Journal of Animal Ecology 2012, 81, 583–593 doi: 10.1111/j.1365-2656.2011.01942.

Quantitative analysis of compensatory and catch-up

growth in diverse taxa
Katie L. Hector1* and Shinichi Nakagawa1,2
1
Department of Zoology, University of Otago, 340 Great King St, Dunedin, 9016, New Zealand; and 2National Research
Centre for Growth and Development (NRCGD), University of Otago, 340 Great King St, Dunedin, 9016, New Zealand

Summary
1. ‘Compensatory growth’ and ‘catch-up growth’ are often used interchangeably to describe the
faster than optimal growth that occurs following a period of dietary restriction in the development
of many animals. Concerns about the statistical analysis of these studies have drawn attention to
the risk of false detection in reports of compensatory and catch-up growth.
2. This study aims to quantify the degree to which these compensatory responses occur across the
animal kingdom. In addition, this study distinguishes the two terms, ‘compensatory growth’ and
‘catch-up growth’, to clarify the fitness consequences of rapid growth. Compensatory growth refers
to a faster than usual growth rate, while catch-up growth implies attainment of control size.
3. Eight meta-analyses and meta-regression analyses were conducted on data extracted from 88
papers, including 11 taxonomic classes. The results confirmed that both growth tactics (i.e. com-
pensatory and catch-up growth) occur across a wide range of taxa and result in decreased direct fit-
ness components.
4. Importantly, the meta-analytic methods used made it possible to identify the specific experi-
mental techniques that most successfully promoted rapid growth after restriction and key differ-
ences in the responses of the four major groups (mammals, birds, fish and arthropods) to dietary
restriction. Endotherms are more likely to show a compensatory growth response because of their
determinate growth; in contrast, the indeterminate and saltatory growth tactics of fish and arthro-
pods reduce the pressure to rapidly achieve a large size.
5. Among the first meta-analyses to be conducted in this field, this study provides valuable support
for the premises of compensatory and catch-up growth and also discusses weaknesses in experi-
mental design, and possible solutions, in compensatory growth research. For example, we recom-
mend conducting the experiment within the most linear phase of an animal’s growth to avoid
analytical complications arising from size-dependent growth, and our results indicate that dietary
dilution more closely resembles quantitative restriction than clutch size and intermittent feeding
restriction methods when normal quantitative restriction is not possible.
Key-words: caloric restriction, developmental plasticity, growth trajectory, meta-analysis,
phenotypic plasticity, quantitative review, systematic review

Stamps 2001). However, growth rates closer to maximal are

Introduction
commonly observed when an organism has previously expe-
Under ideal circumstances, animals grow at an optimal rate, rienced a period of limited growth during development.
limited more by quality control of differentiating tissues than These responses are commonly known as compensatory
by a lack of resources (Metcalfe & Monaghan 2003). Optimal growth or catch-up growth.
growth reduces the negative physiological costs of the accu- Compensatory growth has long been of interest to scien-
mulated cellular damage from rapid growth while maximiz- tists (Jackson 1937), primarily because it begs the question: if
ing fitness potential. Faster than optimal growth (maximal some animals are willing to grow at a maximal rate, why
growth) can decrease cell functioning efficiency, immune don’t they all? The answer seems to lie in the altered cost–
function and resistance to physiological stressors (Mangel & benefit equation of an animal with a poor start in life. An ani-
mal’s adult size is often a major factor in fitness; size is known
*Correspondence author. E-mail: [email protected] to affect mate selection, fecundity and offspring survival

 2012 The Authors. Journal of Animal Ecology  2012 British Ecological Society
584 K. L. Hector & S. Nakagawa

(Blanckenhorn 2005). Therefore, animals with slow growth choice, social dominance and fecundity (Blanckenhorn
during development are at a distinct disadvantage if they 2005). Catching up to the size of normal conspecifics can be
reach a small adult size. If the environmental cause of achieved by extending the developmental period while
restricted growth passes, the opportunity may present itself continuing at an optimal growth rate (Arendt 1997).While
for an animal to increase its growth rate to avoid becoming a there may be some cost to lifetime reproductive success owing
small adult. In this case, it is more beneficial for the growth- to the extra time spent in a non-reproductive phase (Oli,
restricted animal to risk the negative consequences of maxi- Hepp & Kennamer 2002), this strategy can often be favour-
mal growth than bear the fitness costs of small adult size. In able in variable environments and comes at minimal physio-
many circumstances, the cost of maximal growth, in the form logical cost (Wilbur & Rudolf 2006). Thus, it is often unclear
of accumulated cellular damage, is not paid until after the what the authors really mean in reporting results as ‘compen-
reproductive phase. For example, in humans, rapid growth satory’ or ‘catch-up’ growth, and this lack of clarity in termi-
in childhood after a ‘small-for-gestational-age’ birth weight nology means that the associated fitness consequences are
is associated with increased risk of late-onset adult diseases easily confused. For the sake of clarity in this study, ‘catch-
such as heart disease, diabetes and obesity (Cottrell & up growth’ will refer to the attainment of a non-significant
Ozanne 2008). Similar findings have been reported, for exam- difference in size between the control and previously
ple, from the intensive studies of metabolic syndrome in rats restricted animals. This can be achieved at a normal growth
(Bol et al. 2009; Porrello et al. 2009). Compensatory growth rate. In contrast, ‘compensatory growth’ refers to a signifi-
is also known to decrease the maximum life span of individu- cantly steeper growth rate of the previously restricted
als in a number of species (Metcalfe & Monaghan 2003). animals, which may or may not result in catching up to the
Thus, compensatory growth allows animals to reach a larger same size of control-fed animals (Fig. S1, Supporting
size, which increases reproductive fitness before the conse- information).
quences of rapid growth negatively affect them. Compensatory growth is hypothesized to have evolved
The theory of compensatory growth has been reviewed by because the fitness consequences are delayed until after
a number of researchers, with a variety of perspectives. Met- reproduction (Metcalfe & Monaghan 2001). Therefore, selec-
calfe & Monaghan (2001) and Dmitriew (2011) provide a tion favours those individuals capable of compensatory
broad outline of the forms and consequences of growth com- growth because of their increased size-dependent fitness com-
pensation across many taxa. Compensatory growth has also pared to individuals with limited developmental plasticity in
been reviewed with specific taxa in mind (domestic fowl: Nir growth (Yearsley, Kyriazakis & Gordon 2004). This selection
et al. 1996; fish: Ali, Nicieza & Wootton 2003) and with life- is dependent on the delay before health consequences occur
history model simulations to support theoretical assumptions being sufficiently long to allow unimpaired reproductive
(Mangel & Munch 2005). Collectively, these reviews voice effort. The experimental design, in particular the duration of
strong support for compensatory growth as a measurable, the relevant periods of restriction and realimentation (when
repeatable and taxonomically diverse real-world phenome- animals are returned to a normal diet), is often a seemingly
non. However, the review by Nicieza & Alvarez (2009) arbitrary decision. While there are a number of reviews on
addresses concerns about some of the compensatory growth the topic (Wilson & Osbourn 1960; Metcalfe & Monaghan
literature by criticizing the statistical methods used to analyse 2001; Ali, Nicieza & Wootton 2003, Monteiro & Victora
this type of data. They claim that some analyses may not take 2005), there has yet to be a comprehensive meta-analysis of
into account the size dependence of growth rates, meaning compensatory growth across diverse taxa.
that as an animal gets larger, its growth slows. The relation- The main aim of this paper was to use meta-analytical
ship between resource allocation to maintenance and to techniques to explore a number of hypotheses about compen-
growth changes as an animal grows larger, resulting in a size- satory growth. We tested whether: (i) nutritional restriction
dependent growth curve common to most animals when slows the growth rate and affects the size of a variety of taxa
comparing the overall shape of their early growth (West, compared to well-fed conspecifics; (ii) restricted animals are
Brown & Enquist 2001). able to reach the same final size as controls (‘catch-up’) after
The terminology used in this area of research could be con- a subsequent return to normal feeding amounts; (iii)
sidered another impediment to the accurate reporting of restricted animals show faster than normal growth (‘compen-
compensatory growth (Jobling 2010). In many cases, com- satory growth’) after nutritional restriction; and (iv) early
pensatory growth and catch-up growth are used as inter- restriction has an effect on later fitness traits. By analysing
changeable terms, both meaning that the growth rate of a each of these hypotheses with a number of moderators
previously restricted group is significantly higher than a con- included, such as the degree to which an animal was restricted
trol. This is the explanation that has the most relevance to a and the method used for restriction, we aimed to gain a better
trade-off for fitness, owing to the cellular damage incurred understanding of the variation in the literature and compen-
with maximal growth. However, in other studies, the term satory growth as a whole.
catch-up growth is defined more by the actual ‘catching up’ Additionally, a subsequent analysis was performed with
of adult size rather than the growth. Reaching the same final the aim of accounting for the different growth tactics and
size as the controls is important in many species where size- metabolism between the various taxonomic classes. We
dependent fitness traits occur, such as predation, mate selected four groups with sufficient data points to explore

 2012 The Authors. Journal of Animal Ecology  2012 British Ecological Society, Journal of Animal Ecology, 81, 583–593
 Compensatory and catch-up growth 585

these possibilities: mammals, birds, fish and arthropods. (a) (b)

Birds and mammals both have determinate growth tactics so
are most likely to benefit from compensatory growth because
growth ceases at maturation (Metcalfe & Monaghan 2003).
The higher metabolic rate of birds may affect their ability to

Size
grow faster than normal, as a result of the already much
Con Res
higher demands on energy resources of birds compared with
mammals (Speakman 2005). Fish have indeterminate
growth, so may be under only seasonal pressure to rapidly
compensate from restricted growth (Kozlowski 1996).
Arthropods show saltatory growth, which proceeds in bursts
of growth following ecdysis, so rapid growth may not be nec- Time
essary as time between instars can be prolonged (Dmitriew &
Rowe 2007). The higher starvation resistance and lower fat Fig. 1. Hypothetical growth curves of control (Con) and diet-
restricted (Res) treatment groups representing the comparisons used
stores of fish and arthropods as ectotherms may also affect
in the slopes analyses. The shaded area is the restriction period (per-
their energy storage responses compared with endotherms iod A), and the non-shaded area indicates the period post-restriction
(Wang, Hung & Randall 2006). We aimed to identify whether (realimentation, period B). The comparisons were either control per-
these expectations, based on innate growth tactics and meta- iod A (dashed line) against restricted period B (bold line; A vs. B) or
bolic differences, were relevant and detectable signals within control period B (dotted line) against restricted period B (bold line; B
vs. B).
our analyses.

did not include fitness-related traits. This criterion also reduced the
Materials and methods representation of agricultural studies (c. 300 agricultural studies with
no fitness estimate), which would have overwhelmed non-farmed ani-
DATA COLLECTION
mals in the analysis. We note that agricultural papers rarely report fit-
We collected data from papers published in peer-reviewed journals. ness-related trait data, more commonly focussing on meat quality
Primarily, papers were sourced from searches of ISI Web of Science, and characteristics instead.
the most recent of which was conducted on 5th August, 2010. For In summary, we collected data from 88 papers (Table S2, Support-
inclusion in this analysis, a paper must: (i) be an empirical study (i.e. ing information), from which we extracted 226 comparisons of the
no computational simulations), (ii) not include animals that have size of restricted animals with their reported control groups. These
been genetically modified or have a known disorder, (iii) describe an data included 58 species, spanning eight classes within the phyla
experiment with a treatment and a control group, (iv) report body Chordata, Arthropoda and Mollusca. Because some papers reported
mass before and after restriction and after realimentation along with multiple fitness component measures, we collected a total of 207 fit-
estimates of uncertainty (e.g. standard deviation and ⁄ or standard ness component comparisons between restricted and control animals,
error), (v) have quantifiable degrees of diet restriction in a controlled which we categorized into the six broad classes of fitness traits (see
environment (thus excluding correlational data from wild popula- Supporting information for analysis of ‘indirect’ fitness traits). For
tions in different areas of food abundance) and (vi) report the sample use in the main study, only those fitness traits that were considered
size. Only compensatory growth as a result of dietary restriction, as ‘direct’ fitness components (survival, number of offspring, lifetime
opposed to temperature or seasonal effects, was considered to be of reproductive output, fertility and hatchability) were included. This
interest for this study. Most studies used quantitative restriction (i.e. limited the fitness analysis to 94 comparisons.
limiting food amounts); however, we included alternative restriction
methods, namely diet dilution with indigestible material, intermittent
STATISTICAL ANALYSIS
feeding and clutch size manipulations. Diets that limited only one
nutrient, e.g. methionine, were not included because they were con- Data were analysed using R 2.13.1 (R Development Core Team,
sidered less ecologically relevant. 2011) and S-PLUS 8.0.4 (TIBCO; http://www.tibco.com/). Hedges’ d
To investigate the effects of compensatory growth, we collected was considered the most appropriate measure of effect size because it
body size data of restricted and control-fed animals prior to restric- estimates the magnitude of the difference between two groups. Using
tion, at the end of restriction and at the end of the experiment (after this measure, as opposed to Cohen’s d, controls for an upward bias
realimentation). We primarily collected mass data; however, length caused by small sample size (Nakagawa & Cuthill 2007).The variance
data were also recorded where available and ranged from total body of d was used as the weighting value for statistical analysis of the
length to tarsus length. This size estimate would allow us to compare overall effects, such that estimates with lower variance and therefore
compensatory growth in mass with compensatory growth in length, more reliability (either from more consistent results or a greater sam-
to account for the possibility of confounding fat accumulation. ple size) contributed more to the model (i.e. meta-analytical models).
Growth rates were calculated from the three experimental time points Estimating growth slopes from the two periods observed (A: from
indicated, as discussed in more detail below (Fig. 1). the beginning to the end of restriction; and B: from the end of restric-
Studies that used farmed or model laboratory species were also tion to the end of the experiment, i.e. the realimentation period;
required to have some fitness-related trait included, such as mortality Fig. 1) required additional analysis. Details are provided in the Sup-
or fecundity (Table S1, Supporting information). We decided this porting information.
criterion was the best way to optimize the variety of species included, To answer the questions which we had initially posed, three
because some studies of taxonomically diverse wild-caught animals separate sets of meta-analytical models were required: (i) models

 2012 The Authors. Journal of Animal Ecology  2012 British Ecological Society, Journal of Animal Ecology, 81, 583–593
586 K. L. Hector & S. Nakagawa

investigating size differences at each point of interest (before restric- Best models were then selected by running the full model using the
tion, after restriction and after realimentation at the end of the exper- maximum likelihood method, as opposed to restricted maximum
iment); (ii) the effect the dietary treatment had on fitness; (iii) the likelihood (REML), because the changes in the AIC (Akaike infor-
effect of the dietary treatment on growth rates as measured by the mation criterion, measuring model goodness of fit) values are more
growth slopes during the different periods, A or B (Fig. 1). Nicieza & relevant under maximum likelihood (Pinheiro & Bates 2000). Least
Alvarez (2009) identified size-dependent growth as a prominent con- significant moderators were, then, sequentially removed until the
founding variable in analysis of compensatory growth (i.e. control AIC value was no longer lowered, and the best model was then
animals are larger and therefore grow at a slower rate, so comparison reverted to REML so that the effect size estimates could be used in
with post-restriction growth of small restricted animals is biased). As interpreting the data. Full models are reported including all the mod-
such, we compared the ‘compensatory growth’ (period B) of the erators initially included in the supporting information. Some of the
previously restricted animals with both period A and period B of the moderators were not directly given from the original papers and were
control animals (to give two estimates of growth rates initiated at dif- instead calculated by combining data. Namely, the relative age, dura-
ferent sizes; Fig. 1). We also compared differences when the studies tion of restriction and duration of realimentation were all calculated
were limited to those which had controls showing only linear growth as a percentage of the reported maximum longevity of the species
(87 studies, Fig. 1). By only comparing studies where controls and, importantly, were converted to the natural log-scale before
showed linear growth over the entire experiment, we could be certain being scaled and centred. The proportion of the duration of realimen-
that the size dependence of growth would not be a concern. Identifi- tation to restriction was calculated by dividing the length of the
cation of linear growth is described in further detail in the supporting former by the latter.
information. The four-group analysis consisted of the three classes and one phy-
Each meta-analysis was analysed as a linear mixed-effects model, lum, which constituted the majority of our data set: mammals
using a modified version of the method described by Nakagawa et al. (n = 38), birds (n = 39), fish (n = 91) and arthropods (consisting
(2007). In all analyses, we controlled for experimental design effects of classes Insecta, Arachnida, Branchiopoda and Malacostraca,
by including both paper identity and taxonomic class as random fac- n = 30). We analysed these by performing exactly the same model-
tors (the former nested within the later). Both null models (model ling process as above, with study as a random factor, and incorpo-
with the intercept; classically considered and referred to as meta- rated the groups as an interaction term with each moderator. Only
analysis) and scaled best models with moderators (often referred to interaction effects which decreased the AIC value (if any) remained
as meta-regression; Table 1) were used to interpret our results. For in the final best model. As a multilevel factor, the other groups were
scaling, all continuous moderators had the mean subtracted from regressed against ‘fish’, selected because it had the highest sample
each value and were divided by two times the standard deviation. size. For estimating overall intercepts by group, each group was fitted
This method of scaling and centring allows the outputs of the model as the reference level.
to be more fairly interpreted (Gelman 2008; Schielzeth 2010). Binary
variables were left unscaled, and sex (which had the values: both,
VALIDATION OF META-ANALYTICAL TECHNIQUES
male and female) was analysed with ‘both’ as the reference variable
so that the male and female output values could be interpreted as the To ensure that the analyses were not affected by publication bias,
effect of looking at only one sex as opposed to mixed-sex experiments funnel plots of the data sets used in each analysis were visually
(Table 1). inspected (Fig. 2, see Fig. S5, Supporting information for fine scale).

Table 1. Descriptions of the moderators incorporated in the models. ‘Relative’ means calculated as a percentage of the average longevity for
each species (see Table S4, Supporting information)

Moderator Meaning Type Scale

Intercept Overall effect of d Continuous Size measures:

+ large restricted group ) small restricted group
Fitness:
+higher fitness for restricted
)lower fitness for restricted
Slopes:
+slower growth for restricted
)faster growth for restricted
Type Measurement type Binary +Mass )length
DegreeDR Degree of restriction Continuous +More food )less food
Age Relative age of restriction onset Continuous +Older )younger (log-scale)
DurRes Relative duration of restriction Continuous +Longer )shorter (log-scale)
PropReal Proportion of duration of realimentation:restriction Continuous +Longer )shorter
Farmed Farmed or laboratory colony Binary +True )False
Female Female compared to mixed sex Binary +True )False
Male Male compared to mixed sex Binary +True )False
ClutchSize Restricted by clutch size manipulation Binary +True )False
IF Restricted by intermittent feeding Binary +True )False
Dilution Restricted by dilution of diet Binary +True )False
AdLibReal Restricted group fed ad libitum for realimentation Binary +True )False
Mortality Compared to reproduction Binary +True )False

 2012 The Authors. Journal of Animal Ecology  2012 British Ecological Society, Journal of Animal Ecology, 81, 583–593
 Compensatory and catch-up growth 587

(a) Pre-restriction (b) Post-restriction (c) Post-realimentation (d) Fitness

5 5 5 5
0 0 0 0
−5 −5 −5 −5
−10 −10 −10 −10
d

d
−15 −15 −15 −15
−20 −20 −20 −20
−25 −25 −25 −25
0
5
10
15
20
25

0
5
10
15
20
25

0
5
10
15
20
25

0
5
10
15
20
25
Precision (1/SE) Precision (1/SE) Precision (1/SE) Precision (1/SE)

(e) A vs. B (f) A vs. B linear (g) B vs. B (h) B vs. B linear
20 20 20 20
15 15 15 15
10 10 10 10
5 5 5 5
d

d
0 0 0 0
−5 −5 −5 −5
−10 −10 −10 −10
−15 −15 −15 −15
0
5
10
15
20
25

0
5
10
15
20
25

0
5
10
15
20
25

0
5
10
15
20
25
Precision (1/SE) Precision (1/SE) Precision (1/SE) Precision (1/SE)

Fig. 2. Funnel plots showing the distribution of the effect sizes (d) extracted from each study plotted against the precision of the study (1 ⁄ SE).
Asymmetry around the null intercept (dashed line) can indicate either publication bias or true biological heterogeneity. For a-h, n = 212, 218,
226, 94, 226, 87, 226 and 87, respectively.

No obvious asymmetry was detected in the plots, other than that Results
which was expected to reflect true biological heterogeneity in the
post-restriction and post-realimentation data sets (Egger et al. META-ANALYSIS: NULL MODELS
1997; Fig. 2b–c). I2 is a measure of heterogeneity in a meta-analysis
and can be interpreted as the percentage of the total variability in The null models reflect the overall effect size of d without con-
a set of effect sizes owing to between-study variability (Huedo- sidering any moderators but still accounting for paper iden-
Medina et al. 2006). The high I2 values reported reflect the high tity nested within taxonomic class. The pre-restriction
degree of inconsistency across studies (Higgins et al. 2003; intercept (i.e. meta-analytic mean) shows that there was no
Table 2). The large contribution of between-study differences (as size difference between treatment and control groups prior to
opposed to between-class differences) to the overall heterogeneity
restriction (d ± SE: 0Æ02 ± 0Æ03, t = 0Æ642, d.f. = 124,
reflects the importance of moderators in controlling for differing
P = 0Æ522; Fig. 3a). By the end of restriction, the restricted
experimental techniques (Table 2). In most cases, the best model
group was statistically significantly smaller than controls
actually increased the heterogeneity compared to the null model,
although the AIC value of all best models was lower than the null (d ± SE: )1Æ78 ± 0Æ02, t = )9Æ57, d.f. = 130, P < 0Æ0001;
models. Fig. 3a). By the end of the experiment, the animals in the
restricted group had failed to ‘catch-up’ to the size of controls
Table 2. Values of the I2 statistic reflecting the heterogeneity of each and were still statistically significantly smaller (d ± SE:
null and scaled best model. The relative contribution of taxonomic )0Æ54 ± 0Æ10, t = )5Æ416, d.f. = 138, P < 0Æ0001; Fig. 3a)
class and source paper to model heterogeneity is also reported as well
than those in the control group. There was a statistically sig-
as the AIC value calculated by maximum likelihood for each model
nificant negative effect on fitness components as a result of
Model I2 Class Paper AIC the treatment (d ± SE: )0Æ25 ± 0Æ09, t = )2Æ824,
d.f. = 57, P = 0Æ0065; Fig. 3a).
Pre-restriction Null 56Æ57 0Æ00 56Æ57 32 Meta-analytic models using standardized differences
Best 54Æ89 0Æ00 54Æ89 31 (Hedges’ d) between the slopes showed more uncertain esti-
Post-restriction Null 98Æ83 0Æ64 98Æ19 1626
mates (Fig. 3b) than the models using d between point esti-
Best 98Æ9 2Æ80 96Æ10 1429
Post-realimentation Null 95Æ81 1Æ80 94Æ02 909 mates (Fig. 3a). The intercept for period A vs. period B (see
Best 97Æ74 17Æ41 80Æ33 772 Fig. 2a for explanation) shows that there was no detectable
Fitness Null 95Æ63 0Æ00 95Æ63 836 difference between the initial growth of the controls and the
Best 99Æ67 10Æ33 89Æ33 601 growth rate of the restricted animals during realimentation
A vs. B Null 99Æ52 0Æ00 99Æ52 1744
(d ± SE: 0Æ25 ± 0Æ27, t = 0Æ894, d.f. = 124, P = 0Æ373;
Best 99Æ47 0Æ01 99Æ46 1702
A vs. B linear Null 96Æ14 27Æ03 69Æ12 301 Fig. 3b). However, when only experiments during the linear
Best 96Æ01 23Æ77 72Æ23 290 growth phase of the controls were considered (A vs. B linear,
B vs. B Null 95Æ93 8Æ07 87Æ86 665 Fig. 1), the restricted animals appeared to show ‘compensa-
Best 96Æ11 19Æ54 76Æ58 651 tory growth’ at a faster rate than the early growth of controls
B vs. B linear Null 93Æ24 5Æ22 88Æ01 196
(d ± SE: )0Æ45 ± 0Æ13, t = )2Æ090, d.f. = 49, P = 0Æ042;
Best 94Æ71 63Æ25 31Æ46 189
Fig. 3b). This effect was also found to be true when

 2012 The Authors. Journal of Animal Ecology  2012 British Ecological Society, Journal of Animal Ecology, 81, 583–593
588 K. L. Hector & S. Nakagawa

(a) only provided for intercepts, but full statistical information

Pre-restriction for moderators is available in Tables S6–S13 (Supporting
information).
Post-restriction There were no statistically significant moderators for the
pre-restriction analysis, suggesting experiments were indeed
Post-realimentation
fair and both treatment groups started at a similar size
Fitness (d ± SE: )0Æ04 ± 0Æ04, t = )1Æ006, d.f. = 124, P = 0Æ316;
Fig. 4a). Following restriction, the restricted animals were
−2 −1 0 1 2 found to be statistically significantly smaller than controls
Hedges’ d (d ± SE: )1Æ70 ± 0Æ32, t = )5Æ461, d.f. = 125, P <
0Æ0001; Fig. 4b). The size difference between the control and
(b) the treatment groups was heavily dependent on how severe
A vs. B the restriction was, how long the diet lasted and how young
the animals were when restriction began. Restriction also
A vs. B linear showed a greater effect on mass than on length. Intermittent
feeding and clutch size manipulation appeared to be less
B vs. B
effective than other diet methods.
B vs. B linear Contrary to the results of the null model, animals did
achieve ‘catch-up’ growth when taking into account a num-
−2 −1 0 1 2 ber of variables, most significantly the duration and severity
Hedges’ d of the restriction diet (d ± SE: )0Æ08 ± 0Æ31, t = )0Æ244,
d.f. = 134, P = 0Æ807; Fig. 4c). Restricted animals were
Fig. 3. The intercepts of the null models for each of the eight meta- more likely to attain the same length as controls than the
analyses, grouped as (a) standardized difference between point esti- same mass, although this could be because length was not as
mates in size and fitness components and (b) standardized difference
severely affected by restriction.
between growth slopes. For the former, a negative value of d indicates
smaller size or decreased fitness in the restricted group compared with In support of the null model, fitness components were
the control group. For the latter, a negative value indicates that the found to be negatively affected by the diet treatment overall
restricted group grew faster than the control group during the period (d ± SE: )1Æ17 ± 0Æ48, t = )2Æ436, d.f. = 52, P = 0Æ018;
of interest (see Fig. 1). Confidence intervals that span zero indicate Fig. 4d). The effect size of the impact on fitness components
no statistically significant effect at a = 0Æ05.
was even larger when accounting for moderators, especially
duration of restriction. Mortality was more negatively
comparing the period B growth of all the control animals (B affected by treatment than reproduction, and negative fitness
vs. B, Fig. 1), although this result would be expected if size- consequences were more apparent after longer periods of
dependent growth is causing overinflated estimations of com- realimentation. Males and females in single-sex experiments
pensatory growth, as Nicieza & Alvarez (2009) suggest had significantly higher fitness components than mixed-sex
(d ± SE: )0Æ30 ± 0Æ13, t = )2Æ379, d.f. = 124, P = 0Æ019; experiments. Animals on intermittent feeding regimes were
Fig. 3b). Their claim of false detection is supported by the find- less likely to show deficits in fitness components. These
ing that the difference for B vs. B is eliminated when only linear results were extracted from ‘direct’ fitness-related trait mea-
control growth is considered (B vs. B linear, Fig. 1; d ± SE: surements only. The analysis of fitness with ‘indirect’ fitness
)0Æ15 ± 0Æ13, t = )1Æ172, d.f. = 49, P = 0Æ247; Fig. 3b). components found no overall effect on fitness-related traits
(see Supporting information).
With experimental methods taken into account, both A vs.
META-REGRESSION: SCALED BEST MODELS
B and the linear selection of A vs. B (Fig. 1) showed no statis-
These models take into account the effect of various modera- tically significant difference in growth rate between period A
tors, as selected by the methods described earlier. To inter- of the controls and realimentation of the restricted group (A
pret the output of these models, it is necessary to remember vs. B d ± SE: 0Æ64 ± 0Æ34, t = 1Æ913, d.f. = 117, P =
that the models are scaled. The intercepts reflect the overall 0Æ058; Fig. 5a; A vs. B linear d ± SE: )0Æ20 ± 0Æ22,
effect size based on the mean values of continuous variables t = )0Æ894, d.f. = 47, P = 0Æ376; Fig. 5b). Both analyses
and the default level of binary variables. For example, an also agreed that compensatory growth was more likely in
intercept that had dietary restriction and farming as modera- mass than length and a longer duration of restriction was
tors would reflect the overall effect based on a moderate found to decrease the likelihood of compensatory growth.
degree of restriction and non-farmed animals. The modera- For all studies, a longer proportion of realimentation
tor values provided could then be interpreted as the slope of decreased the chance of detecting compensatory growth. Sin-
the degree of restriction centred on this intercept and the gle-sex experiments were more likely to show compensatory
effect when farmed animals are considered. Inclusion in the growth than mixed-sex studies. Intermittent feeding and
best model did not necessarily produce statistical significance clutch size manipulation were found to cause slower growth
in all moderators. For ease of reading, statistical values are of the restricted group compared to initial growth of controls.
 2012 The Authors. Journal of Animal Ecology  2012 British Ecological Society, Journal of Animal Ecology, 81, 583–593
 Compensatory and catch-up growth 589

(a) Smaller Larger (b) Smaller Larger

Pre-restriction Post-restriction

Clutch Size: T
Farmed: T IF: T
Farmed: T
DurRes: longer
Age: older
Intercept Degree DR: more food
Type: mass
Intercept
−3 −2 −1 0 1 2 3 −3 −2 −1 0 1 2 3
Hedges’ d Hedges’ d

(c) Smaller Larger (d) Poorer fitness Higher fitness

Post-realimentation Fitness

AdLibReal: T IF: T
Male
DurRes: longer Female
Degree DR: more food PropReal: longer
Type: mass DurRes: longer
Fitness Trait: mortality
Intercept Intercept
−3 −2 −1 0 1 2 3 −4 −2 0 2 4
Hedges’ d Hedges’ d

Fig. 4. Coefficient plots from meta-regression analyses showing the effects of the treatment on the restricted group in comparison with the con-
trol group (a) before the experiment, (b) after restriction, (c) after realimentation and (d) on fitness components. The intercept shows the overall
model outcome, and the other points describe the contribution of each moderator to the intercept. Confidence intervals that span zero indicate
no significant effect at a = 0Æ05. For graphs a–d, n = 212, 218, 226 and 94, respectively. For binary variables, the coefficient shows the impact
of the named effect in contrast to the default value, upon which the intercept depends. The continuous variables indicate the slopes of these vari-
ables against the effect of d. The intercept is based on intermediate values of these variables. For example, degree of restriction tends to have a
positive slope in relation to d, indicating that more food for the restricted group means a smaller difference in the typically negative relationship
between control and restricted size (d). See also Table 1 for information on moderators and Tables S6–S9 (Supporting information) for relevant
statistics to these analyses.

(a) Faster growth Slower growth (b) Faster growth Slower growth
A vs. B A vs. B linear

IF: T
ClutchSize: T DurRes: longer
Male
Female Type: mass
PropReal: longer
DurRes: longer
Type: mass Intercept
Intercept

−3 −2 −1 0 1 2 3 −3 −2 −1 0 1 2 3
Hedges’ d Hedges’ d

(c) Faster growth Slower growth (d) Faster growth Slower growth
B vs. B B vs. B linear

IF: T AdLibReal: T
IF: T
ClutchSize: T ClutchSize: T
Type: mass DurRes: longer
Type: mass
Intercept Intercept

−3 −2 −1 0 1 2 3 −3 −2 −1 0 1 2 3
Hedges’ d Hedges’ d

Fig. 5. Coefficient plots showing the treatment effects on the growth rate of the restricted group after realimentation (‘compensatory growth’)
compared to (a) the control group during the restriction period (period A) from all studies, (b) the control group during period A where only
control groups in the linear growth phase were included, (c) the control group during the realimentation period (period B) from all studies and
(d) the linear control groups during period B (see Fig. 1). The intercept shows the overall model outcome, and the other points describe the con-
tribution of each moderator to the intercept. Confidence intervals that span zero indicate no statistically significant effect at a = 0Æ05. For
graphs a–d, n = 226, 87, 226 and 87, respectively. For further explanation on interpreting moderator values, see Fig. 4 and Table 1. Full statisti-
cal information for each meta-regression analysis is reported in Tables S10–S13 (Supporting information).

Controlling for experimental methods shows that compen- ment, with compensatory growth in mass being slightly
satory growth is achieved by the restricted group when com- weaker. The evidence for compensatory growth could be
pared with the post-restriction growth of the controls interpreted as false detection, because faster period B growth
(d ± SE: )0Æ57 ± 0Æ17, t = )3Æ312, d.f. = 121, P = 0Æ001; is predicted given size-dependent growth of the controls.
Fig. 5c). This finding was moderated by the type of measure- However, analysis of linear control growth shows that com-

 2012 The Authors. Journal of Animal Ecology  2012 British Ecological Society, Journal of Animal Ecology, 81, 583–593
590 K. L. Hector & S. Nakagawa

pensatory growth is a robust effect (d ± SE: )0Æ80 ± 0Æ37, distinction between mortality and reproduction fitness-
t = )2Æ173, d.f. = 45, P = 0Æ035; Fig. 5d). The linear anal- related traits was no longer important. Full statistical infor-
ysis again demonstrates that mass has weaker compensatory mation is provided in the supplementary material
growth than length and that, like the full analysis, clutch size (Table S14, Supporting information).
and intermittent feeding are unlikely to lead to compensatory
growth. The linear analysis also indicates that longer restric-
Discussion
tion periods are less likely to result in compensatory growth.
These meta-analyses and meta-regression analyses support
the basic assumptions of altered growth patterns in response
FOUR-GROUP ANALYSIS
to early dietary restrictions: animals may reach the same size
The main effects from each of the eight meta-analyses sum- as controls after a period of restriction (catch-up growth) and
marize the relative outcome of dietary restriction for each can grow at a faster than optimal rate to achieve this end
group (mammals, birds, fish and arthropods; Fig. 6). It is (compensatory growth; Fig. S1, Supporting information). In
clear that while all groups are much smaller after restriction, reality, the occurrence of both growth patterns is largely
arthropods are most likely to catch-up. The difference dependent on the experimental methods used, as indicated by
between controls and previously restricted animals is much the large effects of a number of the moderators included and
less substantial after realimentation in all groups. Mammals the high heterogeneity among studies (Table 2). In particu-
were the only group with detectable negative consequences lar, the inclusion of such important moderators as the degree
for fitness components. Compensatory growth in birds was and duration of restriction changed the net result of the post-
statistically significantly faster than control growth in period realimentation analysis, which previously indicated that ani-
A, while mammal compensatory growth was statistically sig- mals do not ‘catch-up’ to the size of controls after restriction
nificantly faster than control growth in period B. Interaction (Fig. 3a). Likewise, the meta-analyses for growth slopes
effects showed that mammals were much more sensitive to showed equivocal results prior to the inclusion of the preemi-
the severity of the diet than other groups, but less sensitive to nent moderators ‘measurement type’ and the duration or
the duration of restriction. The effect of restriction on mass proportion of restriction (Fig. 3b). Biologically, these experi-
was greater than on length in arthropods, while both mea- mental factors have an obvious influence on the growth of
sures were affected in the other groups. After a longer period animals. In meta-regression analyses, we were able to
of realimentation, more negative consequences for fitness account for these interstudy differences by scaling and cen-
components are apparent in mammals, while arthropods are tring data on intermediate values, in order to observe the true
more likely to avoid fitness costs. Longer restriction periods effects of ‘catch-up growth’ and ‘compensatory growth’.
resulted in much faster compensatory growth in birds (com- Our results also support the claim that compensatory
pared to period A controls) than in other groups. Analysis growth comes at a cost to fitness components, at least in
including groups did eliminate some of the moderators ‘direct’ measurements of survival and reproductive output
included in the complete data set analysis. For example, the (Mangel & Munch 2005; Fig. 4d, see Supporting information

Arthropods Birds Fish Mammals

Post-restriction

Post-realimentation

Fitness

A vs. B

A vs. B (linear)

B vs. B

B vs. B (linear)

−8 −6 −4 −2 0 −8 −6 −4 −2 0 −8 −6 −4 −2 0 −8 −6 −4 −2 0
Hedges’ d Hedges’ d Hedges’ d Hedges’ d

Fig. 6. Coefficient plots showing the intercepts of each of the four groups (arthropods, n = 30; birds, n = 39; fish, n = 91; mammals, n = 38)
when included in the original eight meta-regression analyses as interaction terms (pre-restriction size not shown because there no group effects).
Intercepts illustrate the relative strength of size, fitness or growth rate effects between groups. Negative intercepts with confidence intervals not
including zero indicate smaller size (post-restriction, post-realimentation), decreased fitness-related traits (fitness) and faster post-restriction
growth compared with controls during restriction (A vs. B, A vs. B linear) or realimentation (B vs. B, B vs. B linear).

 2012 The Authors. Journal of Animal Ecology  2012 British Ecological Society, Journal of Animal Ecology, 81, 583–593
 Compensatory and catch-up growth 591

for ‘indirect’ measurements). The magnitude of the effect of observed in the controls during period A (A vs. B, Fig. 5a).
duration of restriction on fitness components is particularly Although this difference could be explained by size-depen-
notable. Longer than average periods of restriction appear to dent growth, the linear comparison makes the growth curve
have a tremendous negative impact on the fitness-related of the controls an unlikely explanation (Fig. 5d). Instead, it is
traits of an animal. However, as a cautionary note, all studies possible that time is the important factor.
included in our meta-analyses were conducted in controlled The rapid growth after restriction is faster than optimal
environments to ensure that feeding amounts could be growth for the age of the animals, but not faster than early on
manipulated and monitored. Therefore, the subjects were in life. The importance of age-dependent growth has been
spared the real life trials of predator avoidance, mate selec- shown in response to the natural dietary restriction of alpine
tion and pathogen exposure, which all clearly have great swift nestlings, Apus melba (Bize, Metcalfe & Roulin 2006).
bearing on the fitness outcomes of wild animals. Studies such Chicks that were older during poor weather (and therefore
as Johnsson & Bohlin (2006), where restricted trout regained poor nutrition) delayed fledging to compensate for wing
body condition but showed increased winter mortality in the growth, which had been sacrificed in place of mass during
wild, are much better indicators of the fitness costs of com- restriction. In contrast, younger chicks lost mass during
pensatory growth. undernutrition, but were able to reach the same size as con-
Overall, the analyses show that intermittent feeding and trols at fledging by rapid weight gain. Bize, Metcalfe & Rou-
clutch size manipulation are poor ways of conducting com- lin (2006) explain the difference in strategies on the varying
pensatory research. Manipulating clutch size seems to cause ‘developmental windows’ of tissues, organs and morphologi-
little restrictive effect on offspring and, therefore, does little cal traits, which in turn alter the stress resistance and tissue
to instigate compensatory growth (Alonso-Alvarez et al. preservation hierarchy at different ages. The negative conse-
2006). It is possible that intermittent feeding is ineffective quences of compensatory growth on fitness components are,
because the animals simply gorge on feeding days, a response therefore, due to rapid growth at an age when physiological
known as hyperphagia (Ali & Wootton 2000). Recent restrictions inhibit growth. For example, one proposed mech-
research on the effect of dietary restriction on longevity anism for the later life health consequences of catch-up
found that whether animals were multigeneration laboratory growth in humans is that the critical period for muscle
dwellers or not had a significant impact on their response to growth is around 30 weeks in utero (Robinson & Barker
the diet treatment (Nakagawa et al. in press). It is, therefore, 2002). As there is little cell replication after birth, any gain in
surprising that whether animals were farmed or not had little mass is likely to give a disproportionately high body fat ratio,
impact on their potential for catch-up or compensatory which leads to similar symptoms as in obesity, although an
growth. The heavy dependence of results on the experimental individual may not be obese. Surprisingly, the present study
protocol used suggests that a call be made for unified restric- found that the age (or timing) that restriction was initiated
tion protocols, at least for similar species. Such protocols had little effect on the size and growth rate outcomes.
would allow the evolutionary context of ‘compensatory When interpreting the outcome of our four-group interac-
growth’ to be much more easily interpreted. Sogard & Olla tion models, it is important to remember that the way in
(2002) give a good example of this strategy in their compari- which we incorporated groups into the models is conducive
son of compensatory growth between sablefish, Anoplopoma to detecting trends within our larger study and reflects con-
fimbria, and walleye pollock, Theragra chalcogramma. trasts between the groups rather than outright evidence of
Despite the similar environmental and ecological niches of catching up or not. Most of the heterogeneity in our analysis
these two pelagic fish species, the researchers were able to is derived from interstudy not interclass differences (Table 2),
detect a marked difference in the mechanisms of compensa- and I2 values were mostly over 90% in this additional analy-
tory growth and evolved baseline growth rates by exposing sis (Table S14, Supporting information). As expected, birds
the fish to identical dietary regimes. and mammals invest more in accelerating growth because of
It is interesting that, on the whole, the linear analyses their determinate growth. The severity of the diet for mam-
(Fig. 5b,d) are actually in agreement with the results derived mals and the duration of restriction in birds were important
from the complete slopes analyses (Fig. 5a,c). These results moderators, reflecting their lower tolerance for reduced fat
suggest that the impact of researchers not taking size-depen- stores owing to their higher metabolic rate as endotherms
dent growth into account when analysing compensatory (Wang, Hung & Randall 2006).
growth is less influential when analysing a data set with such In contrast, fish seemed relatively unaffected by restriction,
large statistical power. In smaller, empirical studies like the with little alteration to growth rate and no detectable conse-
papers including in this meta-analysis, researchers need to be quences for fitness components. Fish have indeterminate
aware of this confounding factor, as Nicieza & Alvarez growth and are therefore under less pressure than endo-
(2009) suggest. The conflict between results derived by com- therms to rapidly achieve a large final size. Their lower
parison of the restricted group’s realimentation with part A metabolism, as ectotherms, also makes them more resistant
and part B of the controls relates to size-dependent growth to starvation and reduces the cost of depleting fat stores
concerns (Fig. 1). Compensatory growth after restriction (Wang, Hung & Randall 2006). The saltatory growth of
was faster than the rate of controls at the same time (B vs. B, arthropods (‘leaping’ growth after each ecdysis) may account
Fig. 5c), but, importantly, it was no different than the rate for arthropods being the group most likely catch-up growth
 2012 The Authors. Journal of Animal Ecology  2012 British Ecological Society, Journal of Animal Ecology, 81, 583–593
592 K. L. Hector & S. Nakagawa

in spite of there being no perceptible increase in growth rate the causes and consequences of ‘catch-up growth’ and ‘com-
as developmental time can be increased (Dmitriew & Rowe pensatory growth’ can be compared and contrasted without
2007). Dietary restriction results in a small larval instar, but confusion. Our results confirm that compensatory growth is
if ecdysis is delayed until a threshold size is reached, it stands a measurable, repeatable and widespread response to early
to reason that restricted animals will reach the same size as life dietary restriction. With the implementation of more
controls at the end (Mirth & Riddiford 2007). The cost may standardized restriction protocols, there can be greater
come in terms of seasonal timing; taking longer to reach emphasis on uncovering the proximate mechanisms, as well
the adult phenotype may result in previously restricted as unravelling the ecological and evolutionary consequences,
arthropods missing out on seasonal mating, peak food of compensatory growth across taxa.
abundance or risking pond desiccation (De Block, McPeek
& Stoks 2008). Taken together, these results suggest that
Acknowledgements
what we refer to as ‘compensatory growth’ is best exempli-
fied by mammals, while ‘catch-up growth’ best describes The authors are grateful to A. Senior, E. Santos, M. Lagisz, I. Cleasby, J. M.
Gaillard and one anonymous reviewer for providing helpful comments on the
arthropods. manuscript. K.H. was financially supported by the University of Otago Post-
Our analysis has identified a number of influences at work graduate Award and Publishing Bursary from the Graduate Research Com-
in compensatory growth experiments, and we have noted mittee. S.N. is supported by NRCGD and the Marsden Fund.

some areas that could be improved in future. To truly under-

stand whether it was the compensatory growth causing the
fitness component deficits and not simply the early life experi- References
ence of dietary restriction, it is necessary to compare fitness- Ali, M., Nicieza, A. & Wootton, R.J. (2003) Compensatory growth in fishes: a
response to growth depression. Fish and Fisheries, 4, 147–190.
related traits with permanently restricted animals (a treat-
Ali, M. & Wootton, R.J. (2000) Pattern of hyperphagia in immature three-
ment group that is almost always missing in empirical work). spined sticklebacks after short-term food deprivation. Journal of Fish Biol-
Alternatively, the correlation between growth rate and fitness ogy, 56, 648–653.
Alonso-Alvarez, C., Bertrand, S., Devevey, G., Prost, J., Faivre, B., Chastel, O.
deficits for each individual may disentangle the effects of
& Sorci, G. (2006) An experimental manipulation of life-history trajectories
restriction and compensatory growth (Krause & Naguib and resistance to oxidative stress. Evolution, 60, 1913–1924.
2011). We also recommend that researchers have an estab- Arendt, J.D. (1997) Adaptive intrinsic growth rates: an integration across taxa.
Quarterly Review of Biology, 72, 149–177.
lished understanding of the study animal’s growth patterns
Bize, P., Metcalfe, N.B. & Roulin, A. (2006) Catch-up growth strategies differ
to decide the length of the various treatment periods. The between body structures: interactions between age and structure-specific
study should occur during the most linear phase of an ani- growth in wild nestling alpine swifts. Functional Ecology, 20, 857–864.
Blanckenhorn, W.U. (2005) Behavioral causes and consequences of sexual size
mal’s growth pattern (if possible), aiming to end with the ces-
dimorphism. Ethology, 111, 977–1016.
sation of linear growth in the controls. This procedure Bol, V.V., Delattre, A.I., Reusens, B., Raes, M. & Remacle, C. (2009) Forced
eliminates the influence of size-dependent growth and allows catch-up growth after fetal protein restriction alters the adipose tissue gene
expression program leading to obesity in adult mice. American Journal of
comparisons between species. Further measurements after
Physiology-Regulatory Integrative and Comparative Physiology, 297, 291–
this period are encouraged as indicators of the impact of 299.
compensatory growth on fitness components. Duration and Cottrell, E.C. & Ozanne, S.E. (2008) Early life programming of obesity and
metabolic disease. Physiology & Behavior, 94, 17–28.
degree of restriction were very influential moderators, and we
De Block, M., McPeek, M.A. & Stoks, R. (2008) Life history plasticity to com-
suggest researchers apply ecologically relevant levels of bined time and biotic constraints in Lestes damselflies from vernal and tem-
restriction to best represent likely fitness outcomes. We also porary ponds. Oikos, 117, 908–916.
Dmitriew, C.M. (2011) The evolution of growth trajectories: what limits
suggest that if normal quantitative dietary restriction is not
growth rate? Biological Reviews, 86, 97–116.
possible, researchers use qualitative restriction, such as dilut- Dmitriew, C. & Rowe, L. (2007) Effects of early resource limitation and com-
ing food with cornhusks, because this alternative method pensatory growth on lifetime fitness in the ladybird beetle (Harmonia axyri-
dis). Journal of Evolutionary Biology, 20, 1298–1310.
most closely resembled quantitative restriction (unlike inter-
Egger, M., Smith, G.D., Schneider, M. & Minder, C. (1997) Bias in meta-analy-
mittent feeding and clutch size manipulations). sis detected by a simple, graphical test. British Medical Journal, 315, 629–
In conclusion, these analyses voice strong support for the 634.
Gelman, A. (2008) Scaling regression inputs by dividing by two standard devia-
theoretical concept of compensatory growth as an evolved
tions. Statistics in Medicine, 27, 2865–2873.
type of phenotypic plasticity common to many disparate ani- Higgins, J.P.T., Thompson, S.G., Deeks, J.J. & Altman, D.G. (2003) Measur-
mal taxa. Dependent on the degree of restriction and the per- ing inconsistency in meta-analyses. British Medical Journal, 327, 557–560.
Huedo-Medina, T.B., Sanchez-Meca, J., Marin-Martinez, F. & Botella, J.
iod within which animals were restricted, both catch-up
(2006) Assessing heterogeneity in meta-analysis: Q statistic or I2 index? Psy-
growth and compensatory growth are feasible outcomes fol- chological Methods, 11, 193–206.
lowing realimentation (Figs 4 and 5). As predicted, there is a Jackson, C.M. (1937) Recovery of rats upon refeeding after prolonged suppres-
sion of growth by underfeeding. Anatomical Record, 68, 371–381.
cost to fitness components as a trade-off for the larger size,
Jobling, M. (2010) Are compensatory growth and catch-up growth two sides of
which is manifested as higher mortality and lower reproduc- the same coin? Aquaculture International, 18, 501–510.
tive output. Further work is needed to disentangle how much Johnsson, J.I. & Bohlin, T. (2006) The cost of catching up: increased winter
mortality following structural growth compensation in the wild. Proceedings
of this cost is generated purely by faster than optimal growth
of the Royal Society B-Biological Sciences, 273, 1281–1286.
and not by restriction. An additional contribution of the Kozlowski, J. (1996) Optimal allocation of resources explains interspecific life-
present study may be clarification of the terminology history patterns in animals with indeterminate growth. Proceedings of the
Royal Society B-Biological Sciences, 263, 559–566.
(Fig. S1, Supporting information; cf. Jobling 2010), such that
 2012 The Authors. Journal of Animal Ecology  2012 British Ecological Society, Journal of Animal Ecology, 81, 583–593
 Compensatory and catch-up growth 593

Krause, E.T. & Naguib, M. (2011) Compensatory growth affects exploratory Speakman, J.R. (2005) Body size, energy metabolism and lifespan. Journal of
behaviour in zebra finches, Taeniopygia guttata. Animal Behaviour, 81, 1295– Experimental Biology, 208, 1717–1730.
1300. Wang, T., Hung, C.C.Y. & Randall, D.J. (2006) The comparative physiology
Mangel, M. & Munch, S.B. (2005) A life-history perspective on short- and of food deprivation: from feast to famine. Annual Review of Physiology, 68,
long-term consequences of compensatory growth. American Naturalist, 166, 223–251.
155–176. West, G.B., Brown, J.H. & Enquist, B.J. (2001) A general model for ontoge-
Mangel, M. & Stamps, J. (2001) Trade-offs between growth and mortality and netic growth. Nature, 413, 628–631.
the maintenance of individual variation in growth. Evolutionary Ecology Wilbur, H.M. & Rudolf, V.H.W. (2006) Life-history evolution in uncertain
Research, 3, 583–593. environments: bet hedging in time. American Naturalist, 168, 398–411.
Metcalfe, N.B. & Monaghan, P. (2001) Compensation for a bad start: grow Wilson, P.N. & Osbourn, D.F. (1960) Compensatory growth after undernutri-
now, pay later? Trends in Ecology & Evolution, 16, 254–260. tion in mammals and birds. Biological Reviews of the Cambridge Philosophi-
Metcalfe, N.B. & Monaghan, P. (2003) Growth versus lifespan: perspectives cal Society, 35, 324–363.
from evolutionary ecology. Experimental Gerontology, 38, 935–940. Yearsley, J.M., Kyriazakis, I. & Gordon, I.J. (2004) Delayed costs of growth
Mirth, C.K. & Riddiford, L.M. (2007) Size assessment and growth control: and compensatory growth rates. Functional Ecology, 18, 563–570.
how adult size is determined in insects. Bioessays, 29, 344–355.
Monteiro, P.O.A. & Victora, C.G. (2005) Rapid growth in infancy and child- Received 14 June 2011; accepted 8 December 2011
hood and obesity in later life: a systematic review. Obesity Reviews, 6, 143– Handling Editor: Fanie Pelletier
154.
Nakagawa, S. & Cuthill, I.C. (2007) Effect size, confidence interval and statisti-
cal significance: a practical guide for biologists. Biological Reviews, 82, 591–
605. Supporting Information
Nakagawa, S., Lagisz, M., Hector, K.L. & Spencer, H.G. (in press) Compara-
tive and meta-analytic insights into life-extension via dietary restriction. Additional Supporting Information may be found in the online ver-
Aging Cell. sion of this article.
Nakagawa, S., Ockendon, N., Gillespie, D.O.S., Hatchwell, B.J. & Burke, T. Data S1. Additional methodological details.
(2007) Assessing the function of house sparrows’ bib size using a flexible Data S2. Indirect fitness analysis.
meta-analysis method. Behavioral Ecology, 18, 831–840.
Data S3. Linear slopes selection.
Nicieza, A. & Alvarez, D. (2009) Statistical analysis of structural compensa-
torygrowth: how can we reduce the rate of false detection? Oecologia, 159, Fig. S1. Catch-up and compensatory growth.
27–39. Fig. S2. Fitness coefficient plot.
Nir, I., Nitsan, Z., Dunnington, E.A. & Siegel, P.B. (1996) Aspects of food Fig. S3. Taxonomic class contributions to direct and inclusive fitness
intake restriction in young domestic fowl: metabolic and genetic consider-
analyses.
ations. Worlds Poultry Science Journal, 52, 251–266.
Oli, M.K., Hepp, G.R. & Kennamer, R.A. (2002) Fitness consequences of Fig. S4. Comparison of linear slope selection methods.
delayed maturity in female wood ducks. Evolutionary Ecology Research, 4, Fig. S5. Fine-scale funnel plots.
563–576. Table S1. Fitness traits.
Pinheiro, J.C. & Bates, D.M. (2000) Mixed Effects Models in S and S-Plus. Table S2. Study details.
Springer Verlag, New York, NY.
Table S3. Taxonomic information.
Porrello, E.R., Bell, J.R., Schertzer, J.D., Curl, C.L., McMullen, J.R., Mellor,
K.M., Ritchie, R.H., Lynch, G.S., Harrap, S.B., Thomas, W.G. & Del- Table S4. Longevity data.
bridge, L.M.D. (2009) Heritable pathologic cardiac hypertrophy in adult- Table S5. Collinearity of moderators.
hood is preceded by neonatal cardiac growth restriction. American Journal Tables S6–S13. Statistical information from meta-regression analy-
of Physiology-Regulatory Integrative and Comparative Physiology, 296, 672–
ses.
680.
R Development Core Team (2011) R: A Language and Environment for Statisti- Table S14. Statistical information from four-group analysis.
cal Computing. R Foundation for Statistical Computing, Vienna, Austria. As a service to our authors and readers, this journal provides sup-
ISBN 3-900051-07-0, http://www.R-project.org. porting information supplied by the authors. Such materials may be
Robinson, S.M. & Barker, D.J.P. (2002) Coronary heart disease: a disorder of re-organized for online delivery, but are not copy-edited or typeset.
growth. Proceedings of the Nutrition Society, 61, 537–542.
Technical support issues arising from supporting information (other
Schielzeth, H. (2010) Simple means to improve the interpretability of regression
coefficients. Methods in Ecology and Evolution, 1, 103–113. than missing files) should be addressed to the authors.
Sogard, S.M. & Olla, B.L. (2002) Contrasts in the capacity and underlying
mechanisms for compensatory growth in two pelagic marine fishes. Marine
Ecology-Progress Series, 243, 165–177.

 2012 The Authors. Journal of Animal Ecology  2012 British Ecological Society, Journal of Animal Ecology, 81, 583–593