The adaptive function of hatching asynchrony: an experimental study in great tits

Animal Behaviour 86 (2013) 567e576

Contents lists available at SciVerse ScienceDirect

Animal Behaviour journal homepage: www.elsevier.com/locate/anbehav

The adaptive function of hatching asynchrony: an experimental study in great tits Katarzyna Anna Podlas*, Heinz Richner Institute of Ecology and Evolution, University of Bern, Bern, Switzerland

a r t i c l e i n f o Article history: Received 4 February 2013 Initial acceptance 6 March 2013 Final acceptance 28 May 2013 Available online 26 July 2013 MS. number: 13-00109R Keywords: brood reduction hypothesis great tit hatching asynchrony incubation maternal effects Parus major

In many bird species offspring hatch over hours or days, which leads to an age and size hierarchy within broods. The function of hatching asynchrony is much debated, and it has been suggested that the induced size hierarchy among offspring may be an adaptive maternal mechanism for maximizing reproductive output under variable environmental conditions. The best known hypothesis to explain the adaptive value of hatching asynchrony, the ‘brood reduction’ hypothesis, holds that a size hierarchy among offspring allows for a quick adaptive adjustment of brood size to unpredictable feeding conditions and thus benefits parents. To test the consequences of hatching asynchrony on offspring growth and food provisioning we experimentally manipulated the onset of incubation of eggs within broods of great tits, Parus major, to induce either synchronous or asynchronous hatching, and then manipulated brood size after hatching to simulate favourable, control or harsh conditions. We did not find a difference in offspring mortality between asynchronous and synchronous broods under any of these conditions. In harsh conditions, nestling mass and size were reduced in asynchronous broods compared to synchronous ones. The opposite pattern occurred under control conditions. Although our results showed that induced hatching asynchrony alters offspring phenotype, they do not provide strong support for the brood reduction hypothesis as a mechanism to explain hatching asynchrony. Ó 2013 The Association for the Study of Animal Behaviour. Published by Elsevier Ltd. All rights reserved.

Environmentally induced parental effects include a large range of factors that influence offspring phenotype (Lacey 1998). They are the product of the interaction of a parental genotype with the environment as expressed in the offspring (Mousseau & Fox 1998). Such cross-generational effects have been observed in both animals (Bernardo 1996) and plants (Galloway 1995, 2005). They are based on a variety of transmission pathways, and are potentially powerful means for producing phenotypic variants among offspring that are well adapted to the prevailing or future conditions (Marshall & Uller 2007). Many animals produce more than one offspring at a time, which poses the interesting evolutionary problem of how births should be spaced to maximize reproductive success. Asynchronous birth is a common phenomenon (e.g. amphibians: Ryan & Plague 2004; reptiles: While et al. 2007; insects: Smiseth & Morgan 2009; sharks: Farrell et al. 2010), and has been extensively studied because the incubation pattern of eggs, as a parental effect, can influence hatching patterns and thus competitive hierarchies among offspring. Most studies have been conducted on altricial bird species (e.g. Clark & Wilson 1981; Magrath 1989; Mock &

* Correspondence: K. A. Podlas, Institute of Ecology and Evolution, University of Bern, Baltzerstrasse 6, 3012 Bern, Switzerland. E-mail address: [email protected] (K. A. Podlas).

Parker 1997; Mainwaring et al. 2012), probably because birds offer excellent models to study such parental effects since embryos develop outside the parent’s body and hence facilitate manipulation of incubation patterns, hatching order and other stages of offspring development (Bernardo 1996; Groothuis & Schwabl 2008). Avian parents may control hatching patterns because, in birds, embryonic development typically starts when eggs are actively incubated (Drent 1975; Wiebe et al. 1998). Most birds lay one egg per day, and incubation may begin at any time during the laying sequence (Blackburn & Evans 1986; Bortolotti & Wiebe 1993). Incubation of eggs before the completion of the clutch (Clark & Wilson 1981; Slagsvold 1985; Magrath 1990) leads to hatching asynchrony and in turn to a pronounced age-based competitive hierarchy in which the younger offspring typically face reduced growth (Stokland & Amundsen 1988; Nilsson & Svensson 1996; Rosivall et al. 2005) and increased mortality before or after fledging (Magrath 1990; Forbes et al. 1997; Vinuela 2000). Thus, the differential onset of incubation and hence hatching asynchrony induced by parents (Drent 1975; Wiebe et al. 1998) will also determine offspring growth (Cotton et al. 1999; Clotfelter et al. 2000; Mainwaring et al. 2010) and survival (Forbes et al. 1997; Vinuela 2000), and may have long-term consequences for nestlings (Forbes 2009; Mainwaring et al. 2012). Moreover, it has been

0003-3472/$38.00 Ó 2013 The Association for the Study of Animal Behaviour. Published by Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.anbehav.2013.06.012

568

K. A. Podlas, H. Richner / Animal Behaviour 86 (2013) 567e576

shown previously that the degree of asynchrony can be adjusted to the availability of food (Slagsvold & Lifjeld 1989; Nilsson 1993; Wiebe & Bortolotti 1994; Fleming et al. 1997; Eikenaar et al. 2003; but see Kontiainen et al. 2010; Parejo et al. 2012). Thus, by varying the onset of incubation, mothers may adjust offspring growth via asynchronous hatching in anticipation of the conditions expected during brood rearing (Magrath 1990; Wiebe & Bortolotti 1994; but see Wiehn et al. 2000), and thereby may also enhance their own probability of survival and future reproduction (Mock & Forbes 1994; Horak 1995; but see Stoleson & Beissinger 1997; Szollosi et al. 2007). Many hypotheses have been proposed to explain why females start incubation before clutch completion (reviewed in Nilsson 1993; Stenning 1996). Some hypotheses suggest that the size hierarchy established by hatching asynchrony increases the fitness of both the parents and early hatched offspring. Among these, the best known and the most debated is the brood reduction hypothesis proposed by Lack (1947), which states that hatching asynchrony is advantageous in unpredictable environments. When food is plentiful, all offspring can fledge independently of hatching patterns (Forbes 1990; but see Amundsen & Slagsvold 1991). However, when food is scarce, later hatched offspring might be outcompeted by the older, earlier hatched siblings, and in turn quickly starve to death. Although under unfavourable food conditions hatching asynchrony might be detrimental for last-hatched offspring, it may have no or very small effects upon the early hatched nestlings. If parents invest the saved energy to rear the remaining nestlings, brood survival and nestling condition at fledging may be enhanced. Higher body mass and size of nestlings have been repeatedly shown to predict their subsequent survival and thus might be a good proxy for parental fitness (Pettifor et al. 2001). Moreover, if parents save resources by early elimination of some nestlings, those resources could also be invested in their own survival and future reproduction (e.g. Mock & Ploger 1987; Mock & Forbes 1994). In the majority of empirical studies, two weak trends seem to emerge: first, synchronous broods produce more fledglings than asynchronous broods, and second, asynchronous broods produce heavier nestlings at fledging than synchronous broods (reviewed in Amundsen & Slagsvold 1991, 1998). Studies testing the adaptive significance of hatching asynchrony have shown that the relationship between the prevailing breeding conditions and the benefits of hatching patterns is complex (Forbes 1994; Amundsen & Slagsvold 1996; Szollosi et al. 2007; but see Magrath 1989). Although some experimental studies showed that asynchronous broods are more productive when food is scarce (Magrath 1989; Wiebe & Bortolotti 1994; see also Theofanellis et al. 2008), others have shown that nestlings from asynchronous broods suffered equally under both good and bad conditions compared to synchronous broods (Szollosi et al. 2007), and brood reduction occurred even when food was plentiful (Amundsen & Slagsvold 1991; Parejo et al. 2012). Since the number of starving offspring is sometimes greater than the number of offspring at a competitive disadvantage (e.g. Howe 1978; Horak 1995), it has been suggested that hatching asynchrony may not be essential for facilitating brood reduction (Clark & Wilson 1981). In addition, some studies found that partial brood starvation occurred in both asynchronous and synchronous broods (Howe 1976; Horak 1995; Kontiainen et al. 2010), suggesting some other factors were involved in nestling death, such as sibling despotism (Forbes 1994; Mock & Forbes 1994). Finally, there is some evidence that partial brood reduction increases future female survival (Horak 1995) or parental condition (Slagsvold & Lifjeld 1989), suggesting some advantages for parents rearing asynchronous broods. Other studies, however, found that parental survival is independent of hatching patterns (Stoleson & Beissinger 1997; Szollosi et al. 2007).

The aim of the present study was to test the effects of hatching asynchrony on offspring quality, rates of food provisioning and length of the nestling period as predicted by the brood reduction hypothesis. To this end, we performed a field experiment in a freeliving population of great tits, Parus major, in which we simultaneously manipulated the onset of incubation and the rearing conditions for both parents and offspring by changing brood size. By removing eggs from the nest and returning them for incubation either simultaneously or sequentially, we created clutches that were incubated and hatched either asynchronously or synchronously. We measured the effects of experimentally induced hatching regimes (both asynchrony and synchrony) on average fledgling size, fledgling body mass and fledging mortality, as well as on parental effort and parental body condition in enlarged, control and reduced broods. We used brood size manipulation as a proxy for manipulations of food availability. It has been previously shown that brood size manipulation can successfully alter the foodprovisioning rate to individual nestlings (Dijkstra et al. 1990; Martins & Wright 1993; Pettifor et al. 2001) and change the level of nestling competition (Neuenschwander et al. 2003), thus allowing an indirect but largely adequate test of the brood reduction hypothesis. Based on the predictions of the brood reduction hypothesis, we expected hatching asynchrony and the resulting size hierarchy to be beneficial for parents under suddenly changing harsh conditions. Thus, under harsh conditions as induced by brood enlargement, we expected that hatching asynchrony should result in higher mortality in asynchronous broods owing to the death of the smaller, weaker nestlings compared to synchronous broods. In consequence, the average size and mass of nestlings in asynchronous broods should be greater than in synchronous broods after the death of the smaller, weaker nestlings, so that eventually asynchronous broods should fledge fewer but bigger nestlings than synchronous ones. On the other hand, under favourable conditions we expected hatching asynchrony to have neutral or negative effects, as proposed by Amundsen & Slagsvold (1991). In contrast, under control conditions, we expected either neutral or advantageous effects of hatching asynchrony. METHODS Biological Model The great tit is a small resident passerine, a hole-nester that readily accepts artificial nestboxes for breeding and roosting. Great tits breed between March and July, and the start of breeding depends largely on spring temperature and the peak abundance of caterpillars (Nager & van Noordwijk 1992; van Noordwijk et al. 1995; Visser et al. 1998). Clutch size varies from four to 13 eggs (Slagsvold & Lifjeld 1990; Haywood & Perrins 1992). Only females develop a brood patch and hence incubate the eggs. Full incubation lasts on average 13 days and males provide food to the females during this period (Haftorn 1981). Hatching spread varies from 0 to 4 days, with a mean around 1.5 days (e.g. Haftorn 1981; Slagsvold et al. 1992; Amundsen & Slagsvold 1998; Tilgar & Mand 2006), and leads to a size hierarchy in which the last-hatched nestlings often die before fledging because of starvation (Horak 1995; Amundsen & Slagsvold 1998; Tilgar & Mand 2006). Although the role of hatching asynchrony in great tits has been widely studied and is still highly debated, its consequences for offspring phenotype and its adaptive function remain elusive. Some studies have suggested that increased hatching spread and lower fledging success of nestlings in clutches laid later in the breeding season might facilitate brood reduction if, as is likely, food is scarce late in the season (Barba et al. 1995; Cresswell & McCleery 2003). In

K. A. Podlas, H. Richner / Animal Behaviour 86 (2013) 567e576

agreement with this hypothesis some studies found that asynchronous broods experienced brood reduction, and thus allowed for the production of fewer but heavier nestlings (Amundsen & Slagsvold 1998; Tilgar & Mand 2006; Theofanellis et al. 2008). However, at the same time other studies showed that brood reduction may occur as frequently in asynchronous as in synchronous broods and may not necessarily serve to improve postfledging survival of the first-hatched nestlings but rather may improve the mother’s survival (Horak 1995). General Experimental Protocol The experiment was performed in 2011 in the Könizbergwald, a forest near Bern, Switzerland (46
560 N, 7
240 E). From early March onwards, nestboxes were visited regularly to determine laying date and clutch size. Each egg was marked on the day it was laid, removed from the nest and replaced with a dummy egg. Eggs were stored on the bottom of the same nestbox inside a small wooden box containing nest material. Eggs were turned daily to mimic the natural behaviour of females (Haftorn 1979; Gee et al. 1995) and to prevent the yolk from setting (Drent 1975; Mainwaring et al. 2010). Incubation treatment started from the second day of incubation onwards. The start of incubation was defined as the first day when all dummy eggs in the nest were warm and not covered by nest material. We created two types of experimental clutches. (1) In ‘asynchronous clutches’ (ASYNC), we simulated the onset of incubation before clutch completion by returning eggs to the nest cup for incubation in three steps: on the second day of incubation all eggs apart from the four last-laid ones were returned to the nest; the third and fourth last-laid eggs were returned to the nest on the third day of incubation; the two last-laid eggs were returned to the nest on the fourth day of incubation. (2) In ‘synchronous clutches’ (SYNC), all eggs were simultaneously returned to the nest on the fourth day of incubation, to make incubation effort for females of SYNC and ASYNC broods similar (Deeming 2002). The brood size manipulation treatment was carried out when nestlings were 2 days old (day 0 is hatching date). As found previously, different hatching regimes may lead to differences in hatching success and hence in brood size at hatching (Veiga & Vinuela 1993). Moreover, a brood size manipulation done by adding extra nestlings may perturb family dynamics because of genetic differences between nestlings originating from different parents, and because of potential differences in other maternal effects (e.g. deposition of hormones in certain eggs or laying bigger eggs at the end of the laying sequence; Eising et al. 2001; Muller et al. 2010). Hence, whole broods with similar hatching dates (1 day) were exchanged between nests. Clutches were swapped between pairs to create a difference of two eggs between ‘enlarged’ and ‘reduced’ broods. In ‘control broods’ clutches with the same number of eggs were exchanged. All experimental broods were within the natural range of brood sizes found in great tits (range after manipulation 5e11 nestlings). To preserve sample size, the exchanges were performed between ASYNCeSYNC broods, ASYNCeASYNC broods and SYNCeSYNC broods. Altogether, 158 broods were created (26 ASYNC-reduced, 31 ASYNC-control, 23 ASYNC-enlarged, 27 SYNCreduced, 21 SYNC-control and 30 SYNC-enlarged broods) from which 14 whole broods died before fledging (one ASYNC-enlarged, two ASYNC-control, three ASYNC-reduced, three SYNC-enlarged, two SYNC-control and three SYNC-reduced). There was no difference in nest failure probability between treatments (c22;152 ¼ 0:12, P ¼ 0.94). The death of entire broods is not uncommon and in our case was probably caused by the desertion of parents, given that there were no signs of starvation or predation. The death of broods occurred at different times during the nestling period and not directly after manipulation (one brood was found dead between

569

days 2 and 9 of nestling age, five broods between days 9 and 14, and eight broods between day 14 and date of fledging). Data Collection After the first nestling hatched (nest day 0), nests were visited twice a day to determine the hatching spread, the number of hatchlings and their weight. On the day of hatching, each newly hatched nestling was individually marked. On day 9, birds were ringed with a numbered aluminium ring (Swiss Ornithological Station, Sempach). On day 14 posthatch all nestlings were weighed to the nearest 0.01 g with a portable electronic scale, and their tarsus and wing length were measured to the nearest 0.5 mm with a calliper and a ruler, respectively. On day 10 (24 h before filming) we placed a sticker representing a camera lens on the top inside cover of the nestbox to allow birds to habituate to the later presence of a real camera. On day 11 nestling feeding was recorded (1.5 h) with a video camera installed within the nestbox (for the exact protocol see Helfenstein et al. 2008). To allow individual identification on the videos, each nestling was marked on the head with small spots of dark red acrylic paint 1 day before the recording (Kölliker et al. 1998; Helfenstein et al. 2008). The first 30 min of each film were discarded to exclude potential disturbance effects on behaviour from the camera placing. Owing to technical problems nine of the 144 nests could not be recorded (20 ASYNCreduced, 26 ASYNC-control, 21 ASYNC-enlarged, 24 SYNCreduced, 16 SYNC-control, 27 SYNC-enlarged). This enabled us to estimate the mean per capita feeding rate (calculated at the level of the nest and defined as the total number of parental visits within 1 h divided by the number of nestlings in a brood) as well as individual feeding rate of each nestling and thus within-brood variance in individual feeding rate. Within-brood variance in individual feeding rate was measured to test whether there was a bias in food distribution towards some specific nestlings, such as first or last hatched, especially in asynchronous broods compared to synchronous ones. From day 16 onwards, nests were checked daily to record the number of fledged young and fledging date (Keller & van Noordwijk 1994). On day 12 adults were caught at the nest using door traps to evaluate potential effects of hatching asynchrony on parents. Individuals were identified or ringed and their body mass and tarsus length were measured. We caught 229 adults (148 females: 22 ASYNC-enlarged, 31 ASYNC-control, 23 ASYNC-reduced, 28 SYNC-enlarged, 18 SYNC-control, 26 SYNC-reduced; 81 males: seven ASYNC-enlarged, 14 ASYNC-control, 15 ASYNC-reduced, 18 SYNC-enlarged, 16 SYNC-control, 11 SYNC-reduced). Ethical Note All procedures employed during this field study were approved by the Ethical Committee of the Agricultural Office of the Canton Bern, Switzerland (experimentation permit BE 19/11). Catching and ringing of nestlings and adults were performed under a permit from the Federal Agency for the Environment of the Canton of Bern, Switzerland (ringing permit 2819). All manipulations of eggs and birds lasted between 5 and 15 min and did not cause any desertion or interruption of normal behaviour. Hatching success of manipulated eggs in our study (82%, 1021 nestlings hatched from 1242 manipulated eggs) did not differ from hatching success of nonmanipulated eggs from a neighbouring population. Brood size manipulation is a common method used in bird species to simulate years with bad and good food supply (e.g. Dijkstra et al. 1990; Horak 2003; Raberg et al. 2005; Szollosi et al. 2007). For the transfer of nestlings from nest to nest on day 2 we used small plastic boxes filled with cotton. Both nestlings and parents were handled a few metres away from the nestboxes to minimize the disturbance to

570

K. A. Podlas, H. Richner / Animal Behaviour 86 (2013) 567e576

neighbouring birds. Newly hatched nestlings were individually marked by selectively removing combinations of tuft feathers. Nestlings of great tits have three patches of tuft feathers on their head, one on each wing and one on their back. This allowed a code that contained up to 11 combinations of markings. Tuft feathers are easily removed since they do not seem to be strongly embedded in the skin and have no known function. On the bottom of the boxes, a heating pillow was placed to keep nestlings warm during transfer. The dark red paint colour used for marking nestlings for videos was chosen for its lack of conspicuousness under the dim light inside a nestbox and its contrast with nestling plumage under infrared light on the video recordings (Kölliker et al. 1998). Neither parents nor nestlings were ever seen pecking at the markings. This videorecording method has been successfully used in previous studies without any apparent signs of distress to either the adults or nestlings (Kölliker et al. 1998; Helfenstein et al. 2008; Tanner et al. 2008). Adults were caught at the nest using door traps. This is a common method used in this species because it is quick and noninvasive for both parents and nestlings. A spring-loaded flap door is fixed at the inner side of the entrance hole where the door is kept open by a small wooden stick that is displaced by an entering parent, closing the door. None of the trapped parents deserted and most parents resumed feeding within 15e30 min. Statistical Analyses We used two different estimates of brood mortality, viz. brood size at fledging and whole brood fledging success (proportion of eggs producing fledged young), and also two different estimates of hatching spread. First, hatching spread for each brood was defined as the number of days between first- and last-hatched nestlings in a brood. Second, the hatching spread for each brood was calculated using the relative weight difference between siblings in the brood (method described in Slagsvold et al. 1992; Hillstrom & Olsson 1994). Within-brood variation in the rate at which individual nestlings were fed between ASYNC and SYNC broods was computed as the coefficient of variation (CV ¼ standard deviation/mean, Sokal & Rohlf 1995) for each brood. The CV is statistically more convenient than variance because it is independent of the mean value of the trait (Sokal & Rohlf 1995). All analyses concerning randomization and validation of our treatments, breeding parameters, hatching success or hatching spread, brood size at fledging and whole brood fledging success were conducted on all 158 broods. However, for analyses of body mass, tarsus and wing length, individual offspring survival, parental body condition and parental effort, we used successful broods only (i.e. in which at least one nestling fledged, 144 broods). Patterns found for nestling parameters were similar for both analyses (full sample and successful broods only); however, the analyses on the successful broods revealed more meaningful differences between treatment groups and stronger effect sizes. To test whether our treatments were well randomized with respect to the incubation date (first day on which all dummy eggs in a brood were warm) we used general linear models (GLM) with a normal error distribution. A GLM with normal error was also used to test whether our treatments affected hatching spread (i.e. calculated using the relative weight difference between siblings in the brood), brood size on days 2, 9 and 14, brood size at fledging, mean per capita feeding rate, within-brood variance in nestling’s feeding rates and parental body condition. Body condition as a response variable was obtained by adding parental tarsus length as a predictor variable to the analyses of body mass (Freckleton 2002). Mean per capita feeding rate was square-root transformed to conform to the criterion of normal distribution. To assess the effectiveness of our treatments for differences in hatching spread

(i.e. days between the day of first and last hatching per brood) between SYNC and ASYNC broods we used a GLM with a Poisson error distribution. A GLM with Poisson error was used to test whether our treatments affected the length of the nestling period. Differences in hatching success (number of hatched eggs over clutch size) between both treatments were investigated using a GLM with binomial error. A GLM with binomial error was also used to test the differences in brood mortality between our treatments where the number of fledged nestlings over the number of hatched nestlings per brood was included as the dependent variable. All the models described above included incubation treatment, brood size manipulation treatment and its interactions as fixed factors. The impact of our treatments on offspring phenotypic quality was measured as nestling body size and mass on day 14, that is, 3 days before nestlings could potentially fledge. To this end, we used linear mixed models (LMM) including incubation treatment, brood size manipulation treatment and their two-way interaction as fixed factors and brood identity as a random factor. All models included initial laying date and original clutch size before brood size manipulation as covariates. The model testing differences in mean per capita feeding rate and CV in individual feeding rate included brood size at day 9 (instead of clutch size) as a covariate to correct for potential differences in feeding rates related to the number of nestlings within a brood. All statistical analyses were performed using R version 2.14 (R Development Core Team 2007, package nlme). Residuals were tested for normality and homoscedasticity. Interactions and main effects with a P value higher than 0.1 were backward eliminated using a stepwise elimination procedure in all models. We first excluded the interactions and then the main effects. In models assuming normal distributions with significant interaction terms, we used post hoc t tests to compare groups. To analyse the LMMs we used function lme from the nlme package with restricted maximum likelihood estimation on final models (Zuur et al. 2009). Sample sizes varied slightly between analyses because not all adults or nestlings could be measured. RESULTS Validation of Experimental Set-up There were no differences in incubation date between treatments (GLM: incubation treatment F1,156 ¼ 0.94, P ¼ 0.33; brood size manipulation treatment: F2,154 ¼ 0.03, P ¼ 0.97; incubation treatment * brood size manipulation treatment: F2,152 ¼ 1.64, P ¼ 0.20). Hatching success (i.e. number of hatched eggs over clutch size) did not differ between experimental treatments (GLM: incubation treatment: c21;156 ¼ 0:69, P ¼ 0.40; brood size manipulation treatment: c22;154 ¼ 0:18, P ¼ 0.91; incubation treatment * brood size manipulation treatment: c22;150 ¼ 0:26, P ¼ 0.87; laying date: c21;153 ¼ 0:21, P ¼ 0.64; clutch size: c21;152 ¼ 0:03, P ¼ 0.86; overall hatching success was 82%, calculated as a percentage of young hatched over the number of eggs laid). As predicted, brood size on day 2 differed between the brood size manipulation treatments (GLM: incubation treatment: F1,153 ¼ 2.93, P ¼ 0.089; brood size manipulation treatment: F2,153 ¼ 22.35, P < 0.001; incubation treatment * brood size manipulation treatment: F2,150 ¼ 0.80, P ¼ 0.45; laying date: F1,152 ¼ 0.19, P ¼ 0.66; clutch size: F1,153 ¼ 36.61, P < 0.001; mean brood size on day 2  1SE: 5.72  0.23 for reduced, 6.48  0.25 for control and 7.17  0.27 for enlarged broods). The same pattern was observed for brood size on day 9 (GLM: incubation treatment: F1,151 ¼ 0.15, P ¼ 0.70; brood size manipulation treatment: F2,152 ¼ 13.14, P < 0.001; incubation treatment * brood size manipulation treatment: F2,148 ¼ 0.75, P ¼ 0.47; laying date: F1,150 ¼ 0.04, P ¼ 0.85; clutch size: F1,152 ¼ 18.38, P < 0.001; mean brood size on day

K. A. Podlas, H. Richner / Animal Behaviour 86 (2013) 567e576

Hatching Spread Hatching spread (i.e. days between the day of first and last hatching per brood) increased significantly when the start of incubation of the last-laid eggs was experimentally delayed (GLM: incubation treatment: c21;156 ¼ 24:03, P<0.001; brood size manipulation: c22;154 ¼ 0:41, P ¼ 0.81; incubation treatment * brood size manipulation treatment: c22;150 ¼ 0:65, P ¼ 0.72; laying date: c21;153 ¼ 0:09, P ¼ 0.77; clutch size: c21;152 ¼ 1:89, P ¼ 0.17). Hatching spread was 1 day longer in ASYNC broods (median hatching spread ¼ 2 days. range 0e3) than in SYNC broods (median ¼ 1 day, range 0e2). There was a significant difference in hatching spread (i.e. calculated using the relative weight difference between siblings in the brood) between ASYNC and SYNC broods (GLM: incubation treatment: F1,138 ¼ 27.69, P < 0.001; brood size manipulation: F2,136 ¼ 1.87, P ¼ 0.16; incubation treatment * brood size manipulation treatment: F2,133 ¼ 0.27, P ¼ 0.76; laying date: F1,135 ¼ 0.24, P ¼ 0.62; clutch size: F1,138 ¼ 7.76, P ¼ 0.006). SYNC broods had a significantly shorter hatching spread than ASYNC broods (mean differences  1SE: 0.51  0.10). Brood Mortality The number of fledglings over the number of hatchlings per brood was not affected by the SYNCeASYNC treatment, the brood size treatment or their interaction (GLM: incubation treatment: c21;156 < 0:001, P ¼ 0.99; brood size manipulation: c22;154 ¼ 0:10, P ¼ 0.95; incubation treatment * brood size manipulation treatment: c22;150 ¼ 0:02, P ¼ 0.99; laying date: c21;153 ¼ 3:44, P ¼ 0.06; clutch size: c21;152 ¼ 11:16, P < 0.001; mean difference in number of fledglings between SYNC and ASYNC  1SE: 0.03  0.18 under reduced brood conditions, 0.03  0.17 under control brood conditions, 0.06  0.16 under enlarged brood conditions). Brood size at fledging was only affected by the brood size manipulation treatment (GLM: incubation treatment: F1,152 ¼ 0.23, P ¼ 0.63; brood size manipulation: F2,154 ¼ 3.37, P ¼ 0.037; incubation treatment * brood size manipulation treatment: F2,150 ¼ 0.23, P ¼ 0.79; laying date: F1,154 ¼ 7.62, P ¼ 0.006; clutch size: F1,153 ¼ 0.53, P ¼ 0.47), enlarged broods fledging more nestlings than reduced broods (mean differences  1SE: 1.16 nestling  0.45; post hoc t test: t154 ¼ 2.57, P ¼ 0.011). However, there was no difference in number of fledged nestlings between enlarged and control broods (mean differences  1SE: 0.45 nestling  0.45; post hoc t test: t154 ¼ 1.01, P ¼ 0.31) or between control and reduced broods (mean differences  1SE: 0.70 nestling  0.45; post hoc t test: t154 ¼ 1.57, P ¼ 0.12).

Table 1 LMM testing the effects of incubation treatment and brood size manipulation on the morphological traits of 14-day-old nestlings

Body mass Intercept Incubation treatment* Brood size manipulationy Laying date Clutch size Incubation treatment* brood size manipulation Wing length Intercept Incubation treatment* Brood size manipulationy Laying date Clutch size Incubation treatment* brood size manipulation Tarsus length Intercept Incubation treatment* Brood size manipulationy Laying date Clutch size Incubation treatment* brood size manipulation

Estimates1SE

F

df

P

24.341.33 0.640.41 L0.86±0.39 0.55±0.40 L0.14±0.02 L0.320.09 1.34±0.57 0.96±0.58

18.28 2.45 5.25

1,626 1,136 2,136

<0.001 0.12 0.006

34.49 12.02 2.93

1,136 1,136 2,136

<0.001 <0.001 0.056

60.382.53 L2.28±0.93 1.620.89 0.670.89 L0.26±0.05 0.110.21 4.73±1.30 3.36±1.33

23.84 6.00 1.66

1,626 1,137 2,137

<0.001 0.01 0.19

23.64 0.26 3.88

1,137 1,136 2,137

<0.001 0.61 0.001

20.160.41 0.0020.09 0.030.11 0.140.11 0.020.009 0.040.04 0.400.22 0.340.22

49.32 <0.001 0.96

1,626 1,138 2,140

<0.001 0.98 0.38

3.35 1.31 1.92

1,142 1,139 2,136

0.069 0.25 0.15

Significant terms in the model are highlighted in bold. * Relative to the SYNC broods. y Relative to the enlarged and the reduced broods.

difference  1SE: 0.64  0.41 g; post hoc t test: t136 ¼ 1.56, P ¼ 0.12; Table 1). However, among enlarged broods the nestlings in ASYNC broods tended, although not significantly, to be lighter than nestlings in SYNC broods (mean difference  1SE: 0.71  0.40 g; post hoc t test: t136 ¼ 1.79, P ¼ 0.076; Table 1). The effect of the incubation treatment on wing length of 14-day-old nestlings depended on the brood size treatment (LMM: significant interaction; Table 1;

16.2 P=0.43

Body mass (g)

9  1SE: 5.28  0.23 for reduced, 5.78  0.28 for control and 6.53  0.26 for enlarged broods) and brood size on day 14 (GLM: incubation treatment: F1,147 ¼ 0.16, P ¼ 0.69; brood size manipulation treatment: F2,149 ¼ 3.54, P ¼ 0.031; incubation treatment * brood size manipulation treatment: F2,145 ¼ 0.58, P ¼ 0.56; laying date: F1,148 ¼ 1.08, P ¼ 0.30; clutch size: F1,149 ¼ 7.83, P ¼ 0.006; mean brood size on day 14  1SE: 5.05  0.22 for reduced, 5.49  0.28 for control and 5.64  0.29 for enlarged broods).

571

15.3

P=0.12

P=0.076

Average Body Mass and Size of Nestlings Body mass of 14-day-old nestlings tended, although not significantly, to be differently affected by the incubation treatment in relation to the brood size manipulation treatment (LMM; Table 1, Fig. 1). Among reduced broods there was no difference in body mass between ASYNC and SYNC nestlings (mean difference  1SE: 0.32  0.41 g; post hoc t test: t136 ¼ 0.80, P ¼ 0.43; Table 1). A similar pattern was observed among control broods (mean

14.4 ASYNC

SYNC

Figure 1. Body mass (g) of 14-day-old nestlings in relation to the incubation and brood size manipulation treatments (mean  SE). ASYNC indicates asynchronous broods, SYNC indicates synchronous broods. Circles with ‘dashed’ connecting line indicate reduced broods, squares with ‘dotted’ line control broods and triangles with ‘dotdashed’ line enlarged broods.

572

K. A. Podlas, H. Richner / Animal Behaviour 86 (2013) 567e576

49.5

Wing length (mm)

P=0.25

P=0.008

47.5

P=0.015

45.5 ASYNC

SYNC

Figure 2. Wing length (mm) of 14-day-old nestlings in relation to the incubation and brood size manipulation treatments (mean  SE). ASYNC indicates asynchronous broods, SYNC indicates synchronous broods. Circles with ‘dashed’ connecting line indicate reduced broods, squares with ‘dotted’ line control broods and triangles with ‘dot-dashed’ line enlarged broods.

Fig. 2). Under reduced brood conditions there was no difference in wing length between ASYNC and SYNC nestlings (mean difference  1SE: 1.07  0.94 mm; post hoc t test: t137 ¼ 1.14, P ¼ 0.25). However, in the control groups, nestlings in ASYNC broods had longer wings than nestlings in SYNC broods (mean difference  1SE: 2.28  0.93 mm; post hoc t test: t137 ¼ 2.28, P ¼ 0.015), while under enlarged brood conditions nestlings in ASYNC broods had shorter wings than nestlings in SYNC broods (mean difference  1SE: 2.45  0.91 mm; post hoc t test: t137 ¼ 2.68, P ¼ 0.008). Tarsus length of 14-day-old nestlings was not affected by our treatments (LMM; Table 1). Parental Care The mean per capita feeding rate of ASYNC and SYNC broods was not significantly different between the experimental conditions (GLM: incubation treatment: F1,111 ¼ 0.05, P ¼ 0.82; brood size manipulation treatment: F2,112 ¼ 0.60, P ¼ 0.55; incubation treatment * brood size manipulation: F2,109 ¼ 0.40, P ¼ 0.67; laying date: F1,114 ¼ 0.61, P ¼ 0.43; brood size on day 9: F1,115 ¼ 1.67; P ¼ 0.19; mean per capita feeding rate for ASYNC reduced  1SE: 2.63 visits  0.40; mean per capita feeding rate for ASYNC control  1SE: 2.85 visits  0.33; mean per capita feeding rate for ASYNC enlarged  1SE: 2.94 visits  0.44; mean per capita feeding rate for SYNC reduced  1SE: 2.84 visits  0.32; mean per capita feeding rate for SYNC control  1SE: 2.94 visits  0.52; mean per capita feeding rate for SYNC enlarged  1SE: 2.52 visits  0.35). Coefficient of variation in nestling feeding rates did not differ between treatments (GLM: incubation treatment: F1,115 ¼ 0.23, P ¼ 0.63; brood size manipulation: F2,113 ¼ 0.32, P ¼ 0.73; incubation treatment*brood size manipulation: F2,110 ¼ 2.08, P ¼ 0.13; laying date: F1,112 ¼ 0.02, P ¼ 0.88; brood size on day 9: F1,116 ¼ 1.16, P ¼ 0.28; mean CV  1SE for ASYNC reduced: 0.79  0.14; mean CV  1SE for ASYNC control: 0.68  0.08; mean CV  1SE for ASYNC enlarged: 0.80  0.15; mean CV  1SE for SYNC reduced: 0.59  0.09; mean CV  1SE for SYNC control: 0.58  0.10; mean CV  1SE for SYNC enlarged: 0.75  0.09). The length of the nestling period did not differ between treatments (GLM: incubation treatment: c21;142 ¼ 0:13, P ¼ 0.72; brood

size manipulation: c22;140 ¼ 0:40, P ¼ 0.82; incubation treatment*brood size manipulation: c22;136 ¼ 0:03, P ¼ 0.98; laying date: c21;139 ¼ 0:41, P ¼ 0.52; clutch size: c21;138 ¼ 0:05, P ¼ 0.82; mean length for ASYNC reduced  1SE: 19.39  0.20 days; mean length for ASYNC control  1SE: 18.69  0.20 days; mean length for ASYNC enlarged  1SE: 18.82  0.24 days; mean length for SYNC reduced  1SE: 18.87  0.21 days; mean length for SYNC control  1SE: 18.47  0.34 days; mean length for SYNC enlarged  1SE: 18.67  0.23 days). Body condition of females was also not affected by our treatments (calculated by adding nestling tarsus length as a predictor variable to the analyses of body mass; GLM: incubation treatment: F1,132 ¼ 0.64, P ¼ 0.42; brood size manipulation: F2,129 ¼ 0.05, P ¼ 0.95; tarsus length: F1,134 ¼ 25.69, P < 0.001; incubation treatment * brood size manipulation: F2,127 ¼ 1.31, P ¼ 0.27; laying date: F1,131 ¼ 0.41, P ¼ 0.52; clutch size: F1,133 ¼ 1.05, P ¼ 0.31). The same pattern was observed for males (GLM: incubation treatment: F1,66 ¼ 0.04, P ¼ 0.83; brood size manipulation: F2,69 ¼ 0.41, P ¼ 0.66; tarsus length: F1,71 ¼ 16.09, P < 0.001; incubation treatment * brood size manipulation: F2,64 ¼ 1.50, P ¼ 0.23; laying date: F1,67 ¼ 0.04, P ¼ 0.83; clutch size: F1,68 ¼ 1.06, P ¼ 0.81). DISCUSSION The aim of this study was to test the effects of hatching asynchrony on offspring quality, rates of food provisioning and length of the nestling period as predicted by the brood reduction hypothesis. By removing eggs from nests and returning them for incubation either simultaneously or stepwise, we created clutches that were incubated and hatched either synchronously or asynchronously. Additionally, to simulate favourable, control and harsh conditions after hatching, we conducted a brood size manipulation experiment by transferring nestlings between broods. We did not find a difference in offspring mortality between asynchronous and synchronous broods under any of the experimental conditions. In harsh conditions, nestling mass and size were reduced in asynchronous broods compared to synchronous ones. The opposite pattern, however, occurred under control conditions. Overall, our results do not provide evidence in support of the brood reduction hypothesis and are in line with some previous studies on great tits as well as other passerines (Slagsvold 1982; Hillstrom & Olsson 1994; Horak 1995; Szollosi et al. 2007; but see Magrath 1989; Theofanellis et al. 2008). In contrast to what we predicted, ASYNC broods fledged as many young as SYNC broods independently of the brood size manipulation, and within ASYNC broods the last-hatched nestlings had a lower probability of fledging, independent of rearing conditions (see Appendix). Mortality resulting from the age-based competitive hierarchy in asynchronously hatched broods may be unlikely in great tits, and other passerines, as nestlings lack a sharp beak or claws to kill siblings, and thus mortality is more likely to occur through starvation. In harsh conditions induced by enlarged broods, ASYNC broods produced lighter and smaller nestlings than SYNC broods. Additionally, the last-hatched nestlings in ASYNCenlarged broods were lighter and smaller than last-hatched nestlings from ASYNC-reduced and ASYNC-control broods (see Appendix). These experimental results suggest that hatching asynchrony induced under bad conditions is not advantageous and rather results in detrimental effects for the whole brood, as found in previous studies (Slagsvold 1986; Hillstrom & Olsson 1994; Szollosi et al. 2007; but see Magrath 1989). Experiments involving brood size manipulation have shown that parents in many species of birds can successfully raise additional offspring to independence (reviewed in Dijkstra et al. 1990). However, it has also been shown that in enlarged broods parents are more often confronted with

K. A. Podlas, H. Richner / Animal Behaviour 86 (2013) 567e576

food-dependent trade-offs between offspring number and quality (Smith & Fretwell 1974; Winkler & Wallin 1987; Smith et al. 1988), or trade-offs between rearing larger broods and their own survival (Slagsvold 1982; Dijkstra et al. 1990). It seems from our results that under such conditions parents may incur an even greater cost from hatching asynchrony. One possible explanation is that rearing asynchronously hatched nestlings under bad conditions might be more constrained by complex interactions between prey size, energy demands and foraging efficiency than in good conditions (the feeding constraint hypothesis; Slagsvold & Wiebe 2007). For example, it could be that parents take more time to distribute the food between nestlings in asynchronous broods. It has been shown that females often selectively try to feed the smallest chick and this may take longer when she must ignore the larger and more competitive siblings (Stamps et al. 1989; Clutton-Brock & Vincent 1991). Moreover, a common problem with brood size manipulation experiments is that family dynamics are unintentionally altered. For example, in enlarged broods it is more difficult for a nestling to reach the parental feeding location, the parental choice of which nestling to feed is larger and thus may take more time, and begging contests may become more exaggerated (Kölliker et al. 1998; Kölliker & Richner 2004; Royle et al. 2012). While hatching asynchrony under poor environmental conditions is supposed to benefit parents, it probably does not play an important role under favourable conditions (Forbes 1990). In line with this idea, we found no effects of hatching asynchrony in reduced broods because ASYNC and SYNC broods produced a similar number and quality of nestlings. Studies testing the effects of brood size manipulation on nestling and parental fitness showed that reduced broods in general produce higher quality nestlings with higher recruitment probability the following breeding season, and that parental survival is enhanced when rearing a reduced brood (Dijkstra et al. 1990). Although our results do not support the brood reduction hypothesis, they still suggest that hatching asynchrony might be an adaptive maternal strategy, since under control conditions ASYNC broods produced as many but larger nestlings than SYNC broods. Body mass and size have been repeatedly shown to predict subsequent survival and thus might be good proxies for parental fitness (Tinbergen & Boerlijst 1990; Both et al. 1999). Moreover, longer wings may translate into higher survival after fledging since longer wings confer better flight ability and predator avoidance (Chin et al. 2009). In our study, the clutch size of control broods appeared to be the one best adjusted to parental capacity and environmental conditions. This may therefore suggest that hatching asynchrony is induced in order to facilitate rearing the number of nestlings parents produce and/or to adjust their offspring phenotypes to the conditions parents experience (Mousseau & Fox 1998) rather than to adjust the brood size to unpredictable conditions as in the case of the brood reduction hypothesis. It is worth mentioning that our experiment suffers from some limitations. First, in our study we used a brood size manipulation as a proxy for manipulations of food availability which may not guarantee similar effects as imposed by a food shortage. Thus, we cannot exclude the idea that the lack of a difference in mortality between treatments in our study resulted from the fact that the conditions created by the enlarged brood were not harsh enough to induce nestling mortality. Many studies performing brood size manipulation demonstrated that altricial birds are usually able to raise more nestlings than the number they opted for themselves (e.g. Dijkstra et al. 1990; Saino et al. 1997; Neuenschwander et al. 2003; Garcia-Navas & Sanz 2011). It has been shown that parents may increase activity and energy expenditure in providing more food for the young, they may build up fat reserves (Dijkstra et al. 1990) or they may selectively choose the type of prey delivered

573

to the brood (Wright & Cuthill 1990; but see Neuenschwander et al. 2003). However, these changes may have a negative effect on their survival or future reproduction (Dijkstra et al. 1990). Although it seems that in our study parents adjusted their feeding rates to brood sizes and different-aged chicks (mean feeding rates, length of nestling period and parental body condition at the end of the rearing period were the same in ASYNC and SYNC broods in all the experimental conditions), we cannot exclude the hypothesis that rearing enlarged broods will reduce survival and future reproduction, especially with asynchronous broods. This, however, entails measuring parental survival and condition over several years. Second, we did not manipulate food availability directly. As suggested by Forbes (1994) and Mock & Forbes (1994), the crucial benefits of brood reduction are realized only when food is short relative to brood needs, which may themselves vary stochastically (e.g. owing to changing weather). Therefore, as stated by Mock & Forbes (1994), a strict test of the brood reduction hypothesis requires that food is both in short supply and unpredictable. To summarize, our results from a manipulation of hatching asynchrony combined with a manipulation of brood size shows that under some circumstances hatching asynchrony might be beneficial, but they do not support the brood reduction hypothesis as an explanation for an adaptive function of hatching asynchrony. Further studies on the adaptive significance of hatching asynchrony may investigate the mechanisms and benefits of the phenotypic variation that it induces. Acknowledgments We are grateful to R. Gawlyta and Y. Pottier for help with field work. We also thank P. Smiseth and two anonymous referees for their constructive reviews of our manuscript. The study was financially supported by a Swiss National Science Foundation grant (310030B_138658 to H.R.). References Amundsen, T. & Slagsvold, T. 1991. Asynchronous hatching in the pied flycatcher: an experiment. Ecology, 72, 797e804. Amundsen, T. & Slagsvold, T. 1996. Lack’s brood reduction hypothesis and avian hatching asynchrony: what’s next? Oikos, 76, 613e620. Amundsen, T. & Slagsvold, T. 1998. Hatching asynchrony in great tits: a bethedging strategy? Ecology, 79, 295e304. Barba, E., Gildelgado, J. A. & Monros, J. S. 1995. The costs of being late: consequences of delaying great tit (Parus major) first clutches. Journal of Animal Ecology, 64, 642e651. Bernardo, J. 1996. Maternal effects in animal ecology. American Zoologist, 36, 83e 105. Blackburn, D. G. & Evans, H. E. 1986. Why are there no viviparous birds? The American Naturalist, 128, 165e190. Bortolotti, G. R. & Wiebe, K. L. 1993. Incubation behavior and hatching patterns in the American kestrel (Falco sparverius). Ornis Scandinavica, 24, 41e47. Both, C., Visser, M. E. & Verboven, N. 1999. Density-dependent recruitment rates in great tits: the importance of being heavier. Proceedings of the Royal Society B, 266, 465e469. Chin, E. H., Love, O. P., Verspoor, J. J., Williams, T. D., Rowley, K. & Burness, G. 2009. Juveniles exposed to embryonic corticosterone have enhanced flight performance. Proceedings of the Royal Society B, 276, 499e505. Clark, A. B. & Wilson, D. S. 1981. Avian breeding adaptations: hatching asynchrony, brood reduction and nest failure. Quarterly Review of Biology, 56, 253e277. Clotfelter, E. D., Whittingham, L. A. & Dunn, P. O. 2000. Laying order, hatching asynchrony and nestling body mass in tree swallows (Tachycineta bicolor). Journal of Avian Biology, 31, 329e334. Clutton-Brock, T. H. & Vincent, A. C. J. 1991. Sexual selection and the potential reproductive rates of males and females. Nature, 351, 58e60. Cotton, P. A., Wright, J. & Kacelnik, A. 1999. Chick begging strategies in relation to brood hierarchies and hatching asynchrony. American Naturalist, 153, 412e420. Cresswell, W. & McCleery, R. 2003. How great tits maintain synchronization of their hatch date with food supply in response to long-term variability in temperature. Journal of Animal Ecology, 72, 356e366. Deeming, D. C. 2002. Avian Incubation. Behaviour, Environment and Evolution. Oxford: Oxford University Press.

574

K. A. Podlas, H. Richner / Animal Behaviour 86 (2013) 567e576

Dijkstra, C., Bult, A., Bijlsma, S., Daan, S., Meijer, T. & Zijlstra, M. 1990. Brood size manipulations in the kestrel (Falco tinnunculus): effects on offspring and parent survival. Journal of Animal Ecology, 59, 269e285. Drent, R. 1975. Incubation. New York: Academic Press. Eikenaar, C., Berg, M. L. & Komdeur, J. 2003. Experimental evidence for the influence of food availability on incubation attendance and hatching asynchrony in the Australian reed warbler (Acrocephalus australis). Journal of Avian Biology, 34, 419e427. Eising, C. M., Eikenaar, C., Schwbl, H. & Groothuis, T. G. G. 2001. Maternal androgens in black-headed gull (Larus ridibundus) eggs: consequences for chick development. Proceedings of the Royal Society B, 268, 839e846. Farrell, E. D., Mariani, S. & Clarke, M. W. 2010. Reproductive biology of the starry smooth-hound shark (Mustelus asterias): geographic variation and implications for sustainable exploitation. Journal of Fish Biology, 77, 1505e1525. Fleming, A. S., Ruble, D., Krieger, H. & Wong, P. Y. 1997. Hormonal and experiential correlates of maternal responsiveness during pregnancy and the puerperium in human mothers. Hormones and Behavior, 31, 145e158. Forbes, L. S. 1990. Insurance offspring and the evolution of avian clutch size. Journal of Theoretical Biology, 147, 345e359. Forbes, L. S. 1994. The good, the bad and the ugly: Lack’s brood reduction hypothesis and experimental design. Journal of Avian Biology, 25, 338e343. Forbes, S. 2009. Portfolio theory and how parent birds manage investment risk. Oikos, 118, 1561e1569. Forbes, S., Thornton, S., Glassey, B., Forbes, M. & Buckley, N. J. 1997. Why parent birds play favourites. Nature, 390, 351e352. Freckleton, R. P. 2002. On the misuse of residuals in ecology: regression of residuals vs multiple regression. Journal of Animal Ecology, 71, 542e545. Galloway, L. F. 1995. Response to natural environmental heterogeneity: maternal effects and selection on life-history characters and plasticities in Mimulus guttatus. Evolution, 49, 1095e1107. Galloway, L. F. 2005. Maternal effects provide phenotypic adaptation to local environmental conditions. New Phytologist, 166, 93e100. Garcia-Navas, V. & Sanz, J. J. 2011. Short -term alterations in songbird breeding schedule lead to better synchronization with food availability. Auk, 128, 146e155. Gee, G. F., Hatfield, J. S. & Howey, P. W. 1995. Remote monitoring of parental incubation conditions in the greater sandhill crane. Zoo Biology, 14, 159e172. Groothuis, T. G. G. & Schwabl, H. 2008. Hormone-mediated maternal effects in birds: mechanisms matter but what do we know of them? Philosophical Transactions of the Royal Society B, 363, 1647e1661. Haftorn, S. 1979. Incubation and regulation of egg temperature in the willow tit (Parus montanus). Ornis Scandinavica, 10, 220e234. Haftorn, S. 1981. Incubation during the egg-laying period in relation to clutch size and other aspects of reproduction in the great tit (Parus major). Ornis Scandinavica, 12, 169e185. Haywood, S. & Perrins, C. M. 1992. Is clutch size in birds affected by environmental conditions during growth? Proceedings of the Royal Society B, 249, 195e197. Helfenstein, F., Berthouly, A., Tanner, M., Karadas, F. & Richner, H. 2008. Nestling begging intensity and parental effort in relation to prelaying carotenoid availability. Behavioral Ecology, 19, 108e115. Hillstrom, L. & Olsson, K. 1994. Advantages of hatching synchrony in the pied flycatcher (Ficedula hypoleuca). Journal of Avian Biology, 25, 205e214. Horak, P. 1995. Brood reduction facilitates female but not offspring survival in the great tit. Oecologia, 102, 515e519. Horak, P. 2003. When to pay the cost of reproduction? A brood size manipulation experiment in great tits (Parus major). Behavioral Ecology and Sociobiology, 54, 105e112. Howe, H. F. 1976. Egg size, hatching asynchrony, sex and brood reduction in common grackle. Ecology, 57, 1195e1207. Howe, H. F. 1978. Initial investment, clutch size, and brood reduction in the common grackle (Quiscalus quiscula L.). Ecology, 59, 1109e1122. Keller, L. F. & van Noordwijk, A. J. 1994. Effects of local environmental conditions on nestlings growth in the great tit (Parus major). Ardea, 82, 349e362. Kölliker, M. & Richner, H. 2004. Navigation in a cup: chick positioning in great tit (Parus major) nests. Animal Behaviour, 68, 941e948. Kölliker, M., Richner, H., Werner, I. & Heeb, P. 1998. Begging signals and biparental care: nestling choice between parental feeding locations. Animal Behaviour, 55, 215e222. Kontiainen, P., Pietiainen, H., Karell, P., Pihlaja, T. & Brommer, J. E. 2010. Hatching asynchrony is an individual property of female ural owls which improves nestling survival. Behavioral Ecology, 21, 722e729. Lacey, E. P. 1998. What is an adaptive environmentally induced parental effect? In: Maternal Effects as Adaptations (Ed. by T. A. Mousseau & C. W. Fox), pp. 54e66. Oxford: Oxford University Press. Lack, D. 1947. The significance of clutch size. Ibis, 89, 302e352. Magrath, R. D. 1989. Hatching asynchrony and reproductive success in the blackbird. Nature, 339, 536e538. Magrath, R. D. 1990. Hatching asynchrony in altricial birds. Biological Reviews of the Cambridge Philosophical Society, 65, 587e622. Mainwaring, M. C., Hartley, I. R., Gilby, A. J. & Griffith, S. C. 2010. Hatching asynchrony and growth trade-offs within domesticated and wild zebra finch (Taeniopygia guttata) broods. Biological Journal of the Linnean Society, 100, 763e773. Mainwaring, M. C., Blount, J. D. & Hartley, I. R. 2012. Hatching asynchrony can have long-term consequences for offspring fitness in zebra finches under captive conditions. Biological Journal of the Linnean Society, 106, 430e438.

Marshall, D. J. & Uller, T. 2007. When is a maternal effect adaptive? Oikos, 116, 1957e1963. Martins, T. F. & Wright, J. 1993. Brood reduction in response to manipulated brood sizes in the common swift (Apus apus). Behavioral Ecology and Sociobiology, 32, 61e70. Mock, D. W. & Forbes, L. S. 1994. Life-history consequences of avian brood reduction. Auk, 111, 115e123. Mock, D. W. & Parker, G. A. 1997. The Evolution of Sibling Rivalry. New York: Oxford University Press. Mock, D. W. & Ploger, B. J. 1987. Parental manipulation of optimal hatch asynchrony in cattle egrets: an experimental study. Animal Behaviour, 35, 150e160. Mousseau, T. A. & Fox, C. W. 1998. Maternal Effects as Adaptations. New York: Oxford University Press. Muller, W., Boonen, S., Groothuis, T. G. G. & Eens, M. 2010. Maternal yolk testosterone in canary eggs: toward a better understanding of mechanisms and function. Behavioral Ecology, 21, 493e500. Nager, R. G. & van Noordwijk, A. J. 1992. Energetic limitation in the egg-laying period of great tits. Proceedings of the Royal Society B, 249, 259e263. Neuenschwander, S., Brinkhof, M. W. G., Kölliker, M. & Richner, H. 2003. Brood size, sibling competition, and the cost of begging in great tits (Parus major). Behavioral Ecology, 14, 457e462. Nilsson, J. A. 1993. Energetic constraints on hatching asynchrony. American Naturalist, 141, 158e166. Nilsson, J. A. & Svensson, M. 1996. Sibling competition affects nestling growth strategies in marsh tits. Journal of Animal Ecology, 65, 825e836. van Noordwijk, A. J., McCleery, R. H. & Perrins, C. M. 1995. Selection for the timing of great tit breeding in relation to caterpillar growth and temperature. Journal of Animal Ecology, 64, 451e458. Parejo, D., Aviles, J. M. & Rodriguez, J. 2012. Supplemental food affects egg size but not hatching asynchrony in rollers. Behavioral Ecology and Sociobiology, 66, 1097e1105. Pettifor, R. A., Perrins, C. M. & McCleery, R. H. 2001. The individual optimization of fitness: variation in reproductive output, including clutch size, mean nestling mass and offspring recruitment in manipulated broods of great tits (Parus major). Journal of Animal Ecology, 70, 62e79. Pinheiro, J. C. & Bates, D. M. 2000. Mixed-effects models in S and S-Plus. New York: Springer. Raberg, L., Stjernman, M. & Nilsson, J. A. 2005. Sex and environmental sensitivity in blue tit nestlings. Oecologia, 145, 496e503. R Development Core Team 2007. R: a Language and Environment for Statistical Computing. Vienna: R Foundation for Statistical Computing. http://www.rproject.org. Rosivall, B., Szollosi, E. & Torok, J. 2005. Maternal compensation for hatching asynchrony in the collared flycatcher (Ficedula albicollis). Journal of Avian Biology, 36, 531e537. Royle, N. J., Pike, T. W., Heeb, P., Richner, H. & Kölliker, M. 2012. Offspring social network structure predicts fitness in families. Proceedings of the Royal Society B, 279, 4914e4922. Ryan, T. & Plague, G. 2004. Hatching asynchrony, survival, and the fitness of alternative adult morphs in Ambystoma talpoideum. Oecologia, 140, 46e51. Saino, N., Calza, S. & Moller, A. P. 1997. Immunocompetence of nestling barn swallows in relation to brood size and parental effort. Journal of Animal Ecology, 66, 827e836. Slagsvold, T. 1982. Clutch size, nest site and hatching asynchrony in birds: experiments with the fieldfare (Turdus pilaris). Ecology, 63, 1389e1399. Slagsvold, T. 1985. Asynchronous hatching in passerine birds: influence of hatching failure and brood reduction. Ornis Scandinavica, 16, 81e87. Slagsvold, T. 1986. Asynchronous versus synchronous hatching in birds: experiments with the pied flycatcher. Journal of Animal Ecology, 55, 1115e1134. Slagsvold, T. & Lifjeld, J. T. 1989. Constraints on hatching asynchrony and egg size in pied flycatchers. Journal of Animal Ecology, 58, 837e849. Slagsvold, T. & Lifjeld, J. T. 1990. Influence of male and female quality on clutch size in tits (Parus spp.). Ecology, 71, 1258e1266. Slagsvold, T. & Wiebe, K. L. 2007. Hatching asynchrony and early nestling mortality: the feeding constraint hypothesis. Animal Behaviour, 73, 691e700. Slagsvold, T., Husby, M. & Sandvik, J. 1992. Growth and sex-ratio of nestlings in two species of crowsehow important is hatching asynchrony. Oecologia, 90, 43e49. Smiseth, P. T. & Morgan, K. 2009. Asynchronous hatching in burying beetles: a test of the peak load reduction hypothesis. Animal Behaviour, 77, 519e524. Smith, C. C. & Fretwell, S. D. 1974. The optimal balance between size and number of offspring. The American Naturalist, 108, 499e506. Smith, H. G., Källander, H., Fontell, K. & Ljungström, M. 1988. Feeding frequency and parental division of labour in the double-brooded great tit (Parus major). Behavioral Ecology and Sociobiology, 22, 447e453. Sokal, R. R. & Rohlf, F. J. 1995. Biometry: the Principles and Practice of Statistics in Biological Research. 3rd edn. New York: W.H. Freeman. Stamps, J., Clark, A., Arrowood, P. & Kus, B. 1989. Begging behavior in budgerigars. Ethology, 81, 177e192. Stenning, M. J. 1996. Hatching asynchrony, brood reduction and other rapidly reproducing hypotheses. Trends in Ecology & Evolution, 11, 243e246. Stokland, J. N. & Amundsen, T. 1988. Initial size hierarchy in broods of the shag: relative significance of egg size and hatching asynchrony. The Auk, 105, 308e315. Stoleson, S. H. & Beissinger, S. R. 1997. Hatching asynchrony, brood reduction, and food limitation in a neotropical parrot. Ecological Monographs, 67, 131e154.

K. A. Podlas, H. Richner / Animal Behaviour 86 (2013) 567e576 Szollosi, E., Rosivall, B. & Torok, J. 2007. Is hatching asynchrony beneficial for the brood? Behavioral Ecology, 18, 420e426. Tanner, M., Kolliker, M. & Richner, H. 2008. Differential food allocation by male and female great tit (Parus major) parents: are parents or offspring in control? Animal Behaviour, 75, 1563e1569. Theofanellis, T., Galinou, E. & Akriotis, T. 2008. The role of hatching asynchrony in brood size reduction of the great tit (Parus major) in a Mediterranean pine forest. Journal of Natural History, 42, 375e380. Tilgar, V. & Mand, R. 2006. Sibling growth patterns in great tits: does increased selection on last-hatched chicks favour an asynchronous hatching strategy? Evolutionary Ecology, 20, 217e234. Tinbergen, J. M. & Boerlijst, M. C. 1990. Nestling weight and survival in individual great tit (Parus major). Journal of Animal Ecology, 59, 1113e1127. Veiga, J. P. & Vinuela, J. 1993. Hatching asynchrony and hatching success in the house sparrow: evidence for the egg viability hypothesis. Ornis Scandinavica, 24, 237e242. Vinuela, J. 2000. Opposing selective pressures on hatching asynchrony: egg viability, brood reduction, and nestling growth. Behavioral Ecology and Sociobiology, 48, 333e343. Visser, M. E., van Noordwijk, A. J., Tinbergen, J. M. & Lessells, C. M. 1998. Warmer springs lead to mistimed reproduction in great tits (Parus major). Proceedings of the Royal Society B, 265, 1867e1870. While, G. M., Jones, S. M. & Wapstra, E. 2007. Birthing asynchrony is not a consequence of asynchronous offspring development in a non-avian vertebrate, the Australian skink (Egernia whitii). Functional Ecology, 21, 513e519. Wiebe, K. L. & Bortolotti, G. R. 1994. Food supply and hatching spans of birds: energy constraints of facultative manipulation. Ecology, 75, 813e823. Wiebe, K. L., Wiehn, J. & Korpimaki, E. 1998. The onset of incubation in birds: can females control hatching patterns? Animal Behaviour, 55, 1043e1052. Wiehn, J., Ilmonen, P., Korpimaki, E., Pahkala, M. & Wiebe, K. L. 2000. Hatching asynchrony in the Eurasian kestrel (Falco tinnunculus): an experimental test of the brood reduction hypothesis. Journal of Animal Ecology, 69, 85e95. Winkler, D. W. & Wallin, K. 1987. Offspring size and number: a life history model linking effort per offspring and total effort. The American Naturalist, 129, 708e720. Wright, J. & Cuthill, I. 1990. Biparental care: short-term manipulation of partner contribution and brood size in the starling, Sturnus vulgaris. Behavioral Ecology, 1, 116e124. Zuur, A. F., Ieno, E. N., Walker, N. J., Saveliev, A. A. & Smith, G. M. 2009. Mixed Effects Models and Extensions in Ecology with R. New York: Springer.

APPENDIX Additional Predictions Concerning Specific Nestlings For nestlings, the brood reduction hypothesis predicts that, especially under harsh conditions, in asynchronous broods firsthatched nestlings should grow to be heavier and bigger, and have higher survival probability than first-hatched nestlings in synchronous broods. In asynchronous broods under harsh conditions last-hatched nestlings should grow to be lighter and smaller, as well as have lower survival probability, than last-hatched nestlings in asynchronous broods under favourable conditions. Statistical Analyses To evaluate size and body mass differences between nestlings of different hatching order, we used LMM including brood size manipulation treatment, hatching order (three-level factor: nestlings hatched on day 0, hatched on day 1 and hatched on day 2 or later) and their two-way interaction as fixed factors and brood identity as a random factor. To evaluate differences in individual nestling mortality in ASYNC broods with respect to hatching order, we used generalized linear mixed models (GLMMs) with binomial distribution of errors including brood size manipulation treatment, hatching order and their two-way interaction as fixed factors and brood identity as a random factor. This was done for ASYNC broods only since in SYNC broods most of the nestlings had the same hatching order, that is, hatched on the same day. To evaluate size and body mass differences between first-hatched offspring in ASYNC and SYNC broods in different rearing conditions, we used an LMM including incubation treatment, brood size manipulation treatment and their two-way interaction as fixed factors and brood identity as a random factor. A GLMM with binomial distribution

575

of errors was used to test differences in individual mortality of firsthatched nestlings between ASYNC and SYNC broods with respect to the brood size manipulation treatment. This was done for nestlings only with the same hatching day, that is, hatched on day 0. The variance structures of the hatching order error in the analysis of nestling traits in ASYNC broods were modelled with varIdent, that is, with different variances for each level of stratification (Pinheiro & Bates 2000). To analyse the LMMs we used function lme from the nlme package, whereas for GLMMs with binomial error structure the function lmer from the lme4 package was used (Zuur et al. 2009). To estimate P values for interactions in GLMMs, we used the ANOVA command which compares the two models that include or exclude the interaction (Zuur et al. 2009). Results Body mass and size of nestlings according to their hatching order in ASYNC broods Within ASYNC broods, nestlings’ body mass and wing length but not tarsus length on day 14 were differently affected by the brood size manipulation treatment in relation to their hatching order (LMM; Table A1). Nestlings hatched on day 0 under reduced brood conditions were heavier (LMM; Table A2) but had similar wing length as nestlings in the same hatching order under both control and enlarged brood conditions (LMM; Table A2). However, nestlings hatched on day 0 in control brood conditions were heavier and had longer wings than nestlings in the same hatching order in enlarged brood conditions (LMM; Table A2). Nestlings hatched on day 1 under reduced brood conditions were heavier but had similar wing length as nestlings in the same hatching order under both control and enlarged brood conditions (LMM; Table A2). Nestlings hatched on day 1 under control brood conditions were heavier but had similar wing length as those hatched on day 1 under enlarged brood conditions (LMM; Table A2). Nestlings hatched on day 2 under reduced brood conditions did not differ in their body mass and wing length from nestlings in the same hatching order under control conditions (LMM; Table A2), but were heavier and had longer wings than those under enlarged brood conditions (LMM; Table A2). Nestlings hatched on day 2 under control brood conditions were heavier and had longer wings than nestlings hatched on day 2 under enlarged conditions (LMM; Table A2). Individual offspring survival in ASYNC broods was only affected by the nestling’s hatching order (GLMM: brood size manipulation treatment * hatching order: c24 ¼ 0:79, P ¼ 0.94; all z < 1.33, all P > 0.18). In ASYNC broods, nestlings hatched on day 1 and on day 2 had lower fledging probability than those hatched on day 0 (GLMM: mean difference  1SE for hatched on day 0 versus hatched on day 1: 1.63  0.35; z ¼ 4.63, P < 0.001; mean difference  1SE for hatched on day 0 versus hatched on day 2: 2.29  0.50; z ¼ 4.57, P < 0.001). However, there was no difference in fledging probability between nestlings hatched on days 1 and 2 (mean difference  1SE: 2.29  0.46; z ¼ 1.43, P ¼ 0.15). Body mass and size of first-hatched nestlings Within nestlings hatched on day 0, only body mass and wing length of nestlings differed significantly between ASYNC and SYNC broods according to the brood manipulation treatment (LMM; Table A3). Under reduced and enlarged brood conditions nestlings hatched on day 0 did not differ in their body mass and wing length between ASYNC and SYNC broods (mean differences  1SE under reduced conditions: body mass: 0.03  0.39 g; post hoc t test: t136 ¼ 0.07, P ¼ 0.94; wing length: 0.34  0.94 mm; post hoc t test: t137 ¼ 0.37, P ¼ 0.71; mean differences  1SE under enlarged conditions: body mass: 0.52  0.38 g; post hoc t test: t136 ¼ 1.39, P ¼ 0.17; wing length: 1.27  0.91 mm; post hoc t test: t137 ¼ 1.40, P ¼ 0.16). In

576

K. A. Podlas, H. Richner / Animal Behaviour 86 (2013) 567e576

contrast, under control brood conditions nestlings in ASYNC broods tended to be heavier and had longer wings than those hatched in SYNC broods (mean differences  1SE: body mass: 0.74  0.38 g; post hoc t test: t136 ¼ 1.92, P ¼ 0.056; wing length: 3.04  0.93 mm; post hoc t test: t137 ¼ 3.28, P ¼ 0.001). Individual offspring survival between ASYNC and SYNC broods was similar in all experimental conditions (GLMM: brood size manipulation treatment * hatching order: c22 ¼ 0:38, P ¼ 0.83; all z < 0.89, all P > 0.37). Table A1 LMM testing the effects of brood size manipulation and hatching order on the morphological traits of 14-day-old nestlings in ASYNC broods Estimates1SE Body mass Intercept Brood size manipulation* Hatching ordery Laying date Clutch size Brood size manipulation* hatching order

Wing length Intercept Brood size manipulation* Hatching ordery Laying date Clutch size Brood size manipulation* hatching order

Tarsus length Intercept Brood size manipulation* Hatching ordery Laying date Clutch size Brood size manipulation* hatching order

F

df

P

25.071.82 L0.77±0.37 0.79±0.39 L0.58±0.10 L0.32±0.24 L0.12±0.03 L0.49±0.13 L0.17±0.14 0.07±0.16 L1.94±0.36 L0.18±0.31

189.64 6.12

1,321 2,69

<0.001 0.004

15.01

2,321

<0.001

14.75 14.42 9.16

1,69 1,69 4,321

<0.001 <0.001 <0.001

63.753.49 1.520.74 0.270.74 L3.88±0.21 5.430.48 L0.22±0.06 0.360.25 1.04±0.31 0.41±0.32 L1.78±0.67 0.78±0.59

332.71 2.24

1,321 2,70

<0.001 0.11

205.01

2,321

<0.001

12.98 2.16 9.27

1,70 1,69 4,321

<0.001 0.15 <0.001

20.490.58 L0.23±0.12 0.15±0.13 L0.28±0.04 L0.32±0.07 0.0050.01 L0.08±0.04 0.170.08 0.080.09 0.050.18 0.190.19

1247.3 3.75

1,321 2,70

<0.001 0.028

33.40

2,321

<0.001

0.29 2.74 2.02

1,69 1,70 4,321

0.59 0.10 0.092

Estimates1SE

t

df

L1.52±0.74 0.270.74 1.250.80

2.04 0.37 1.56

1,70 1,70 1,70

0.044 0.71 0.12

2.50 2.16 4.01

1,69 1,69 1,69

0.015 0.034 <0.001

0.63 0.18 0.76

1,70 1,70 1,70

0.53 0.86 0.45

5.56 1.29 6.32

1,69 1,69 1,69

<0.001 0.20 <0.001

3.52 0.56 4.06

1,70 1,70 1,70

<0.001 0.58 <0.001

Nestlings hatched on day 1 Body mass Enlarged vs Control L0.94±0.37 Reduced vs Control 0.86±0.40 Reduced vs Enlarged 1.79±0.45 Wing length Enlarged vs Control 0.480.76 Reduced vs Control 0.130.75 Reduced vs Enlarged 0.610.81 Nestlings hatched on day 2 Body mass Enlarged vs Control L2.71±0.49 Reduced vs Control 0.610.47 Reduced vs Enlarged 3.32±0.52 Wing length Enlarged vs Control L3.30±0.94 Reduced vs Control 0.500.89 Reduced vs Enlarged 3.80±0.94

P

Table A3 LMM testing the effects of incubation and brood size manipulation treatment on the morphological traits of 14-day-old nestlings hatched on day 0 Estimates1SE Body mass Intercept Incubation treatment* Brood size manipulationy Laying date Clutch size Incubation treatment* Brood size manipulation Wing length Intercept Incubation treatment* Brood size manipulationy Laying date Clutch size Incubation treatment* Brood size manipulation

Table A2 Post hoc t tests testing the effects of brood size manipulation and hatching order on the morphological traits of 14-day-old nestlings in ASYNC broods

Nestlings hatched on day 0 Body mass Enlarged vs Control L0.77±0.37 Reduced vs Control 0.79±0.39 Reduced vs Enlarged 1.55±0.44

Wing length Enlarged vs Control Reduced vs Control Reduced vs Enlarged

Significant terms in the model are highlighted in bold.

Significant terms in the model are highlighted in bold. * Relative to the enlarged and reduced rearing conditions. y Relative to the nestlings hatched on day 1 and day 2.

Estimates1SE

(continued )

t

df

P

2.07 1.99 3.49

1,69 1,69 1,69

0.042 0.049 <0.001

Tarsus length Intercept Incubation treatment* Brood size manipulationy Laying date Clutch size Incubation treatment* Brood size manipulation

F

df

P

23.641.26 L0.74±0.38 L0.85±0.37 0.43±0.39 L0.12±0.02 L0.27±0.09 1.26±0.54 0.77±0.55

350.35 3.72 4.86

1,317 1,136 2,136

<0.001 0.056 0.009

30.35 9.29 2.77

1,136 1,136 2,136

<0.001 0.003 0.066

60.432.99 3.040.93 1.520.89 0.570.90 L0.22±0.05 0.030.21 4.31±1.30 2.69±1.32

407.54 10.76 1.46

1,317 1,137 2,137

<0.001 0.001 0.24

17.17 0.02 5.58

1,137 1,136 2,137

<0.001 0.89 0.005

1753.6 3.06 0.56

1,317 1,142 2,138

<0.001 0.082 0.57

2.11 0.03 1.45

1,141 1,140 2,136

0.15 0.87 0.24

20.400.49 0.150.09 0.060.11 0.060.11 0.010.008 0.0050.03 0.340.21 0.260.21

Significant terms in the model are highlighted in bold. * Relative to the SYNC broods. y Relative to the enlarged and reduced rearing conditions.