1Most animals do not grow at their maximal rate. This might appear puzzling because the early attainment of a large body size incurs several selective benefits, such as reduced risk of predation and earlier reproductive output. Several hypotheses have been suggested to explain this paradox. Among them, the cost due to high levels of oxidative stress, as the consequence of sustained metabolic activity during growth, has been put forward.
2In this study, we wished to assess this cost in captive zebra finches (Taeniopygia guttata). In order to manipulate access to food and consequently early growth, hatchlings were randomly assigned to reduced or enlarged broods. Even though nestlings raised in enlarged broods were smaller when 20-days-old compared to nestlings raised in reduced broods, they grew faster during the following 20-day period.
3When 60-days-old, we measured the resistance of red blood cells against a free radical attack and correlated these values to nestling growth rate. In agreement with the prediction, we found that red blood cell resistance to free radicals was negatively correlated with growth rate, nestlings that grew faster being those with the weakest capacity to resist a free radical attack. These results support the hypothesis that oxidative damage might constrain growth rate. Furthermore, the fact that the relationship was only established during the period of accelerated growth after the initial delay also suggests that compensatory growth can negatively affect the individual resistance to oxidative damage.
If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
However, the nature of the mechanisms that give rise to the correlation between oxidative stress and growth rate remain poorly understood. Our current knowledge comes mostly from the experimental manipulation of growth factors, mutant strains and transgenic models (i.e. rodents). For instance, it has been shown that mutations that affect growth factor pathways (GH/IGF-1/insulin) also influence the expression of antioxidant defences (Rollo 2002 and references therein). Moreover, whereas the enzymatic antioxidant defences are up-regulated in GH-deficient dwarf mice, their activity is decreased when GH is artificially restored (Brown-Borg & Rakoczy 2003; Bartke & Brown-Borg 2004).
Surprisingly, simple experimental approaches have been overlooked. As far as we know, no study has experimentally explored the relationship between the among-individual variability in growth rate and susceptibility to oxidative damage in any vertebrate species.
In an attempt to fill this gap, in this study we assessed the effect of experimentally manipulated growth rate of developing zebra finches (Taeniopygia guttata) on red blood cell resistance to free radicals. Growth rate was manipulated by the mean of a brood size manipulation experiment. This experiment was set up to investigate the long-term effects of initial conditions on life-history trajectories as reported in Alonso-Alvarez et al. (2006). Birds reared in enlarged broods showed lower red blood cell resistance to free radicals at the age at maturity (i.e. 60-days-old; Alonso-Alvarez et al. 2006). Here, we specifically focused on the effect of accelerated growth rate, after a period of reduced somatic growth, on red blood cell resistance to free radicals.
Materials and methods
The present study was conducted on a population of captive zebra finches. The findings reported here are based on an experiment aimed at exploring the long-term life-history consequences of early developmental conditions (see Alonso-Alvarez et al. 2006 for the full description of the experiment). Randomly formed pairs were placed in breeding cages with both food (a commercial mix of seeds) and water ad libitum. For all birds (all less than 1-year-old) this was the first breeding attempt. Breeding cages were maintained in a room with constant temperature and light–dark daily cycles (21 °C ± 1 °C and 13 L : 11 D, respectively). Breeding pairs were randomly allocated into two groups and allowed to raise either a two-chick brood or a six-chick brood (within the range of natural brood sizes; Zann 1996). Thirty-four pairs reared reduced broods and 18 pairs raised enlarged broods. To disrupt the natural covariation between brood size and parental quality, brood size was manipulated when nestlings were 2-days-old. Initial clutch size did not differ between the groups that further raised two or six chicks (Alonso-Alvarez et al. 2006). Most birds had left the nestbox at 20 days of age, acquiring the capacity to feed by themselves during the next 20 days (Zann 1996). Birds were separated from parents when they were 40-days-old.
body mass gain measurement
Body mass (±0·1 g) was assessed at 1, 20, 40 and 60 days of age. When 60-days-old all birds were released in a large outdoor aviary and their lifetime breeding success and lifespan carefully monitored (details in Alonso-Alvarez et al. 2006).
We computed growth rate (in mass) for three periods. The first period corresponded to the time when nestlings are most dependent on parents (1–20 days of age); the second period corresponded to the time when nestlings begin to be trophically independent (21–40 days of age); the third period corresponded to the time of full nestling independence (41–60 days of age). Growth rate was computed for each of these periods as the increment in body mass (body mass 2 minus body mass 1). The results were identical when using other measurements of growth rate such as the residuals of a linear regression between (body mass 2 minus body mass 1) and (body mass 1), which corrects for the among-individual differences in body mass at the beginning of each period. We restricted our analysis to growth in body mass, because skeletal growth (tarsus growth) is mostly achieved by the age of 20 days, with very little gain in tarsus length occurring between day 21 and day 60 (Alonso Alvarez et al. 2006).
muscle growth and fat reserves
In order to evaluate the development of muscles and fat reserves, we used muscle and fat scores. The muscle score estimates the relative size of the pectoral muscles, observing the cross-sectional contour at mid-sternum and rating the shape on a scale from 0 (slightly concave) to 3 (well-rounded convex) (see Geller & Temple 1983). The fat score estimates the size of the intrafurcular fat also on a 0–3 scale (e.g. Dall & Witter 1998). Scoring was performed blindly with respect to the brood size treatment. Muscle and fat scores were assessed at the age of 20, 40 and 60 days.
assessment of red blood cell resistance to oxidative damage
A blood sample was taken from each bird at 60-days-old to assess the resistance of red blood cells to a controlled free radical attack (N = 50 and 68, for birds reared in reduced and enlarged broods, respectively). We used the KRL test (Brevet Spiral V02023, Couternon, France; http://www.nutriteck.com/sunyatakrl.html) adapted to bird physiological parameters (osmolarity, temperature) (Alonso-Alvarez et al. 2004a, 2004b). This assay provides a quantitative measurement of the overall antioxidant status of whole blood as it assesses the time required to haemolyse 50% of red blood cells of the sample when exposed to a controlled free radical attack. Briefly, 20 µL of whole blood were immediately diluted in 730 µL of KRL buffer (150 mm Na+, 120 mm Cl−, 6 mm K+, 24 mm , 2 mm Ca2+, 340 mOsM, pH 7·4) and stored at 4 °C before analysis that occurred within 24 h.
A 80 µL of the whole blood diluted in the KRL buffer were added to each well of a 96-well microplate. We subsequently added to each well 136 µL of a 150 mm solution of 2,2′-azobis-(amidinopropane)hydrochloride (646 mg of [2,2′-azobis-(amidinopropane)hydrochloride] (AAPH; a free radical generator), that is, 646 mg of AAPH diluted in 20 mL of KRL buffer (Rojas Wahl et al. 1998). The microplate was subsequently read with a microplate titrator (iEMS Reader MF, Kirial SA, Couternon, France) at 40 °C. The rate of haemolysis is determined by the change in optical density measured at 540 nm (e.g. Alonso-Alvarez et al. 2004a, 2004b; Bertrand et al. 2006).
AAPH acts mostly by producing peroxyl radicals, which induce lipid and protein peroxidation in the cell membrane (see Dai et al. 2006; Yang et al. 2006 and references therein). Therefore, we can expect that if the amount of peroxidised molecules in red blood cells at the time of the assay is already high, the time required to induce haemolysis should be reduced (e.g. Brzezinska-Slebodzinska 2001). In addition, previous work has shown that, if at least one component of the anti-radical detoxification system is impaired, the haemolysis curve is shifted towards shorter times (Blache & Prost 1992; Esterbauer & Ramos 1996; Girard et al. 2005). In summary, this assay reflects the current availability of antioxidant defences as well as the past oxidative insult experienced by red blood cells.
Data were analysed using mixed models and repeated-measurements mixed models of covariance (proc mixed in sas software; SAS Institute 2001). In order to control for random effects of birds sharing the same environment as nestlings or being genetically related, data were cross-classified by including the original and foster broods where birds were reared as random factors. The identity of each bird (N = 118) was included as a third random factor in the repeated-measurements mixed models. Experimental brood size (two- vs six-chicks) was included as fixed factor in every model. The sex of each bird was also tested in the models, but was removed as a non-significant term (all P-values > 0·10). To explore the effect of compensatory growth on red blood cell resistance to free radicals, we ran a model where the three growth periods were included as covariates. This was done to disentangle the possible confounding effect of depressed growth during the first growth period and enhanced growth during the second period.
Fat and muscle scores and their increments (score at the end of the period minus score at the beginning of the period) were analysed with ordinal analyses of variance with a multinomial distribution of errors and cumulative logit link function (genmod procedure in sas software). The influence of fat and muscle scores on body mass gain and red blood cell resistance to free radicals was assessed by including them, as categorical variables (i.e. fixed factors), in the mixed models.
In order to take into account random effects, the reported means and SEMs are least square means and SEs obtained from the mixed model (‘lsmeans’ statement in sas software; e.g. Gil et al. 2004). Degrees of freedom were approximated by Satterthwaite correction (Littell et al. 1996).
body mass gain
Growth trajectories were significantly different between the two brood size groups (Table 1). As expected, nestlings in the enlarged broods grew slower during their first 20 days of age (Fig. 1). The pattern was reversed during the following growth period (between 21 and 40 days of age), with birds from enlarged broods growing at a faster rate than nestlings from reduced broods (Fig. 1). Finally, there was almost no increment in body mass during the third period (between 41 and 60 days of age) and, thus, the two brood size groups did not differ with respect to growth in this period (Fig. 1). Therefore, birds from enlarged broods had slow growth during the first period, accelerated growth during the second period and similar growth during the third period when compared to reduced-brood nestlings. To corroborate the idea that birds with fast early growth (during the 1–20 days of age) had slower growth afterwards and that nestlings with slow early growth had an accelerated growth afterwards, we found that growth rate during the 1–20 days of age period was negatively correlated with growth rate during the 21–40 days of age period (mixed model with growth rate between 21 and 40 days of age as dependent variable and growth rate between 1 and 20 days of age as independent variable: F 1,112 = 5·08, P = 0·026, parameter estimate ± SE: –0·131 ± 0·0578, R2 = –0·135).
Table 1. Repeated-measurements mixed model reporting the effect of period (between 1 and 20, 21 and 40, 41 and 60 days of age) and the experimental brood size on increment in body mass (g)
Experimental brood size
Experimental brood size × Period
Although the difference in body mass between birds issued from enlarged and reduced broods clearly decreased with time, the two groups were still statistically different at the age of 60 days (F1,13·7 = 7·01, P = 0·020; 15·77 ± 0·27 g and 14·88 ± 0·28 g for birds in reduced and enlarged broods, respectively). However, when birds were recaptured about 3 months after their release into the aviary, body mass of brood reduced and brood enlarged birds was statistically indistinguishable (P = 0·401).
To test the effect of compensatory growth on red blood cell resistance to free radicals, we ran a model with experimental brood size as factor and the three growth rate periods as covariates. We found that only growth rate during the 21–40 days of age period was negatively correlated with red blood cell resistance to free radicals (Table 2 and Fig. 2). Interestingly, growth rate during the first age period (1–20 days) was not significantly correlated with red blood cell resistance to oxidative damage, suggesting that the difference in resistance to oxidative stress between brood size groups (i.e. Alonso Alvarez et al. 2006) was not due to the unfavourable conditions experienced during the first days of life, but more likely to the compensatory growth that took place when birds began to be independent from their parents. Thus, if the experimental brood size is maintained alone into the model, its effect is significant (P < 0·05). Nevertheless, if the increase in body mass during the 21–40 day period is added as a covariate, the effect of the experimental brood size loses its significance (P = 0.43). In fact, only body mass gain during the 21–40 day period remains as significant term after a backward step-wise procedure (note in Table 2). This supports the idea that the difference in oxidative stress resistance between brood size groups was indeed due to the effect of accelerated growth in some birds, mostly those in enlarged broods.
Table 2. Mixed model reporting the effect of experimental brood size and the three growth periods on the red blood cell resistance to free radicals assessed at the age of 60 days
Experimental brood size
Increment in body mass during the 1–20 day period
Increment in body mass during the 21–40 day period
Increment in body mass during the 41–60 day period
In a backward procedure, only 21–40 day period remains in the model (P = 0·003).
Since birds from enlarged broods were those with the highest growth during the 21–40 day period (Figs 1 and 2), we might expect that the relationship between red blood cell resistance and growth should be mostly due to birds raised in enlarged broods. In agreement with this hypothesis, although the interaction between experimental brood size and growth rate was not statistically significant (F1,101 = 0·36, P = 0·550), we found that, when the correlation between red blood cell resistance to free radicals and growth rate was tested separately for the two groups of birds, it was statistically significant for enlarged-brood individuals (F1,64·1= 5·58, P = 0·021) but not for reduced-brood birds (F1,46·6 = 2·12, P = 0·152). In fact, the slope of the regression between mass gain and red blood cell resistances to free radicals was 36% steeper for birds of the enlarged brood compared to birds from reduced broods (estimated slopes ± SE: –1·29 ± 0·54 and –0·82 ± 0·56, respectively).
As mentioned above, using the residuals of a regression of body mass gain on the body mass at the beginning of each period (therefore controlling for among individual variation in previous body mass) provided the same results, with the growth rate during the 21–40 days period being the only significant predictor of red blood cell resistance to free radicals.
muscle growth and fat reserves
To understand the nature of the body mass gain, we analysed muscle and fat scores. Not surprisingly, both muscle and fat scores increased during the 21–40 day period, showing that birds continued to add resources to their muscular tissue and lipid stocks. However the rate of development was double for muscles (40d–21d scores; mean: ± SE: 0·41 ± 0·06) than for fat (0·23 ± 0·06). As for body mass, both muscle and fat scores stayed constant during the 41–60 day period (–0·05 ± 0·04 and 0·01 ± 0·04, respectively). Body mass gain between 21 and 40 days was positively correlated with the increase of muscle scores during that period (F3,72·8 = 4·50, P = 0·006) but not with the fat score increment (F3,72·8 = 1·49, P = 0·225). A similar result was also found for the 41–60 day period (muscle score: P = 0·035; fat score: P = 0·091). Birds from reduced broods showed higher fat scores from the beginning to the end of the sampling period (all P < 0·05). Muscle scores were higher in birds from reduced broods at 20 and 60 days of age (both P < 0·010), but not at 40 days of age (χ2 = 0·75, df = 1, P = 0·39). Chicks from enlarged broods showed a higher increase of muscle score during the 21–40 day period than chicks from reduced broods (means ± SE: +0·59 ± 0·08 and +0·16 ± 0·07, respectively; χ2 = 18·71, df = 1, P < 0·0001). The increase of fat score during the same period was slightly lower in birds from enlarged broods (means ± SE: +0·21 ± 0·08 and +0·26 ± 0·08, enlarged and reduced broods, respectively; χ2 = 3·93, df = 1, P = 0·047). There was no statistically significant difference in fat and muscle increments between brood size groups in the 41–60 day period (P > 0·05). Fat and muscle scores were not correlated with red blood cell resistance to free radicals (all P > 0·18).
This study suggests that accelerated growth rate, following a period of depressed growth, can negatively affect a measurement of organismal capacity to resist a free radical attack. The results, therefore, indicate that increased susceptibility to oxidative damage can be a cost of compensatory growth in zebra finches. Our results also suggest that the effect of compensatory growth was mostly due to a compensatory development of pectoral muscles.
Several studies have tried to explain the adaptive nature of compensatory growth as well as the conditions favouring its evolution (see a recent review in Mangel & Much 2005). In a recent theoretical study, Mangel & Munch (2005) worked out a life-history model that incorporates the mortality consequences of both size and damage, in order to understand the evolution of compensatory growth. The model explicitly assumes that the key to understanding compensatory growth resides in the fact that growth leads to the accumulation of cellular damage, which is ultimately expressed at the organismal level (Mangel & Munch 2005). Although the nature of the cellular damage is still debated, several authors have suggested that oxidative stress might play a primary role (e.g. Metcalfe & Monaghan 2001; Mangel & Munch 2005 and references therein).
Metcalfe & Monaghan (2001, 2003) suggested that the costs induced by a ‘catch-up’ growth, that is, a compensatory growth after a transitory delay, probably differ from those induced by a simple fast growth period. Their argument is based on the fact that the number of cells in key organs is fixed relatively early during development (Robinson, Sinclair & McEvoy 1999). If energy deficit coincides with such a critical period, organisms would develop with fewer differentiated cells. A subsequent catch-up growth would imply an increased metabolic loading (Ozanne & Hales 2002, 2004), which could have consequences in terms of higher levels of oxidative stress.
We measured red blood cell resistance to free radicals at the age of 60 days, which is about 20 days after the period of compensatory growth (21–40 days) had occurred. The negative correlation between the growth rate between 21 and 40 days, and red blood cell resistance to free radicals indicates that accelerated growth had a long lasting effect on the antioxidant status. This is not surprising given that a few studies have shown that early environmental conditions have a very long lasting effect on the level of a number of antioxidant molecules. For instance, zebra finch nestlings fed with a poor quality diet during their first 15 days of life experienced substantially decreased amounts of circulating vitamin A, E and carotenoids when 100-days-old (Blount et al. 2003). Similarly, we have found that zebra finch nestlings raised in enlarged broods have much lower levels of circulating carotenoids compared with nestlings raised in reduced broods at the age of 60 days, which is well after the period of trophic independence (Alonso Alvarez et al. 2006).
Because of our experimental design, compensatory growth occurred in birds from enlarged broods, since they were those with depressed growth during the first 20 days of life. We should have expected therefore that any effect of growth on red blood cell resistance to free radicals should be evident in birds from enlarged broods. In agreement with this hypothesis, we found that the correlation between growth and red blood cell resistance to free radicals was significant only for the group of birds from experimentally enlarged broods. Nevertheless, we acknowledge that the interaction between brood size and growth was not statistically significant, although the slope for enlarged-brood birds was 36% steeper than for reduced-brood birds. This is however not so surprising, given that we expected negative slopes for the two groups of brood size and therefore the statistical power to detect a significant interaction in this particular case has to be very high.
Recently, using an elegant experimental design, Fisher et al. (2006) showed that compensatory growth had a negative impact on cognitive performance of adult zebra finches. Birds that had the strongest compensatory growth following a period of nutritional deficit were those with the poorer learning performance in adulthood. The mechanism underlying such a link is unknown; however, oxidative stress might play a role. The brain is known to be particularly susceptible to oxidative damage, and oxidative stress can induce substantial neurodegenerative disorders (Coyle & Puttfarcken 1993; Lin & Beal 2006). Although we did not assess the oxidative damage of compensatory growth on the brain of our birds, the assay we adopted is supposed to indicate the organismal capacity to face a free radical attack. As such it could be that zebra finches with accelerated growth during the 21–40 days of age period also paid a cost in terms of oxidative-induced neurodegenerative disorders. This obviously deserves further work.
In summary, our results suggest that increased susceptibility to free radicals could be a cost of accelerated growth. This cost would contribute to explain why most animals do not grow at their maximal rate (Metcalfe & Monaghan 2003). Furthermore, the fact that the relationship between resistance to free radicals and body mass gain was only established during the period of accelerated growth after the initial delay also suggests that compensatory growth (also catch-up growth) can negatively affect the individual resistance to oxidative damage.
We are grateful to Pat Monaghan and Diego Gil for their valuable comments on the manuscript. We thank the staff of the Station Biologique de Foljuif (École Normale Supérieure). Financial support was provided by the Ministère de la Recherche (ACI Jeunes Chercheurs to GS), and the Université de Bourgogne (BQR to GD, JP and BF). We also thank the ANR, Projet number ANR-NT05–2_45491 for financial support to BF and SB. CA-A was supported by a Ramon y Cajal fellowship (Ministerio de Educación, Cultura y Deporte, Spain).