SEARCH

SEARCH BY CITATION

Keywords:

  • fleas;
  • LPS;
  • parasites;
  • sibling competition;
  • resource allocation

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Material and methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References
  10. Supporting Information
  • 1
    Carotenoids are fat-soluble pigments that stimulate the immune system and can act as antioxidants. Carotenoids are thus expected to buffer the effects of environmental stressors on health. As carotenoids are a limited resource, the ability of an individual to use and metabolize carotenoids is assumed to influence its stress-resistance. Accordingly, it has been found that nestlings hatched from eggs with increased carotenoid concentration, show an enhanced ability to use carotenoids and a lower susceptibility of tissues to lipid peroxidation.
  • 2
    We tested the prediction that nestling great tits (Parus major), hatched from eggs laid by carotenoid-supplemented mothers, cope better with a transient stressor encountered after hatching. We supplemented half of the breeders with carotenoids during egg production (C+), used the other half as a control (C–), and cross-fostered the eggs between nests after clutch completion. Three days after hatching, we applied a stressor in two-third of the nests either by increasing brood size, or by infesting nests with hen fleas (Ceratophyllus gallinae) during five consecutive days. A third group was kept as a control. We then assessed the responses of C+ and C– nestlings to each stressor by measuring mass gain, body condition, plumage coloration, humoral immune response and fever response to a lipopolysaccharide injection.
  • 3
    In control nests, C+ and C– nestlings showed similar body condition but C+ nestlings had a higher increase in body temperature and tended to have a higher wing web swelling in response to lipopolysaccharide injection. Under stress, however, there were no differences in overall condition between C+ and C– nestlings. The two stressors led to different responses: when sibling competition was increased, C– nestlings favoured immune development, whereas C+ nestlings favoured mass gain and body condition, while under parasite exposure C+ and C– nestlings seemed to invest in immune development and body growth similarly.
  • 4
    Our results support the hypothesis that carotenoid-induced maternal effects provide developmental benefits under natural conditions without additional stressors. Additionally, we show that the response to sudden environmental changes depends on the environment during the initial phases of development, which thus shape phenotype and individual variation.

Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Material and methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References
  10. Supporting Information

In the wild, animals face a large range of stressful factors such as predation, parasitism, food shortage, intra- and inter-specific competition, and seasonal changes. Stressors induce physiological responses that can alter health and thereby reduce fitness (Kitaysky et al. 2001; El-Lethey, Huber-Eicher & Jungi 2003; Padgett & Glaser 2003; Romero 2004; Bubliy & Loeschcke 2005; Love et al. 2005). For example, the hypothalamic–pituitary–adrenal axis of stressed animals releases elevated quantities of stress hormones such as glucocorticoids (Romero & Wikelski 2001). Subsequent to a period of stress some individuals can show chronic elevation of these stress hormones (Kitaysky et al. 1999; Padgett & Glaser 2003; Saino et al. 2003b). Chronic elevation of corticosterone is known to suppress the immune system (El-Lethey et al. 2003; Padgett & Glaser 2003), decrease endogenous lipid reserves and promote loss of muscle tissue (Kitaysky et al. 1999; Love et al. 2005). Additionally, under some stressful situations such as intense cold, increased competition and parasite infestation, the energy expenditure (Svensson et al. 1998; Sunardi et al. 2007), and thus the metabolic activity, may increase (Svensson et al. 1998; Carere & van Oers 2004; Arens & Cooper 2005). This could in some instances lead to an elevated production of reactive oxygen species such as free radicals. Free radicals have deleterious effects on the DNA, cellular proteins and lipids (Fang, Yang & Wu 2002). Thus, individuals that are able to mitigate the effects of those stress-related ‘disorders’ could hypothetically be in better condition and have higher survival rates.

Different physiological mechanisms have been suggested to influence stress-resistance: (i) the capacity of individuals to rapidly and efficiently reduce glucocorticoid production after stress (El-Lethey et al. 2003; Romero 2004); (ii) the ability to keep metabolic activity at low rates under stress (Vleck & Vleck 1979; Hoffmann & Parsons 1989; Tieleman & Williams 2000); and (iii) the antioxidant status of an animal that influences the degree of oxidative stress (e.g. Finkel & Holbrook 2000; Alonso-Alvarez et al. 2004a; Bertrand et al. 2006a; Horak et al. 2007). In fact, antioxidants such as carotenoids can help animals to resist oxidative stress by scavenging free radicals produced under high metabolic activity (Koutsos et al. 2003; Alonso-Alvarez et al. 2004a). Carotenoids are not only antioxidants but can also stimulate the immune system (Blount et al. 2003b; McGraw & Ardia 2003; Saino et al. 2003a). They might then counteract the suppressive effect of stress hormones on the immune system as suggested by Berthouly, Helfenstein & Richner (2007). Carotenoids can only be obtained via the food and are thus expected to be limited in nature. A trade-off is thus hypothesized in their use for their different functions (i.e. detoxification, immunity and coloration Horak et al. 2000; Tschirren, Fitze & Richner 2003a; Blount et al. 2003b; Hill, Farmer & Beck 2004). Thus, animals having access to higher quantities of carotenoids, or animals being able to optimally use carotenoids may be better able to cope with environmental stressors and thereby enjoy higher fitness. In other words, the strength of the effect of a stress on ‘health’ might be influenced by the availability and the efficiency of antioxidant defences and immunostimulant molecules that might be enhanced when availability of carotenoids in the diet increases.

Early-life conditions shape the development of morphological and physiological traits. In rats, for example, a low protein diet during pregnancy and lactation causes sex-specific changes in organ and body mass in offspring (Desai et al. 1995). In adult pheasants, sexual ornaments reflect nutritional status during early growth (Ohlsson et al. 2002). The environment during early life stages can be actively modified by parents depending on environmental conditions at breeding (e.g. as a function of food abundance, quality of the resources, predation, prevalence of parasites, ambient temperature Ojanen 1983; Verhulst, Vanbalen & Tinbergen 1995; Saino et al. 2004; Tschirren, Richner & Schwabl 2004; Love et al. 2005; Nooker, Dunn & Whittingham 2005; Berthouly et al. 2007). In birds, mothers have been found to modify the allocation of resources in their eggs in relation to habitat characteristics (Verhulst et al. 1995; Tschirren et al. 2004; Nooker et al. 2005; Berthouly et al. 2007). In the great tit, females exposed to ectoparasites deposited less testosterone in their eggs (Tschirren et al. 2004), which has been suggested to reduce the negative effects of parasites on offspring. Also, carotenoid availability has been found to influence the breeders’ condition (Alonso-Alvarez et al. 2004b; Blount et al. 2004) and the carotenoid contents of eggs (Blount et al. 2002; Biard, Surai & Møller 2005; Berthouly et al. 2007). When carotenoid availability to breeders was experimentally enhanced, mothers increased the quantities of carotenoids deposited into the egg yolk (Blount et al. 2002; Biard et al. 2005; Berthouly et al. 2007). Nestlings that hatched from eggs with higher carotenoid levels showed a better immune response (Saino et al. 2003a), a more intense colouring of carotenoid-based traits such as the beak (McGraw, Adkins-Regan & Parker 2005), and an enhanced ability to incorporate carotenoids into immune tissues (Koutsos et al. 2003). Carotenoid richness of neonatal nutrition in zebra finches affects their capacity to assimilate carotenoids as adults (Blount et al. 2003a), suggesting that different quantities of carotenoids in yolks could also influence the ability to use carotenoids later in life. These nestlings may subsequently also be able to better cope with negative effect of stressors on health by neutralizing free radicals more efficiently, and most of all by being able to mount a stronger immune response.

In this study we tested whether nestling great tits hatched from eggs laid by carotenoid-supplemented mothers cope better with a transient stressor arising a few days after hatching. We manipulated carotenoid availability during egg production and cross-fostered the eggs to disentangle the effects of egg quality from the effects of post-hatching environment on nestling performance. Over a period of 5 days, we induced two different physiological stressors either by increasing sibling competition in one group of birds, which has been shown to impair nestling immunity (Horak et al. 1999) and body condition (Sanz & Tinbergen 1999), or by infesting nests with hen fleas (Ceratophyllus gallinae) in another group, which have been shown to impair nestling growth (Nilsson 2003), body condition (Richner, Oppliger & Christe 1993; Tschirren, Fitze & Richner 2003b) and immunity (Berthouly et al. 2007). Another group was kept as a control. Sibling competition was enhanced by adding two nestlings to the nest. We hypothesized that sibling competition and flea infestation would (i)increase metabolic activity and thus generate oxidative stress (Bertrand et al. 2006b; Costantini et al. 2006), (ii)increase baseline corticosterone levels (Saino et al. 2003b; Raouf et al. 2006), and (iii) increase energy requirements (Ots et al. 2001). Such stressors may intensify the trade-offs in the allocation of limited resources, such as carotenoids, to different physiological functions, for example, immune development, detoxification, and plumage development (Horak et al. 2001; Alonso-Alvarez et al. 2004a). We thus expected nestlings hatched from eggs laid by carotenoid-supplemented mothers to be less impaired by the stressors, and to show higher immunocompetence and/or higher growth and/or more intensive plumage colour depending on the type of stressor to which they were subjected.

Material and methods

  1. Top of page
  2. Summary
  3. Introduction
  4. Material and methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References
  10. Supporting Information

food supplementation

The experiment was carried out in 2006 in a nest-box breeding population of great tits in the Bremgartenwald, a forest near Bern, Switzerland. The study area contained 336 nest-boxes distributed over 24 plots (14 nests per plot). In March, the nest boxes were cleaned by removing old nest material and thorough brushing to eliminate all ectoparasites. At the same time we hung an inverted medium-sized flowerpot near each nest box. We then randomly assigned one of the two food treatments to each plot: nests supplemented with carotenoid-enriched (C+) or with non-enriched food (C–). We supplemented the birds with commercial bird fatballs mixed with sunflower seeds that were first crumbled, and heated to 60 °C to denaturate the majority of the carotenoids naturally present in the mixture (as advised by a representative of Hoffmann-La Roche). After cooling, we added 0·132 mg of lutein and 0·005 mg of zeaxanthin both in oil prepared by Roche Vitamins Inc., and 0·028 mg of β-carotene in powder (Roche) per gram to half of the food mixture. The other half of the mixture was used as control. These relative quantities of each carotenoid species correspond to the approximate ratio found in lepidopteran larvae eaten by great tits, i.e. 80% of lutein, 3% of zeaxanthin, and 17% of β,β-carotene (Partali et al. 1987). Finally, we moulded the mixture into two types of fatballs, i.e. with and without the added carotenoids (thereafter called C+ and C– respectively). The mixtures were finally moulded into either enriched or non-enriched balls. Each ball weighed c. 40 grams, and C+ and C– fatballs were consumed at similar rates by great tits and other passerine species (assessed by observation and weighing). We estimated that the quantities of daily carotenoids ingested by each bird were about the same they would have ingested if their diet mainly consisted of lepidopteran caterpillars (de Ayala, Martinelli & Saino 2006; Isaksson & Andersson 2007), as it is the case during the nestling period later in the breeding season when insect abundance is the highest . In order to minimize the influence of our food treatment on breeding habitat choice, supplementation started only when the nest box was already occupied (i.e. floor covered with a layer of c. 2 cm of nesting material). Nests were visited every fourth day to determine the stage of nest building, to replenish the food stock if necessary (i.e. replacement of the former fatball by a new one when at least half of it had been consumed) and to determine the start of egg-laying. We stopped supplying nests with food on the first day of full incubation. Onset of full-time incubation was judged to have started when eggs were warm and uncovered during three consecutive days.

cross-fostering of eggs

On the third day of full incubation, whole clutches were cross-fostered between nests to disentangle the effects of maternally derived carotenoids from post-hatching environmental effects. Eggs were exchanged between nests of different or equal food-supplementation treatments (C– inline imageC–, C– inline imageC+, C+ inline imageC+). From the first expected day of hatching, nests were inspected daily to determine hatching date precisely.

stress induction

In order to investigate the effect of a period of increased physiological stress early in life on nestling performance, and the ability of nestlings hatched from eggs laid by carotenoid-supplemented (C+) or control mothers (C–) respectively to cope with stress, we applied a stressor to some nests. We induced two different types of stressors: we either increased sibling competition by increasing original brood size by two nestlings, or infested nests with hen fleas. Findings of several experimental studies carried out on the great tit/flea system suggest that responses to flea infestation involve physiological as well as immunological defence (Buechler et al. 2002; Tschirren et al. 2004, 2007). Additionally, cat fleas show higher mortality when feeding on blood of rabbits immunized with flea antigens (Nisbet & Huntley 2006), and bird fleas most likely also induce an immune response in their host. Treatments were applied 4 days post-hatching, after all eggs had hatched, and ended 9 days post-hatching (see Appendix S1 in Supplementary Material for more details). In summary we had three groups of nests: (i) control nests (11 C+ nests with 47 nestlings and 14 C– nests with 62 nestlings) where no additional stressor was applied; (ii) nests with a temporarily increased brood size (9 C+ nests with 39 nestlings and 13 C– nests with 55 nestlings); (iii) nests temporarily infested with hen fleas (11 C+ nests with 41 nestlings and 15 C– nests with 65 nestlings).

nestling measurements: body mass, immunity assessment and plumage chroma

All new hatchlings were weighed to the nearest 0·1 g and marked by partially removing tuft feathers from their heads, backs and wings. We took blood samples from each hatchling for molecular sexing (Griffiths et al. 1998). Eight days post-hatching, nestlings were permanently marked with a metal ring. At the same time, they were weighed and their tarsus and wing length measured. Fifteen days post-hatching, the same measurements were taken again.

Fifteen days after hatching, we assessed immune performance of three or four nestlings per brood, the number depending on the size of the brood. Nestlings were weighed and ranked according to their body mass. We then selected every second nestling in the mass hierarchy. Nestlings below 14 g (31 nestlings of 589) were not considered for the LPS injection because the LPS-induced immune response could have potentially harmed them. We assessed immune performance by an injection of lipopolysaccharide (LPS) (Parmentier, Reilingh & Nieuwland 1998). LPS is a mitogen that induces an antigen-mediated infiltration and aggregation of inflammatory cells and T-cells at the site of injection. We injected the wing web of the nestlings with 0·01 mg of LPS (LPS, Sigma-Aldrich) dissolved in 0·02 mL of phosphate buffered saline. It has been found by Parmentier et al. (1998) in chicken that the response to LPS is highest 4 h following the injection. Nestling great tits did not show any response after 4 or 12 h but had the strongest response 24 h later. We measured the wing web thickness before and 24 h after the injection to the nearest 0·01 mm with a constant-tension dial micrometer (Mitotuyo, Type 2046FB-60). A greater swelling reflects a better immune response (Parmentier et al. 1998). It has already been found that birds can develop fever in response to LPS injection (Maloney & Gray 1998). We thus also measured rectal temperature of the nestlings in randomly chosen nests before and after injection with a digital thermometer (Omega HH41 from Exacon) fitted with a temperature probe with a diameter of 0·08 mm. Finally, we measured the skin temperature of all nestlings (i.e. injected and not injected) by placing an auricular thermometer (ThermoScan from Braun) on the skin of the stomach. Four measurements were made for each nestling and the mean recorded. The fever is a mechanism that helps animals to fight pathogens. It causes an unbearable environment for some pathogens and enhances immunological functions (Ostberg & Repasky 2006). Thus, individuals being able to respond to a bacterial infection by raising body temperature are assumed to fight pathogens more efficiently and be less harmed by them. A total of 256 nestlings from 73 nests were injected with LPS.

Sixteen days post-hatching, we took c. 12 breast feathers on both the left and right sides of the injected nestlings in order to measure the carotenoid chroma of their plumage. Nine feathers from the same side were overlaid and fixed on black felt. We then measured the chroma of the right and the left plumage of each nestling and used the mean of these two values as a measure of the yellow coloration of a nestling. Repeatability within feather measurements was equal to 0·51 (F(75,76) = 3·085, P < 0·0001) (Lessells & Boag 1987) (see Supplementary Appendix S2).

egg mass and carotenoid analysis of yolk

An important assumption is that carotenoid-fed mothers would deposit higher quantities of carotenoids into theirs eggs than control mothers. We verified this assumption in our population by collecting a sample of eggs before incubation. We collected the third egg in the laying sequence (n = 11), and also in a parallel study on another topic in the same forest the third and the fifth eggs (n = 96; performed by F. Helfenstein). These eggs were replaced by dummy eggs to avoid desertion, and stored in a freezer. As expected, carotenoid-supplemented mothers deposited significantly higher quantities of carotenoids into their eggs compared to control-fed mothers (F1,60= 7·08, P = 0·001; C+ females: 25·79 ± 9·91 µg g−1 of egg yolk; C– females: 20·27 ± 9·21 µg g−1 of egg yolk) (see also (Berthouly et al. 2007). Additionally, we measured individual carotenoids (i.e. lutein, zeaxanthin and β-carotene) in the yolk. Total and individual carotenoid concentrations were assessed using High Performance Liquid Chromatography following the protocol described in (Surai & Speake 1998; Karadas et al. 2005).

statistical procedures

The effect of carotenoid-supplementation on lutein, zeaxanthine, and β-carotene concentration in the yolk was assessed using mixed models including the nest as random factor. The carotenoid treatment and egg rank were included as fixed factors in the model and clutch size and laying date as covariates. The two-factorial interactions between factors and covariates were also included.

In all analyses of dependent variables corresponding to a measurement on nestlings, that is, mass gain between days 3 and 15, tarsus on day 15, body condition on day 15, swelling response to LPS injection, plumage chroma, body temperature before and 24 h after LPS injection, we included only nestlings that were immune-challenged. We used linear mixed models for these analyses. The following explanatory variables were included in all models: the food treatment of the genetic mother, the stress treatment applied to the entire brood (i.e. no stress, increased sibling competition, flea infestation), the hatching date of first hatched nestlings, the nestling sex, and the two-factorial interactions. We controlled for the real age of nestlings by including hatching sequence as a factor. Nestlings could either have hatched on the day where the first hatching occurred in the nest or 1 to 2 days thereafter (i.e. two modalities: hatched the first day (A) or hatched the second or the third day (B)). The nest of rearing was included as a random factor to correct for non-independence of siblings (Pinheiro & Bates 2000). For the analysis of mass gain between days 3 and 15 we used the body mass on day 15 as the dependent variable and included the body mass on day 3 as a covariate. By including previous body weight as a covariate, the dependent variable represents a measure of mass gain over the given time period. For the analysis of body condition on day 15, the body mass on day 15 was taken as the dependent variable and structural body size (here the tarsus length) included as a covariate into the analysis. Finally, in the analyses of responses to LPS injection (swelling response, febrile response) we included the external temperature on the day of injection.

parental condition

Body mass was taken as the dependent variable and tarsus length included as a covariate into the model. We used a linear mixed model and included the following explanatory variables: the food treatment during egg-laying, the stress treatment applied to the entire brood, the date of capture, the sex, and the two-factorial interactions. The nest of rearing was included as a random factor to correct for non-independence of the two parents (Pinheiro & Bates 2000).

All the analyses were performed using the software r (R Development Core Team 2006) (Maindonald & Braun 2003). We used the package nlme to perform the mixed model analysis (Pinheiro & Bates 2000; Pinheiro & Bates 2006). Responses were transformed, if necessary, in order to meet the assumption of normal distribution of residuals (i.e. the swelling response was log-transformed). We used restricted maximum likelihood estimation in all mixed effect models. Interactions and main factors showing a P-value > 0·1 were backward eliminated using a stepwise elimination procedure. We first excluded the interactions and then the main factors. In the results and the tables, the P-values given for interactions and main effects that have a P-value < 0·1 were extracted from the model containing only interactions and main factors with a P-value lower than 0·1. P-values of main factors with a P-value > 0·1 were derived from an intermediate model containing all main factors but only the interactions with a P-values ≤ 0·1. Tests are two-tailed with a significance level set to α = 0·05.

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Material and methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References
  10. Supporting Information

egg yolk carotenoids

We found that the concentration of lutein and zeaxantin was higher in C+ eggs (i.e. laid by carotenoid-supplemented mothers) than in C– eggs (i.e. laid by control mothers) (respectively: F1,58 = 9·87, P = 0·0026; F1,58 = 14·62, P = 0·0003) (see Supplementary Table S1). However there were no differences in β-carotene concentrations between C+ and C– eggs (F1,58 = 0·36, P = 0·55) (see Supplementary Table S1).

nestling performance

Before the stress was applied, that is, 3 days after hatching, nestling mass between the two groups assigned to be experimentally stressed or controls did not differ (F2,67 = 0·95, P = 0·39). Also, it did not differ between C+ and C– nestlings (F1,67 = 1·15, P = 0·29).

We found a significant interaction between food treatment of the genetic mother and stress treatment on body mass gain between days 3 and 15 (F2,66 = 3·96, P = 0·024) (Fig. 1), indicating that the stress treatments affected C+ and C– nestlings differently. There were no mass gain differences between C+ and C– nestlings in nests infested with fleas and controls (respectively: F1,22 = 1·08, P = 0·31; F1,21 = 1·05 , P = 0·32). In experimentally enlarged broods, however, C+ nestlings gained significantly more mass than C– nestlings (F1,19 = 7·13; P = 0·015). We also found that nestlings gained more mass when they were raised by control parents (F1,66 = 4·37, P = 0·040) (see Supplementary Fig. S1). Male nestlings gained more mass than female nestlings (F1,180 = 112·36, P < 0·0001). Finally, there were no effect of hatching date on body mass gain over this time period (F1,65 = 0·0007, P = 0·98).

image

Figure 1. Mean (± SE) mass gain between days 3 and 15 of C+ nestlings (i.e. hatched from eggs laid by carotenoid-supplemented females) (points and solid lines in black) and C– nestlings (i.e. hatched from eggs laid by control females) (triangle and dashed lines in grey), raised either in nests where brood size was increased, in nests infested with fleas, or in control nests.

Download figure to PowerPoint

As for body mass gain between days 3 and 15, we found a significant interaction between food treatment of the genetic mother and stress treatment on body condition on day 15 (F2,66 = 4·35, P = 0·017). There were no body condition differences between C+ and C– nestlings in nests infested with fleas and controls (respectively: F1,22 = 0·67, P = 0·42; F1,21 = 1·12, P = 0·3). However, C+ nestlings were in better condition than C– nestlings in enlarged broods (F1,19 = 6·33, P = 0·021). Additionally, we found that nestlings were in better condition when they were raised by control parents (F1,66 = 4·6, P = 0·035). Males were in better condition than females (F1,181 = 112·36, P < 0·0001). Finally, there was no effect of hatching date on body condition on day 15 (F1,65 = 0·001, P = 0·97). Differences in body condition are due to differences in body mass as tarsus length was neither influenced by food treatment of the genetic mother nor by the stress treatment applied to the nestlings (respectively: F1,65 ≈ 0, P = 0·94 ; F2,65 = 0·2, P = 0·82). Males had a longer tarsus than females (F1,180 = 168·38, P < 0·0001). Finally, the hatching date and the food treatment of the rearing parents did not influence tarsus length (respectively: F1,65 = 0·19, P = 0·66; F1,65 = 1·32, P = 0·25).

Again, we found a significant interaction between food treatment of the genetic mother and stress treatment on the swelling response to LPS injection (F2,65 = 4·68, P = 0·013) (Fig. 2). There was no difference in swelling between C+ and C– nestlings in nests infested with fleas (F1,22 = 1·65, P = 0·21), however when brood size was increased C– nestlings tended to have a bigger swelling than C + nestlings (F1,19 = 4·07, P = 0·058), whereas C+ nestlings tended to show a bigger swelling than C– nestlings in control nests (F1,23 = 3·43, P = 0·076). We also found that stress treatments affected male and female nestlings differently as shown by the significant interaction between sex and stress treatment (F2,177 = 4·07, P = 0·019). In fact, there were no differences between males and females in flea-infested and control nests (respectively: F1,63 = 0·005, P = 0·94; F1,57 = 0·07, P = 0·79), however, when brood size was increased males had a significantly smaller swelling than females (F1,54 = 7·78, P = 0·0072). We found that nestlings had a stronger swelling response when raised by control parents (F1,65 = 9·29, P = 0·0033) (see Supplementary Fig. S2). Finally there was no effect of hatching date on swelling response over this time period (F1,65 = 2·978, P = 0·089).

image

Figure 2. Mean (± SE) log values of the swelling response to LPS injection of C+ nestlings (i.e. hatched from eggs laid by carotenoid-supplemented females) (points and solid lines in black) and C– nestlings (i.e. hatched from eggs laid by control females) (triangle and dashed lines in grey), raised either in nests where brood size was increased, in nests infested with fleas, or in control nests.

Download figure to PowerPoint

Neither food treatments of genetic parents nor stress treatment influenced nestling plumage chroma (respectively: F1,63 = 1·18, P = 0·28, F2,63 = 0·28 , P = 0·75). Also, plumage chroma did not differ between nestlings of different weight or between nestlings hatched at different dates (respectively: F1,162 = 0·41, P = 0·52; F1,66 = 2·90, P = 0·093). However we found that males were significantly yellower than females (F1,164 = 6·85, P = 0·0097).

Rectal and skin temperature were significantly correlated (r = 0·75; permutation test: P = 0·0009). We thus decided to use skin temperature as the dependent variable because skin temperature was measured in all the nestlings, can be measured quickly and is least stressful to the birds. The following results concern nestling temperature on day 15, before LPS was injected. Carotenoid-supplementation of genetic parents did not influence nestling body temperature (F1,49 = 2·70, P = 0·11). There was also no effect of stress treatment on nestling temperature (F2,49 = 0·29 , P = 0·75). Finally, nestling temperature strongly depended on external temperature (F1,53 = 15·76, P = 0·0002): nestling temperature increased when environmental temperature increased, indicating that nestlings do not thermo-regulate perfectly at that age. Not surprisingly, we then found that overall nestlings showed an increase in temperature between day 15 and 16, thereby becoming closer to adult temperature, which is around 41 °C (personal measurements) (Carere & van Oers 2004). Because nestlings seemed not to thermo-regulate perfectly at that age, the finding that nestling temperature increases between day 15 and 16 could be due to the experimentally induced immune response, but also to the fact that nestlings were older and thus had more feathers and maybe a more efficient system of thermoregulation. To test whether an increase in nestling temperature after LPS injection would be at least partially due to an ‘immune’ response to fight against infection, we took the temperature of all nestlings, also the ones that were not immune challenged, and ran a model including the change in temperature of all nestlings. The dependent variable was the change in temperature, which corresponded to the difference between the temperature recorded 24 h after injection (i.e. on day 16) and the temperature recorded before the injection (i.e. recorded on day 15): nestling temperature on day 16 – nestling temperature on day 15. We included the food treatment of the genetic mother, the stress treatment applied to the entire brood, the hatching date of first hatched nestlings, the nestling sex, the hatching sequence, the LPS treatment (i.e. nestlings immune challenged or not), the environmental temperature on day 15, the change in environmental temperature between the two measurements and the two-factorials interactions as explanatory variables. Again, the nest of rearing was included as a random factor. We found a significant effect of the interaction between the food treatment of the genetic parents and the LPS treatment on the temperature change (F1,318 = 3·97, P = 0·047) (Fig. 3): there were no differences between C+ and C– (F1,54 = 0·008 , P = 0·93) nestlings that were not immune challenged, whereas C+ nestlings showed a higher increase in temperature than C– nestlings when they were injected with LPS (F1,55 = 4·84, P = 0·032). In other words, C+ nestlings injected with LPS had a higher increase in temperature than non-injected C+ nestlings (F1,128 = 6·23, P = 0·014) while there were no differences between injected and non-injected C– nestlings (F1,192 = 0·055, P = 0·82). We also found that the change in temperature between days 15 and 16 depended on the change in environmental temperature between the two measurements (F1,59 = 73·18, P < 0·0001). We did not find an effect of our stress treatment on nestling change in temperature (F2,55 = 0·35, P = 0·70). Finally, there were no differences in temperature change between males and females, between nestlings of different weight, and between nestlings hatched at different periods of the breeding season (respectively: F1,317 = 1·24, P = 0·27; F1,318 = 2·78, P = 0·96; F1,55 = 0·52, P = 0·47). We also tested whether the change in temperature between the time of injection and the measurement of the swelling had an effect on the swelling response and found no effect (F1,143 = 0·43, P = 0·52).

image

Figure 3. Mean (± SE) values of temperature increase between days 15 and 16 of C+ nestlings (i.e. hatched from eggs laid by carotenoid-supplemented females) (points and solid lines in black) and C– nestlings (i.e. hatched from eggs laid by control females) (triangle and dashed lines in grey), with and without an immune challenge.

Download figure to PowerPoint

parental condition

We found that C+ parents tended to be in better condition than C– parents (F1,36 = 3·7, P = 0·060). Adult males were in better condition than females (F1,36 = 25·55, P < 0·0001). Stress treatment applied to the brood did not affect parental condition (F1,36 = 0·23, P = 0·79). Late breeders were not in worse condition than early ones (F1,35 = 0·95, P = 0·34).

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Material and methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References
  10. Supporting Information

nestling growth, condition and immunity

First, our results show that C+ nestlings are in better overall condition than C– nestlings when raised under conditions where no stress was applied. In those nests, C+ and C– nestlings were in similar body condition, but C+ nestlings tended to have a bigger swelling and showed a higher temperature-increase in response to LPS injection than C– nestlings. Under experimentally induced stress, C+ nestlings still responded to LPS injection by increasing their body temperature whereas C– nestlings did not. However, C+ nestlings did not seem to be in higher overall condition than C– nestlings under induced stress regimes. Our results rather suggest that C+ and C– nestlings responded differently to the stressors by adopting different resource allocation strategies depending on rearing conditions. When competition was increased, C– nestlings seemed to favour the development of immunity, whereas C+ nestlings favoured mass gain and body condition. However, in flea infested nests, C+ and C– nestlings seemed to invest in immune development and body growth similarly. By allocating more resources to mass gain and body condition under high sibling competition, C+ nestlings may become larger and heavier more rapidly and then obtain food more easily from parents than small nestlings (Oddie 2000). Large nestlings may manifest more vigorous visual behaviours when soliciting food (Rodriguez-Girones, Zuniga & Redondo 2002) and/or may be better in jockeying for a feeding positions near an arriving parent (McRae, Weatherhead & Montgomerie 1993; Ostreiher 2001). This result is in line with previous findings (Berthouly et al. 2007; Helfenstein et al. 2008) where nestlings hatched from eggs laid by mothers supplemented with carotenoids were more competitive (i.e. begged more intensely) than nestlings hatched from control eggs. In flea-infested nests, there were no differences between C+ and C– nestlings, maybe because the selective pressures of a parasite exposure that act on the different physiological functions are stronger. On the one hand, it must be important to develop an efficient immune system in the presence of fleas to fight parasite attacks. On the other hand, however, it may also be important to leave the infested nest environment as soon as possible in order to reduce the duration of the contact with the fleas, and thereby their negative effects on health. In such a situation, it may be optimal to invest in all physiological functions similarly for any individual.

In summary, we found that, independent of the rearing conditions, C+ nestlings showed a stronger temperature increase than C– nestlings, suggesting a better defence against bacterial infestations of C+ nestlings (Long 1996; Ostberg & Repasky 2006). Moreover, when no additional stress was applied, C+ nestlings tended to have a stronger swelling response to LPS injection than C– nestlings. These findings support the idea of beneficial effects of carotenoid availability during egg-laying on nestling immunity and thereafter performance (Saino et al. 2003a). However, we found no evidence that C+ nestlings coped better with the increased sibling competition or with the flea infestation than C– nestlings. Yet, we rather found that a short transient stress (i.e. presence of fleas, high sibling competition) already shapes the development of the phenotype, depending on the type of stressor and on previous experience (i.e. environment during embryonic development). In line with this finding we found that C+ eggs contained more carotenoids than C– eggs. Also, Blount et al. (2002) found that carotenoid-supplemented lesser black-backed gull (Larus fuscus) females produced eggs containing higher carotenoid but lower Ig concentrations. Together, these results indicate that carotenoid supplementation affected egg composition such as carotenoid content, but also egg content of other metabolites such as immunoglobulins, hormones, nutrients. It suggest that egg composition can modulate the programming of nestling physiology, and thereby its energetic metabolism, and produce nestlings of different phenotypes (Royle et al. 1999; Koutsos et al. 2003; Saino et al. 2003a) that react differently to environmental changes. Additionally, our findings give strong evidence that a given maternal effect does not have an obligate effect on certain nestling traits but rather interacts with the environment to shape offspring phenotype and will thus be context-dependent (Marshall & Uller 2007). It questions the adaptive significance of maternal effects in unstable and heterogeneous environments, and points to the importance of long term studies when investigating the influence of maternal effects on offspring fitness before interpreting different maternal effects as adaptive, or as beneficial, as suggested by Marshall et al. (2007).

nestling plumage chroma

We did not find that C+ nestlings were more yellow than C– nestlings, indicating that additional carotenoids into the eggs did not influence plumage chroma at that age. It does not necessarily mean that carotenoid-induced maternal effects did not influence plumage colour, but could suggest that investing in plumage coloration is not a priority at this developmental stage, especially when pigments are scarce and may be used for other important functions for several reasons. First, it is a period of intense growth and development, and it might therefore be more important to use carotenoids as antioxidants and immunostimulants. Second, a role for plumage coloration in parent–offspring communication inside the nest is unlikely in the great tit because parents may not be able to differentiate among plumage colorations in poorly lit nestling cavities (Tschirren, Fitze & Richner 2005) (Tanner, M. & Richner H., unpublished). Also nestlings do not have a completely developed plumage at that age. Interestingly, under controlled conditions in the laboratory, McGraw et al. (2005) have found in zebra finches that sons hatched from eggs laid by carotenoid-supplemented mothers displayed a more brightly coloured beak as adults, which is sexually more attractive than a drab break. Additionally, Blount et al. (2003a) have shown that zebra finches with better nutrition at early developmental stages maintain a higher capacity to assimilate dietary antioxidants as adults. Thus, we can not exclude the possibility that C+ nestlings in our study will display more intense plumage chroma as adults when sexual selection may be operating.

parental investment into current reproduction

Concerning the effect of carotenoid supplementation on parental investment and parental condition, we found that nestlings raised by carotenoid-supplemented parents were in poorer overall condition than nestlings raised by control parents (i.e. nestlings raised by carotenoid-supplemented parents showed a lower mass gain between days 3 and 15, a lower body condition on day 15, and a lower response to LPS injection), but carotenoid-supplemented parents tended themselves to be in better body condition than control parents (P = 0·06) during the second half of the rearing period, when parents were caught and measured. It suggests that carotenoid-availability early in the breeding cycle influences the trade-off for parents between investment in current reproduction vs. investment in self-maintenance and thus future reproduction. A strategic decision to invest in current rather than future reproduction would make sense if carotenoids had a disproportionally stronger effect on future survival than on current reproductive success.

Conclusion

  1. Top of page
  2. Summary
  3. Introduction
  4. Material and methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References
  10. Supporting Information

In conclusion, our results suggest that carotenoid-mediated maternal effects enhance overall nestling condition when no additional stressors are applied to the nestlings. Yet, C+ nestlings did not cope better with added stressors. Our results rather show that the two types of nestlings responded differently to increased sibling competition by adopting different resource allocation strategies. They responded, however, in the same way to flea infestation by allocating similar amounts of resources into immunity and growth. This provides evidence that pre-hatching environment influences the response to sudden environmental changes due for example to parasite infections, drops in temperature or food abundance, and supports the hypothesis that maternal effects are context-dependent. It thus shows how the pre-hatching environment can shape phenotypic development and thereby maximize maternal fitness by enhancing phenotypic variation in offspring (Mousseau & Fox 1998; Marshall & Uller 2007). Finally, the results suggest that carotenoid availability during nest construction and egg production can influence the parental trade-off between current and future reproduction.

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Material and methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References
  10. Supporting Information

Authors thank Stéphanie Bonnaure, and Christelle Bugeaud for field assistance; Verena Saladin and Danielle Bonfils for the work in the laboratory; Katharina Gallizzi, Fabrice Helfenstein and Marion Tanner for their statistical help and comments on the manuscript. Authors also thank Filiz Karadas and Peter Surai for carotenoid analyses, and Doris Gomez for providing the AVICOL program and giving advices on measurements of plumage colour. Thanks also to the two anonymous referees for their valuable comments on the manuscript. Finally, authors thank Roche Vitamins Inc. for providing carotenoids and Erbo Agro AG for providing fat-balls. This work was financially supported by the Swiss National Science Foundation (grant 3100A0-102017 to H.R.) and conducted under a licence provided by the Ethical Committee of the Office of Agriculture of the canton of Bern, Switzerland.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Material and methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References
  10. Supporting Information

Supporting Information

  1. Top of page
  2. Summary
  3. Introduction
  4. Material and methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References
  10. Supporting Information

Appendix S1. Details on stress induction procedure.

Appendix S2. Measurement of nestling plumage coloration.

Table S1. Effect of carotenoid-supplementation on yolk concentrations of lutein, zeaxanthin, and β-carotene.

Fig. S1. Nestling mass gain between days 3 and 15 in relation to the food treatment of the rearing parents.

Fig. S2. Nestling swelling response in relation to the food treatment of the rearing parents.

Please note: Blackwell Publishing is not responsible for the content or functionality of any supplementary materials supplied by the authors. Any queries (other than missing material) should be directed to the corresponding author for the article.

FilenameFormatSizeDescription
FEC_1439_sm_Supmat.doc71KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.