Metabolism correlates with variation in post-natal growth rate among songbirds at three latitudes
Summary
- Variation in post-natal growth rates is substantial among organisms and especially strong among latitudes because tropical and south temperate species typically have slower growth than north temperate relatives. Metabolic rate is thought to be a critical mechanism underlying growth rates after accounting for allometric effects of body mass. However, comparative tests on a large spatial scale are lacking, and the importance of metabolism for growth rates remains unclear both within and particularly across latitudes.
- Songbirds exhibit strong interspecific variation in growth rates across geographic space, although within latitudes an association between metabolic rate and growth rate has not always been observed. Moreover, the hypothesis that differences in growth rates across latitudes reflect underlying differences in metabolism is untested. Here, we investigate these possibilities across north temperate, south temperate and tropical study sites.
- Phylogenetic analyses showed that, for a given body mass, metabolic rates of north temperate nestlings were higher than tropical and south temperate species. Metabolic rates controlled for body mass correlated with post-natal growth rates both within and among latitudes. Offspring body mass explained substantial residual variation in growth rates as expected under classic allometric theory.
- Our results suggest that variation in metabolic rates has an important influence on broad patterns of avian growth rates at a global scale. We suggest further studies that address the ecological and physiological costs and consequences of variation in metabolism and growth rates.
Introduction
Growth rates of offspring vary extensively among taxa, especially when comparing slow-growing tropical versus fast-growing north temperate organisms (Case1978; Roff 1992; Arendt 1997). Fast growth can decrease the risk of time-dependent mortality (Skutch 1949; Case 1978; Ricklefs 1993; Benrey & Denno 1997; Remeš & Martin 2002), allow earlier access to food and other resources (Conover & Present 1990) and increase opportunities for repeated reproductive events (Sibly & Calow 1986). However, fast growth can also negatively affect a broad array of traits related to organismal quality (Arendt 2003; Alonso-Alvarez et al. 2007; Arriero, Majewska & Martin 2013) and can reduce longevity (Metcalfe & Monaghan 2003; Hulbert et al. 2007; Lee, Monaghan & Metcalfe 2013; but see Martin et al. 2015). Despite these important ramifications for organismal quality and fitness (Starck & Ricklefs 1998), our understanding of the physiological mechanisms underlying broad interspecific and latitudinal variation in growth remains limited (Dmitriew 2011; Flatt & Heyland 2011).
Body mass, body temperature and metabolic rate are thought to be responsible for extensive interspecific variation in rates of cellular proliferation (von Bertalanffy 1957). In particular, ontogenetic models predict that for a given temperature and body mass, the rate of somatic growth increases with metabolic rate (West, Brown & Enquist 1997, 2001). While this relationship has long been accepted, recent results have questioned the generality of metabolic rate as a pacemaker for growth within (Burton et al. 2011) and among species (Glazier 2015). This uncertainty is highlighted in songbirds (Fig. 1) where the influence of metabolism and body mass on growth rates remains unclear.
Previous interspecific tests of the relationship between growth and metabolism in birds within latitudes have found no correlation (Konarzewski et al. 2000), a weak positive correlation (Williams et al. 2007) and a positive correlation (Drent & Klaassen 1989; Klaassen & Drent 1991). However, the latter two studies did not directly measure metabolic rate but projected it from allometric equations. Moreover, reanalysis of their data yielded conclusions opposite to those of the original tests (Konarzewsky 1995), increasing the uncertainty surrounding the metabolism-growth relationship. The ability of metabolic rate to explain differences in growth rates of songbirds is further questioned by the absence of a correlation between offspring body mass and growth rate (Martin et al. 2011), opposite to expectations under scaling theory (West, Brown & Enquist 2001). Overall, this inconclusive evidence begs for studies directly testing if metabolism explains interspecific variation in avian growth.
While the role of metabolism in explaining growth rate variation within latitudes has been unclear, the role of metabolism in determining latitudinal differences in growth rates is untested. Metabolic rates of adults are lower in tropical songbirds compared to temperate relatives (Wikelski et al. 2003; Wiersma et al. 2007). If nestlings exhibit the same latitudinal pattern, then this may explain geographic variation in growth rates. Interestingly, embryonic metabolism did not differ among latitudes (Martin, Ton & Niklison 2013), raising questions about latitudinal patterns of metabolism in offspring. Of course, metabolic rates can change across life stages (Glazier 2005), which emphasizes the need for direct measurements of post-natal metabolism. Here, we test the hypothesis that metabolic rates underlie interspecific growth rates of songbirds both within and among latitudes.
Materials and Methods
Study areas and species
Data were collected in a high elevation (2350 m) mixed forest in north temperate Arizona, USA (34° 34′ N, 111°14 W); in a tropical mid-elevation forest (1450–1950 m) in Malaysia (5° 59′N, 116° 34′ E); and in a south temperate dwarf shrubland located at sea level in South Africa (33° 41′ S, 18° 26′ E) (see Martin et al. 2015). Our sample included 59 species from 52 genera and 25 families within the order Passeriformes spanning substantial variation in body mass and rates of post-natal growth (see data accessibility).
Metabolic measurements
We measured metabolic rates for 436 nestlings of 59 passerine species (see Table S1, supporting information). We aimed for 10 samples per species, but sample size varied between 1 and 13 with a mean ± SE of 7·4 ± 0·44 individuals measured per species. Only one nestling per clutch was tested to ensure independence among samples. Measurements were taken for 22 species in Arizona from May through July between 2011 and 2014; 23 species in Malaysia from February through April in 2012–2014; and 14 species in South Africa between August and November 2014. To account for possible effects of circadian rhythms on metabolic rates, measurements were taken between 11 a.m. and 5 p.m.
We recorded oxygen consumption [VO2 (mL h−1)] at 39·0 °C in an open-flow respirometry system using a Foxbox field gas analyzer (Sable System, Las Vegas, NV, USA). The temperature of 39·0 °C was selected because it appears to match with the thermoneutral zone (temperature of lower oxygen consumption) for our species at this age. We reached this conclusion based on extensive metabolic measurements we performed along five steps of temperatures in a Q10 interval from 31 to 41 °C (unpublished data). Also 39·0 °C best approximated the average value of body temperature recorded in the field for the species studied here (38·95 ± 0·0586 °C). We measured a nestling's internal body temperature as soon as it was removed from a nest and before the metabolic measurements using a HH506RA Multilogger Thermometer (Omega, Stamford, CT, USA). After inserting a 0·8-mm diameter thermocouple in the nestling's cloaca, we monitored core body temperature for at least 10 seconds and recorded the highest temperature value for each individual.
Metabolic rates were measured at pin break, a standardized developmental stage when primary feathers break their sheaths and thermoregulatory capacities are comparable among species (Fig. 1, Sogge et al. 1991; Pereyra & Morton 2001; Cheng & Martin 2012). Therefore, recording VO2 at pin break allowed us to control for interspecific variation in thermoregulation and its effect on metabolism during growth.
Each nestling was put in a 3·2-L stainless steel airtight metabolism chamber where it could sit on a cup-shaped piece of iron mesh that prevented extensive movements of the nestling while still allowing normal airflow. The chamber sat in a large, dark, insulated cabinet with a Peltier device (Pelt-4; Sable Systems) maintaining temperature at 39 ± 0·1 °C. The chamber was connected to an open-flow system and flushed with 200–300 mL min-1 flow of atmospheric air scrubbed of CO2 and water vapour. These flow rates guaranteed a stable proportion of oxygen available to birds within the sampled body range (Table S1). Air was filtered through scrubbers with soda lime, magnesium perchlorate and drierite to remove water and CO2 (Lighton 2008). After allowing the nestling to become accustomed to the chamber for 30 min, an initial baseline was recorded for about 10 min. Subsequently, VO2 was measured continuously every 0·5 s until a plateau was reached (stable oxygen readings for at least 10 min). Lastly, a final baseline was recorded for 10 more minutes. The two baselines were later used to correct for potential drift in ambient O2 during measurements and thus maximize the accuracy of our estimates. VO2 was calculated as the most stable 5 min of oxygen consumption within the plateau. The total amount of time needed to complete a measurement ranged from 70 to 110 min depending on nestling size. After completion of VO2 measurements, nestlings were fed with commercial food for altricial birds and returned to their nest unharmed. VO2 (mL h−1) was calculated in expedata (ver. 1·3·2) software from Sable Systems using the formula VO2 = FRi(FiO2–FeO2)/(1−FeO2). Where FRi is the incurrent mass flow rate scrubbed from water vapour and CO2, FiO2 is the incurrent fractional concentration of oxygen, and Feo2 is the excurrent fractional concentration of oxygen (Lighton 2008).
Growth rate
Nestling growth rates were estimated for 53 species. Small and incomplete sample sizes prevented us from obtaining robust data on growth rates for two species in Arizona, two in Malaysia and two in South Africa. Our growth rate estimates are based on extensive sampling lasting 15 years (1999–2013) in Arizona, 6 years (2009–2014) in Malaysia and 5 years (2000–2004) in South Africa (see Martin et al. 2011, 2015). To calculate growth rates, we measured nestling body mass at the same time (±1 h) every day for the first 3 days after hatch and then every other day until fledge. We used GemPro 250 portable electronic scales (My Weigh, Phoenix, Arizona, USA) with an accuracy of ±0·001 g. We calculated the growth rate constant (k) for each species using logistic regression; a standardized and widely used method that allows interspecific comparisons independent of absolute development time and body mass (Ricklefs 1968; Remeš & Martin 2002). Growth rate data are available online at the following link http://dx.doi.org/10.5061/dryad.ks62j.
Statistical analysis
We produced our estimates for metabolic rate (VO2 at 39 °C) and body mass using a linear mixed model to capture within and among species variation for each parameter of interest. To explain the scaling relationship between metabolic rate (VO2 at 39 °C) and body mass, we ran a generalized linear model with body mass, site and the interaction between body mass and site as fixed factors and metabolic rate as the response variable. Both nestling metabolic rate and body mass were log10-transformed to meet assumptions of normality. To evaluate whether these relationships varied across sites, we conducted post hoc tests for differences in slopes among sites. To control for possible phylogenetic effects (Felsenstein 1985), we conducted a phylogenetic generalized least squares (PGLS) analysis using the Caper package in r (Orme 2013). To create the phylogenetic tree used to constrain the analysis, we sampled 1000 trees containing our study species from www.birdtree.org (Jetz et al. 2012) using the Hackett backbone (Hackett et al. 2008). We then produced a majority-rules consensus tree using the program Mesquite (Appendix S1 in supporting information, Maddison & Maddison 2001). We used the scaling parameter Pagel's lambda (λ) as a measure of phylogenetic signal, which can range from 0 to 1 (Pagel 1999). A value closer to 0 indicates lower similarity in traits among species than strictly expected by their phylogenetic relationships based on a Brownian motion model of evolution (Pagel 1999). The λ values produced by our PGLS analysis were based on maximum likelihood optimization.
To assess how metabolic rates may influence growth rates, we ran an additional PGLS model, with growth rate as the response factor and metabolic rate, body mass, site, site × body mass and site × metabolic rate interaction terms as fixed effects. We used backward-stepwise selection criteria to pick the best model and dropped nonsignificant variables from our analysis (Table S2). Full model outputs including all variables are listed in Tables S3 and S4. All statistical analyses were done in r v.3·0·3 (R Core Team 2014).
Results
Metabolic rate increased with nestling body mass for 59 species of songbirds at all three latitudes (Fig. 2). The slope of this scaling relationship was 0·77 ± 0·022 (β ± SE), very close to the ¾ exponent expected under allometric theory (Kleiber 1932). For a given body mass, metabolic rates were significantly higher for north temperate species compared to south temperate and tropical species (Table 1). Also, metabolic rates of nestlings in the south temperate appeared slightly higher than in the tropics, but the differences were not significant (Table 1).
| Variable | β (SE) | F | d.f. | P-value |
|---|---|---|---|---|
| Body mass | 0·771 (0·022) | 1240·4 | 1 | <0·001 |
| Site | – | 9·347 | 2 | <0·001 |
| Malaysia vs Arizona | −0·065 (0·022) | – | – | <0·001 |
| South Africa vs Arizona | −0·046 (0·015) | – | – | 0·003 |
| South Africa vs Malaysia | 0·018 (0·013) | – | – | 0·164 |
| Error | 55 |
- Offspring metabolic rate as the dependent variable, (R2 = 0·94), (λ = 0·710).
The residual variation in metabolic rate unexplained by body mass strongly correlated with nestling growth rates (Fig. 3b; Table 2). Moreover, metabolism explained latitudinal differences in growth rate, as reflected by the fact that latitudinal differences in growth rate disappeared (site: P = 0·814) when metabolism was included in the model (Table S2).
| Variable | β (SE) | F | d.f. | P-value |
|---|---|---|---|---|
| Residual metabolic rate | 0·780 (0·152) | 24·95 | 1 | <0·001 |
| Body mass | −0·720 (0·125) | 33·06 | 1 | <0·001 |
| Error | 50 |
- Offspring growth rate (k) as the dependent variable (R2 = 0·44), (λ = 0·573).
Allometric rules predict that growth rates should decrease with increasing body mass. After accounting for metabolism effects, nestling body mass explained 33% of the residual variation in growth rates (Fig. 3a; Table 2). None of the interactions terms tested in our models were significant indicating similar relationships among sites and were therefore dropped from the analysis and reported only in the supporting information (Tables S1 and S2). The λ values produced by our PGLS analyses were all greater than 0 but <1 (Tables 1 and 2) as seen in other comparative studies among birds and passerines in particular (Freckleton, Harvey & Pagel 2002).
Discussion
Predictions that body mass and metabolism should underlie growth rate variation have not been well supported in interspecific avian studies within latitudes (Dunn 1980; Klaassen & Drent 1991; Konarzewsky 1995; Konarzewski et al. 2000; Williams et al. 2007) and have never been tested across latitudes. We directly measured growth, body mass and metabolic rates from three bird communities across the world and documented strong correlations among these traits. Moreover, we demonstrated that a geographic pattern in metabolism coincided with a known latitudinal gradient in growth rates (Martin et al. 2011). North temperate nestlings had higher metabolic rates and faster growth than south temperate and tropical species. Previous studies have argued that metabolic variation in adult songbirds among latitudes reflected differences in ‘pace of life’ (Wikelski et al. 2003; Wiersma et al. 2007). This argument revolves on theoretical expectations (Ricklefs & Wikelski 2002) and direct tests (Williams et al. 2010) suggesting that life-history and physiological traits are expressed in coordination along a slow-fast continuum. Our study supported these previous findings by offering correlative evidence that a physiological trait (metabolic rate) underlies broad interspecific variation in a major life-history trait (rate of growth).
The latitudinal metabolic differences we found between tropical and north temperate sites were not detected for embryonic metabolic rates for the same species (Martin, Ton & Niklison 2013). However, embryos are ectothermic while nestlings are first poikilothermic and finally endothermic as they grow. These radically different thermoregulatory conditions are known to be associated with equally different metabolic regimes (Peterson, Nagy & Diamond 1990). Therefore, dissimilarities between studies may reflect ontogenetic changes in metabolism across life stages of the same organism (White & Kearney 2013).
This stage-dependent nature of metabolic processes may also explain differences in the scaling relationship between metabolic rate and body mass among studies (Glazier 2005). Our slope for nestlings closely approximated the traditional ¾ exponent expected under classic scaling theories (Kleiber 1932; West, Brown & Enquist 1997; Banavar et al. 2014). However, this result did not match predictions of shallow slopes close to ⅔ for organisms of small body size (Kolokotrones et al. 2010). Our scaling exponent was higher than for embryos (Martin, Ton & Niklison 2013) and adults (Wiersma et al. 2007) of songbirds, but lower than for growing young in other avian studies (Klaassen & Drent 1991). Similar ontogenetic changes in allometric scaling have been previously documented (Czarnołeski et al. 2008; Peng et al. 2010; Glazier, Hirst & Atkinson 2015) and contribute to the debate over the existence of a universal slope for the metabolism–body mass correlation (West, Brown & Enquist 1997; White, Cassey & Blackburn 2007).
Previous studies have found weak effects or failed to detect the predicted negative relationship between body mass and growth rate (Remeš & Martin 2002; Martin et al. 2011). Here, we found that, when taking into account interspecific differences in metabolic rate, body mass explained part of growth rate variation. This result unveils the masking effect that metabolic rate can have on allometric relationships in growth rate and further emphasizes the importance of interspecific differences in metabolism for broad geographic patterns of growth. Metabolic variation unexplained by body mass may be favoured by environmental mortality during offspring growth (Rose 1991), but this possibility needs testing.
Understanding the selective pressures that influence metabolic variation would be fruitful because of the potential role of metabolism in life-history evolution and ‘rate of living’ theory (Pearl 1928). For example, high metabolic activity favouring fast growth may induce high production of reactive oxygen species (ROS) (Harman 1955). These metabolic by-products have been hypothesized to carry physiological costs paid off later in life via oxidative damage that increase probability of adult mortality (Harman 2001; Hulbert et al. 2007). Yet, the actual connection between high metabolism, ROS production and costs for adult survival is still debated (Barja 2007). Some evidence also suggests that the intrinsic costs of growth rate variation may be of less importance than extrinsic sources of mortality in passerines (Martin et al. 2015). Future studies could elucidate if the variation in metabolic rate underlying post-natal growth is also a possible mediator of trade-offs between offspring and adult mortality.
Acknowledgements
We are thankful to our laboratory and two anonymous reviewers for helpful comments on the manuscript. We thank many field assistants that helped with this project, in particular M. L. Markowsky, J. Olson, K. Corder, J.A. Brandauer, N. Gobbo, O. M. Sutto and V.J. Kerrigan. Sabah Parks and the Sabah Biodiversity Council in Malaysia, and Western Cape Nature Conservation Board in South Africa helped to make the present study possible. The U. S. Geological Survey Climate Change Research Program, the National Science Foundation (DEB-0841764, DEB-1241041, IOS-1349178) and the Drollinger-Dial Foundation supported this work. We performed this research under the auspices of University of Montana IACUC protocol #059-10TMMCWRU. Any use of trade, firm or product names is for descriptive purposes only and does not imply endorsement by the U.S. Government. The authors of this paper declare no conflict of interest.
Data accessibility
Metabolic data used in this manuscript can be accessed in the supplementary information (Appendix S1). Growth rate data are available at the following link http://dx.doi.org/10.5061/dryad.ks62j (Martin et al. 2015)
