack (1947, 1954) was among the first
did not support Lack's hypothesis. Williams (1966) ar-
to point out the existence of a negative covariation
between life history variables, so that increased gued
in- that the effects outside the breeding season also
could influence the optimal clutch size, because current
vestment in one life history component usually in-
parental effort should be balanced against the risk of
volves decreased investment in another. He suggested
future survival. This trade-off would result in a smaller
that avian clutch sizes have evolved to maximize fitness
of the parents. Large clutches may be selected against than expected optimal clutch size if the total breeding
because offspring size, and thereby offspring survival, history were taken into account (Charnov and Krebs
would be reduced, resulting in fewer recruits to 1974). the Thus, the optimal clutch size will be the result
next generation than from the optimal intermediate of a balance of investment in current reproduction
clutch size (the Lackian clutch size). Thus, this hy- against investment in future survival or reproduction.
pothesis assumes a negative relationship between Such a balance in the allocation of resources may affect
clutch size and the fitness of each offspring (Godfray long-lived species particularly, where a negative im-
1987). pact of current reproduction on future survival will
However, a general observation of avian clutch size have greater consequences for the optimal clutch size
studies has been that the median clutch size is less than (Goodman 1974, Charlesworth 1980, Wooller et al.
the most productive size. For instance, in 40 of 53 1992).
studies reviewed by Linden and M0ller (1989) and Even though a large number of clutch size manip-
ulations have been made, most of them involve short-
' Manuscript received 16 February 1994; revised 14 No- lived passerines (Linden and M0ller 1989, Roff 1992,
vember 1994; accepted 20 November 1994; final version re- Stearns 1992). In these species, the probability that an
ceived 12 December 1994.
individual will survive between two breeding seasons
2 Address correspondence to: Kjell Einar Erikstad, NINA,
Troms0 Museum, University of Troms0, N-9005 often is so low that a large investment in the current
Troms0,
Norway. breeding attempt will be expected (Charlesworth

This content downloaded from 134.117.10.200 on Tue, 14 Nov 2017 18:51:16 UTC
All use subject to http://about.jstor.org/terms
July 1995 COST OF REPRODUCTION IN THE KITTIWAKE 1637

1980). Furthermore, few manipulative studies include 1- Lost one or more chicks * Chicks fledged

data on effects of the manipulation on body condition

100' (22)
of the adults, which may impact optimal clutch size
(16)
(Price and Liou 1989). o

The clutch size of the Kittiwake Rissa tridactyla var- -

6 60
ies from one to three eggs; however, in northern Nor-
way >90% of the nests contain only two eggs = 40(Barrett (32)

1978; T Tveraa, unpublished data). The purpose of our (66)

20 -.
study was to examine experimentally whether two eggs
AM

represent the optimal clutch size according to Lack's .

hypothesis, or whether a survival cost of increased Reduced re- Control Enlarged

productive investment is important in determining
clutch size. According to Lack's hypothesis,chicks
expect that by removing or adding an egg control,
clutches, fewer offspring should be fledged
FIG. 1. Percentage of Kittiwake nests losing one or more
we wouldand percentage of chicks that fledged from reduced,
to two-egg and enlarged broods. Sample size is given above each
bar.in manip-

ulated nests than in control nests. Alternatively, if a

survival cost of reproductive investment is important
age, Kittiwake chicks in normal broods show a linear
for clutch size determination, we would expect higher
mortality among parents with artificially-enlarged and rapid growth rate (Barrett 1978).
Adults were caught on the nest with a noose on the
clutches, thus reducing their contribution of individuals
to the next generation relative to the controls. end of a 5 m long fishing rod and were sexed using
the discriminant function described by Barrett et al.
STUDY AREA (1985). At Horn0y in Norway, 87.3% of adult Kitti-
The study was carried out at Horn0y, Finnmark are correctly classified using this method (Bar-
wakes
(70022' N, 31°10' E), in northern Norway in rett 1990 et al. 1985). We measured head length (from the
(May-July) and 1991-1993 (May-June). Horn0yback is aof the head to the tip of the bill parallel to the
cutting edge of the lower bill; Coulson et al. 1983) to
sandstone and shale island with steep cliff faces on the
the
west coast, particularly suitable for cliff-nesting sea- nearest 0.1 mm using a specially adapted vernier
birds. The size of the island is =1000 X 700 m. caliper.
ManyFor most of the pairs, one of the mates was
species of seabirds breed on Horn0y, and several dyedpop- using picric acid on the back of its head, making
ulations have shown recent increases. The Kittiwake it easier to identify individuals during catching.
population increased from 9000 pairs in 1974 to 21 000
Like many other birds (Moreno 1989), Kittiwakes
pairs in 1983 (Furness and Barrett 1985). Since lose1983,
body mass during the breeding season (Barrett et
al. 1985). To examine the possible effect of brood-size
the population has been relatively stable (R. T. Barrett,
personal communication). manipulation, we weighed adults after they had spent
-12 d rearing chicks. They were weighed to the nearest
METHODS 1 g using a spring balance.
To standardize the effect of brood manipulation, only
Sixty-seven nests were used in the 1990 experiment.
Broods at 22 nests were enlarged from two to three broods that maintained their manipulated or control
chicks, 29 broods were reduced from two to one chick,>12 d were used in the analysis.
size for
and 16 broods were assigned as controls. During the 1991-1993 breeding seasons, work was
concentrated on the return rate and mortality of adults
Brood sizes were manipulated during the 1st wk after
from the different experimental groups. Searches were
hatching. We did not know the sex of the chicks, but
the size of the adopted chick was always close madeto,for orcolor-banded individuals in the colony in
between, the sizes of its new siblings. This was done and July, using binoculars and telescopes.
May, June,
Statistical tests for all analyses were carried out using
to minimize any problem with dominance among the
siblings. There is no parent-offspring recognition the SAS package
in (SAS Institute 1990). P values < 0.05
were
Kittiwakes, and parents are willing to accept any chick
considered significant. If nothing else is stated,
all
they find in their nest (Storey et al. 1992). To identify
P values are one-tailed. All values are expressed as
mean _ 1 SE.
the individuals in multiple-chick broods, we marked
the chicks with colored tape on one leg until they were RESULTS
large enough to be banded with one metal and one color
ring. Barriers were built between some nests to avoid Chick loss and fledging success
mixing of neighboring chicks when nests wereThere ap- was considerable chick loss during the nestling
proached for handling chicks and adults. period (Fig. 1). Many chicks were preyed upon by Her-
The nests were visited every 3rd d to estimate ring Gulls Larus argentatus, Great Black-backed Gulls
chick
mortality. Chicks 19-22 d post-hatching were weighedL. marinus, and Common Ravens Corvus corax, but
to the nearest 1 g using a spring balance. Until
siblicide
this and/or starvation was probably the cause of

This content downloaded from 134.117.10.200 on Tue, 14 Nov 2017 18:51:16 UTC
All use subject to http://about.jstor.org/terms
1638 KARL-OTTO JACOBSEN ET AL. Ecology, Vol. 76, No. 5

l Reduced (29)
50

H Control (32)
40
1 Enlarged (66)
-

(18)
? 30 FIG. 2. Chick loss in relation to age of
0
chicks in reduced, control, and enlarged Kit-
.4 20 (3) tiwake broods. Sample size is given above
(2) each bar.
o
10
(4) (2)(2)
(0 1) ( (0) (1) (1 31(2)36
0
m . - 1 _
0 -
0
......... l _
I -

0-6 7-12 13-18 19-24 25-30 31 - 36

Days after hatching of first chick

death for those chicks found dead in or under the nests. number of days after hatching when brood size was
manipulated (F = 2.62, df = 1, P > 0.05; PROC GLM
The frequency of nests which lost one chick was similar
among reduced (32%, n = 29), control (33%, n = 16), based on type III sum of squares). We also performed
and enlarged broods (21%, n = 22); (X2 = 1.11, df =
pairwise comparisons of least square estimates for all
2, P > 0.05). The percentage of nests that lost one three
or groups (reduced vs. control, P = 0.002, enlarged
more chicks was lower among reduced broods (47.1%) vs. control, P = 0.78; enlarged vs. reduced, P = 0.02).
than among control (80%) and enlarged (100%) broods Furthermore, chicks that survived to fledging were
heavier at the age of 19-22 d (321.7 ± 9.0 g, n = 44)
(X2 = 12.9, df = 2, P < 0.001; Fig. 1). The percentage
of chicks fledging decreased from 72.4% among re- than those that died before fledging (275.8 ± 16.3 g,
duced broods to 22.7% among enlarged broods (X2 n= = 12); (t = 12.9, P < 0.001).
21.1, df = 2, P < 0.001; Fig. 1). Body mass of adults
During the first 12 d after hatching, few chicks were
lost, but chick loss increased in the second half of the
Females rearing enlarged broods had a lower body
mass (356.3 ± 3.8 g) late in the chick-rearing period
nestling period, peaking at 19-24 d for all groups (Fig.
compared with those rearing control (370.6 ± 5.3 g)
2). However, there was no significant difference among
and reduced (391.3 ± 6.9 g) broods, but there was no
the different groups in when chick loss occurred (Ko-
similar trend for males (Fig. 4). We tested the differ-
lomogorov-Smirnov, Dmax = 9.8, P = 0.08).
ences first by performing a covariance analysis (PROC
Nests in all three groups produced, on average, the
GLM) relating body mass to category and time after
same number of young (ranging from 0.63 among en-
day 12 in the chick-rearing period when the adults were
larged to 0.68 among reduced broods (Kruskal Wallis
weighed. For females, there was a significant effect of
test: x2 = 0.7, df = 2, P > 0.05; Table 1). When those
treatment (F = 3.92, df = 2, P = 0.003), but not of
nests that lost all young were excluded from the anal-
time (F = 0.13, df = 1, P = 0.72). For males, there
ysis, the mean brood size at fledging was higher among
was neither a treatment effect (F = 1.56, df = 2, P =
control (1.4) and enlarged (1.5) than among reduced
0.23) nor time effect (F = 0.00, df = 1, P = 0.89).
(1.0) broods (Kruskal Wallis test: x2 = 11.6, df = 2,
For females, we also performed pairwise comparisons
P < 0.003; Table 1.
of least square estimates (PROC GLM) for all three
Chick growth groups (reduced vs. control, P = 0.03; control vs. en-
larged, P = 0.18; reduced vs. enlarged, P = 0.01). Thus
At the age of 19-22 d, chicks in reduced broods were
females with enlarged clutch sizes lost more mass dur-
heavier (339.0 ± 11.0 g) than control (312.0 + 11.0
ing the fledging period than other females.
g) and enlarged (276.9 ± 16.9 g) broods (F = 5.44,
Adult survival
df = 2, P < 0.007; Fig. 3). These differences were
significant even after controlling for the day when Females rearing enlarged broods had a lower prob-
chicks were weighed (F = 0.73, df = 1, P > 0.1), and ability of surviving (55.6%) than did those rearing con

TABLE 1. Mean (+ 1 SE) number of fledged young in experimentally reduced (from two chicks to one), control (t
and enlarged (from two chicks to three) Kittiwake broods. n = number of broods.

Reduced Control Enlarged

Mean SE n Mean SE n Mean SE n

All nests 0.68 0.09 31 0.67 0.18 18 0.63 0.17 24

Nests that fledged young 1.0 21 1.4 0.10 9 1.5 0.15 10

This content downloaded from 134.117.10.200 on Tue, 14 Nov 2017 18:51:16 UTC
All use subject to http://about.jstor.org/terms
 July 1995 COST OF REPRODUCTION IN THE KITTIWAKE 1639

400
4" Males * Females

(20)
350
100 (8) (25)
I (13)
(7)
80

W 300' (16)
f ' 60 (9)
.!
o 250- t 40
m

20
200 .
Reduced Control Enlarged 0 - --

Reduced Control Enlarged

FIG. 3. Mean + 1 SE body mass (grams) of chicks in
reduced, control, and enlarged broods at ages of 19-22 d. FIG. 5. Survival (%) of adult male and female Kittiwakes
Sample size is given above each bar. to the following breeding season after rearing reduced, con-
trol, and enlarged broods.

trol (78.6%) and reduced (92.0%) broods (Fig. 5;

PROC CATMOD with binary response. Maximum a single clutch. It has become customary to refer to
likelihood estimates: x2 = 5.0, df = 2, P = 0.04). For
this clutch size as "Lack's solution clutch size" (Char-
males, there was no similar trend (X2 = 0.65, df = 2,
nov and Skinner 1984, Godfray 1987). This hypothesis
P = 0.36).
was later extended by Perrins and Moss (1975) to in-
Contribution to the next generation corporate the idea that a female should lay the clutch
size that maximizes the number of new recruits to the
The number of individuals recruited to the popula-
tion in the next generation can be used as an index population.
of This optimal clutch size would depend on
fitness of the parents in the different experimental the female's and the male's ability to raise offspring to
groups. When all nests are included, females in theindependence (H6gstedt 1980, Saether 1990, Pettifor
reduced group had the highest fitness, whereas the low- 1993a). A prediction from this hypothesis is that, when
est contribution was found among females in the en- clutch size is experimentally altered, the number
the
larged group (Table 2). However, when only pairs that of fledglings or recruits produced should decrease. In
were able to produce a fledgling were included, the the present study, when all nests were included, fledg-
control females were most successful. For clutch sizes ling production was very similar in the control and the
larger than one, male fitness was, in general, higher two manipulated groups (Table 1). However, when only
than that of females, whereas the reverse trend was the pairs that managed to rear a fledgling were consid-
found in pairs with reduced clutch size. Among males, ered, removal of one egg significantly decreased the
the controls were most successful (Table 2). Thus, thenumber of fledglings produced. In contrast, only a small
manipulation had a different effect on fitness in the two difference was found between the enlarged and control
sexes due to the sex-specific difference in survival (Fig.group in the number of fledglings per successful pair.
5). These patterns show that a large component of the vari-
ance in reproductive success among Kittiwakes at Hor-
DISCUSSION
n0ya is due to a complete loss of fledglings, which is
Productivity in relation to clutch size independent of clutch size. If a pair managed to raise
Lack (1947) first suggested that selection may anact
offspring to independence, the fledging success of
upon birds to maximize the number of fledglings one-egg from clutches was lower than that of two- or three-

500' * Males
TABLE 2. The contribution of individuals to the next gen-
o Females
(3) eration by adult Kittiwakes in the three experimental groups
450 (8) (see Table 1) in relation to sex (M, male; F, female). Fitness
(15)
(W) is computed as W = N-Pj.S + PA, where N is the number
of offspring produced, S is the sex ratio (S = 0.5), and Pj
(19)
r 400 is the juvenile survival rate 1 yr after fledging. PA is adult
350
(9) (4)
survival. We assume Pj = 0.79 (Coulson and White 1959),
1 independent of experimental treatment and sex.
o 350- t
Adult fitness (W)
300 Re- En-

Reduced Control Enlarged Sex duced Control larged

All nests M 1.15 1.22 1.11
FIG. 4. Mean (+ 1 SE body mass) of adult male and F 1.19 1.06 0.81
female Kittiwakes late in the chick-rearing period in reduced,
Nests that fledged young M 1.28 1.50 1.45
control, and enlarged broods. Sample size is given above each F 1.32 1.34 1.15
bar.

This content downloaded from 134.117.10.200 on Tue, 14 Nov 2017 18:51:16 UTC
All use subject to http://about.jstor.org/terms
1640 KARL-OTTO JACOBSEN ET AL. Ecology, Vol. 76, No. 5

egg clutches. Even though the chick lossthe was return

higher rate of individual offspring, body mass at
in enlarged clutches than in the other groups fledging was2significantly lower for chicks in the en-
(Figs.
and 3), the fledging rate per successful pair larged than
was in the control group. Such a trade-off be-
still
as high as in the controls. Thus, Lack'stween hypothesis
the number and quality of the offspring has pre-
(1947) may explain why there are so fewviously been recorded in the Kittiwake (Coulson in
Kittiwakes
laying only one egg, but not why the frequency Lackof 1966), as well as in a great number of other stud-
three-
egg clutches is so low. ies (Stearns 1992). From other Kittiwake studies, it ha
An ability to raise extra fledglings hasbeen alsoshown
been that the probability of recruitment to th
documented in other studies of gulls. A small breeding population is dependent on the size of the
number
(n = 4) of Kittiwakes manipulated by Coulson chick, (aswhich
citedis dependent on its hatching sequence
in Lack 1966) was able to raise even four (Coulson offspring. and Porter 1985). A third chick in the clutch
Other gull species have also raised artificially has aenlarged
low growth rate and a small chance of subse-
broods (Glaucous-winged Gull Larus glaucescens, quently beingVer-recorded in the colony.
meer 1963, Ward 1973; Swallow-tailed Gull Creagus If low fledgling mass has an effect on offspring fit-
furcatus, Harris 1970; Herring Gull, Haymes and Mor- ness, as recorded by Coulson and Porter (1985), the
ris 1977). It should be noted, however, that in two of small proportion of three-egg clutches can be explained
these studies (Vermeer 1963, Haymes and Morris by the extended version of Lack's hypothesis (Perrins
1977), the abundance of food was artificially high dueand Moss 1975, Godfray 1987), which includes the
to supplemental feeding or access to human garbage. effect of offspring quality on their chances for future
Increased losses in the enlarged group may be relatedrecruitment to the population. This is in accordance
to changes in the activity budgets of the parents, mak- with findings in studies of both Great Tit Parus major
ing the contents of the nest more vulnerable to pre- (Pettifor et al. 1988) and Blue Tit Parus caeruleus (Pet
dation by Herring Gulls, Great Black-backed Gulls, andtifor 1993a), where the number of surviving offspring
Common Ravens. Seabirds are able to change their was highest in the control broods.
feeding behavior in response to food stress during the
Effects on the parents
breeding season (Furness and Ainley 1984, Burger and
Piatt 1990). For instance, several studies of Kittiwake As originally pointed out by Williams (1966), Lack's
have demonstrated that both parents forage simulta- hypothesis does not incorporate the effects on adult
neously in years with limited food resources (Barrettsurvival. If there is a relationship between the numbe
1978, Barrett and Runde 1980, Galbraith 1983, Hamer of offspring raised and the future life expectancy of
et al. 1993). Normally, parental nest attendance is 100%parents, a smaller number of eggs may be laid than
during the nestling period, but Kittiwake nest atten-predicted by Lack's solution clutch size (Charnov and
dance dropped to 35% in a year with low chick survivalKrebs 1974, Partridge 1989). However, such a survival
on Runde, a colony in western Norway (Barrett and cost has been difficult to document in birds (Lessells
Runde 1980). Larger broods were left unattended more1991, Roff 1992, Pettifor 1993b), particularly in small
often than smaller broods; larger broods may also bepasserines (Gustaffson and Sutherland 1988, Pettifor
more vulnerable to predation, as the parents are not1993b). In the present study, we were able to dem-
able to cover all the chicks properly during attacks byonstrate reduced return rate of females in the enlarged
gulls (Galbraith 1983). However, in our study, some ofgroup (Fig. 5).
the chicks lost were found dead under the nest and This effect can be explained by two mechanisms. (1)
most probably had been pushed out by larger siblings. The reduced return rate can be due to the increased

Siblicide is common in periods of food shortage investment

or needed for rearing one additional chi
during bad weather (Cullen 1957, Braun and Hunt This increased parental effort may result in a high
1983), and may be an efficient mechanism by which risk of dying after the breeding season. According
to adjust brood size to food resources during the nest- females with an enlarged brood size were in poor
ling period (Lack 1954, Howe 1976). In this study, the body condition at the end of the nesting season th
lightest chicks died. The cause was most probably a were females in the other groups. (2) As a conseque
combination of low food availability, low nest atten- of the increased investment or the relatively hig
dance, predation, and siblicide. chick loss, a higher proportion of females with enlar
broods may have changed their nesting sites the f
Effects on offspring
lowing year. This explanation is less likely, howev
The Lack's solution clutch size maximizes the fitness because most parts of Horn0ya and nearby islands w
return per clutch for the parent (Charnov and Skinner carefully searched for banded Kittiwakes in the th
1984, Godfray 1987). An optimal clutch size lower following
than breeding seasons. It is unlikely that lar
the maximal number of offspring the parents can numbers
suc- of Kittiwakes have remained undiscovered.
cessfully raise is found when the fitness per offspring Thus, we suggest that the reduced return rate amo
is negatively related to the total number of offspring females with enlarged brood size does represent a m
produced by the parents. Although we have no data talityoncost of increased breeding effort. A similar ef

This content downloaded from 134.117.10.200 on Tue, 14 Nov 2017 18:51:16 UTC
All use subject to http://about.jstor.org/terms
July 1995 COST OF REPRODUCTION IN THE KITTIWAKE 1641

of clutch size manipulation was also foundcreasing

in the the clutch size is, however, most likely t
Glau-
cous-winged Gull (Reid 1987). Furthermore, several
the major mechanism counteracting larger clutch
studies of gulls raising unmanipulated broods have
in the population.
shown that survival is inversely related to the number
ACKNOWLEDGMENTS
of fledglings produced, due to a higher reproductive
The study
effort in large clutches (Wooller and Coulson was supported by the Norwegian Council fo
1977,
Science and Humanities and the Directorate for Nature Man-
Pugesek 1981, 1983, 1987, Pugesek and Diem 1990).
agement. We are very grateful for comments by B. Pugesek,
In general, survival costs of clutch size enlargements
J. Swenson, T. Tveraa, and an anonymous reviewer, which
have been difficult to document, at least within the greatly improved the manuscript.
normal range of clutch size variation (Linden and M011-
LITERATURE CITED
er 1989). We suggest that reason for this is that such
inverse relationships between reproductive investment Barrett, R. T 1978. The breeding biology of the Kittiw
and survival are most likely to appear in species with Rissa tridactyla (L.) in Troms, North Norway. Thesis.
Troms0 Museum, University of Troms0, Troms0, Norw
intermediate (60-85%) adult survival rates. In small
Barrett, R. T., R. Fieler, T. Anker-Nilssen, and F Rikardsen.
temperate passerines, the adult survival rate is low 1985. Measurements and weight changes of Norwegian
(Saether 1989); any costs of reproductive investment adult Puffins Fratercula arctica and Kittiwakes Rissa tri-
will be difficult to record (Nur 1988) and are not likely dactyla during the breeding season. Ringing and Migratio
6:102-112.
to have any significant impact on variation among in-
Barrett, R. T., and 0. J. Runde. 1980. Growth and survival
dividuals in their lifetime reproductive success (Gus- of nestling Kittiwakes Rissa tridactyla in Norway. Ornis
tafsson and Sutherland 1988). In contrast, in long-lived Scandinavica 11:228-235.
species with a fixed clutch size of one, even a small Boyce, M. S., and C. M. Perrins. 1987. Optimizing Gre
reduction in adult survival rate will have great con- Tit clutch size in a fluctuating environment. Ecology
142-153.
sequences for the lifetime reproductive output (Wooller
Braun, B. M., and G. L. Hunt. 1983. Brood reduction in
et al. 1992). Accordingly, such species do not respond Black-legged Kittiwake. Auk 100:469-476.
readily to changes in offspring demands (Ricklefs and Burger, A. E., and J. F Piatt. 1990. Flexible time budgets in
Minot 1991, Saether et al. 1993, but see Johnsen et al. Common Murres: buffers against variable prey abundance.
1994). In support of this hypothesis, species with an Studies in Avian Biology 14:71-83.
Charlesworth, B. 1980. Evolution in age-structured popu-
intermediate adult survival rate predominate in those lations. Cambridge University Press, Princeton, New Jer-
cases where an experimentally increased clutch size sey, USA.
decreased the probability of future adult survival (cf. Charnov, E. L., and J. R. Krebs. 1974. On clutch size and
fitness. Ibis 116:217-219.
Linden and M0ller 1989, Roff 1992, Stearns 1992).
Charnov, E. L., and S. W. Skinner. 1984. Evolution of host
The reduced survival rate in the enlarged group oc-
selection and clutch size in parasitoid wasps. Florida En-
curred only among the females. This suggests that fe- tomologist 67:5-21.
males respond more readily to changes in food requi- Clutton-Brock, T. H. 1991. The evolution of parental care.
rements of the chicks than do males. Similar sexual Princeton University Press, Princeton, New Jersey, USA.
Coulson, J. C., and J. M. Porter. 1985. Reproductive success
differences in the willingness to increase reproductive
of the Kittiwake Rissa tridactyla: the roles of clutch size,
effort also have been recorded in other species (Clut-
chick growth rates and parental quality. Ibis 127:450-466.
ton-Brock 1991). Coulson, J. C., C. S. Thomas, J. E. L. Butterfield, N. Duncan,
The increased mortality of females in the enlarged
P. Monaghan, and C. Shedden. 1983. The use of head and
group probably will lead to a smaller numberbill oflength
in- to sex live gulls (Laridae). Ibis 125:549-557.
Coulson, J. C., and E. White. 1959. The post-fledging mor-
dividuals being recruited to the population in the fol-
tality of the Kittiwake. Bird Study 6:97-102.
lowing generation, relative to the control and reduced
Cullen, E. 1957. Adaptions in the Kittiwake to cliff nesting.
groups (Table 2). In fact, the highest fitness was esti-
Ibis 99:275-302.
Dijkstra, C., A. Bult, S. Bijlsma, S. Daan, T. Meijer, an
mated for females in the reduced group. In contrast,
males in the control group were estimated to haveZijlstra.
the 1990. Brood size manipulations in the Kest
Falco tinnunculus: effects on offspring and parent survi
highest fitness (Table 2). Fitness differences are prob-
Journal of Animal Ecology 59:269-285.
ably underestimated, since we assume no survival
Furness,cost R. W., and D. G. Ainley. 1984. Threats to sea
of lower fledging mass. This suggests a different op-
populations presented by commercial fisheries. Pages
708 in J. P. Croxall, P. G. H. Evans, and R. W. Schreiber,
timal clutch size for males and females. Sex-specific
editors. Status and conservation of the world's seabirds.
mortality in response to clutch size manipulations has
International Council for Bird Preservation, Cambridge
also been recorded in the Great Tit (Boyce and UK.Perrins
1987) and the Blue Tit (Nur 1984, Pettifor 1993b).
Furness, R. W., and R. T. Barrett. 1985. The food require-
In conclusion, Lack's clutch size solution can ments
explain and ecological relationships of a seabird community
why single-egg clutches are not common among in North
the Norway. Ornis Scandinavica 16:305-313.
Galbraith, H. 1983. The diet and feeding ecology of breeding
Kittiwakes at Horn0ya. If the reduced fledgling mass
Kittiwakes Rissa tridactyla. Bird Study 30:109-120.
of chicks from enlarged broods influences offspring
Godfray, H. C. J. 1987. The evolution of clutch size in in-
fitness, this hypothesis also can explain the low fre-
vertebrates. Oxford Survey of Evolutionary Biology 4:
quency of three-egg clutches. The survival cost of in-
117-154.

This content downloaded from 134.117.10.200 on Tue, 14 Nov 2017 18:51:16 UTC
All use subject to http://about.jstor.org/terms
1642 KARL-OTTO JACOBSEN ET AL. Ecology, Vol. 76, No. 5

Goodman, D. 1974. Natural selection and a cost dividual

ceiling optimisation
on hypothesis. Nature (London) 336:
160-162.
reproductive effort. American Naturalist 108:247-268.
Gustafsson, L., and W. J. Sutherland. 1988. The cost of re- Price, T., and L. Liou. 1989. Selection on clutch-size in birds.
production in the Collared Flycatcher Ficedula albicollis. American Naturalist 134:950-959.
Nature (London) 335:813-815. Pugesek, B. H. 1981. Increased reproductive effort with a
Hamer, K. C., P. Monaghan, J. D. Uttley, P. Walton, and M. in the California Gull Larus californicus. Science 212:82
D. Burns. 1993. The influence of food supply on the breed- 823.
ing ecology of Kittiwakes Rissa tridactyla in Shetland. Ibis . 1983. The relationship between parental age and
135:255-263. reproductive effort in the California Gull Larus californ
Harris, M. P. 1970. Breeding ecology of the Swallow-tailed cus. Behavioral Ecology and Sociobiology 13:161-171.
Gull, Creagrusfurcatus. Auk 87:215-243. . 1987. Age-specific survivorship in relation to clutc
Haymes, G. T, and R. D. Morris. 1977. Brood size manip- size and fledging success in California Gulls. Behavior
ulation in Herring Gulls. Canadian Journal of Zoology Ecology55: and Sociobiology 21:217-221.
1762-1766. Pugesek, B. H., and K. L. Diem. 1990. The relationship
Howe, A. F 1976. Egg size, hatching asynchrony, sex, and reproduction and survival in known-aged Cali-
between
brood reduction in the Common Grackle. Ecologyfornia 57: Gulls. Ecology 71:811-817.
1195-1207. Reid, W. V. 1987. The cost of reproduction in the Glaucous-
Hogstedt, G. 1980. Evolution of clutch size in birds: adaptive winged Gull. Oecologia 74:458-467.
variation in relation to territory quality. Science Ricklefs,
210:1148- R. E., and E. O. Minot. 1991. Concluding remarks:
1150. parent-offspring interactions in the evolution of avian life
Johnsen, I., K. E. Erikstad, and B.-E. Saether. 1994. Regu- histories. Proceedings of the International Ornithological
lation of parental investment in a long-lived seabird, the Congress XX:992-997.
Puffin Fratercula arctica: an experiment. Oikos 71:273- Roff, D. A. 1992. The evolution of life histories. Chapman
278. and Hall, New York, New York, USA.
Lack, D. 1947. The significance of clutch-size. Ibis 89:302- SAS Institute. 1990. SAS/STAT user's guide, release 6.04
352. edition. SAS Institute, Cary, North Carolina, USA.
. 1954. The natural regulation of animal numbers.Stearns, S. C. 1992. The evolution of life histories. Oxford
Oxford University Press, Oxford, UK. University Press, Oxford, UK.
.1966. Population studies of birds. Clarendon Press,Storey, A. E., R. E. Anderson, J. M. Porter, and A. M.
Oxford, UK. MacCharles. 1992. Absence of parent-young recognition
Lessells, K. 1991. Evolution of life history. Pages 32-68 in in Kittiwakes: a re-examination. Behaviour 120:302-323.
J. R. Krebs, and N. Davies, editors. Behavioural ecology. Saether, B.-E. 1989. Survival rates in relation to body weigh
Blackwell Scientific, Oxford, UK. in European birds. Ornis Scandinavica 20:13-21.
Linden, M., and A. P. M0ller. 1989. Cost of reproduction . 1990. Age-specific variation in reproductive per-
and covariation of life history traits in birds. Trends in formance of birds. Current Ornithology 7:251-283.
Ecology and Evolution 4:367-371. Saether, B.-E., R. Andersen, and H. C. Pedersen. 1993. Reg
Moreno, J. 1989. Strategies of mass changes in breeding ulation of parental effort in a long-lived seabird: an ex
birds. Biological Journal of the Linnaean Society 37:297- perimental manipulation of the cost of reproduction in the
310. Antarctic Petrel Thalassoica antarctica. Behavioral Ecol-
Nur, N. 1984. The consequences of brood size for breedingogy and Sociobiology 33:147-150.
Blue Tits. I. Adult survival, weight change and cost ofVermeer, K. 1963. The breeding biology of the Glaucou
reproduction. Journal of Animal Ecology 53:479-496. winged Gull Larus glaucescens on Mandarte Island. Oc
1988. The cost of reproduction in birds: an exam- casional Papers of the British Columbia Provincial Muse
ination of the evidence. Ardea 76:155-168. 13:1-104.
Partridge, L. 1989. Lifetime reproductive success andWard,lifeJ. G. 1973. Reproductive success, food supply, a
history evolution. Pages 421-440 in I. Newton, editor. Life-
the evolution of clutch size in the Glaucous-winged Gul
time reproduction in birds. Academic Press, London, UK.
Dissertation. University of British Columbia, Vancouver
Perrins, C. M., and D. Moss. 1975. Reproductive rates in
Canada.
the Great Tit. Journal of Animal Ecology 44:695-706.
Williams, G. C. 1966. Natural selection, the cost of repr
Pettifor, R. A. 1993a. Brood-manipulation experiments. I.
duction, and a refinement of Lack's principle. Americ
The number of offspring surviving per nest in Blue Naturalist
Tits 100:687-690.
Parus caeruleus. Journal of Animal Ecology 62:131-144.
Wooller, R. D., J. S. Bradley, and J. P. Croxall. 1992. Lon
1993b. Brood-manipulation experiments. II. A term cost population studies of seabirds. Trends in Ecology a
of reproduction in Blue Tits Parus caeruleus. Journal of 7:111-114.
Evolution
Animal Ecology 62:145-159. Wooller, R. D., and J. C. Coulson. 1977. Factors affe
Pettifor, R. A., C. M. Perrins, and R. H. McCleery.the 1988.age of first breeding of the Kittiwake Rissa trida
Variation in clutch-size in Great Tits: evidence for the in- Ibis 119:339-349.

This content downloaded from 134.117.10.200 on Tue, 14 Nov 2017 18:51:16 UTC
All use subject to http://about.jstor.org/terms