Naturwissenschaften DOI 10.1007/s00114-014-1190-2
ORIGINAL PAPER
Nestling rearing is antioxidant demanding in female barn swallows (Hirundo rustica) David Costantini & Andrea Bonisoli-Alquati & Diego Rubolini & Manuela Caprioli & Roberto Ambrosini & Maria Romano & Nicola Saino
Received: 21 March 2014 / Revised: 24 April 2014 / Accepted: 20 May 2014 # Springer-Verlag Berlin Heidelberg 2014
Abstract Reproduction is a demanding activity, since organisms must produce and, in some cases, protect and provision their progeny. Hence, a central tenet of life-history theory predicts that parents have to trade parental care against body maintenance. One physiological cost thought to be particularly important as a modulator of such trade-offs is oxidative stress. However, evidence in favour of the hypothesis of an oxidative cost of reproduction is contradictory. In this study, we manipulated the brood size of wild barn swallows Hirundo rustica soon after hatching of their nestlings to test whether an increase in nestling rearing effort translates into an increased oxidative damage and a decreased antioxidant protection at the end of the nestling rearing period. We found that, while plasma oxidative damage was unaffected by brood size enlargement, females rearing enlarged broods showed a decrease in plasma non-enzymatic antioxidants during the nestling rearing period. This was not the case among females rearing Communicated by: Alexandre Roulin D. Costantini (*) Department of Biology, University of Antwerp, Universiteitsplein 1, 2610 Wilrijk, Belgium e-mail: [email protected] D. Costantini Institute for Biodiversity, Animal Health and Comparative Medicine, University of Glasgow, Glasgow G12 8QQ, UK A. Bonisoli-Alquati Department of Biological Sciences, University of South Carolina, 715 Sumter Street, 29208 Columbia, SC, USA D. Rubolini : M. Caprioli : M. Romano : N. Saino Dipartimento di Bioscienze, Università degli Studi di Milano, Milan, Italy R. Ambrosini Dipartimento di Biotecnologie e Bioscienze, Università degli Studi di Milano-Bicocca, Milan, Italy
reduced broods and among males assigned to either treatment. Moreover, individuals with higher plasma oxidative damage soon after the brood size manipulation had lower plasma nonenzymatic antioxidants at the end of the nestling rearing period, suggesting that non-enzymatic antioxidants were depleted to buffer the negative effects of high oxidative damage. Our findings point to antioxidant depletion as a potential mechanism mediating the cost of reproduction among female birds. Keywords Antioxidants . Life history . Oxidative damage . Oxidative stress . Parental effort . Trade-off
Introduction A central concept in evolutionary ecology is that all fitnessrelated traits cannot be simultaneously maximised, but instead they are traded-off against each other because of physiological constraints (e.g., stress, energy consumption) that arise as a consequence of resource investment (Stearns 1992). Reproduction is a demanding phase of organisms’ lives, since high investment in reproduction may lead to increased senescence or reduced survival and future fecundity (Williams 1966; Stearns 1992; Boonekamp et al. 2014). Although this paradigm is well grounded into evolutionary ecology, very little is known about how such costs of reproduction are actually incurred, since the majority of the studies have focused on the outcomes rather than the mechanisms. It has been recently proposed that oxidative stress may be one cost associated with reproductive investment (Costantini 2008; Metcalfe and Alonso-Alvarez 2010). Oxidative stress is the rate at which oxidative damage to biomolecules is generated, which is dependent on a complex balance between prooxidants and antioxidant mechanisms (Halliwell and Gutteridge 2007; Costantini and Verhulst 2009). Such a
Naturwissenschaften
damage can contribute to cell senescence and loss in organismal performance (Weinert and Timiras 2003), and may influence life-history strategies (Costantini 2008; Metcalfe and Alonso-Alvarez 2010). Although some studies have provided evidence that oxidative stress increases during reproduction, others have not (Stier et al. 2012; Metcalfe and Monaghan 2013; Speakman and Garratt 2014). Due to the relative ease of altering parental effort via clutch or brood size manipulation experiments, many studies have focused on altricial avian species. For example, in an experiment manipulating brood size of zebra finches (Taeniopygia guttata), males experiencing a higher workload were found to show no change in daily energy expenditure, but a lower antioxidant enzyme activity, whereas females showed increased energy expenditure but no change in antioxidant enzyme activity (Wiersma et al. 2004). Further studies have also found decreases in blood antioxidant protection in reproducing birds when rearing an enlarged brood (Alonso-Alvarez et al. 2004; Losdat et al. 2011; Christie et al. 2012) or when engaging in a higher number of reproductive episodes (Bertrand et al. 2006). However, the results of these studies are difficult to interpret, especially because they report only measures of components of antioxidant defenses, while measures of oxidative damage are largely lacking. Two studies have found evidence supporting the hypothesis of an oxidative cost of parental effort. Eurasian kestrel (Falco tinnunculus) males, but not females, had higher levels of oxidative damage while rearing nestlings (Casagrande et al. 2011). Similarly, Florida scrub jays (Aphelocoma coerulescens) males, but not females, had higher levels of oxidative damage after breeding (Heiss and Schoech 2012). Conversely, breeding barn swallow (Hirundo rustica) females had higher plasma protein oxidative damage than their mates, possibly because of the higher costs associated with gamete production and egg incubation (Rubolini et al. 2012). Studies of mammals have also found mixed support for the link between reproduction and oxidative stress (Nussey et al. 2009; Bergeron et al. 2011; Garratt et al. 2011), as well as for differences among tissues in sensitivity to oxidative stress (Yang et al. 2013). What is lacking so far in iteroparous species, especially in birds, is therefore a conclusive demonstration that parental effort increases oxidative damage. In this study, we used a brood size manipulation approach to test the hypothesis that an increase in nestling feeding effort results in an increase in oxidative damage and/or depletion in antioxidant defenses. To this end, we increased or decreased the brood size of freeliving barn swallows by unbalanced cross-fostering of nestlings. We then assessed the effects of brood manipulation on plasma oxidative damage and plasma non-enzymatic antioxidant capacity of parents during the nestling rearing period. We assessed the change in plasma oxidative stress levels of parents during the nestlings’ pre-hatching period, from day 1
post-hatching (baseline levels soon after brood size manipulation) to day 15 post-hatching, a few days before fledging of nestlings (Møller 1994). We predicted a higher workload (nestling feeding rate) and oxidative stress in birds rearing enlarged broods vs. those rearing reduced broods. Moreover, we tested the hypothesis that levels of non-enzymatic antioxidant defenses at day 15 post-hatching were lower if birds had initially higher baseline levels of oxidative damage, especially among parents rearing enlarged broods, which should experience higher workload. This is expected if antioxidant defenses are used up during the demanding nestling rearing phase to buffer high levels of oxidative damage. Furthermore, this mechanism implies that the baseline oxidative status before the nestling rearing period might constrain the future investment in parental care and, consequently, the reproductive success (Stier et al. 2012). In breeding birds, body mass shows a progressive decline through the breeding season, which is generally more pronounced in females than in males and reflects a short-term cost of reproduction (Rands et al. 2006). We therefore assessed the effects of brood size manipulation on the seasonal decline of body mass, predicting a greater decline in birds rearing enlarged vs. reduced broods.
Material and methods Brood size manipulation experiment Female barn swallows lay up to three clutches of two to seven eggs (modal clutch size is five) at 1-day intervals (Møller 1994). Both parents feed altricial nestlings up to several days after fledging, which occurs at 18–21 days of age (Møller 1994). Nestlings solicit parental feeding by conspicuous begging displays (Møller 1994), and sibling competition throughout the nestling stage is intense (Saino et al. 1997). The study was carried out in five barn swallow colonies east of Milan (Italy), during April–July 2009. After completion of hatching (nestlings 1–2 days old), we performed a partial cross-fostering of nestlings within pairs of synchronous broods (hereafter ‘dyads’), so as to either increase or decrease the original brood size by one nestling. Two broods were considered synchronous when hatching of at least one nestling in both nests occurred on the same day. Brood size before manipulation did not differ between enlarged and reduced broods [paired samples t-test; reduced: 4.35 (0.18 SE); enlarged: 4.70 (0.17 SE); t22 =1.50, P=0.15]. Post-manipulation brood size was 5.65 (0.18 SE) in enlarged broods and 3.30 (0.18 SE) in reduced broods. Within 2 days of cross-fostering (hereafter day 1 posthatching), we captured both parents using a specially designed nest-trap that allowed the capture of the target breeding pair while not disturbing other birds in the colony. Within a few
Naturwissenschaften
minutes after capture, we collected a blood sample in capillary tubes by puncturing the brachial vein and recorded body mass using a Pesola spring balance (±0.1 g). We repeated the capture of both parents, collected a second blood sample and measured their body mass again when nestlings were ca. 15 days old, i.e., at the end of the nestling rearing period (see above) (hereafter day 15 post-hatching). We captured the parents of the enlarged and reduced brood of a dyad on the same day or in 2 consecutive days, in random order. Blood samples were kept refrigerated until brought to the laboratory (within 10 h from collection), where they were centrifuged (10 min, 11,500 rpm). Both plasma and erythrocytes were frozen at −20 °C until we performed the analyses of plasma oxidative status. To verify whether our experimental manipulation of brood size effectively resulted in a higher workload of parents rearing enlarged vs. reduced broods, in a subsample of 13 dyads (26 broods) we assessed nestling rearing effort by recording the number of feeding visits to the nests in a single 2-h observation period per nest. Two observers hidden near the nest independently recorded visits of parents to the nest when nestlings were 12–17 days old, using markings on the parents that were previously captured to assign feeding visits to either the male or the female parent. To control for time of day and weather effects on nestling feeding rates, parents rearing an enlarged and a reduced brood of a given dyad were observed in rapid succession, while randomising whether the enlarged or the reduced brood was observed first. Although for some analyses we used the number of feeding visits in 2 h as the dependent variable (see below), throughout the results, for ease of interpretation, nestling feeding rate was expressed as the number of nest feeding visits per hour of observation. Analyses of plasma oxidative status Early oxidative damage products (mostly hydroperoxides) were measured using the d-ROMs assay (Diacron International, Grosseto, Italy), according to established protocols (e.g., Costantini et al. 2006). Briefly, the plasma (10 μl) was diluted with 200 μl of a solution containing 0.01 m acetic acid/sodium acetate buffer (pH 4.8) and N,N-diethyl-pphenylenediamine as chromogen. We then incubated the solution for 75 min at 37 °C. At the end of incubation, the supernatants of each sample, standard and blank were transferred to a fresh microplate to remove the precipitated protein present in sample wells. The absorbance was read using a spectrophotometer (Microplate Reader Model 550; Bio-Rad, Tokyo, Japan) at a wavelength of 490 nm. Levels of oxidative damage products were calculated using a calibration curve and were expressed as mM of H2O2 equivalents. The non-enzymatic antioxidant capacity of plasma was quantified using the OXY-adsorbent assay (Diacron International srl, Grosseto, Italy), according to established
protocols (e.g., Costantini et al. 2006). This assay quantifies the in vitro plasma antioxidant barrier to cope with the oxidant action of hypochlorous acid (HOCl; oxidant of pathological relevance in biological systems). The plasma (5 μl) was diluted 1:100 with distilled water. We incubated 200 μl aliquot of a HOCl solution with 5 μl of the diluted plasma for 10 min at 37 °C. We then added 5 μl of the same chromogen solution used for the d-ROMs assay. The alkyl-substituted aromatic amine dissolved in the chromogen is oxidized by the residual HOCl and transformed into a pink derivative, the intensity of which is inversely related to OXY. The absorbance was read using a spectrophotometer (Microplate Reader Model 550; Bio-Rad) at a wavelength of 490 nm. OXY levels were calculated using a reference standard and were expressed as mM of HOCl neutralised. The intra-assay and inter-assay coefficient of variation were 3.58 and 8.90 for the d-ROMs assay, and 4.93 and 8.45 for OXY assay, respectively. Statistical analyses The effect of brood size manipulation on nestling rearing effort was assessed by running a negative binomial linear mixed model with the total number of feeding visits over the 2 h observation period as the dependent variable and treatment (reduced vs. enlarged brood), sex and their interaction as predictors, while dyad and nest (nested within dyad) were included as random effects. For the analyses of treatment effects on OD, OXY and body mass, we relied on linear mixed models with a repeated measures design. We included the dyad and the nest (nested within dyad) as random effects in the analyses. The individual was included as a further random factor to account for repeated measures from the same individual. Treatment, sex, sampling day (day 1 vs. day 15 post-hatching) and their two- and three-way interactions were included as fixed factors. OD, OXY and body mass were included as dependent variables. Residuals of all models were normally distributed (details not shown). Post-hoc comparisons were performed when we found a statistically significant interaction effect. We calculated differences between least-square means for biologically meaningful comparisons, and adjusted their significance for multiple statistical tests according to the False Discovery Rate procedure (Benjamini and Hochberg 1995). In order to test whether OD and OXY levels at day 1 predicted OXY at day 15 post-hatching, we ran a linear mixed model with OXY at day 15 as the response variable and OD and OXY at day 1 as covariates. We also included sex, treatment and the two-way interactions between factors and between factors and covariates as predictors, while dyad and nest (nested within dyad) were included as random effects.
Naturwissenschaften
Results Effects of brood size manipulation on parental workload Observations of parental feeding visits in a subsample of 13 dyads indicated that our manipulation of brood size affected parental workload. In fact, parents rearing enlarged broods returned more often to the nest to feed their nestlings than those rearing reduced broods [sexes pooled; reduced broods: 4.60 (0.91 SE) feedings×h−1; enlarged: 9.10 (1.04 SE) feedings×h−1; estimate=0.74 (0.18 SE), z=4.03, P
ORIGINAL PAPER
Nestling rearing is antioxidant demanding in female barn swallows (Hirundo rustica) David Costantini & Andrea Bonisoli-Alquati & Diego Rubolini & Manuela Caprioli & Roberto Ambrosini & Maria Romano & Nicola Saino
Received: 21 March 2014 / Revised: 24 April 2014 / Accepted: 20 May 2014 # Springer-Verlag Berlin Heidelberg 2014
Abstract Reproduction is a demanding activity, since organisms must produce and, in some cases, protect and provision their progeny. Hence, a central tenet of life-history theory predicts that parents have to trade parental care against body maintenance. One physiological cost thought to be particularly important as a modulator of such trade-offs is oxidative stress. However, evidence in favour of the hypothesis of an oxidative cost of reproduction is contradictory. In this study, we manipulated the brood size of wild barn swallows Hirundo rustica soon after hatching of their nestlings to test whether an increase in nestling rearing effort translates into an increased oxidative damage and a decreased antioxidant protection at the end of the nestling rearing period. We found that, while plasma oxidative damage was unaffected by brood size enlargement, females rearing enlarged broods showed a decrease in plasma non-enzymatic antioxidants during the nestling rearing period. This was not the case among females rearing Communicated by: Alexandre Roulin D. Costantini (*) Department of Biology, University of Antwerp, Universiteitsplein 1, 2610 Wilrijk, Belgium e-mail: [email protected] D. Costantini Institute for Biodiversity, Animal Health and Comparative Medicine, University of Glasgow, Glasgow G12 8QQ, UK A. Bonisoli-Alquati Department of Biological Sciences, University of South Carolina, 715 Sumter Street, 29208 Columbia, SC, USA D. Rubolini : M. Caprioli : M. Romano : N. Saino Dipartimento di Bioscienze, Università degli Studi di Milano, Milan, Italy R. Ambrosini Dipartimento di Biotecnologie e Bioscienze, Università degli Studi di Milano-Bicocca, Milan, Italy
reduced broods and among males assigned to either treatment. Moreover, individuals with higher plasma oxidative damage soon after the brood size manipulation had lower plasma nonenzymatic antioxidants at the end of the nestling rearing period, suggesting that non-enzymatic antioxidants were depleted to buffer the negative effects of high oxidative damage. Our findings point to antioxidant depletion as a potential mechanism mediating the cost of reproduction among female birds. Keywords Antioxidants . Life history . Oxidative damage . Oxidative stress . Parental effort . Trade-off
Introduction A central concept in evolutionary ecology is that all fitnessrelated traits cannot be simultaneously maximised, but instead they are traded-off against each other because of physiological constraints (e.g., stress, energy consumption) that arise as a consequence of resource investment (Stearns 1992). Reproduction is a demanding phase of organisms’ lives, since high investment in reproduction may lead to increased senescence or reduced survival and future fecundity (Williams 1966; Stearns 1992; Boonekamp et al. 2014). Although this paradigm is well grounded into evolutionary ecology, very little is known about how such costs of reproduction are actually incurred, since the majority of the studies have focused on the outcomes rather than the mechanisms. It has been recently proposed that oxidative stress may be one cost associated with reproductive investment (Costantini 2008; Metcalfe and Alonso-Alvarez 2010). Oxidative stress is the rate at which oxidative damage to biomolecules is generated, which is dependent on a complex balance between prooxidants and antioxidant mechanisms (Halliwell and Gutteridge 2007; Costantini and Verhulst 2009). Such a
Naturwissenschaften
damage can contribute to cell senescence and loss in organismal performance (Weinert and Timiras 2003), and may influence life-history strategies (Costantini 2008; Metcalfe and Alonso-Alvarez 2010). Although some studies have provided evidence that oxidative stress increases during reproduction, others have not (Stier et al. 2012; Metcalfe and Monaghan 2013; Speakman and Garratt 2014). Due to the relative ease of altering parental effort via clutch or brood size manipulation experiments, many studies have focused on altricial avian species. For example, in an experiment manipulating brood size of zebra finches (Taeniopygia guttata), males experiencing a higher workload were found to show no change in daily energy expenditure, but a lower antioxidant enzyme activity, whereas females showed increased energy expenditure but no change in antioxidant enzyme activity (Wiersma et al. 2004). Further studies have also found decreases in blood antioxidant protection in reproducing birds when rearing an enlarged brood (Alonso-Alvarez et al. 2004; Losdat et al. 2011; Christie et al. 2012) or when engaging in a higher number of reproductive episodes (Bertrand et al. 2006). However, the results of these studies are difficult to interpret, especially because they report only measures of components of antioxidant defenses, while measures of oxidative damage are largely lacking. Two studies have found evidence supporting the hypothesis of an oxidative cost of parental effort. Eurasian kestrel (Falco tinnunculus) males, but not females, had higher levels of oxidative damage while rearing nestlings (Casagrande et al. 2011). Similarly, Florida scrub jays (Aphelocoma coerulescens) males, but not females, had higher levels of oxidative damage after breeding (Heiss and Schoech 2012). Conversely, breeding barn swallow (Hirundo rustica) females had higher plasma protein oxidative damage than their mates, possibly because of the higher costs associated with gamete production and egg incubation (Rubolini et al. 2012). Studies of mammals have also found mixed support for the link between reproduction and oxidative stress (Nussey et al. 2009; Bergeron et al. 2011; Garratt et al. 2011), as well as for differences among tissues in sensitivity to oxidative stress (Yang et al. 2013). What is lacking so far in iteroparous species, especially in birds, is therefore a conclusive demonstration that parental effort increases oxidative damage. In this study, we used a brood size manipulation approach to test the hypothesis that an increase in nestling feeding effort results in an increase in oxidative damage and/or depletion in antioxidant defenses. To this end, we increased or decreased the brood size of freeliving barn swallows by unbalanced cross-fostering of nestlings. We then assessed the effects of brood manipulation on plasma oxidative damage and plasma non-enzymatic antioxidant capacity of parents during the nestling rearing period. We assessed the change in plasma oxidative stress levels of parents during the nestlings’ pre-hatching period, from day 1
post-hatching (baseline levels soon after brood size manipulation) to day 15 post-hatching, a few days before fledging of nestlings (Møller 1994). We predicted a higher workload (nestling feeding rate) and oxidative stress in birds rearing enlarged broods vs. those rearing reduced broods. Moreover, we tested the hypothesis that levels of non-enzymatic antioxidant defenses at day 15 post-hatching were lower if birds had initially higher baseline levels of oxidative damage, especially among parents rearing enlarged broods, which should experience higher workload. This is expected if antioxidant defenses are used up during the demanding nestling rearing phase to buffer high levels of oxidative damage. Furthermore, this mechanism implies that the baseline oxidative status before the nestling rearing period might constrain the future investment in parental care and, consequently, the reproductive success (Stier et al. 2012). In breeding birds, body mass shows a progressive decline through the breeding season, which is generally more pronounced in females than in males and reflects a short-term cost of reproduction (Rands et al. 2006). We therefore assessed the effects of brood size manipulation on the seasonal decline of body mass, predicting a greater decline in birds rearing enlarged vs. reduced broods.
Material and methods Brood size manipulation experiment Female barn swallows lay up to three clutches of two to seven eggs (modal clutch size is five) at 1-day intervals (Møller 1994). Both parents feed altricial nestlings up to several days after fledging, which occurs at 18–21 days of age (Møller 1994). Nestlings solicit parental feeding by conspicuous begging displays (Møller 1994), and sibling competition throughout the nestling stage is intense (Saino et al. 1997). The study was carried out in five barn swallow colonies east of Milan (Italy), during April–July 2009. After completion of hatching (nestlings 1–2 days old), we performed a partial cross-fostering of nestlings within pairs of synchronous broods (hereafter ‘dyads’), so as to either increase or decrease the original brood size by one nestling. Two broods were considered synchronous when hatching of at least one nestling in both nests occurred on the same day. Brood size before manipulation did not differ between enlarged and reduced broods [paired samples t-test; reduced: 4.35 (0.18 SE); enlarged: 4.70 (0.17 SE); t22 =1.50, P=0.15]. Post-manipulation brood size was 5.65 (0.18 SE) in enlarged broods and 3.30 (0.18 SE) in reduced broods. Within 2 days of cross-fostering (hereafter day 1 posthatching), we captured both parents using a specially designed nest-trap that allowed the capture of the target breeding pair while not disturbing other birds in the colony. Within a few
Naturwissenschaften
minutes after capture, we collected a blood sample in capillary tubes by puncturing the brachial vein and recorded body mass using a Pesola spring balance (±0.1 g). We repeated the capture of both parents, collected a second blood sample and measured their body mass again when nestlings were ca. 15 days old, i.e., at the end of the nestling rearing period (see above) (hereafter day 15 post-hatching). We captured the parents of the enlarged and reduced brood of a dyad on the same day or in 2 consecutive days, in random order. Blood samples were kept refrigerated until brought to the laboratory (within 10 h from collection), where they were centrifuged (10 min, 11,500 rpm). Both plasma and erythrocytes were frozen at −20 °C until we performed the analyses of plasma oxidative status. To verify whether our experimental manipulation of brood size effectively resulted in a higher workload of parents rearing enlarged vs. reduced broods, in a subsample of 13 dyads (26 broods) we assessed nestling rearing effort by recording the number of feeding visits to the nests in a single 2-h observation period per nest. Two observers hidden near the nest independently recorded visits of parents to the nest when nestlings were 12–17 days old, using markings on the parents that were previously captured to assign feeding visits to either the male or the female parent. To control for time of day and weather effects on nestling feeding rates, parents rearing an enlarged and a reduced brood of a given dyad were observed in rapid succession, while randomising whether the enlarged or the reduced brood was observed first. Although for some analyses we used the number of feeding visits in 2 h as the dependent variable (see below), throughout the results, for ease of interpretation, nestling feeding rate was expressed as the number of nest feeding visits per hour of observation. Analyses of plasma oxidative status Early oxidative damage products (mostly hydroperoxides) were measured using the d-ROMs assay (Diacron International, Grosseto, Italy), according to established protocols (e.g., Costantini et al. 2006). Briefly, the plasma (10 μl) was diluted with 200 μl of a solution containing 0.01 m acetic acid/sodium acetate buffer (pH 4.8) and N,N-diethyl-pphenylenediamine as chromogen. We then incubated the solution for 75 min at 37 °C. At the end of incubation, the supernatants of each sample, standard and blank were transferred to a fresh microplate to remove the precipitated protein present in sample wells. The absorbance was read using a spectrophotometer (Microplate Reader Model 550; Bio-Rad, Tokyo, Japan) at a wavelength of 490 nm. Levels of oxidative damage products were calculated using a calibration curve and were expressed as mM of H2O2 equivalents. The non-enzymatic antioxidant capacity of plasma was quantified using the OXY-adsorbent assay (Diacron International srl, Grosseto, Italy), according to established
protocols (e.g., Costantini et al. 2006). This assay quantifies the in vitro plasma antioxidant barrier to cope with the oxidant action of hypochlorous acid (HOCl; oxidant of pathological relevance in biological systems). The plasma (5 μl) was diluted 1:100 with distilled water. We incubated 200 μl aliquot of a HOCl solution with 5 μl of the diluted plasma for 10 min at 37 °C. We then added 5 μl of the same chromogen solution used for the d-ROMs assay. The alkyl-substituted aromatic amine dissolved in the chromogen is oxidized by the residual HOCl and transformed into a pink derivative, the intensity of which is inversely related to OXY. The absorbance was read using a spectrophotometer (Microplate Reader Model 550; Bio-Rad) at a wavelength of 490 nm. OXY levels were calculated using a reference standard and were expressed as mM of HOCl neutralised. The intra-assay and inter-assay coefficient of variation were 3.58 and 8.90 for the d-ROMs assay, and 4.93 and 8.45 for OXY assay, respectively. Statistical analyses The effect of brood size manipulation on nestling rearing effort was assessed by running a negative binomial linear mixed model with the total number of feeding visits over the 2 h observation period as the dependent variable and treatment (reduced vs. enlarged brood), sex and their interaction as predictors, while dyad and nest (nested within dyad) were included as random effects. For the analyses of treatment effects on OD, OXY and body mass, we relied on linear mixed models with a repeated measures design. We included the dyad and the nest (nested within dyad) as random effects in the analyses. The individual was included as a further random factor to account for repeated measures from the same individual. Treatment, sex, sampling day (day 1 vs. day 15 post-hatching) and their two- and three-way interactions were included as fixed factors. OD, OXY and body mass were included as dependent variables. Residuals of all models were normally distributed (details not shown). Post-hoc comparisons were performed when we found a statistically significant interaction effect. We calculated differences between least-square means for biologically meaningful comparisons, and adjusted their significance for multiple statistical tests according to the False Discovery Rate procedure (Benjamini and Hochberg 1995). In order to test whether OD and OXY levels at day 1 predicted OXY at day 15 post-hatching, we ran a linear mixed model with OXY at day 15 as the response variable and OD and OXY at day 1 as covariates. We also included sex, treatment and the two-way interactions between factors and between factors and covariates as predictors, while dyad and nest (nested within dyad) were included as random effects.
Naturwissenschaften
Results Effects of brood size manipulation on parental workload Observations of parental feeding visits in a subsample of 13 dyads indicated that our manipulation of brood size affected parental workload. In fact, parents rearing enlarged broods returned more often to the nest to feed their nestlings than those rearing reduced broods [sexes pooled; reduced broods: 4.60 (0.91 SE) feedings×h−1; enlarged: 9.10 (1.04 SE) feedings×h−1; estimate=0.74 (0.18 SE), z=4.03, P
