Annual timing and life-history variation in free-living stonechats

The stonechat has one of the widest breeding distributions among
the Old World passerines. It breeds under tropical, subtropical,
north- and south-temperate, and sub-arctic conditions. Also its
migratory strategies vary, ranging from year-round residency in
areas such different as the British Isles and equatorial Africa, to
long-distance migration in northern Asia. These characteristics and
the fact that it can be successfully bred in captivity make the
stonechat an ideal model species for studying life-history
variation under varying seasonal regimes, both in the field and in
the lab. In Chapter 1 I develop the conceptual framework for the
study of annual timing and life-history variation in stonechats. My
studies focus on free-living stonechats from two different
populations, Siberian stonechats breeding in northern Kazakhstan
and European stonechats breeding in central Slovakia. The taxonomic
status and the basic biology of the stonechat and the
characteristics of the two study sites are outlined in Chapter 2.
Both sites are situated in the temperate zones and hence the local
photoperiodic conditions are similar. However, due to the
continental climate in Kazakhstan the summer is shorter and hotter
and the precipitation lower than in Slovakia, and as a result the
breeding season is about two months longer at the European site.
The two populations differ in their migratory strategies (Chapter
3). The Siberian stonechats are long-distance migrants that travel
up to 6500 km from their breeding to their wintering grounds and
pass major ecological barriers like deserts and high mountain
ranges on their way. The stonechats from eastern Europe are
short-distance migrants that travel about 1500 km. Ringing recovery
data suggest that a large proportion of them pass the Mediterranean
Sea on their way. There are no ringing recoveries for Siberian
stonechats, and information on their migratory routes is sparse and
dispersed in different sources. I tried to fill this gap by
reconstructing the spatio-temporal pattern of the stonechat
migrations within Central Asia and southern Siberia from published
passage data. Stonechats enter Central Asia in spring from the
south-west, which suggests that they largely avoid the high
inner-Asian mountain ranges. From there they spread northwards into
their Central Asian and Siberian breeding areas. In autumn they
retrace this pattern backwards, however the movement is less
concerted than in spring. The passage data can be used to calculate
the average speed of migration within the area and compare it to
migration speed in the European stonechats. The European stonechats
move at higher speeds during spring than during autumn. This is
consistent with data on other European passerines. Within Central
Asia, however, the stonechats seem to be constrained in their
spring movements because they have to stop and wait for
ameliorating local conditions. Increasing spring temperatures due
to global warming may therefore have a high potential to change
migratory phenology in the Siberian stonechats. Birds shut down
their reproductive system during the nonbreeding seasons. Migratory
birds face a potential conflict between the preparation for
breeding and migration, because gonadal maturation takes several
weeks. I investigated whether the migratory strategy affects the
gonadal state at the time of arrival in the breeding areas (Chapter
4). Males of both populations arrive at the breeding sites with
gonads that are not fully developed. Neither the gonadal state in
relation to the onset of reproduction, nor the rate of development
to the mature state differ in the two populations. This could
indicate that the late stages of reproductive development are
mainly affected by the targeted date of reproduction and/or by
physiological determinants of gonadal maturation rates, whereas the
migratory period has less effect. More knowledge about reproductive
development along the migratory route, particularly in
long-distance migrants, is required. In my study populations local
breeders and passage migrants are not distinguishable on the basis
of their gonadal development. However, the data in the passage
migrants is more variable, which could reflect differences in
reproductive timing due to different migratory destinations. Due to
the local conditions at the breeding site and the temporal
requirements of migration the stonechats in Kazakhstan have a
shorter breeding season than their conspecifics in Slovakia
(Chapter 5). The Siberian stonechats produce generally one clutch
per season, but lost clutches are often replaced. European
stonechats lay up to three clutches per season. The interval
between clutch loss and relaying does not differ between the two
populations. Breeding synchrony is significantly higher in
Kazakhstan, both over the whole breeding period and when only the
first clutches are considered. Overall clutch size is higher in the
Siberian stonechats. Clutch size decreases from first to
replacement clutches in Kazakhstan, and from first to second
clutches in Slovakia. The clutch size of the third clutches in the
European stonechats increases again, which may indicate that only
high-quality parents initiate a third seasonal breeding attempt, or
that a strategy of terminal investment is involved. Fledging
success per egg is higher in Kazakhstan and does not differ between
breeding attempts. In Slovakia fledging success drops markedly from
the first to the later breeding attempts. Predation during the egg
stage plays a greater role in the European stonechats, whereas in
the Siberian stonechats predation on nestlings is more important.
As a result of smaller clutch sizes later in the season, and of
higher failure rates in later clutches, particularly in Slovakia,
fledgling output per clutch decreases in both populations. A Kazakh
breeding pair produces on average about five fledglings per season,
a Slovak breeding pair about seven fledglings per season. In one
study season 9% of the Kazakh males mated simultaneously with two
females. Facultative polygyny has previously been reported in
European, but not in Siberian stonechats. The primary clutches of
polygynous males were initiated earlier, contained more eggs and
produced more fledglings than the comparable first clutches of
monogamous males. This may indicate that polygynous males were of
superior parental quality. The secondary polygynous clutches were
initiated later, contained less eggs and produced less fledglings,
indicating costs in terms of reproductive success for secondary
females. According to the challenge hypothesis the length of the
breeding period, the degree of inter-male competition, and the
degree of parental care affect the seasonal profiles of circulating
androgens (Chapter 6). Male Siberian stonechats have elevated
circulating androgen levels during May and June with a peak in
mid-May. In the European stonechats androgens are elevated from
March until the end of May, they peak in mid-April and mid-May. In
both populations androgen levels are highest in males during the
territorial stages and when the females are fertile, and decrease
during the parenting stages. Overall androgen levels are higher in
the European stonechat males, which could be explained by higher
male-male competition and/or a lower degree of paternal care in
this population. Male aggression during simulated territorial
intrusions is high throughout the season in both populations and
therefore apparently unrelated to the circulating androgen levels.
In the Siberian stonechat males dihydrotestosterone (DHT)
contributes relatively more to the total circulating androgens.
This could be related to different roles of DHT and testosterone
(T) in regulating male reproductive behaviours, or to different
costs associated with high circulating levels of these androgens.
Estradiol (E2) is basal throughout the season in Siberian and
European stonechat males. In female stonechats the levels of
circulating gonadal steroids (E2, T, DHT) are basal throughout the
season. T is higher in the European stonechat females than in the
Kazakh females, however there is no apparent seasonal pattern. The
adrenal glucocorticoids (GCs) are steroid hormones that act in
general energy metabolism and as part of the response of an
organism to unpredictable threats to its physiological homeostasis
(Chapter 7). Two hypotheses try to explain seasonal and individual
variation in circulating GCs. The reproductive limitation
hypothesis predicts that, because high GCs may cause nest
desertion, GC response to stress is reduced when breeding
opportunities are limited, as in the Siberian stonechats. Contrary
to this predictions, overall GC levels are higher in the Siberian
than in the European stonechats. Hence there seems to be no simple
relationship between the potential number of breeding attempts and
GC levels in stonechats. The energy mobilisation hypothesis
predicts higher GC levels during periods of increased energy
demand, such as breeding. The energetic costs may differ between
breeding stages, at least in males. However, GC levels do not vary
with breeding stage in the stonechats. GC levels decrease in
Kazakhstan when the birds start to moult. Feather replacement is an
energetically costly task; therefore a simple relationship between
energy demand and circulating GCs is not supported. It is possible
that GCs interfere with the physiology of feather replacement and
are therefore reduced during this period. Time-dependent mortality,
the degree of sibling competition, and internal constraints on
growth have been discussed as the major factors affecting the
evolution of developmental rates (Chapter 8). The length of the
breeding season may also affect juvenile development, particularly
in migrants, where the juveniles have to gain a certain level of
maturity to meet the demands of the autumn movements. Incubation
periods are slightly shorter in Kazakhstan than in Slovakia,
indicating that embryonic development may proceed at a higher rate
in the Siberian stonechats. Postnatal increase rate in body size,
but not in wing length, is also higher in the Siberian stonechats.
These measures of postnatal growth are not affected by the hatching
date. Runt nestlings, which are at a competitive disadvantage
because they have hatched later than their nestmates, are initially
heavier and bigger but loose this head start during the growth
period due to lower growth rates. Postjuvenile moult is initiated
very early in the life of Siberian stonechats. as in their captive
conspecifics. It is shifted forward by a few days in the
late-hatched chicks of replacement clutches, however the effect of
hatch date on the onset of moult is much lower than in captive
European stonechats. The period from the start of incubation until
the end of postjuvenile moult lasts about half as long in the
Siberian stonechats than in the European stonechats, mainly due to
differences in the onset and duration of moult between the
populations. Siberian stonechats lay only one clutch per season,
but lost clutches are often replaced (Chapter 9). Normal breeders
(those birds that raised their first clutch successfully) initiate
postnuptial moult shortly after their young have left the nest.
Late breeders (birds with replacement clutches) moult later than
normal breeders. However, they initiate moult earlier in relation
to the age of their offspring than normal breeders. As a result,
they overlap breeding and moult more than normal breeders.
Simultaneous reproduction and feather replacement is thought to be
costly and therefore generally avoided. This is not always possible
in time-constrained breeders. In the Siberian stonechats the degree
of moult-breeding overlap increases the later the offspring hatches
in the season. Scarce data suggests that late breeding males
initiate moult earlier than their female partners. This may imply
that they reduce their share in parental care, as has been found in
other species. Postponing moult of body feathers, which serves
mainly in insulation, may have less severe consequences in a
migratory species, than postponing wing moult. Late breeders
postpone body moult more than wing moult. Siberian stonechats in
Kazakhstan and European stonechats in Slowakia are closely related
and breed both in the north-temperate zones in the same latitude
and under similar photoperiodic conditions. However, due to the
different local climatic conditions, the migratory distance and the
length of the breeding season differ. This brings about consistent
differences between the two populations in the migratory behaviour,
the breeding performance, the hormonal regulation of reproduction,
the hormonal response to environmental challenges, and the juvenile
development. Because the Siberian and Slovak stonechat populations
are closely related, a divergent genetic background and/or
species-specific physiological constraints probably play a minor
role in creating these differences. They rather reflect differences
in annual timing, which affect the trade-off between current and
future reproductive success in different life-history stages. The
climatic changes that are observed in recent years have been
associated with changing migratory and reproductive schedules and
shifts in species’ distribution ranges. The example of the
stonechats shows that a multitude of systemic changes is required
to change the annual cycle. The success of an organism in a
changing environment will depend on its ability to successfully
integrate all these physiological and behavioural alterations.