SEARCH

SEARCH BY CITATION

Keywords:

  • egg composition;
  • egg mass;
  • maternal effects;
  • parental investment;
  • reproductive strategies

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  • 1
    Egg size is an important maternal trait that can have major consequences on offspring phenotype. However, the effects of the variation of different components of cleidoic eggs have been little investigated.
  • 2
    Here, we addressed whether a reduction of the relative egg albumen content within the natural range of variation affects viability, time to hatching, early post-natal begging displays, morphology and immune response of yellow-legged gull Larus michahellis chicks.
  • 3
    Egg mass strongly positively predicted chick size and mass at all ages, while time to hatching positively predicted tarsus length and immune response, irrespective of albumen removal. Variation in time to hatching may thus affect immune system maturation.
  • 4
    Albumen removal resulted in a lower embryonic viability and increased time to hatching of individual eggs. The probability that an egg originated a chick surviving until 8 days of age increased with original egg mass among controls, but not among chicks hatching from eggs with reduced albumen content (‘albumen chicks’).
  • 5
    Begging rate increased with laying order among albumen chicks while it decreased among controls. Concomitantly, begging rate decreased with egg mass among controls while it did not vary among albumen chicks. Surprisingly, albumen removal did not affect body mass or tarsus length except at 8 days of age, when control chicks were lighter than albumen chicks.
  • 6
    In conclusion, our study indicates that a reduction of the relative egg albumen content can have complex effects on offspring development, behaviour and viability of a semiprecocial bird, suggesting that the relative albumen content of the eggs represents an important mechanism of maternal effects.

Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Maternal effects via propagule size are traditionally recognized as a major class of parental effects in ecological studies (Bernardo 1996; Mousseau & Fox 1998). According to Kirkpatrick & Lande (1989), hatchling size can be regarded as ‘maternally inherited’, as phenotypic values of the offspring are determined by a mother's contribution beyond direct genetic effects (i.e. via the egg). As a result, propagule size is a peculiar character, in that it is a joint feature of the mother and of the offspring (Bernardo 1996).

A considerable proportion of the studies on maternal effects mediated by egg features have been carried out in birds. These studies have mostly focused on the correlation between egg mass and offspring phenotypic traits, including putative fitness correlates (e.g. body size and tarsus length), during early ontogenetic stages (reviews in Williams 1994; Christians 2002).

It could be argued, however, that the investigation of the individual egg components, rather than whole egg mass, may better unravel the mechanisms of maternal effects mediated via the egg. The yolk and albumen fractions, for example, have distinctive roles during development (Carey, Rahn & Parisi 1980; Sotherland & Rahn 1987). The yolk is the main source of lipids and is implicated more in mass gain than in body size development, whereas the albumen is a key source of water and proteins (Romanoff & Romanoff 1949; Carey et al. 1980; Sotherland & Rahn 1987). In particular, albumen appears to be a functionally overlooked component of the egg (but see Saino et al. 2005; Ferrari, Martinelli & Saino 2006), despite some authors having suggested that albumen mass may represent the most crucial portion of the egg in determining hatchling performance, as the yolk's participation in tissue formation is less relevant (Nisbet 1978).

In this framework, experimental manipulation of egg components may provide useful insights (Sinervo et al. 1992). Previous studies of poultry have shown that albumen removal could result in reduced body mass and tarsus length of chicks (Hill 1993; Finkler, van Orman & Sotherland 1998). The only study carried out in the wild, in the barn swallow Hirundo rustica, showed that albumen removal reduced nestling growth and survival, mainly of late-hatched chicks (Ferrari et al. 2006). Thus, in that species, the observed increase in both absolute and relative albumen content of the eggs along the laying sequence may mediate a maternal ‘brood survival’ strategy (sensu Lack 1954) – whereby parents increase the number of viable offspring by mitigating the negative consequences of hatching late by, for example, laying the larger eggs last. Conversely, ‘brood reduction’ strategies have been widely documented among bird species, where within-brood size hierarchies are often established by asynchronous egg hatching coupled with reduced size of the last laid eggs. In these species, no increase in relative albumen content along the egg laying sequence should be expected, unlike the pattern observed in barn swallows (see above).

In the present study of the yellow-legged gull Larus michahellis, a species adopting a brood reduction strategy, with markedly smaller last laid eggs (Cramp 1998), we first analysed variation in relative albumen content of the eggs according to laying order, while predicting no increase in last laid eggs. In addition, we investigated the consequences of a reduction in albumen content within the natural range of variation of albumen mass relative to total egg mass on a suite of phenotypic traits (including body size, intensity of begging behaviour and immune response) of the chicks under natural conditions. We predicted that nestlings from experimentally manipulated eggs hatched later and attained smaller body size and mass (Ferrari et al. 2006). We also expected positive effects of albumen removal on begging, based on the assumption that the intensity of begging display honestly reflects offspring need (Wright & Leonard 2002). Moreover, we speculated that immune response could also be negatively influenced by albumen removal because of expected generalized negative effects on offspring growth and development. Such negative consequences on phenotype were predicted to be more severe among chicks hatched from smaller last laid compared with earlier laid eggs within each clutch, because of the close association between hatching and laying order and the lower competitive ability of late hatched chicks.

Materials and methods

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

study species, field procedures and albumen content manipulation

The yellow-legged gull is a large, semicolonial bird (Cramp 1998), which lays clutches of one to three eggs, weighing 80–100 g each. Eggs are laid at 1–3-day intervals, and egg size declines with the laying sequence (Rubolini et al. 2005; see also below). Hatching takes place 27–30 days after laying and is typically asynchronous (see Rubolini et al. 2005). The semiprecocial chicks remain near the nest for the first few days after hatching.

The experiment was carried out during spring 2005 at a monospecific colony (c.300 pairs) in the Comacchio lagoon (north-east Italy). The colony was visited daily or every second day (depending on the breeding stage and weather conditions), and all eggs and nests were marked. Clutches were randomly assigned to either of two treatments (albumen removal or control, hereafter albumen or control clutches, respectively), with a ratio of 3 : 2. Egg mass at laying did not differ between treatment groups (mixed model anova with nest identity as a random effect, F1,69 = 0·01, P = 0·93), and no interaction between laying order and treatment was detected (F2,127 = 1·27, P = 0·29). However, egg mass strongly declined with laying order (F2,129 = 83·4, P < 0·0001; all post-hoc pairwise comparisons P < 0·001). Owing to clutch desertion or predation, the final sample size consisted of 44 albumen and 27 control clutches (127 and 75 eggs, respectively). Once marked, the eggs were temporarily removed from the nest for a maximum of 4–5 h, during which they were replaced with dummy gull eggs, to extract the albumen. Albumen removal therefore occurred within the first day after laying. All eggs were weighed (nearest 1 g) before being subjected to the albumen removal treatment. The experiment was designed in order to reduce the proportion of albumen relative to egg mass within the natural range of variation. We therefore decided to remove an albumen fraction equal to 1 SD of the mean proportion of albumen, calculated for eggs of each laying order (a-, b-, or c-eggs). The mean proportion of albumen was obtained based on a sample of 16 three-egg clutches collected in the same area at the beginning of the season. The yolk and albumen were carefully separated and weighed (accuracy 0·01 g). The amount of albumen removed was calculated as the product of the estimated albumen mass for a given egg mass [as obtained from linear regressions of albumen mass on egg mass for eggs of each laying order (a-eggs: r2 = 0·89; b-eggs: r2 = 0·79; c-eggs: r2 = 0·90)] and the SD of the mean albumen proportion (see also Ferrari et al. 2006 for a similar procedure). This resulted in a mean proportion of removed albumen equal to 2·29% (0·34 SD) of the estimated egg albumen content.

Before removing the albumen, we carefully disinfected the eggshell. During extraction, eggs were kept horizontal, with the acute pole pointing slightly upwards, and a hole was drilled into the eggshell at c.1 cm from the acute pole by means of a sterile pin of the same diameter of the syringe needle. Albumen was extracted by means of sterile 2·5 mL disposable syringes (22 gauge needle), and the hole sealed immediately afterwards. The syringe content was weighed (accuracy 0·01 g) to check for extraction accuracy. The difference between the planned and extracted amounts of albumen did not significantly deviate from 0 [0·003 (0·07 SD), one sample t-test: t126 = 0·47, P = 0·64], indicating that our removal procedure was accurate. Control eggs were subjected to exactly the same procedures, including insertion of the needle in the albumen, but no albumen was extracted.

chick phenotypic traits

Nests were checked daily around the time of hatching. In order to correctly link each chick to its original egg, when the embryo produced a small opening in the eggshell we injected a small amount of blue or green food dye, which allowed us to assign every chick to its egg of origin (see also Rubolini et al. 2005). When first found hatched, at the age of 0·61 [(0·05 SE); 0 = day of hatching of the individual (egg)] days (hereafter, age 1), chicks were marked with combinations of coloured elastic leg bands, weighed (to the nearest 1 g), and the length of their right tarsus measured (to the nearest 0·01 mm).

On day 1, we recorded the intensity of begging behaviour according to a standard protocol (see Rubolini et al. 2005). Briefly, chicks were presented with a realistic dummy head of an adult gull, to which they respond by pecking towards the red patch on the lower mandible, a behaviour that triggers food regurgitation by parents (Tinbergen 1967). The intensity of this display was expressed as the number of pecks directed to the dummy head in 1-min trials (hereafter ‘begging rate’). Throughout this study, we will assume that the intensity of begging positively reflects hunger, although we cannot rule out the possibility that some chicks in very poor state could not perform vigorous begging despite not being satiated. We could not assess satiation levels of the chicks because of practical reasons (e.g. prolonged observer disturbance at the colony site).

At an age of 4·52 (0·05 SE) and 8·58 (0·05 SE) days (hereafter ages 4 and 8, respectively), we recorded again body mass and tarsus length. At age 8, we also assessed the intensity of the T-cell-mediated immune response by means of a standard in vivo test (Lochmiller, Vestey & Boren 1993; Saino, Calza & Møller 1997; Tella, Scheuerlein & Ricklefs 2002). The wing web of the right wing was injected subcutaneously with 0·2 mg of a lectin (phytohaemagglutinin) dissolved in 0·05 mL phosphate-buffered saline (PBS). As a control, we injected the same amount of PBS in the left wing web. The thickness of both webs at the site of inoculation was measured using a pressure sensitive micrometer prior to the inoculation and approximately 24 h afterwards. The difference in variation in thickness between the right and the left wing webs was assumed to be an index of the T-cell-mediated immune response.

statistical analyses

The effects of albumen manipulation on egg hatchability, time to hatching (days elapsed between laying and hatching), chick traits and viability were analysed using mixed model analyses of variance, with nest identity as a random factor, as implemented by the SAS system (ver. 9·0). In all analyses, treatment and laying order were included as factors, while original egg mass and age at measurement (to account for small variations in the day at measurement relative to hatching date) as covariates, together with all the two-way interactions between treatment, laying order and egg mass. In specific analyses, other predictors were also included (see Results). Nonsignificant terms (P > 0·05) were sequentially removed (starting from interactions) until minimal adequate models, containing only significant terms, were obtained (Crawley 1993). For brevity, results of nonsignificant terms at removal are not presented, apart from the effect of treatment. Hatching success and viability were analysed assuming a binomial error distribution by means of the SAS GLIMMIX macro. Means and parameter estimates are reported together with their standard errors, unless stated otherwise.

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

variation in relative egg albumen content

As predicted, the proportion of albumen did not vary according to laying order [mean: a-eggs, 0·634 (0·025 SD); b-eggs, 0·632 (0·025 SD); c-eggs, 0·639 (0·025 SD); mixed model analysis of variance, effect of laying order: F2,30 = 0·92, P = 0·41]. Overall, the mean proportion of egg albumen across all eggs was 0·635 (0·023), and showed a marked among-clutch variation (F15,32 = 3·30, P = 0·002).

hatching success, time to hatching and post-hatching viability

Overall, the eggs that were subjected to albumen removal (‘albumen eggs’) had a lower hatching success than controls, but the effect of egg treatment on the probability that an egg reached the pipping stage depended on laying order (Fig. 1; treatment × laying order interaction, F2,127 = 8·90, P = 0·0002). In fact, viability of first and second eggs was reduced by 35% or, respectively, 39% among albumen eggs compared with control eggs, whereas viability of third eggs was slightly lower among control compared with albumen eggs (Fig. 1). When the interaction term was removed, a significant effect of treatment emerged (F1,77·3 = 12·7, P = 0·0006). The effect of nest identity was highly significant (Z = 3·50, P = 0·0002).

image

Figure 1. (a) Hatching success (proportion of eggs that hatched) and (b) probability that an egg originated a chick surviving until age 8 in relation to treatment and laying order. Samples size (n of eggs or chicks) is reported above bars. Open bars represent control eggs (n = 60) or chicks (n = 28), black bars albumen eggs (n = 72) or chicks (n = 43). In panel (a), different letters denoting significant differences (P < 0·05) at post-hoc pairwise comparisons have been added above bars.

Download figure to PowerPoint

Time to hatching varied according to treatment (F1,49 = 7·19, P = 0·009), albumen eggs hatching approximately 1 day later than controls [least squares mean of control eggs = 27·73 (0·205); albumen eggs = 28·48 (0·184)]. Laying order also influenced time to hatching (F1,93·3 = 47·4, P < 0·0001), a-eggs hatching approximately 2 days earlier than b- and c-eggs (P < 0·001 at post-hoc tests), while no significant difference existed between b- and c-eggs (P > 0·10). Time to hatching increased by 0·038 (0·019) days for each gram of increase in fresh original egg mass (F1,82·8 = 4·04, P = 0·048). The between-nest variation was significant (Z = 2·56, P = 0·005).

The probability that an egg originated a chick surviving to age 8 was 36% (28 of 75) for control chicks and 35% (45 of 127) for albumen chicks (Fig. 1). The effect of treatment on the probability that an egg originated a chick surviving to age 8 varied according to the mass of the original egg (egg mass × treatment interaction, F1,183 = 4·33, P = 0·039). The slope (in logits) for the albumen group was −0·020 (0·035) and not significantly different from 0 (t155 = −0·57, P = 0·57) while the slope for the control group was 0·098 (0·04) and significantly larger than 0 (t155 = 2·20, P = 0·029), implying that among control chicks viability was positively predicted by the mass of their egg.

begging behaviour, morphology and immunity

We analysed the effects of egg treatment on chick phenotype in mixed analysis of variance models where laying order, egg mass and time to hatching were included as predictors, together with two-way interactions.

Begging rate declined monotonically with laying order among control chicks, whereas the pattern of variation was opposite among albumen chicks (Table 1; Fig. 2). The difference in begging rate between treatment groups was most pronounced among chicks hatched from c-eggs (Fig. 2). In addition, begging rate declined with egg mass among chicks from control eggs [coefficient = −0·83 (0·39), t79·9 = −2·07, P = 0·041], whereas it slightly increased among albumen chicks [coefficient = 0·39 (0·33), t65·6 = 1·19, P = 0·24] (see Table 1).

Table 1.  Minimal adequate mixed analysis of variance model of the intensity of chick begging displays (see Statistical analyses and Fig. 2 for details). Brood identity was included as a random factor (see Results). Denominator d.f. are estimated according to Satterthwaithe's approximation
Source of variationFd.f.P
Treatment5·401, 74·60·023
Laying order0·022, 90·40·98
Egg mass0·731, 740·39
Treatment × laying order3·922, 90·40·023
Egg mass × treatment5·571, 740·021
image

Figure 2. Mean (SE) intensity of chick begging displays (pecks min−1, see Methods) in relation to egg laying order and experimental treatment. Samples size (n of chicks) is reported above bars. Open bars represent control chicks (n = 57), black bars albumen chicks (n = 69). Different letters denote significant differences (P < 0·05) at post-hoc pairwise comparisons.

Download figure to PowerPoint

The effect of egg treatment on chick morphology at ages 1 and 4 was nonsignificant (tarsus length: age 1, F1,58·8 = 0·13, P = 0·72; age 4, F1,52·8 = 0·01, P = 0·93; body mass: age 1, F1,54·6 = 0·05, P = 0·83; age 4, F1,56= 0·08, P = 0·78) (Fig. 3). At age 8, the effect of treatment was nonsignificant on tarsus length (F1,37·3 = 3·65, P = 0·064), while it was significant on body mass (Table 2). Contrary to our expectation, body mass was 15% smaller among control chicks compared with albumen ones [least-squares means, control chicks = 201·77 g (8·55); albumen chicks = 233·12 g (6·48)] (Fig. 3). This effect persisted also when tarsus length was included in the model to account for the effect of body size on body mass (effect of treatment, F1,40·1= 5·15, P = 0·028).

image

Figure 3. Mean (SE) (a) body mass and (b) tarsus length of albumen and control chicks at different ages. Samples size (n of chicks) is reported above bars. Open bars represent control chicks, black bars albumen chicks.

Download figure to PowerPoint

Table 2.  Minimal adequate mixed analysis of variance models of nestling phenotype measured at different ages (see Statistical analyses). Brood identity was included as a random factor (see Results). The number of chicks is reported in parentheses. Denominator d.f. are estimated according to Satterthwaithe's approximation.
Source of variationFd.f.PCoefficient (SE)
  1. Note: all significant effects would remain as such even when the α-level is corrected for multiple testing according to the Sidak's procedure (see Uitenbroek 1997).

Tarsus length (age 1) (n = 128)
 Egg mass 29·111, 109< 0·001 0·076 (0·014)
 Time to hatching  7·361, 1190·008 0·147 (0·054)
 Age at measurement 73·351, 109< 0·001 1·28 (0·15)
Tarsus length (age 4) (n = 105)
 Egg mass  9·061, 76·80·004 0·098 (0·033)
 Laying order  4·312, 76·80·017
 Age at measurement 27·991, 85·7< 0·001 1·89 (0·36)
Tarsus length (age 8) (n = 73)
 Egg mass 25·731, 70< 0·001 0·251 (0·049)
 Age at measurement 23·291, 43·3< 0·001 2·77 (0·57)
Body mass (age 1) (n = 128)
 Egg mass110·451, 116< 0·001 0·686 (0·065)
 Age at measurement 47·471, 118< 0·001 5·49 (0·79)
Body mass (age 4) (n = 105)
 Egg mass 17·781, 93·8< 0·001 1·12 (0·27)
 Age at measurement 37·681, 95·7< 0·00122·1 (3·6)
Body mass (age 8) (n = 73)
 Treatment  8·441, 350·006
 Egg mass 22·861, 67·3< 0·001 2·85 (0·59)
 Age at measurement 27·141, 48·1< 0·00137·7 (7·2)
Immune response (n = 66)
 Laying order  5·862, 480·005
 Time to hatching  5·941, 57·70·01817·6 (6·4)

Immune response was unaffected by albumen removal [F1,60 = 0·08, P = 0·78; least-squares means for control chicks = 1·73 mm (0·10); albumen chicks = 1·77 mm (0·09)].

Original egg mass had a positive effect on both morphological variables at all ages but not on immune response (Table 2). Time to hatching positively predicted tarsus length at age 1, and immune response (Table 2). Variation in tarsus length due to variation in time to hatching between eggs with extreme values of time to hatching (24 and 32 days) was 0·82 SDs of mean tarsus length at age 1. The effect of variation in time to hatching, however, was considerably more pronounced on immune response, as it corresponded to 2·48 SDs of the mean immune response.

Finally, we found a different pattern of variation of tarsus length and immune response in relation to egg laying order. Tarsus length at age 4 of chicks originating from a- and b-eggs was similar (P > 0·90 at post-hoc tests), but larger than that of chicks from c-eggs (P < 0·02 for both comparisons). T-cell-mediated immune response was significantly larger in chicks from c-eggs than in chicks from either a- or b-eggs (both P < 0·01).

Brood identity had a significant effect on nestling morphological variables (all P < 0·05, details not shown) but not on immune response.

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Traditionally, studies of maternal effects via egg size in birds have primarily focused on the total amount of resources included in eggs (e.g. Ricklefs, Hahn & Montevecchi 1978; Ricklefs 1984), while little is known about the function of variation in yolk and albumen content in natural populations. Current evidence is mostly correlative (reviews in Williams 1994; Bernardo 1996), and only a few studies have adopted a manipulative approach in the analysis of the importance of yolk and albumen egg content for offspring performance in birds or other vertebrates (Sinervo & Huey 1990; Hill 1993; Sinervo 1993; Finkler et al. 1998; Ferrari et al. 2006).

In this study of a gull species in the wild, we have shown that experimental removal of albumen may have complex consequences on diverse chick traits, including time to hatching, begging behaviour, morphology and survival. Some of these effects, however, varied according to laying order and egg mass.

Albumen removal reduced egg hatchability and increased the time to hatching. In addition, albumen removal differentially affected the probability that an egg originated a chick surviving until age 8 in relation to egg mass, as among control chicks viability depended on original egg mass, while no relationship existed for albumen chicks. Furthermore, begging intensity was differentially affected by albumen removal in relation to laying order and original egg mass, as it increased with laying order among albumen chicks, but decreased among controls, and increased with increasing egg mass among controls, whereas it decreased among albumen chicks. Finally, and unexpectedly, albumen chicks were heavier than controls on day 8 post-hatching, while albumen removal did not affect body mass at earlier ages or tarsus length at any age. However, as expected, egg mass strongly positively predicted body mass and size of the chicks, at any age.

The reduced egg hatching success following albumen removal, which occurred among a- and b- but not c-eggs, indicates that even a small variation in albumen content can have large consequences, and particularly so for the early laid eggs in a clutch, which generate chicks with greater survival prospects (Hillström, Kilpi & Lindström 2000).

The increase in time to hatching following albumen removal is consistent with the results obtained by Ferrari et al. (2006) in the barn swallow, indicating that variation in albumen content may affect embryo developmental trajectories in birds. This discloses the possibility that mothers adjust hatching times of their clutches in relation to extrinsic factors (e.g. resource availability or egg predation pressure) via a modification of the proportion of the albumen in their eggs. In addition, the different pattern of variation of relative albumen content along the laying sequence in the yellow-legged gull compared with the barn swallow, as well as the difference in the consequences of the experimental albumen removal, may represent adaptations to their reproductive strategies. While the barn swallow adopts a brood survival strategy, which could be partly mediated by the increase in the proportion of albumen along the laying sequence (Ferrari et al. 2006), the yellow-legged gull exhibits a facultative brood reduction strategy, and its eggs show no within-clutch variation in the relative albumen content. Thus, albumen provisioning to the eggs may match differences in an important life-history trait, an idea that could be widely tested in comparative studies of variation in egg composition among species showing different reproductive strategies.

Contrary to our expectation, albumen removal did not negatively affect body size or mass in the first days after hatching, differently from previous studies of birds (Hill 1993; Finkler et al. 1998; Ferrari et al. 2006). However, apart from the study by Ferrari et al. (2006), which subtracted a proportion of albumen similar to the present one (up to 5·6% vs. 2·3%, respectively), the other studies, all carried out on poultry, removed a far greater proportion of albumen from the eggs, i.e. up to 16% in the study of Hill (1993), and 15% in the study of Finkler et al. (1998). Moreover, our experimental design differed from that adopted by Ferrari et al. (2006), in that we assigned whole clutches to the same treatment, whereas Ferrari et al. (2006) adopted a within-clutch design, where chicks of different treatments hatched in a same nest. In the present study, the adoption of an intraclutch design was not feasible, due to the smaller clutch size of gulls and much higher post-hatching mortality due to stochastic factors (see Rubolini et al. 2005). Thus, in our study, albumen chicks did not compete with control siblings, which could have exacerbated the consequences of the albumen removal on body mass and size in the barn swallow experiment.

Albumen chicks were heavier than controls on day 8 post-hatching, while albumen removal did not affect body mass at earlier ages or tarsus length at any age. This counterintuitive result may be a by-product of a reduced egg hatchability of a- and b-eggs and of the consequent variation in the proportion of chicks from a-, b- or c-eggs in albumen compared with control broods. The larger body mass of albumen chicks was not mediated by a different number of chicks in albumen nests compared with control nests (Mann–Whitney U-test; age 4: Z = 0·87; age 8: Z = 0·06; P > 0·35). Thus, the alteration of brood composition in terms of chick quality following differential hatching success of eggs according to laying order in control and albumen clutches may have generated the observed pattern.

A novel result emerging from this study is that albumen removal may affect begging display, depending on the laying order of the original egg and on original egg mass. This finding may be due to a greater sensitivity of a- and b-eggs and chicks to albumen removal, which may have consequences on their behavioural development or performance, and suggests the existence of an albumen-mediated interaction between pre-natal resource allocation and post-natal signals of need.

The analysis of the effect of albumen removal disclosed significant patterns of association of chick phenotype with some of the covariates we considered. Size at hatching and immune response were found to positively covary with time to hatching. This may indicate that chicks hatching from eggs that are incubated longer hatch at an advanced developmental stage compared with eggs that are incubated for a shorter period (e.g. Ricklefs 1992). This result is consistent with comparative evidence showing that, across both altricial and precocial species, the intensity of the T-cell-mediated immune response positively covaries with the duration of the incubation period (Tella et al. 2002; see also Ricklefs 1992).

The larger T-cell-mediated immune response we found in chicks from c- compared with a- or b-eggs could be explained by the higher antioxidant capacity we documented in c-eggs in a companion study of the same population (see Rubolini et al. 2006). In fact, egg antioxidants are known to enhance immunity of chicks during their early post-hatching life (see Surai 2003 for a review).

Finally, we found that chick morphological traits were positively predicted by egg mass. This result is consistent with previous findings in the same as well as in other species (e.g. Rubolini et al. 2006; Ferrari et al. 2006), suggesting that variation in egg mass can have persistent effects on chick quality.

In conclusion, the results of this study show that variation in egg albumen content may affect egg and chick viability, and modulate hatching times in a semiprecocial gull species. The patterns of variation in the proportion of albumen in the eggs of different species may parallel differences in reproductive strategies. Thus, our study calls for a major consideration of the functional role of the albumen in ecological and evolutionary investigations of maternal effects via egg composition.

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

We thank the Parco Regionale del Delta del Po for permission to work in the study area. We are also sincerely grateful to Valentina De Simone and all those who helped during fieldwork. The thorough comments by two referees and the Editor helped to improve a previous draft of the manuscript.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  • Bernardo, J. (1996) The particular maternal effect of propagule size, especially egg size: patterns, models, quality of evidence and interpretations. American Zoologist 36, 216236.
  • Carey, C., Rahn, H. & Parisi, P. (1980) Calories, water, lipid and yolk in avian eggs. Condor 82, 335343.
  • Christians, J.K. (2002) Avian egg size: variation within species and inflexibility within individuals. Biology Reviews 77, 126.
  • Cramp, S. (1998) The Birds of the Western Paleartic on CD Rom. Oxford University Press, Oxford.
  • Crawley, M.J. (1993) GLIM for Ecologists. Blackwell Science, Oxford.
  • Ferrari, R.P., Martinelli, R. & Saino, N. (2006) Differential effects of egg albumen content on barn swallow nestlings in relation to hatch order. Journal of Evolutionary Biology 19, 981993.
  • Finkler, M.S., Van Orman, J.B. & Sotherland, P.R. (1998) Experimental manipulation of egg quality in chickens: influence of albumen and yolk on the size and body composition of near-term embryos in a precocial birds. Journal of Comparative Physiology 168, 1724.
  • Hill, W.L. (1993) Importance of prenatal nutrition to the development of a precocial chick. Developmental Psychobiology 26, 237249.
  • Hillström, L., Kilpi, M. & Lindström, K. (2000) Is asynchronous hatching adaptive in herring gulls (Larus argentatus)? Behavioral Ecology and Sociobiology 47, 304311.
  • Kirkpatrick, M. & Lande, R. (1989) The evolution of maternal characters. Evolution 43, 485503.
  • Lack, D. (1954) The Natural Regulation of Animal Numbers. Clarendon, Oxford.
  • Lochmiller, R.L., Vestey, M.R. & Boren, J.C. (1993) Relationship between protein nutritional status and immunocompetence in Northern Bobwhite chicks. Auk 110, 503510.
  • Mousseau, T.A. & Fox, C.W. (1998) Maternal Effects as Adaptations. Oxford University Press, New York.
  • Nisbet, I.C.T. (1978) Dependence of fledging success on egg-size, parental performance and egg-composition among Common and Roseate terns, Sterna hirundo and S. dougallii. Ibis 120, 207215.
  • Ricklefs, R.E. (1984) Egg dimensions and neonatal mass of shorebirds. Condor 86, 711.
  • Ricklefs, R.E. (1992) Embryonic development period and the prevalence of avian blood parasites. Proceedings of the National Academy of Sciences of the USA 89, 47224725.
  • Ricklefs, R.E., Hahn, C.D. & Montevecchi, W.A. (1978) The relationship between egg size and chick size in the Laughing Gull and Japanese Quail. Auk 95, 135144.
  • Romanoff, A.L. & Romanoff, A.J. (1949) The Avian Egg. Wiley, New York.
  • Rubolini, D., Romano, M., Boncoraglio, G., Ferrari, R.P., Martinelli, R., Galeotti, P., Fasola, M. & Saino, N. (2005) Effects of elevated egg corticosterone levels on behavior, growth, and immunity of yellow-legged gull (Larus michaellis) chicks. Hormones and Behavior 47, 592605.
  • Rubolini, D., Romano, M., Bonisoli Alquati, A. & Saino, N. (2006) Early maternal, genetic and environmental components of antioxidant protection, morphology and immunity of yellow-legged gull (Larus michahellis) chicks. Journal of Evolutionary Biology 19, 15711584.
  • Saino, N., Calza, S. & Møller, A.P. (1997) Immunocompetence of nestling Barn Swallows in relation to brood size and paternal effort. Journal of Animal Ecology 66, 827836.
  • Saino, N., Romano, M., Ferrari, R.P., Martinelli, R. & Møller, A.P. (2005) Stressed mothers lay eggs with high corticosterone levels which produce low-quality offspring. Journal of Experimental Zoology 11, 9981006.
  • Sinervo, B. (1993) The effect of offspring size on physiology and life history. Bioscience 43, 210218.
  • Sinervo, B. & Huey, R.B. (1990) Allometric engineering: an experimental test of the causes of interpopulation differences in locomotor performance. Science 252, 13001302.
  • Sinervo, B., Doughty, P., Huey, R.B. & Zamudio, K. (1992) Allometric engineering – A causal analysis of natural selection on offspring size. Science 258, 19271930.
  • Sotherland, P.R. & Rahn, H. (1987) On the composition of bird eggs. Condor 89, 4865.
  • Surai, P.F. (2003) Natural Antioxidants in Avian Nutrition and Reproduction. Nottingham University Press, Nottingham.
  • Tella, J.L., Scheuerlein, A. & Ricklefs, R.E. (2002) Is cell-mediated immunity related to the evolution of life-history strategies in birds? Proceedings of the Royal Society of London B 269, 10591066.
  • Tinbergen, N. (1967) The Herring Gull's World. Anchor Books, New York.
  • Uitenbroek, D.G. (1997) SISA-Bonferroni (http://home.clara.net/sisa/bonhlp.htm).
  • Williams, T.D. (1994) Intraspecific variation in egg size and egg composition: effects on offspring fitness. Biology Reviews 68, 3559.
  • Wright, J. & Leonard, M.L. (2002) The Evolution of Nestling Begging: Competition, Cooperation and Communication. Kluwer Academic Press, Dordrecht.