Oecologia DOI 10.1007/s00442-015-3429-9
PHYSIOLOGICAL ECOLOGY - ORIGINAL RESEARCH
Starting with a handicap: effects of asynchronous hatching on growth rate, oxidative stress and telomere dynamics in free-living great tits Antoine Stier1 · Sylvie Massemin2,3 · Sandrine Zahn2,3 · Mathilde L. Tissier2,3 · François Criscuolo2,3
Received: 1 October 2014 / Accepted: 14 August 2015 © Springer-Verlag Berlin Heidelberg 2015
Abstract A trade-off between resource investment into growth rate and body self-maintenance is likely to occur, but the underlying molecular mediators of such a trade-off remain to be determined. In many altricial birds, hatching asynchrony creates a sibling competitive hierarchy within the brood, with first-hatched nestlings enjoying substantial advantages compared to last-hatched nestlings. We used this opportunity to test for a trade-off between growth and self-maintenance processes (oxidative stress, telomere erosion) in great tit nestlings, since resource availability and allocation are likely to differ between first-hatched and last-hatched nestlings. We found that despite their starting competitive handicap (i.e. being smaller/lighter before day 16), last-hatched nestlings exhibited growth rate and mass/ size at fledging similar to first-hatched ones. However, lasthatched nestlings suffered more in terms of oxidative stress, and ended growth with shorter telomeres than first-hatched ones. Interestingly, growth rate was positively related to plasma antioxidant capacity and early life telomere length (i.e. at 7 days old), but among last-hatched nestlings, those exhibiting the faster body size growth were also those exhibiting the greatest telomere erosion. Last-hatched nestlings exhibited elevated levels of plasma testosterone (T), but only at day 7. T levels were positively associated with
Communicated by Mark A. Chappell. * Antoine Stier
[email protected] 1
GECCO (Groupe écologie et conservation des vertébrés), University of Angers, Angers, France
2
IPHC, Université de Strasbourg, 23 rue Becquerel, 67087 Strasbourg Cedex 2, France
3
UMR7178, CNRS, 67087 Strasbourg, France
oxidative damage levels and plasma antioxidant capacity, the latter being only significant for first-hatched nestlings. Our results suggest that last-hatched nestlings present a specific trade-off between growth rate and self-maintenance processes, which is possibly driven by their need to compete with their older siblings and potentially mediated by elevated levels of T. Keywords Antioxidant · Testosterone · Intra-brood competition · Trade-off · Self-maintenance · Oxidative damage
Introduction A bad start in early development could have a profound impact on individuals’ life history trajectories (Metcalfe and Monaghan 2001). It is well established that individuals are able to compensate for an early deficit by accelerating their growth rate once favourable conditions are restored [i.e. compensatory growth (Metcalfe and Monaghan 2001)]. However, this accelerated growth might be associated with both ecological [e.g. predation risk (Gotthard 2001)] and physiological costs [e.g. higher resting metabolic rate (Criscuolo et al. 2008); altered physical performances (Criscuolo et al. 2011); or reduced lifespan (Lee et al. 2013)], leading to the idea that a trade-off in resource allocation exists between growth rate and self-maintenance processes. In the last decade, more attention has been given to the specific nature of the mechanisms underlying the global idea of trade-offs (Zera and Harshman 2001). For instance, the deleterious impact of one trait on another might be mediated through oxidative stress (Dowling and Simmons 2009; Metcalfe and Alonso Alvarez 2010). Oxidative stress is defined as the breakdown in the equilibrium between
13
Oecologia
the generation of pro-oxidants (mainly by the mitochondria during normal energy processing) and the antioxidant defences, which leads to oxidative damage to biomolecules and potentially to ageing (Finkel and Holbrook 2000; Halliwell and Gutteridge 2007). Oxidative stress could contribute to an acceleration of telomere erosion (i.e. protective non-coding nucleoprotein structures located at the end of eukaryotic chromosomes), which ultimately leads to cell death or replicative senescence (von Zglinicki 2002; Monaghan and Haussmann 2006). More particularly, oxidative stress and telomere erosion during growth have been identified as two factors potentially underlying the negative impact of fast growth on other life history traits, with a special focus on longevity (Mangel and Munch 2005; Monaghan and Haussmann 2006). Compensatory or fast growth was found to increase oxidative stress, either measured as a decreased resistance to an oxidative challenge (Alonso-Alvarez et al. 2007b; Kim et al. 2011) or as an increase in oxidative damage levels (Tarry-Adkins et al. 2008; Geiger et al. 2012; Stier et al. 2014a). Similarly, compensatory or accelerated growth rate has been associated with short telomeres or increased telomere erosion, both in laboratory (Tarry-Adkins et al. 2008) and wild conditions (Geiger et al. 2012; but see Foote et al. 2011b). We showed in a previous study that starting life with a handicap (i.e. being born late in the breeding season) was associated with elevated levels of oxidative stress, shorter telomeres and low survival probability in young king penguin chicks (Stier et al. 2014b). Still, to test the universality of such a relationship between early life adverse conditions and impaired self-maintenance, we need to investigate the potential trade-off between growth rate and self-maintenance parameters under various natural conditions expected to affect growth trajectories [e.g. altitude (Stier et al. 2014a)]. In many altricial birds, hatching occurs asynchronously, which creates a sibling competitive hierarchy within the brood due to large size differences between first- and lasthatched nestlings (Magrath 1990). Hatching asynchrony is thought to be adaptive for the parents, among other things by facilitating brood reduction under poor environmental conditions (Magrath 1990; but see Podlas and Richner 2013). First-hatched nestlings enjoy substantial advantages compared to last-hatched nestlings. For instance, it has been shown that first-hatched nestlings might benefit from an increased food intake despite a reduced begging rate (Cotton et al. 1999). Therefore, asynchronous hatching offers a good opportunity to test for a trade-off between growth and self-maintenance processes within a natural context, since resource availability and allocation are likely to differ between first-hatched and last-hatched nestlings (Nilsson
13
and Svensson 1996; Nilsson and Gårdmark 2001). It was shown in many cases that last-hatched nestlings might suffer from reduced growth rate or lower mass/size at fledging (e.g. Rubolini et al. 2006; Kilgas et al. 2010). However, altered growth rate or reduced mass/size at fledging is not always apparent (e.g. Clotfelter et al. 2000), might be sex specific and/or year dependent (Tilgar and Mänd 2006), or might also vary according to the morphological trait measured (Nilsson and Svensson 1996; Nilsson and Gårdmark 2001; Mainwaring et al. 2010; Kilgas et al. 2010). Moreover, last-hatched nestlings are able in some cases to catch up with first-hatched ones during the growth period (Clotfelter et al. 2000; Wegrzyn 2012). This raises the question of the physiological and ageing costs that could be paid by last-hatched nestlings to overcome their starting competitive handicap. Mothers are known to modulate offspring phenotype through adjusting testosterone (T) content within egg yolk, with for instance last-laid eggs presenting higher T levels in great tits (Podlas et al. 2013). High T levels might promote competitiveness (Magrath 1990) and fast growth (Muller et al. 2007), but have also been shown to elevate oxidative stress levels (Alonso-Alvarez et al. 2007a), or even to impair DNA repair mechanisms (Treidel et al. 2013). In this context, we might hypothesize that higher T levels may modulate last-hatched nestling ability to compete and catch up with their older sibling, but probably at the expense of physiological costs associated with high levels of this hormone (Alonso-Alvarez et al. 2007a). To date, few studies have examined the impact of hatching asynchrony on oxidative stress and telomere erosion, but it was shown for instance that last-hatched nestlings might exhibit lower plasma antioxidant capacity (Rubolini et al. 2006 but see Kilgas et al. 2010; Saino et al. 2011). Information about the impact of natural hatching asynchrony on oxidative damage and telomere erosion is thus limited. However, recent experimental studies have demonstrated the importance of within-brood hierarchy in determining telomere loss (Nettle et al. 2013, 2015). In this study, we investigated naturally asynchronous hatching in a passerine bird, the great tit (Parus major), and more specifically its impact on the following variables: (1) growth rates and body mass/size at fledging; (2) physiological parameters (i.e. oxidative stress, plasma T levels and telomere dynamics); and (3) the relationships between growth rate and physiological parameters. Given the potential advantage of growing fast to compensate for their starting handicap, we predicted that last-hatched nestlings could exhibit higher T levels and faster growth than first-hatched ones, but hypothetically at the expense of high oxidative stress levels and/or faster telomere erosion.
Oecologia
Materials and methods Fieldwork and bird sampling We monitored breeding activity and nestling growth of great tits in artificial nest boxes, within the Forêt de la Wantzenau (Strasbourg, France) during the 2012 season (April–May). During the early season, nests were checked regularly before hatching to determine the approximate hatching date (i.e. 13 days after clutch completion) and clutch size. In 2012, twenty-two nest boxes were successfully occupied by great tits pairs (mean clutch size 9.9 ± 0.4 eggs). Nests were visited once a day around the approximate hatching date to estimate hatching order. We selected four nestlings per nest (i.e. 88 in total): two first-hatched and two last-hatched nestlings, based on body mass and body size at the time of nest visit [i.e. day 1 or 2; see Clotfelter et al. (2000) for the close association between hatching order and nestling body mass shortly after hatching). In most cases, it was impossible to distinguish the real hatching order between the two first-hatched, or between the two last-hatched nestlings. Nestlings were individually marked with a waterproof marker on their rump. Overall mortality was very low (7.7 %), but five last-hatched and two firsthatched nestlings died before the first blood sampling (days 2–5) and were subsequently replaced by the closest nestling in regard to hatching rank. We did not observe mortality after day 5. Last-hatched nestlings hatched on average 1.4 ± 0.1 days later than first-hatched ones (range 1–3 days). Body mass and body size (wing length, tarsus length, head size) were recorded every 2 days following hatching (hatching = day 1) using an electronic balance (0.1-g precision) and a digital caliper rule (0.1-mm precision). Nestlings were followed until they fledged (mean estimated fledging time ± SE = 17.55 ± 0.13 days after the first hatching). One small blood sample (≈50 µL) was taken from the brachial vein with a heparinised glass capillary at day 7 (mean ± SE 7.46 ± 0.08 days) and day 16 (16.35 ± 0.08 days), and we paid attention to sample firstand last-hatched nestlings at the same age relative to hatching. Blood was kept on ice until return to the laboratory. Plasma was separated from blood cells by centrifugation (10 min, 3000 g) and samples were subsequently stored at −80 °C until analysis. Growth analysis As an estimate of body size, we used the first principal axis (PC1) resulting from a principal component analysis (PCA) of wing, tarsus and head measurements, which explained 96.4 % of the total variance of body size. Therefore body size is expressed in arbitrary units (AU).
Body mass and body size (PC1) growth were fitted with A , the following logistic equation: Y (x) = [1+exp(−K×x−B)] 2 which was the best fitting model based on R . Y(x) represents the size or mass of a nestling at age x (in AU or g), A is the asymptotic mass or size (i.e. mass or size at fledging), K is the growth rate constant and B is a constant determining the initial mass/size. Growth fitting was performed with the nonlinear regression procedure in SPSS (SPSS 20.0 1989–2011; SPSS, USA) for each nestling. Using the parameters (A, K and B) obtained for each nestling, we were able to estimate mass and size at any given age. We used this opportunity to present growth data in Fig. 1. We calculated both the mass/size according to the real age of the nestling (Fig. 1a, b) and according to the time elapsed since the first hatching (i.e. taking into account hatching asynchrony; Fig. 1c, d), the latter being more straightforward in revealing the starting handicap of last-hatched nestlings (Wegrzyn 2012). Plasma oxidative stress markers The non-enzymatic antioxidant capacity and the concentration of reactive oxygen metabolites (ROMs) in the plasma were measured using the OXY-Adsorbent (5 µL of 1:100 diluted plasma) and d-ROM tests (5 µL of plasma; Diacron International, Italy), respectively, following the manufacturer’s instructions (optical density measurement at 510 nm using a Tecan Infinite 200 microplate reader). The OXY adsorbent test quantifies the ability of the plasma antioxidant compounds to buffer a massive oxidation through hypochlorous acid, while the d-ROM test measures mostly hydroperoxide as a marker of potential oxidative damage [but see Kilk et al. (2014) for recent criticisms about this assay]. Antioxidant capacity is expressed as mM HClO neutralised and ROMs as mg H2O2 equivalent/dL. Intraassay variation based on duplicate samples was low [coefficients of variation (CV) 4.25 ± 0.43 % for the OXY test and CV 5.37 ± 0.78 % for the d-ROMs test] as well as inter-assays variation based on a standard sample repeated over plates in duplicate (CV 6.02 % for OXY and 5.60 % for the d-ROM test). Telomere length Telomere length was determined with DNA extracted from blood cells (Macherey–Nagel Nucleospin Blood QuickPure extraction kit) using a quantitative real-time polymerase chain reaction (qPCR) amplification protocol previously used in birds (Criscuolo et al. 2009) and Bio-Rad CXF384 apparatus. Primer sequences for telomere amplification (Tel1b and Tel2b) were similar to those used by Criscuolo et al. (2009). As a single control gene (S), we used the
13
Oecologia
Fig. 1 Body mass (a, c) and body size (b, d) dynamics of firsthatched (white dots) and last-hatched (black diamonds) nestlings, according to the real age of individuals (a, b) or to the time elapsed since the first hatching (c, d). Body mass and body size values are those predicted according to the logistic growth curve of each individual (see “Materials and methods”), for a period of 19 days after the first hatching. Consequently, body mass/size of last-hatched nest-
lings are not presented at a real age of 19 days (a, b), and at 1 day after the first hatching (c, d). Statistics for logistic growth curve parameters (real age; a, b) are presented in Table 1. Statistics regarding the starting handicap (c, d) are presented in the text, with significant differences between first- and last-hatched nestling at a given time represented as an asterisk. Means are expressed ±SE, but error bars are not visible because of the low dispersion around the mean
chicken zinc finger protein, with primer sequences defined by Primer 3 software as—ZENK1 (5′-TACATGTGCCATGGTTTTGC-3′), ZENK2 (5′-AAGTGCTGCTCCCAAAGAAG-3′). This gene was checked to be a non-variable copy number (non-VCN) gene within our population (Smith et al. 2011) using a standard electrophoresis on a 3 % agarose gel run in standard Tris–borate-ethylenediaminetetraacetic acid buffer (50 V for 15 min and 100 V thereafter for 30 min). The qPCR amplicon size of eight different and randomly chosen individuals was found to be 167 bp (one single band). Primer concentrations in the final mix were 100 nM for telomere length determination and 500 nM for the control gene. Telomere and control gene PCR conditions were 2 min at 95 °C followed by 40 cycles of 15 s at 95 °C, 30 s at 56 °C, 30 s at 72 °C and 60 s at 95 °C. We used 2.5 ng DNA per reaction and the Power SYBR Green PCR Master Mix fluorescent probe (Applied Biosystems, USA). Mean amplification efficiencies of the qPCR runs for telomere and control gene were: 99.86 ± 0.22 and 100.16 ± 0.31 %, respectively. Intraplate mean CVs for raw values were 1.74 ± 0.08 % for the telomere assay and 0.81 ± 0.03 % for the control gene assay. Samples were randomly distributed and run on five
different plates and inter-plate CVs based on five repeated samples were 2.15 ± 0.11 % for the telomere assay and 1.04 ± 0.12 % for non-VCN assay (raw values again). CVs for the final telomere/control gene ratio (T/S) were 13.38 ± 0.82 % (intra-plate) and 12.36 ± 1.33 % (interplate). A negative control was used on each plate to check for non-specific amplification and a melting curve ended each qPCR run to control for the absence of primer-dimer amplification (presence of one single melt peak). Final calculation of telomere length (T/S ratio) was done following Pfaffl (2001) using the telomere and control gene efficiencies of each plate. Sex determination was also done on DNA extracted from blood cells, following an adapted method from Griffiths et al. (1998).
13
T assays Circulating T levels were determined at day 7 and day 16. Plasma T levels were determined by immunoassay according to guidelines provided by the manufacturer (Testosterone EIA Kit, ADI-901-065; Enzo Life Sciences, NY). Intra-assay variation based on five duplicates per plate was low (7.11 ± 0.97 %), as well as inter-assay variation based
Oecologia Table 1 Summary of the most parsimonious linear mixed models explaining the variability in logistic growth curve parameters: asymptotic body mass and body size, as well as growth rate constant for body mass and body size Estimate
SE
df
F
p-value
Asymptotic mass (A) Random effect Nest 0.68 Fixed effects and covariates
Statistics
0.25
16.61
0.55
1,25.9
915.01
